diff --git a/.gitattributes b/.gitattributes
index bed0738c7eeb449bca98b5d2f33c89a1ee56349a..55e3a15d0b58cf70db47b29f4f36992187827302 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -58,3 +58,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
# Video files - compressed
*.mp4 filter=lfs diff=lfs merge=lfs -text
*.webm filter=lfs diff=lfs merge=lfs -text
+runs/test-model/checkpoint-16/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+runs/test-model/checkpoint-160/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+runs/test-model/checkpoint-240/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+runs/test-model/checkpoint-320/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+runs/test-model/checkpoint-80/tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/runs/test-model/checkpoint-16/chat_template.jinja b/runs/test-model/checkpoint-16/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..699ff8df401fe4788525e9c1f9b86a99eadd6230
--- /dev/null
+++ b/runs/test-model/checkpoint-16/chat_template.jinja
@@ -0,0 +1,85 @@
+{%- 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 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.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set content = message.content %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in message.content %}
+ {%- set content = message.content.split('')[-1].lstrip('\n') %}
+ {%- set reasoning_content = message.content.split('')[0].rstrip('\n').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' }}
+ {{- message.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/runs/test-model/checkpoint-16/config.json b/runs/test-model/checkpoint-16/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..44f7b99e220689bde520b89c11fdd83d1b5348de
--- /dev/null
+++ b/runs/test-model/checkpoint-16/config.json
@@ -0,0 +1,32 @@
+{
+ "architectures": [
+ "LlamaForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "float32",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 512,
+ "initializer_range": 0.02,
+ "intermediate_size": 1536,
+ "max_position_embeddings": 2048,
+ "mlp_bias": false,
+ "model_type": "llama",
+ "num_attention_heads": 4,
+ "num_hidden_layers": 20,
+ "num_key_value_heads": 4,
+ "pad_token_id": 151645,
+ "pretraining_tp": 1,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 10000.0,
+ "rope_type": "default"
+ },
+ "tie_word_embeddings": true,
+ "transformers_version": "5.5.0",
+ "use_cache": false,
+ "vocab_size": 151671
+}
diff --git a/runs/test-model/checkpoint-16/generation_config.json b/runs/test-model/checkpoint-16/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f962c4bfc66159cdeb7ba836cbbd79365e9304f3
--- /dev/null
+++ b/runs/test-model/checkpoint-16/generation_config.json
@@ -0,0 +1,11 @@
+{
+ "_from_model_config": true,
+ "eos_token_id": [
+ 151645
+ ],
+ "output_attentions": false,
+ "output_hidden_states": false,
+ "pad_token_id": 151645,
+ "transformers_version": "5.5.0",
+ "use_cache": true
+}
diff --git a/runs/test-model/checkpoint-16/model.safetensors b/runs/test-model/checkpoint-16/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..372973f6d4e308faf967872e0c9d7fc682c13a70
--- /dev/null
+++ b/runs/test-model/checkpoint-16/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad234226fb19745b417959595347233d3cecf4c6a81089093b1e374c24cc6f18
+size 583356232
diff --git a/runs/test-model/checkpoint-16/optimizer.pt b/runs/test-model/checkpoint-16/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..08a1b8c58b1deea38f79a9ef63063bf331f7ed91
--- /dev/null
+++ b/runs/test-model/checkpoint-16/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c89956656d7ed8be0b6ebffc0830e7a4c57a38efa25bcef43371fa6067c7fddb
+size 1166825803
diff --git a/runs/test-model/checkpoint-16/rng_state.pth b/runs/test-model/checkpoint-16/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0a969df10527f1f8224738f4b67111c3ae7c6989
--- /dev/null
+++ b/runs/test-model/checkpoint-16/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f90a5f2ed6d30ebfd28acaade74bb026468970204fa5b02eda67c20566c1a648
+size 14709
diff --git a/runs/test-model/checkpoint-16/scheduler.pt b/runs/test-model/checkpoint-16/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c1a7a9a760d4aac6c7cbd71e2284b245fc87440b
--- /dev/null
+++ b/runs/test-model/checkpoint-16/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86479ad780523296a76407f01b89fed8d9f0bbf000e92b244362f11d7085b30f
+size 1465
diff --git a/runs/test-model/checkpoint-16/tokenizer.json b/runs/test-model/checkpoint-16/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..b83d93986de8ecfc4343943be1c86e9532682c15
--- /dev/null
+++ b/runs/test-model/checkpoint-16/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
+size 11424079
diff --git a/runs/test-model/checkpoint-16/tokenizer_config.json b/runs/test-model/checkpoint-16/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1d6bb391bb8fbc446de576db5220ec91c98a150
--- /dev/null
+++ b/runs/test-model/checkpoint-16/tokenizer_config.json
@@ -0,0 +1,18 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|l2r_pred|>",
+ "<|r2l_pred|>"
+ ],
+ "is_local": false,
+ "model_max_length": 131072,
+ "pad_token": "<|im_end|>",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/runs/test-model/checkpoint-16/trainer_state.json b/runs/test-model/checkpoint-16/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..b596c20d863c77d98a9a3fc4d6762d2aa891b948
--- /dev/null
+++ b/runs/test-model/checkpoint-16/trainer_state.json
@@ -0,0 +1,146 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 1.0,
+ "eval_steps": 500,
+ "global_step": 16,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0625,
+ "grad_norm": 0.4655202031135559,
+ "learning_rate": 0.0,
+ "loss": 2.00244402885437,
+ "step": 1
+ },
+ {
+ "epoch": 0.125,
+ "grad_norm": 0.4847956895828247,
+ "learning_rate": 3e-05,
+ "loss": 2.008800506591797,
+ "step": 2
+ },
+ {
+ "epoch": 0.1875,
+ "grad_norm": 0.4177771806716919,
+ "learning_rate": 6e-05,
+ "loss": 1.9685651063919067,
+ "step": 3
+ },
+ {
+ "epoch": 0.25,
+ "grad_norm": 0.3550562858581543,
+ "learning_rate": 8.999999999999999e-05,
+ "loss": 1.9168976545333862,
+ "step": 4
+ },
+ {
+ "epoch": 0.3125,
+ "grad_norm": 0.23293137550354004,
+ "learning_rate": 0.00012,
+ "loss": 1.8660516738891602,
+ "step": 5
+ },
+ {
+ "epoch": 0.375,
+ "grad_norm": 0.13890701532363892,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.8209998607635498,
+ "step": 6
+ },
+ {
+ "epoch": 0.4375,
+ "grad_norm": 0.08890970796346664,
+ "learning_rate": 0.00017999999999999998,
+ "loss": 1.806854248046875,
+ "step": 7
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.061611492186784744,
+ "learning_rate": 0.00021000000000000004,
+ "loss": 1.793172836303711,
+ "step": 8
+ },
+ {
+ "epoch": 0.5625,
+ "grad_norm": 0.05786389857530594,
+ "learning_rate": 0.00024,
+ "loss": 1.7882877588272095,
+ "step": 9
+ },
+ {
+ "epoch": 0.625,
+ "grad_norm": 0.03947344794869423,
+ "learning_rate": 0.00027,
+ "loss": 1.7696826457977295,
+ "step": 10
+ },
+ {
+ "epoch": 0.6875,
+ "grad_norm": 0.029783710837364197,
+ "learning_rate": 0.00030000000000000003,
+ "loss": 1.7767844200134277,
+ "step": 11
+ },
+ {
+ "epoch": 0.75,
+ "grad_norm": 0.025729672983288765,
+ "learning_rate": 0.00033,
+ "loss": 1.7827998399734497,
+ "step": 12
+ },
+ {
+ "epoch": 0.8125,
+ "grad_norm": 0.020592188462615013,
+ "learning_rate": 0.00035999999999999997,
+ "loss": 1.766276478767395,
+ "step": 13
+ },
+ {
+ "epoch": 0.875,
+ "grad_norm": 0.01877395436167717,
+ "learning_rate": 0.00039000000000000005,
+ "loss": 1.7644124031066895,
+ "step": 14
+ },
+ {
+ "epoch": 0.9375,
+ "grad_norm": 0.015196867287158966,
+ "learning_rate": 0.00042000000000000007,
+ "loss": 1.771270990371704,
+ "step": 15
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.014262701384723186,
+ "learning_rate": 0.00045,
+ "loss": 1.755500316619873,
+ "step": 16
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 800,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 50,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": true,
+ "should_log": false,
+ "should_save": false,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 214470802538496.0,
+ "train_batch_size": 8,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/runs/test-model/checkpoint-16/training_args.bin b/runs/test-model/checkpoint-16/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1c106baba014e8cee356170e7acaf136a1e2efb8
--- /dev/null
+++ b/runs/test-model/checkpoint-16/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
+size 5265
diff --git a/runs/test-model/checkpoint-160/chat_template.jinja b/runs/test-model/checkpoint-160/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..699ff8df401fe4788525e9c1f9b86a99eadd6230
--- /dev/null
+++ b/runs/test-model/checkpoint-160/chat_template.jinja
@@ -0,0 +1,85 @@
+{%- 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 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.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set content = message.content %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in message.content %}
+ {%- set content = message.content.split('')[-1].lstrip('\n') %}
+ {%- set reasoning_content = message.content.split('')[0].rstrip('\n').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' }}
+ {{- message.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/runs/test-model/checkpoint-160/config.json b/runs/test-model/checkpoint-160/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..44f7b99e220689bde520b89c11fdd83d1b5348de
--- /dev/null
+++ b/runs/test-model/checkpoint-160/config.json
@@ -0,0 +1,32 @@
+{
+ "architectures": [
+ "LlamaForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "float32",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 512,
+ "initializer_range": 0.02,
+ "intermediate_size": 1536,
+ "max_position_embeddings": 2048,
+ "mlp_bias": false,
+ "model_type": "llama",
+ "num_attention_heads": 4,
+ "num_hidden_layers": 20,
+ "num_key_value_heads": 4,
+ "pad_token_id": 151645,
+ "pretraining_tp": 1,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 10000.0,
+ "rope_type": "default"
+ },
+ "tie_word_embeddings": true,
+ "transformers_version": "5.5.0",
+ "use_cache": false,
+ "vocab_size": 151671
+}
diff --git a/runs/test-model/checkpoint-160/generation_config.json b/runs/test-model/checkpoint-160/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f962c4bfc66159cdeb7ba836cbbd79365e9304f3
--- /dev/null
+++ b/runs/test-model/checkpoint-160/generation_config.json
@@ -0,0 +1,11 @@
+{
+ "_from_model_config": true,
+ "eos_token_id": [
+ 151645
+ ],
+ "output_attentions": false,
+ "output_hidden_states": false,
+ "pad_token_id": 151645,
+ "transformers_version": "5.5.0",
+ "use_cache": true
+}
diff --git a/runs/test-model/checkpoint-160/model.safetensors b/runs/test-model/checkpoint-160/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c8ff471a063db3760368f91de9b4996dc3be5e31
--- /dev/null
+++ b/runs/test-model/checkpoint-160/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6899b9b051c84815d868fe05ac75c958ebdf75f3fe4035b6e492d85d995cf3f9
+size 583356232
diff --git a/runs/test-model/checkpoint-160/optimizer.pt b/runs/test-model/checkpoint-160/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5a16b970c734eaf28231a6a49a8f679d70b788a2
--- /dev/null
+++ b/runs/test-model/checkpoint-160/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50f9bf0dcffceaad8da7969f8eda8b324301d2b259b53016e9f6574ec729641e
+size 1166825803
diff --git a/runs/test-model/checkpoint-160/rng_state.pth b/runs/test-model/checkpoint-160/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d0540ee8ecbdbb40e958e769ceec4a39e1a9e872
--- /dev/null
+++ b/runs/test-model/checkpoint-160/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbc34b8b1329e714980cee98a4611b1026e736159845ee7101938dae8e999619
+size 14709
diff --git a/runs/test-model/checkpoint-160/scheduler.pt b/runs/test-model/checkpoint-160/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..440e0e394c5fb6e3a8b26d9cd65559667b396df9
--- /dev/null
+++ b/runs/test-model/checkpoint-160/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a1f53431b5ed2197da9763c32428013d318c9c2ab414a9f1e06b94d72caad84
+size 1465
diff --git a/runs/test-model/checkpoint-160/tokenizer.json b/runs/test-model/checkpoint-160/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..b83d93986de8ecfc4343943be1c86e9532682c15
--- /dev/null
+++ b/runs/test-model/checkpoint-160/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
+size 11424079
diff --git a/runs/test-model/checkpoint-160/tokenizer_config.json b/runs/test-model/checkpoint-160/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1d6bb391bb8fbc446de576db5220ec91c98a150
--- /dev/null
+++ b/runs/test-model/checkpoint-160/tokenizer_config.json
@@ -0,0 +1,18 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|l2r_pred|>",
+ "<|r2l_pred|>"
+ ],
+ "is_local": false,
+ "model_max_length": 131072,
+ "pad_token": "<|im_end|>",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/runs/test-model/checkpoint-160/trainer_state.json b/runs/test-model/checkpoint-160/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..efd5c55a93d0e2818a3b6d2007779329e50241ae
--- /dev/null
+++ b/runs/test-model/checkpoint-160/trainer_state.json
@@ -0,0 +1,1226 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 10.0,
+ "eval_steps": 500,
+ "global_step": 160,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0625,
+ "grad_norm": 0.4655202031135559,
+ "learning_rate": 0.0,
+ "loss": 2.00244402885437,
+ "step": 1
+ },
+ {
+ "epoch": 0.125,
+ "grad_norm": 0.4847956895828247,
+ "learning_rate": 3e-05,
+ "loss": 2.008800506591797,
+ "step": 2
+ },
+ {
+ "epoch": 0.1875,
+ "grad_norm": 0.4177771806716919,
+ "learning_rate": 6e-05,
+ "loss": 1.9685651063919067,
+ "step": 3
+ },
+ {
+ "epoch": 0.25,
+ "grad_norm": 0.3550562858581543,
+ "learning_rate": 8.999999999999999e-05,
+ "loss": 1.9168976545333862,
+ "step": 4
+ },
+ {
+ "epoch": 0.3125,
+ "grad_norm": 0.23293137550354004,
+ "learning_rate": 0.00012,
+ "loss": 1.8660516738891602,
+ "step": 5
+ },
+ {
+ "epoch": 0.375,
+ "grad_norm": 0.13890701532363892,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.8209998607635498,
+ "step": 6
+ },
+ {
+ "epoch": 0.4375,
+ "grad_norm": 0.08890970796346664,
+ "learning_rate": 0.00017999999999999998,
+ "loss": 1.806854248046875,
+ "step": 7
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.061611492186784744,
+ "learning_rate": 0.00021000000000000004,
+ "loss": 1.793172836303711,
+ "step": 8
+ },
+ {
+ "epoch": 0.5625,
+ "grad_norm": 0.05786389857530594,
+ "learning_rate": 0.00024,
+ "loss": 1.7882877588272095,
+ "step": 9
+ },
+ {
+ "epoch": 0.625,
+ "grad_norm": 0.03947344794869423,
+ "learning_rate": 0.00027,
+ "loss": 1.7696826457977295,
+ "step": 10
+ },
+ {
+ "epoch": 0.6875,
+ "grad_norm": 0.029783710837364197,
+ "learning_rate": 0.00030000000000000003,
+ "loss": 1.7767844200134277,
+ "step": 11
+ },
+ {
+ "epoch": 0.75,
+ "grad_norm": 0.025729672983288765,
+ "learning_rate": 0.00033,
+ "loss": 1.7827998399734497,
+ "step": 12
+ },
+ {
+ "epoch": 0.8125,
+ "grad_norm": 0.020592188462615013,
+ "learning_rate": 0.00035999999999999997,
+ "loss": 1.766276478767395,
+ "step": 13
+ },
+ {
+ "epoch": 0.875,
+ "grad_norm": 0.01877395436167717,
+ "learning_rate": 0.00039000000000000005,
+ "loss": 1.7644124031066895,
+ "step": 14
+ },
+ {
+ "epoch": 0.9375,
+ "grad_norm": 0.015196867287158966,
+ "learning_rate": 0.00042000000000000007,
+ "loss": 1.771270990371704,
+ "step": 15
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.014262701384723186,
+ "learning_rate": 0.00045,
+ "loss": 1.755500316619873,
+ "step": 16
+ },
+ {
+ "epoch": 1.0,
+ "eval_loss": 0.8736820816993713,
+ "eval_runtime": 30.2743,
+ "eval_samples_per_second": 3.369,
+ "eval_steps_per_second": 0.429,
+ "step": 16
+ },
+ {
+ "epoch": 1.0625,
+ "grad_norm": 0.01042763702571392,
+ "learning_rate": 0.00048,
+ "loss": 1.7556875944137573,
+ "step": 17
+ },
+ {
+ "epoch": 1.125,
+ "grad_norm": 0.009123577736318111,
+ "learning_rate": 0.00051,
+ "loss": 1.753890037536621,
+ "step": 18
+ },
+ {
+ "epoch": 1.1875,
+ "grad_norm": 0.009028254076838493,
+ "learning_rate": 0.00054,
+ "loss": 1.7515414953231812,
+ "step": 19
+ },
+ {
+ "epoch": 1.25,
+ "grad_norm": 0.009630529209971428,
+ "learning_rate": 0.00057,
+ "loss": 1.7465155124664307,
+ "step": 20
+ },
+ {
+ "epoch": 1.3125,
+ "grad_norm": 0.008235524408519268,
+ "learning_rate": 0.0006000000000000001,
+ "loss": 1.7533702850341797,
+ "step": 21
+ },
+ {
+ "epoch": 1.375,
+ "grad_norm": 0.009377758949995041,
+ "learning_rate": 0.00063,
+ "loss": 1.738031029701233,
+ "step": 22
+ },
+ {
+ "epoch": 1.4375,
+ "grad_norm": 0.008636261336505413,
+ "learning_rate": 0.00066,
+ "loss": 1.735835313796997,
+ "step": 23
+ },
+ {
+ "epoch": 1.5,
+ "grad_norm": 0.008001700975000858,
+ "learning_rate": 0.0006900000000000001,
+ "loss": 1.7505712509155273,
+ "step": 24
+ },
+ {
+ "epoch": 1.5625,
+ "grad_norm": 0.007517208810895681,
+ "learning_rate": 0.0007199999999999999,
+ "loss": 1.7330820560455322,
+ "step": 25
+ },
+ {
+ "epoch": 1.625,
+ "grad_norm": 0.00952192209661007,
+ "learning_rate": 0.00075,
+ "loss": 1.7318685054779053,
+ "step": 26
+ },
+ {
+ "epoch": 1.6875,
+ "grad_norm": 0.006763546261936426,
+ "learning_rate": 0.0007800000000000001,
+ "loss": 1.7329007387161255,
+ "step": 27
+ },
+ {
+ "epoch": 1.75,
+ "grad_norm": 0.006364013999700546,
+ "learning_rate": 0.0008100000000000001,
+ "loss": 1.7285535335540771,
+ "step": 28
+ },
+ {
+ "epoch": 1.8125,
+ "grad_norm": 0.007770394906401634,
+ "learning_rate": 0.0008400000000000001,
+ "loss": 1.7320656776428223,
+ "step": 29
+ },
+ {
+ "epoch": 1.875,
+ "grad_norm": 0.006639651022851467,
+ "learning_rate": 0.00087,
+ "loss": 1.7304966449737549,
+ "step": 30
+ },
+ {
+ "epoch": 1.9375,
+ "grad_norm": 0.008190952241420746,
+ "learning_rate": 0.0009,
+ "loss": 1.7399260997772217,
+ "step": 31
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 0.008188036270439625,
+ "learning_rate": 0.00093,
+ "loss": 1.7375996112823486,
+ "step": 32
+ },
+ {
+ "epoch": 2.0,
+ "eval_loss": 0.860374927520752,
+ "eval_runtime": 30.3049,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 32
+ },
+ {
+ "epoch": 2.0625,
+ "grad_norm": 0.005768604576587677,
+ "learning_rate": 0.00096,
+ "loss": 1.7309706211090088,
+ "step": 33
+ },
+ {
+ "epoch": 2.125,
+ "grad_norm": 0.006822533905506134,
+ "learning_rate": 0.00099,
+ "loss": 1.7321696281433105,
+ "step": 34
+ },
+ {
+ "epoch": 2.1875,
+ "grad_norm": 0.009974546730518341,
+ "learning_rate": 0.00102,
+ "loss": 1.7290048599243164,
+ "step": 35
+ },
+ {
+ "epoch": 2.25,
+ "grad_norm": 0.00856814906001091,
+ "learning_rate": 0.00105,
+ "loss": 1.7231603860855103,
+ "step": 36
+ },
+ {
+ "epoch": 2.3125,
+ "grad_norm": 0.011694240383803844,
+ "learning_rate": 0.00108,
+ "loss": 1.728348731994629,
+ "step": 37
+ },
+ {
+ "epoch": 2.375,
+ "grad_norm": 0.005738088861107826,
+ "learning_rate": 0.00111,
+ "loss": 1.7308874130249023,
+ "step": 38
+ },
+ {
+ "epoch": 2.4375,
+ "grad_norm": 0.010038000531494617,
+ "learning_rate": 0.00114,
+ "loss": 1.7276837825775146,
+ "step": 39
+ },
+ {
+ "epoch": 2.5,
+ "grad_norm": 0.01586989127099514,
+ "learning_rate": 0.00117,
+ "loss": 1.7368175983428955,
+ "step": 40
+ },
+ {
+ "epoch": 2.5625,
+ "grad_norm": 0.16198667883872986,
+ "learning_rate": 0.0012000000000000001,
+ "loss": 1.730246663093567,
+ "step": 41
+ },
+ {
+ "epoch": 2.625,
+ "grad_norm": 0.01653525047004223,
+ "learning_rate": 0.00123,
+ "loss": 1.7403852939605713,
+ "step": 42
+ },
+ {
+ "epoch": 2.6875,
+ "grad_norm": 0.01241418905556202,
+ "learning_rate": 0.00126,
+ "loss": 1.7306722402572632,
+ "step": 43
+ },
+ {
+ "epoch": 2.75,
+ "grad_norm": 0.022560030221939087,
+ "learning_rate": 0.00129,
+ "loss": 1.7223260402679443,
+ "step": 44
+ },
+ {
+ "epoch": 2.8125,
+ "grad_norm": 0.019568368792533875,
+ "learning_rate": 0.00132,
+ "loss": 1.7304983139038086,
+ "step": 45
+ },
+ {
+ "epoch": 2.875,
+ "grad_norm": 0.016755228862166405,
+ "learning_rate": 0.00135,
+ "loss": 1.7223992347717285,
+ "step": 46
+ },
+ {
+ "epoch": 2.9375,
+ "grad_norm": 0.02981281839311123,
+ "learning_rate": 0.0013800000000000002,
+ "loss": 1.7094981670379639,
+ "step": 47
+ },
+ {
+ "epoch": 3.0,
+ "grad_norm": 0.02506650611758232,
+ "learning_rate": 0.00141,
+ "loss": 1.7054555416107178,
+ "step": 48
+ },
+ {
+ "epoch": 3.0,
+ "eval_loss": 0.8416222333908081,
+ "eval_runtime": 30.3027,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 48
+ },
+ {
+ "epoch": 3.0625,
+ "grad_norm": 0.02260257676243782,
+ "learning_rate": 0.0014399999999999999,
+ "loss": 1.6888771057128906,
+ "step": 49
+ },
+ {
+ "epoch": 3.125,
+ "grad_norm": 0.02620658650994301,
+ "learning_rate": 0.00147,
+ "loss": 1.6639578342437744,
+ "step": 50
+ },
+ {
+ "epoch": 3.1875,
+ "grad_norm": 0.047354575246572495,
+ "learning_rate": 0.0015,
+ "loss": 1.6736350059509277,
+ "step": 51
+ },
+ {
+ "epoch": 3.25,
+ "grad_norm": 0.04594860598444939,
+ "learning_rate": 0.0015300000000000001,
+ "loss": 1.6675360202789307,
+ "step": 52
+ },
+ {
+ "epoch": 3.3125,
+ "grad_norm": 0.08131621032953262,
+ "learning_rate": 0.0015600000000000002,
+ "loss": 1.6574077606201172,
+ "step": 53
+ },
+ {
+ "epoch": 3.375,
+ "grad_norm": 0.028789900243282318,
+ "learning_rate": 0.00159,
+ "loss": 1.6299148797988892,
+ "step": 54
+ },
+ {
+ "epoch": 3.4375,
+ "grad_norm": 0.2623404860496521,
+ "learning_rate": 0.0016200000000000001,
+ "loss": 1.6315665245056152,
+ "step": 55
+ },
+ {
+ "epoch": 3.5,
+ "grad_norm": 0.056828465312719345,
+ "learning_rate": 0.0016500000000000002,
+ "loss": 1.6279325485229492,
+ "step": 56
+ },
+ {
+ "epoch": 3.5625,
+ "grad_norm": 0.0549125075340271,
+ "learning_rate": 0.0016800000000000003,
+ "loss": 1.6549303531646729,
+ "step": 57
+ },
+ {
+ "epoch": 3.625,
+ "grad_norm": 0.07287958264350891,
+ "learning_rate": 0.00171,
+ "loss": 1.606963872909546,
+ "step": 58
+ },
+ {
+ "epoch": 3.6875,
+ "grad_norm": 0.07426343113183975,
+ "learning_rate": 0.00174,
+ "loss": 1.6250534057617188,
+ "step": 59
+ },
+ {
+ "epoch": 3.75,
+ "grad_norm": 0.1364986002445221,
+ "learning_rate": 0.0017699999999999999,
+ "loss": 1.5964961051940918,
+ "step": 60
+ },
+ {
+ "epoch": 3.8125,
+ "grad_norm": 0.15007159113883972,
+ "learning_rate": 0.0018,
+ "loss": 1.6019010543823242,
+ "step": 61
+ },
+ {
+ "epoch": 3.875,
+ "grad_norm": 0.08447311073541641,
+ "learning_rate": 0.00183,
+ "loss": 1.5763543844223022,
+ "step": 62
+ },
+ {
+ "epoch": 3.9375,
+ "grad_norm": 0.1681874394416809,
+ "learning_rate": 0.00186,
+ "loss": 1.5769858360290527,
+ "step": 63
+ },
+ {
+ "epoch": 4.0,
+ "grad_norm": 0.09677319973707199,
+ "learning_rate": 0.00189,
+ "loss": 1.534682273864746,
+ "step": 64
+ },
+ {
+ "epoch": 4.0,
+ "eval_loss": 0.7682511806488037,
+ "eval_runtime": 30.4228,
+ "eval_samples_per_second": 3.353,
+ "eval_steps_per_second": 0.427,
+ "step": 64
+ },
+ {
+ "epoch": 4.0625,
+ "grad_norm": 0.42635980248451233,
+ "learning_rate": 0.00192,
+ "loss": 1.5509562492370605,
+ "step": 65
+ },
+ {
+ "epoch": 4.125,
+ "grad_norm": 0.8467249274253845,
+ "learning_rate": 0.0019500000000000001,
+ "loss": 1.5001531839370728,
+ "step": 66
+ },
+ {
+ "epoch": 4.1875,
+ "grad_norm": 0.22631719708442688,
+ "learning_rate": 0.00198,
+ "loss": 1.4554829597473145,
+ "step": 67
+ },
+ {
+ "epoch": 4.25,
+ "grad_norm": 0.05312464386224747,
+ "learning_rate": 0.00201,
+ "loss": 1.4732775688171387,
+ "step": 68
+ },
+ {
+ "epoch": 4.3125,
+ "grad_norm": 0.050442785024642944,
+ "learning_rate": 0.00204,
+ "loss": 1.4176194667816162,
+ "step": 69
+ },
+ {
+ "epoch": 4.375,
+ "grad_norm": 0.18886858224868774,
+ "learning_rate": 0.00207,
+ "loss": 1.4325997829437256,
+ "step": 70
+ },
+ {
+ "epoch": 4.4375,
+ "grad_norm": 0.05258706212043762,
+ "learning_rate": 0.0021,
+ "loss": 1.39802086353302,
+ "step": 71
+ },
+ {
+ "epoch": 4.5,
+ "grad_norm": 0.07482657581567764,
+ "learning_rate": 0.00213,
+ "loss": 1.4000320434570312,
+ "step": 72
+ },
+ {
+ "epoch": 4.5625,
+ "grad_norm": 0.4853312075138092,
+ "learning_rate": 0.00216,
+ "loss": 1.4199151992797852,
+ "step": 73
+ },
+ {
+ "epoch": 4.625,
+ "grad_norm": 0.48888301849365234,
+ "learning_rate": 0.00219,
+ "loss": 1.3049702644348145,
+ "step": 74
+ },
+ {
+ "epoch": 4.6875,
+ "grad_norm": 0.284348726272583,
+ "learning_rate": 0.00222,
+ "loss": 1.3884068727493286,
+ "step": 75
+ },
+ {
+ "epoch": 4.75,
+ "grad_norm": 0.20248942077159882,
+ "learning_rate": 0.0022500000000000003,
+ "loss": 1.3502120971679688,
+ "step": 76
+ },
+ {
+ "epoch": 4.8125,
+ "grad_norm": 0.2714093327522278,
+ "learning_rate": 0.00228,
+ "loss": 1.3195693492889404,
+ "step": 77
+ },
+ {
+ "epoch": 4.875,
+ "grad_norm": 0.2544061243534088,
+ "learning_rate": 0.00231,
+ "loss": 1.3194304704666138,
+ "step": 78
+ },
+ {
+ "epoch": 4.9375,
+ "grad_norm": 0.18475523591041565,
+ "learning_rate": 0.00234,
+ "loss": 1.3006361722946167,
+ "step": 79
+ },
+ {
+ "epoch": 5.0,
+ "grad_norm": 0.6583278775215149,
+ "learning_rate": 0.00237,
+ "loss": 1.2868517637252808,
+ "step": 80
+ },
+ {
+ "epoch": 5.0,
+ "eval_loss": 0.6618342399597168,
+ "eval_runtime": 30.1233,
+ "eval_samples_per_second": 3.386,
+ "eval_steps_per_second": 0.432,
+ "step": 80
+ },
+ {
+ "epoch": 5.0625,
+ "grad_norm": 0.04499061033129692,
+ "learning_rate": 0.0024000000000000002,
+ "loss": 1.25992751121521,
+ "step": 81
+ },
+ {
+ "epoch": 5.125,
+ "grad_norm": 0.05609860643744469,
+ "learning_rate": 0.0024300000000000003,
+ "loss": 1.2764487266540527,
+ "step": 82
+ },
+ {
+ "epoch": 5.1875,
+ "grad_norm": 0.03665148839354515,
+ "learning_rate": 0.00246,
+ "loss": 1.2908766269683838,
+ "step": 83
+ },
+ {
+ "epoch": 5.25,
+ "grad_norm": 0.17949481308460236,
+ "learning_rate": 0.00249,
+ "loss": 1.253080129623413,
+ "step": 84
+ },
+ {
+ "epoch": 5.3125,
+ "grad_norm": 0.14031393826007843,
+ "learning_rate": 0.00252,
+ "loss": 1.2648870944976807,
+ "step": 85
+ },
+ {
+ "epoch": 5.375,
+ "grad_norm": 0.08193210512399673,
+ "learning_rate": 0.00255,
+ "loss": 1.296234130859375,
+ "step": 86
+ },
+ {
+ "epoch": 5.4375,
+ "grad_norm": 0.056413356214761734,
+ "learning_rate": 0.00258,
+ "loss": 1.2784960269927979,
+ "step": 87
+ },
+ {
+ "epoch": 5.5,
+ "grad_norm": 0.09004318714141846,
+ "learning_rate": 0.00261,
+ "loss": 1.2054948806762695,
+ "step": 88
+ },
+ {
+ "epoch": 5.5625,
+ "grad_norm": 0.044756047427654266,
+ "learning_rate": 0.00264,
+ "loss": 1.2707666158676147,
+ "step": 89
+ },
+ {
+ "epoch": 5.625,
+ "grad_norm": 0.049932755529880524,
+ "learning_rate": 0.00267,
+ "loss": 1.2369787693023682,
+ "step": 90
+ },
+ {
+ "epoch": 5.6875,
+ "grad_norm": 0.09624820947647095,
+ "learning_rate": 0.0027,
+ "loss": 1.2419278621673584,
+ "step": 91
+ },
+ {
+ "epoch": 5.75,
+ "grad_norm": 0.17648059129714966,
+ "learning_rate": 0.0027300000000000002,
+ "loss": 1.2992033958435059,
+ "step": 92
+ },
+ {
+ "epoch": 5.8125,
+ "grad_norm": 0.07027823477983475,
+ "learning_rate": 0.0027600000000000003,
+ "loss": 1.2255799770355225,
+ "step": 93
+ },
+ {
+ "epoch": 5.875,
+ "grad_norm": 0.06040462478995323,
+ "learning_rate": 0.0027900000000000004,
+ "loss": 1.136023759841919,
+ "step": 94
+ },
+ {
+ "epoch": 5.9375,
+ "grad_norm": 0.09845872968435287,
+ "learning_rate": 0.00282,
+ "loss": 1.1807364225387573,
+ "step": 95
+ },
+ {
+ "epoch": 6.0,
+ "grad_norm": 0.09086572378873825,
+ "learning_rate": 0.00285,
+ "loss": 1.2171341180801392,
+ "step": 96
+ },
+ {
+ "epoch": 6.0,
+ "eval_loss": 0.6200892329216003,
+ "eval_runtime": 30.1252,
+ "eval_samples_per_second": 3.386,
+ "eval_steps_per_second": 0.432,
+ "step": 96
+ },
+ {
+ "epoch": 6.0625,
+ "grad_norm": 0.11787034571170807,
+ "learning_rate": 0.0028799999999999997,
+ "loss": 1.242072582244873,
+ "step": 97
+ },
+ {
+ "epoch": 6.125,
+ "grad_norm": 0.20093177258968353,
+ "learning_rate": 0.00291,
+ "loss": 1.2056899070739746,
+ "step": 98
+ },
+ {
+ "epoch": 6.1875,
+ "grad_norm": 0.03221331909298897,
+ "learning_rate": 0.00294,
+ "loss": 1.1774635314941406,
+ "step": 99
+ },
+ {
+ "epoch": 6.25,
+ "grad_norm": 0.06112038344144821,
+ "learning_rate": 0.00297,
+ "loss": 1.1831618547439575,
+ "step": 100
+ },
+ {
+ "epoch": 6.3125,
+ "grad_norm": 0.05896962434053421,
+ "learning_rate": 0.003,
+ "loss": 1.219610571861267,
+ "step": 101
+ },
+ {
+ "epoch": 6.375,
+ "grad_norm": 0.4068014323711395,
+ "learning_rate": 0.0029999850445531283,
+ "loss": 1.231442928314209,
+ "step": 102
+ },
+ {
+ "epoch": 6.4375,
+ "grad_norm": 3.013302803039551,
+ "learning_rate": 0.0029999401785137437,
+ "loss": 1.2512195110321045,
+ "step": 103
+ },
+ {
+ "epoch": 6.5,
+ "grad_norm": 0.07003403455018997,
+ "learning_rate": 0.0029998654027855406,
+ "loss": 1.1912157535552979,
+ "step": 104
+ },
+ {
+ "epoch": 6.5625,
+ "grad_norm": 1.1927955150604248,
+ "learning_rate": 0.002999760718874652,
+ "loss": 1.1158745288848877,
+ "step": 105
+ },
+ {
+ "epoch": 6.625,
+ "grad_norm": 0.11684458702802658,
+ "learning_rate": 0.0029996261288896225,
+ "loss": 1.1835001707077026,
+ "step": 106
+ },
+ {
+ "epoch": 6.6875,
+ "grad_norm": 0.07812155783176422,
+ "learning_rate": 0.0029994616355413666,
+ "loss": 1.1712734699249268,
+ "step": 107
+ },
+ {
+ "epoch": 6.75,
+ "grad_norm": 0.19128946959972382,
+ "learning_rate": 0.0029992672421431115,
+ "loss": 1.0906574726104736,
+ "step": 108
+ },
+ {
+ "epoch": 6.8125,
+ "grad_norm": 0.033625371754169464,
+ "learning_rate": 0.002999042952610332,
+ "loss": 1.115525245666504,
+ "step": 109
+ },
+ {
+ "epoch": 6.875,
+ "grad_norm": 0.2518763244152069,
+ "learning_rate": 0.0029987887714606723,
+ "loss": 1.1433348655700684,
+ "step": 110
+ },
+ {
+ "epoch": 6.9375,
+ "grad_norm": 0.037159573286771774,
+ "learning_rate": 0.002998504703813852,
+ "loss": 1.1867601871490479,
+ "step": 111
+ },
+ {
+ "epoch": 7.0,
+ "grad_norm": 0.04564391076564789,
+ "learning_rate": 0.0029981907553915674,
+ "loss": 1.1302828788757324,
+ "step": 112
+ },
+ {
+ "epoch": 7.0,
+ "eval_loss": 0.5897142887115479,
+ "eval_runtime": 31.3671,
+ "eval_samples_per_second": 3.252,
+ "eval_steps_per_second": 0.414,
+ "step": 112
+ },
+ {
+ "epoch": 7.0625,
+ "grad_norm": 0.05205775797367096,
+ "learning_rate": 0.0029978469325173717,
+ "loss": 1.182741641998291,
+ "step": 113
+ },
+ {
+ "epoch": 7.125,
+ "grad_norm": 0.12080052495002747,
+ "learning_rate": 0.002997473242116551,
+ "loss": 1.1457600593566895,
+ "step": 114
+ },
+ {
+ "epoch": 7.1875,
+ "grad_norm": 0.05600045621395111,
+ "learning_rate": 0.0029970696917159833,
+ "loss": 1.1101720333099365,
+ "step": 115
+ },
+ {
+ "epoch": 7.25,
+ "grad_norm": 0.0507950522005558,
+ "learning_rate": 0.0029966362894439874,
+ "loss": 1.1785833835601807,
+ "step": 116
+ },
+ {
+ "epoch": 7.3125,
+ "grad_norm": 0.6396550536155701,
+ "learning_rate": 0.002996173044030159,
+ "loss": 1.1538987159729004,
+ "step": 117
+ },
+ {
+ "epoch": 7.375,
+ "grad_norm": 0.06849977374076843,
+ "learning_rate": 0.0029956799648051934,
+ "loss": 1.1528857946395874,
+ "step": 118
+ },
+ {
+ "epoch": 7.4375,
+ "grad_norm": 0.0685896947979927,
+ "learning_rate": 0.002995157061700702,
+ "loss": 1.1049470901489258,
+ "step": 119
+ },
+ {
+ "epoch": 7.5,
+ "grad_norm": 1.027722716331482,
+ "learning_rate": 0.002994604345249006,
+ "loss": 1.1308050155639648,
+ "step": 120
+ },
+ {
+ "epoch": 7.5625,
+ "grad_norm": 0.21406109631061554,
+ "learning_rate": 0.00299402182658293,
+ "loss": 1.0960861444473267,
+ "step": 121
+ },
+ {
+ "epoch": 7.625,
+ "grad_norm": 0.0503656342625618,
+ "learning_rate": 0.002993409517435574,
+ "loss": 1.1096141338348389,
+ "step": 122
+ },
+ {
+ "epoch": 7.6875,
+ "grad_norm": 0.04800846800208092,
+ "learning_rate": 0.002992767430140078,
+ "loss": 1.1386680603027344,
+ "step": 123
+ },
+ {
+ "epoch": 7.75,
+ "grad_norm": 0.13091202080249786,
+ "learning_rate": 0.002992095577629376,
+ "loss": 1.1129441261291504,
+ "step": 124
+ },
+ {
+ "epoch": 7.8125,
+ "grad_norm": 0.04949796572327614,
+ "learning_rate": 0.0029913939734359306,
+ "loss": 1.106520175933838,
+ "step": 125
+ },
+ {
+ "epoch": 7.875,
+ "grad_norm": 0.07686582207679749,
+ "learning_rate": 0.0029906626316914655,
+ "loss": 1.1572234630584717,
+ "step": 126
+ },
+ {
+ "epoch": 7.9375,
+ "grad_norm": 0.053587451577186584,
+ "learning_rate": 0.0029899015671266773,
+ "loss": 1.125717282295227,
+ "step": 127
+ },
+ {
+ "epoch": 8.0,
+ "grad_norm": 0.08031050115823746,
+ "learning_rate": 0.0029891107950709406,
+ "loss": 1.153022050857544,
+ "step": 128
+ },
+ {
+ "epoch": 8.0,
+ "eval_loss": 0.5910838842391968,
+ "eval_runtime": 30.7073,
+ "eval_samples_per_second": 3.322,
+ "eval_steps_per_second": 0.423,
+ "step": 128
+ },
+ {
+ "epoch": 8.0625,
+ "grad_norm": 0.3520950675010681,
+ "learning_rate": 0.0029882903314519997,
+ "loss": 1.1582988500595093,
+ "step": 129
+ },
+ {
+ "epoch": 8.125,
+ "grad_norm": 0.07153584063053131,
+ "learning_rate": 0.002987440192795645,
+ "loss": 1.1344678401947021,
+ "step": 130
+ },
+ {
+ "epoch": 8.1875,
+ "grad_norm": 0.06336807459592819,
+ "learning_rate": 0.0029865603962253833,
+ "loss": 1.120319128036499,
+ "step": 131
+ },
+ {
+ "epoch": 8.25,
+ "grad_norm": 0.11782127618789673,
+ "learning_rate": 0.0029856509594620912,
+ "loss": 1.1245379447937012,
+ "step": 132
+ },
+ {
+ "epoch": 8.3125,
+ "grad_norm": 0.2592264413833618,
+ "learning_rate": 0.0029847119008236576,
+ "loss": 1.1454271078109741,
+ "step": 133
+ },
+ {
+ "epoch": 8.375,
+ "grad_norm": 0.21255695819854736,
+ "learning_rate": 0.0029837432392246185,
+ "loss": 1.231805443763733,
+ "step": 134
+ },
+ {
+ "epoch": 8.4375,
+ "grad_norm": 15.057058334350586,
+ "learning_rate": 0.0029827449941757698,
+ "loss": 1.1380956172943115,
+ "step": 135
+ },
+ {
+ "epoch": 8.5,
+ "grad_norm": 0.17857126891613007,
+ "learning_rate": 0.00298171718578378,
+ "loss": 1.1401441097259521,
+ "step": 136
+ },
+ {
+ "epoch": 8.5625,
+ "grad_norm": 0.18281924724578857,
+ "learning_rate": 0.002980659834750781,
+ "loss": 1.171494722366333,
+ "step": 137
+ },
+ {
+ "epoch": 8.625,
+ "grad_norm": 0.10490825027227402,
+ "learning_rate": 0.0029795729623739555,
+ "loss": 1.1622605323791504,
+ "step": 138
+ },
+ {
+ "epoch": 8.6875,
+ "grad_norm": 0.344013512134552,
+ "learning_rate": 0.002978456590545104,
+ "loss": 1.1702141761779785,
+ "step": 139
+ },
+ {
+ "epoch": 8.75,
+ "grad_norm": 0.05058368295431137,
+ "learning_rate": 0.0029773107417502055,
+ "loss": 1.2006211280822754,
+ "step": 140
+ },
+ {
+ "epoch": 8.8125,
+ "grad_norm": 0.09775760769844055,
+ "learning_rate": 0.002976135439068965,
+ "loss": 1.1407063007354736,
+ "step": 141
+ },
+ {
+ "epoch": 8.875,
+ "grad_norm": 0.0785030648112297,
+ "learning_rate": 0.0029749307061743475,
+ "loss": 1.1778209209442139,
+ "step": 142
+ },
+ {
+ "epoch": 8.9375,
+ "grad_norm": 0.053056441247463226,
+ "learning_rate": 0.0029736965673321032,
+ "loss": 1.1302082538604736,
+ "step": 143
+ },
+ {
+ "epoch": 9.0,
+ "grad_norm": 0.4020994305610657,
+ "learning_rate": 0.0029724330474002743,
+ "loss": 1.1655206680297852,
+ "step": 144
+ },
+ {
+ "epoch": 9.0,
+ "eval_loss": 0.6083483099937439,
+ "eval_runtime": 30.0453,
+ "eval_samples_per_second": 3.395,
+ "eval_steps_per_second": 0.433,
+ "step": 144
+ },
+ {
+ "epoch": 9.0625,
+ "grad_norm": 0.04755420982837677,
+ "learning_rate": 0.002971140171828702,
+ "loss": 1.1900575160980225,
+ "step": 145
+ },
+ {
+ "epoch": 9.125,
+ "grad_norm": 0.2549609839916229,
+ "learning_rate": 0.002969817966658505,
+ "loss": 1.1891400814056396,
+ "step": 146
+ },
+ {
+ "epoch": 9.1875,
+ "grad_norm": 0.11641769111156464,
+ "learning_rate": 0.0029684664585215618,
+ "loss": 1.1930484771728516,
+ "step": 147
+ },
+ {
+ "epoch": 9.25,
+ "grad_norm": 0.13052672147750854,
+ "learning_rate": 0.00296708567463997,
+ "loss": 1.1905772686004639,
+ "step": 148
+ },
+ {
+ "epoch": 9.3125,
+ "grad_norm": 0.15126636624336243,
+ "learning_rate": 0.0029656756428255017,
+ "loss": 1.1417806148529053,
+ "step": 149
+ },
+ {
+ "epoch": 9.375,
+ "grad_norm": 0.08826905488967896,
+ "learning_rate": 0.00296423639147904,
+ "loss": 1.1802985668182373,
+ "step": 150
+ },
+ {
+ "epoch": 9.4375,
+ "grad_norm": 0.01381592359393835,
+ "learning_rate": 0.002962767949590008,
+ "loss": 1.1525968313217163,
+ "step": 151
+ },
+ {
+ "epoch": 9.5,
+ "grad_norm": 0.04648139327764511,
+ "learning_rate": 0.0029612703467357866,
+ "loss": 1.169856309890747,
+ "step": 152
+ },
+ {
+ "epoch": 9.5625,
+ "grad_norm": 0.014019235037267208,
+ "learning_rate": 0.0029597436130811153,
+ "loss": 1.1442453861236572,
+ "step": 153
+ },
+ {
+ "epoch": 9.625,
+ "grad_norm": 0.03701096400618553,
+ "learning_rate": 0.0029581877793774885,
+ "loss": 1.112680435180664,
+ "step": 154
+ },
+ {
+ "epoch": 9.6875,
+ "grad_norm": 0.030421797186136246,
+ "learning_rate": 0.0029566028769625334,
+ "loss": 1.223894715309143,
+ "step": 155
+ },
+ {
+ "epoch": 9.75,
+ "grad_norm": 0.12336083501577377,
+ "learning_rate": 0.00295498893775938,
+ "loss": 1.220346212387085,
+ "step": 156
+ },
+ {
+ "epoch": 9.8125,
+ "grad_norm": 0.038585737347602844,
+ "learning_rate": 0.0029533459942760174,
+ "loss": 1.2078583240509033,
+ "step": 157
+ },
+ {
+ "epoch": 9.875,
+ "grad_norm": 0.4653286039829254,
+ "learning_rate": 0.002951674079604639,
+ "loss": 1.2035815715789795,
+ "step": 158
+ },
+ {
+ "epoch": 9.9375,
+ "grad_norm": 0.020640619099140167,
+ "learning_rate": 0.0029499732274209777,
+ "loss": 1.1313148736953735,
+ "step": 159
+ },
+ {
+ "epoch": 10.0,
+ "grad_norm": 0.04939777031540871,
+ "learning_rate": 0.002948243471983625,
+ "loss": 1.1253995895385742,
+ "step": 160
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 800,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 50,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": true,
+ "should_log": false,
+ "should_save": false,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2144708025384960.0,
+ "train_batch_size": 8,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/runs/test-model/checkpoint-160/training_args.bin b/runs/test-model/checkpoint-160/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1c106baba014e8cee356170e7acaf136a1e2efb8
--- /dev/null
+++ b/runs/test-model/checkpoint-160/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
+size 5265
diff --git a/runs/test-model/checkpoint-240/chat_template.jinja b/runs/test-model/checkpoint-240/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..699ff8df401fe4788525e9c1f9b86a99eadd6230
--- /dev/null
+++ b/runs/test-model/checkpoint-240/chat_template.jinja
@@ -0,0 +1,85 @@
+{%- 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 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.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set content = message.content %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in message.content %}
+ {%- set content = message.content.split('')[-1].lstrip('\n') %}
+ {%- set reasoning_content = message.content.split('')[0].rstrip('\n').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' }}
+ {{- message.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/runs/test-model/checkpoint-240/config.json b/runs/test-model/checkpoint-240/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..44f7b99e220689bde520b89c11fdd83d1b5348de
--- /dev/null
+++ b/runs/test-model/checkpoint-240/config.json
@@ -0,0 +1,32 @@
+{
+ "architectures": [
+ "LlamaForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "float32",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 512,
+ "initializer_range": 0.02,
+ "intermediate_size": 1536,
+ "max_position_embeddings": 2048,
+ "mlp_bias": false,
+ "model_type": "llama",
+ "num_attention_heads": 4,
+ "num_hidden_layers": 20,
+ "num_key_value_heads": 4,
+ "pad_token_id": 151645,
+ "pretraining_tp": 1,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 10000.0,
+ "rope_type": "default"
+ },
+ "tie_word_embeddings": true,
+ "transformers_version": "5.5.0",
+ "use_cache": false,
+ "vocab_size": 151671
+}
diff --git a/runs/test-model/checkpoint-240/generation_config.json b/runs/test-model/checkpoint-240/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f962c4bfc66159cdeb7ba836cbbd79365e9304f3
--- /dev/null
+++ b/runs/test-model/checkpoint-240/generation_config.json
@@ -0,0 +1,11 @@
+{
+ "_from_model_config": true,
+ "eos_token_id": [
+ 151645
+ ],
+ "output_attentions": false,
+ "output_hidden_states": false,
+ "pad_token_id": 151645,
+ "transformers_version": "5.5.0",
+ "use_cache": true
+}
diff --git a/runs/test-model/checkpoint-240/model.safetensors b/runs/test-model/checkpoint-240/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..6e49e2a01b6a5bdc2fd077f9c0ca20a3db7f2992
--- /dev/null
+++ b/runs/test-model/checkpoint-240/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc0adfeb06ac15912b39f89b2490ede0c5e934af3c98e9e56d6b7f4c10dce44e
+size 583356232
diff --git a/runs/test-model/checkpoint-240/optimizer.pt b/runs/test-model/checkpoint-240/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a12e2cf4c7c8ba7c905ba49b2aa63877ca4cbcf8
--- /dev/null
+++ b/runs/test-model/checkpoint-240/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12d8c230552978e462c54219226052b16dc20d6d949aed1dabee656cf7950546
+size 1166825803
diff --git a/runs/test-model/checkpoint-240/rng_state.pth b/runs/test-model/checkpoint-240/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2726b3edfe08ad43025995f27bed930c424b533c
--- /dev/null
+++ b/runs/test-model/checkpoint-240/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b83457a5d388987769679ffdec584d99ee95c32f932be11346ca60064470ca6e
+size 14709
diff --git a/runs/test-model/checkpoint-240/scheduler.pt b/runs/test-model/checkpoint-240/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cc10d484ecbc83c3f9a9b39e67e369b42f0c2cbe
--- /dev/null
+++ b/runs/test-model/checkpoint-240/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfedbe6d0f52f2cf3df90f0461bcecd255a3bff553a3fe44f96dbec4eb2d024d
+size 1465
diff --git a/runs/test-model/checkpoint-240/tokenizer.json b/runs/test-model/checkpoint-240/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..b83d93986de8ecfc4343943be1c86e9532682c15
--- /dev/null
+++ b/runs/test-model/checkpoint-240/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
+size 11424079
diff --git a/runs/test-model/checkpoint-240/tokenizer_config.json b/runs/test-model/checkpoint-240/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1d6bb391bb8fbc446de576db5220ec91c98a150
--- /dev/null
+++ b/runs/test-model/checkpoint-240/tokenizer_config.json
@@ -0,0 +1,18 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|l2r_pred|>",
+ "<|r2l_pred|>"
+ ],
+ "is_local": false,
+ "model_max_length": 131072,
+ "pad_token": "<|im_end|>",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/runs/test-model/checkpoint-240/trainer_state.json b/runs/test-model/checkpoint-240/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..021ab3feaafec90a04ff99d3e73c997919ca5621
--- /dev/null
+++ b/runs/test-model/checkpoint-240/trainer_state.json
@@ -0,0 +1,1826 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 15.0,
+ "eval_steps": 500,
+ "global_step": 240,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0625,
+ "grad_norm": 0.4655202031135559,
+ "learning_rate": 0.0,
+ "loss": 2.00244402885437,
+ "step": 1
+ },
+ {
+ "epoch": 0.125,
+ "grad_norm": 0.4847956895828247,
+ "learning_rate": 3e-05,
+ "loss": 2.008800506591797,
+ "step": 2
+ },
+ {
+ "epoch": 0.1875,
+ "grad_norm": 0.4177771806716919,
+ "learning_rate": 6e-05,
+ "loss": 1.9685651063919067,
+ "step": 3
+ },
+ {
+ "epoch": 0.25,
+ "grad_norm": 0.3550562858581543,
+ "learning_rate": 8.999999999999999e-05,
+ "loss": 1.9168976545333862,
+ "step": 4
+ },
+ {
+ "epoch": 0.3125,
+ "grad_norm": 0.23293137550354004,
+ "learning_rate": 0.00012,
+ "loss": 1.8660516738891602,
+ "step": 5
+ },
+ {
+ "epoch": 0.375,
+ "grad_norm": 0.13890701532363892,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.8209998607635498,
+ "step": 6
+ },
+ {
+ "epoch": 0.4375,
+ "grad_norm": 0.08890970796346664,
+ "learning_rate": 0.00017999999999999998,
+ "loss": 1.806854248046875,
+ "step": 7
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.061611492186784744,
+ "learning_rate": 0.00021000000000000004,
+ "loss": 1.793172836303711,
+ "step": 8
+ },
+ {
+ "epoch": 0.5625,
+ "grad_norm": 0.05786389857530594,
+ "learning_rate": 0.00024,
+ "loss": 1.7882877588272095,
+ "step": 9
+ },
+ {
+ "epoch": 0.625,
+ "grad_norm": 0.03947344794869423,
+ "learning_rate": 0.00027,
+ "loss": 1.7696826457977295,
+ "step": 10
+ },
+ {
+ "epoch": 0.6875,
+ "grad_norm": 0.029783710837364197,
+ "learning_rate": 0.00030000000000000003,
+ "loss": 1.7767844200134277,
+ "step": 11
+ },
+ {
+ "epoch": 0.75,
+ "grad_norm": 0.025729672983288765,
+ "learning_rate": 0.00033,
+ "loss": 1.7827998399734497,
+ "step": 12
+ },
+ {
+ "epoch": 0.8125,
+ "grad_norm": 0.020592188462615013,
+ "learning_rate": 0.00035999999999999997,
+ "loss": 1.766276478767395,
+ "step": 13
+ },
+ {
+ "epoch": 0.875,
+ "grad_norm": 0.01877395436167717,
+ "learning_rate": 0.00039000000000000005,
+ "loss": 1.7644124031066895,
+ "step": 14
+ },
+ {
+ "epoch": 0.9375,
+ "grad_norm": 0.015196867287158966,
+ "learning_rate": 0.00042000000000000007,
+ "loss": 1.771270990371704,
+ "step": 15
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.014262701384723186,
+ "learning_rate": 0.00045,
+ "loss": 1.755500316619873,
+ "step": 16
+ },
+ {
+ "epoch": 1.0,
+ "eval_loss": 0.8736820816993713,
+ "eval_runtime": 30.2743,
+ "eval_samples_per_second": 3.369,
+ "eval_steps_per_second": 0.429,
+ "step": 16
+ },
+ {
+ "epoch": 1.0625,
+ "grad_norm": 0.01042763702571392,
+ "learning_rate": 0.00048,
+ "loss": 1.7556875944137573,
+ "step": 17
+ },
+ {
+ "epoch": 1.125,
+ "grad_norm": 0.009123577736318111,
+ "learning_rate": 0.00051,
+ "loss": 1.753890037536621,
+ "step": 18
+ },
+ {
+ "epoch": 1.1875,
+ "grad_norm": 0.009028254076838493,
+ "learning_rate": 0.00054,
+ "loss": 1.7515414953231812,
+ "step": 19
+ },
+ {
+ "epoch": 1.25,
+ "grad_norm": 0.009630529209971428,
+ "learning_rate": 0.00057,
+ "loss": 1.7465155124664307,
+ "step": 20
+ },
+ {
+ "epoch": 1.3125,
+ "grad_norm": 0.008235524408519268,
+ "learning_rate": 0.0006000000000000001,
+ "loss": 1.7533702850341797,
+ "step": 21
+ },
+ {
+ "epoch": 1.375,
+ "grad_norm": 0.009377758949995041,
+ "learning_rate": 0.00063,
+ "loss": 1.738031029701233,
+ "step": 22
+ },
+ {
+ "epoch": 1.4375,
+ "grad_norm": 0.008636261336505413,
+ "learning_rate": 0.00066,
+ "loss": 1.735835313796997,
+ "step": 23
+ },
+ {
+ "epoch": 1.5,
+ "grad_norm": 0.008001700975000858,
+ "learning_rate": 0.0006900000000000001,
+ "loss": 1.7505712509155273,
+ "step": 24
+ },
+ {
+ "epoch": 1.5625,
+ "grad_norm": 0.007517208810895681,
+ "learning_rate": 0.0007199999999999999,
+ "loss": 1.7330820560455322,
+ "step": 25
+ },
+ {
+ "epoch": 1.625,
+ "grad_norm": 0.00952192209661007,
+ "learning_rate": 0.00075,
+ "loss": 1.7318685054779053,
+ "step": 26
+ },
+ {
+ "epoch": 1.6875,
+ "grad_norm": 0.006763546261936426,
+ "learning_rate": 0.0007800000000000001,
+ "loss": 1.7329007387161255,
+ "step": 27
+ },
+ {
+ "epoch": 1.75,
+ "grad_norm": 0.006364013999700546,
+ "learning_rate": 0.0008100000000000001,
+ "loss": 1.7285535335540771,
+ "step": 28
+ },
+ {
+ "epoch": 1.8125,
+ "grad_norm": 0.007770394906401634,
+ "learning_rate": 0.0008400000000000001,
+ "loss": 1.7320656776428223,
+ "step": 29
+ },
+ {
+ "epoch": 1.875,
+ "grad_norm": 0.006639651022851467,
+ "learning_rate": 0.00087,
+ "loss": 1.7304966449737549,
+ "step": 30
+ },
+ {
+ "epoch": 1.9375,
+ "grad_norm": 0.008190952241420746,
+ "learning_rate": 0.0009,
+ "loss": 1.7399260997772217,
+ "step": 31
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 0.008188036270439625,
+ "learning_rate": 0.00093,
+ "loss": 1.7375996112823486,
+ "step": 32
+ },
+ {
+ "epoch": 2.0,
+ "eval_loss": 0.860374927520752,
+ "eval_runtime": 30.3049,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 32
+ },
+ {
+ "epoch": 2.0625,
+ "grad_norm": 0.005768604576587677,
+ "learning_rate": 0.00096,
+ "loss": 1.7309706211090088,
+ "step": 33
+ },
+ {
+ "epoch": 2.125,
+ "grad_norm": 0.006822533905506134,
+ "learning_rate": 0.00099,
+ "loss": 1.7321696281433105,
+ "step": 34
+ },
+ {
+ "epoch": 2.1875,
+ "grad_norm": 0.009974546730518341,
+ "learning_rate": 0.00102,
+ "loss": 1.7290048599243164,
+ "step": 35
+ },
+ {
+ "epoch": 2.25,
+ "grad_norm": 0.00856814906001091,
+ "learning_rate": 0.00105,
+ "loss": 1.7231603860855103,
+ "step": 36
+ },
+ {
+ "epoch": 2.3125,
+ "grad_norm": 0.011694240383803844,
+ "learning_rate": 0.00108,
+ "loss": 1.728348731994629,
+ "step": 37
+ },
+ {
+ "epoch": 2.375,
+ "grad_norm": 0.005738088861107826,
+ "learning_rate": 0.00111,
+ "loss": 1.7308874130249023,
+ "step": 38
+ },
+ {
+ "epoch": 2.4375,
+ "grad_norm": 0.010038000531494617,
+ "learning_rate": 0.00114,
+ "loss": 1.7276837825775146,
+ "step": 39
+ },
+ {
+ "epoch": 2.5,
+ "grad_norm": 0.01586989127099514,
+ "learning_rate": 0.00117,
+ "loss": 1.7368175983428955,
+ "step": 40
+ },
+ {
+ "epoch": 2.5625,
+ "grad_norm": 0.16198667883872986,
+ "learning_rate": 0.0012000000000000001,
+ "loss": 1.730246663093567,
+ "step": 41
+ },
+ {
+ "epoch": 2.625,
+ "grad_norm": 0.01653525047004223,
+ "learning_rate": 0.00123,
+ "loss": 1.7403852939605713,
+ "step": 42
+ },
+ {
+ "epoch": 2.6875,
+ "grad_norm": 0.01241418905556202,
+ "learning_rate": 0.00126,
+ "loss": 1.7306722402572632,
+ "step": 43
+ },
+ {
+ "epoch": 2.75,
+ "grad_norm": 0.022560030221939087,
+ "learning_rate": 0.00129,
+ "loss": 1.7223260402679443,
+ "step": 44
+ },
+ {
+ "epoch": 2.8125,
+ "grad_norm": 0.019568368792533875,
+ "learning_rate": 0.00132,
+ "loss": 1.7304983139038086,
+ "step": 45
+ },
+ {
+ "epoch": 2.875,
+ "grad_norm": 0.016755228862166405,
+ "learning_rate": 0.00135,
+ "loss": 1.7223992347717285,
+ "step": 46
+ },
+ {
+ "epoch": 2.9375,
+ "grad_norm": 0.02981281839311123,
+ "learning_rate": 0.0013800000000000002,
+ "loss": 1.7094981670379639,
+ "step": 47
+ },
+ {
+ "epoch": 3.0,
+ "grad_norm": 0.02506650611758232,
+ "learning_rate": 0.00141,
+ "loss": 1.7054555416107178,
+ "step": 48
+ },
+ {
+ "epoch": 3.0,
+ "eval_loss": 0.8416222333908081,
+ "eval_runtime": 30.3027,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 48
+ },
+ {
+ "epoch": 3.0625,
+ "grad_norm": 0.02260257676243782,
+ "learning_rate": 0.0014399999999999999,
+ "loss": 1.6888771057128906,
+ "step": 49
+ },
+ {
+ "epoch": 3.125,
+ "grad_norm": 0.02620658650994301,
+ "learning_rate": 0.00147,
+ "loss": 1.6639578342437744,
+ "step": 50
+ },
+ {
+ "epoch": 3.1875,
+ "grad_norm": 0.047354575246572495,
+ "learning_rate": 0.0015,
+ "loss": 1.6736350059509277,
+ "step": 51
+ },
+ {
+ "epoch": 3.25,
+ "grad_norm": 0.04594860598444939,
+ "learning_rate": 0.0015300000000000001,
+ "loss": 1.6675360202789307,
+ "step": 52
+ },
+ {
+ "epoch": 3.3125,
+ "grad_norm": 0.08131621032953262,
+ "learning_rate": 0.0015600000000000002,
+ "loss": 1.6574077606201172,
+ "step": 53
+ },
+ {
+ "epoch": 3.375,
+ "grad_norm": 0.028789900243282318,
+ "learning_rate": 0.00159,
+ "loss": 1.6299148797988892,
+ "step": 54
+ },
+ {
+ "epoch": 3.4375,
+ "grad_norm": 0.2623404860496521,
+ "learning_rate": 0.0016200000000000001,
+ "loss": 1.6315665245056152,
+ "step": 55
+ },
+ {
+ "epoch": 3.5,
+ "grad_norm": 0.056828465312719345,
+ "learning_rate": 0.0016500000000000002,
+ "loss": 1.6279325485229492,
+ "step": 56
+ },
+ {
+ "epoch": 3.5625,
+ "grad_norm": 0.0549125075340271,
+ "learning_rate": 0.0016800000000000003,
+ "loss": 1.6549303531646729,
+ "step": 57
+ },
+ {
+ "epoch": 3.625,
+ "grad_norm": 0.07287958264350891,
+ "learning_rate": 0.00171,
+ "loss": 1.606963872909546,
+ "step": 58
+ },
+ {
+ "epoch": 3.6875,
+ "grad_norm": 0.07426343113183975,
+ "learning_rate": 0.00174,
+ "loss": 1.6250534057617188,
+ "step": 59
+ },
+ {
+ "epoch": 3.75,
+ "grad_norm": 0.1364986002445221,
+ "learning_rate": 0.0017699999999999999,
+ "loss": 1.5964961051940918,
+ "step": 60
+ },
+ {
+ "epoch": 3.8125,
+ "grad_norm": 0.15007159113883972,
+ "learning_rate": 0.0018,
+ "loss": 1.6019010543823242,
+ "step": 61
+ },
+ {
+ "epoch": 3.875,
+ "grad_norm": 0.08447311073541641,
+ "learning_rate": 0.00183,
+ "loss": 1.5763543844223022,
+ "step": 62
+ },
+ {
+ "epoch": 3.9375,
+ "grad_norm": 0.1681874394416809,
+ "learning_rate": 0.00186,
+ "loss": 1.5769858360290527,
+ "step": 63
+ },
+ {
+ "epoch": 4.0,
+ "grad_norm": 0.09677319973707199,
+ "learning_rate": 0.00189,
+ "loss": 1.534682273864746,
+ "step": 64
+ },
+ {
+ "epoch": 4.0,
+ "eval_loss": 0.7682511806488037,
+ "eval_runtime": 30.4228,
+ "eval_samples_per_second": 3.353,
+ "eval_steps_per_second": 0.427,
+ "step": 64
+ },
+ {
+ "epoch": 4.0625,
+ "grad_norm": 0.42635980248451233,
+ "learning_rate": 0.00192,
+ "loss": 1.5509562492370605,
+ "step": 65
+ },
+ {
+ "epoch": 4.125,
+ "grad_norm": 0.8467249274253845,
+ "learning_rate": 0.0019500000000000001,
+ "loss": 1.5001531839370728,
+ "step": 66
+ },
+ {
+ "epoch": 4.1875,
+ "grad_norm": 0.22631719708442688,
+ "learning_rate": 0.00198,
+ "loss": 1.4554829597473145,
+ "step": 67
+ },
+ {
+ "epoch": 4.25,
+ "grad_norm": 0.05312464386224747,
+ "learning_rate": 0.00201,
+ "loss": 1.4732775688171387,
+ "step": 68
+ },
+ {
+ "epoch": 4.3125,
+ "grad_norm": 0.050442785024642944,
+ "learning_rate": 0.00204,
+ "loss": 1.4176194667816162,
+ "step": 69
+ },
+ {
+ "epoch": 4.375,
+ "grad_norm": 0.18886858224868774,
+ "learning_rate": 0.00207,
+ "loss": 1.4325997829437256,
+ "step": 70
+ },
+ {
+ "epoch": 4.4375,
+ "grad_norm": 0.05258706212043762,
+ "learning_rate": 0.0021,
+ "loss": 1.39802086353302,
+ "step": 71
+ },
+ {
+ "epoch": 4.5,
+ "grad_norm": 0.07482657581567764,
+ "learning_rate": 0.00213,
+ "loss": 1.4000320434570312,
+ "step": 72
+ },
+ {
+ "epoch": 4.5625,
+ "grad_norm": 0.4853312075138092,
+ "learning_rate": 0.00216,
+ "loss": 1.4199151992797852,
+ "step": 73
+ },
+ {
+ "epoch": 4.625,
+ "grad_norm": 0.48888301849365234,
+ "learning_rate": 0.00219,
+ "loss": 1.3049702644348145,
+ "step": 74
+ },
+ {
+ "epoch": 4.6875,
+ "grad_norm": 0.284348726272583,
+ "learning_rate": 0.00222,
+ "loss": 1.3884068727493286,
+ "step": 75
+ },
+ {
+ "epoch": 4.75,
+ "grad_norm": 0.20248942077159882,
+ "learning_rate": 0.0022500000000000003,
+ "loss": 1.3502120971679688,
+ "step": 76
+ },
+ {
+ "epoch": 4.8125,
+ "grad_norm": 0.2714093327522278,
+ "learning_rate": 0.00228,
+ "loss": 1.3195693492889404,
+ "step": 77
+ },
+ {
+ "epoch": 4.875,
+ "grad_norm": 0.2544061243534088,
+ "learning_rate": 0.00231,
+ "loss": 1.3194304704666138,
+ "step": 78
+ },
+ {
+ "epoch": 4.9375,
+ "grad_norm": 0.18475523591041565,
+ "learning_rate": 0.00234,
+ "loss": 1.3006361722946167,
+ "step": 79
+ },
+ {
+ "epoch": 5.0,
+ "grad_norm": 0.6583278775215149,
+ "learning_rate": 0.00237,
+ "loss": 1.2868517637252808,
+ "step": 80
+ },
+ {
+ "epoch": 5.0,
+ "eval_loss": 0.6618342399597168,
+ "eval_runtime": 30.1233,
+ "eval_samples_per_second": 3.386,
+ "eval_steps_per_second": 0.432,
+ "step": 80
+ },
+ {
+ "epoch": 5.0625,
+ "grad_norm": 0.04499061033129692,
+ "learning_rate": 0.0024000000000000002,
+ "loss": 1.25992751121521,
+ "step": 81
+ },
+ {
+ "epoch": 5.125,
+ "grad_norm": 0.05609860643744469,
+ "learning_rate": 0.0024300000000000003,
+ "loss": 1.2764487266540527,
+ "step": 82
+ },
+ {
+ "epoch": 5.1875,
+ "grad_norm": 0.03665148839354515,
+ "learning_rate": 0.00246,
+ "loss": 1.2908766269683838,
+ "step": 83
+ },
+ {
+ "epoch": 5.25,
+ "grad_norm": 0.17949481308460236,
+ "learning_rate": 0.00249,
+ "loss": 1.253080129623413,
+ "step": 84
+ },
+ {
+ "epoch": 5.3125,
+ "grad_norm": 0.14031393826007843,
+ "learning_rate": 0.00252,
+ "loss": 1.2648870944976807,
+ "step": 85
+ },
+ {
+ "epoch": 5.375,
+ "grad_norm": 0.08193210512399673,
+ "learning_rate": 0.00255,
+ "loss": 1.296234130859375,
+ "step": 86
+ },
+ {
+ "epoch": 5.4375,
+ "grad_norm": 0.056413356214761734,
+ "learning_rate": 0.00258,
+ "loss": 1.2784960269927979,
+ "step": 87
+ },
+ {
+ "epoch": 5.5,
+ "grad_norm": 0.09004318714141846,
+ "learning_rate": 0.00261,
+ "loss": 1.2054948806762695,
+ "step": 88
+ },
+ {
+ "epoch": 5.5625,
+ "grad_norm": 0.044756047427654266,
+ "learning_rate": 0.00264,
+ "loss": 1.2707666158676147,
+ "step": 89
+ },
+ {
+ "epoch": 5.625,
+ "grad_norm": 0.049932755529880524,
+ "learning_rate": 0.00267,
+ "loss": 1.2369787693023682,
+ "step": 90
+ },
+ {
+ "epoch": 5.6875,
+ "grad_norm": 0.09624820947647095,
+ "learning_rate": 0.0027,
+ "loss": 1.2419278621673584,
+ "step": 91
+ },
+ {
+ "epoch": 5.75,
+ "grad_norm": 0.17648059129714966,
+ "learning_rate": 0.0027300000000000002,
+ "loss": 1.2992033958435059,
+ "step": 92
+ },
+ {
+ "epoch": 5.8125,
+ "grad_norm": 0.07027823477983475,
+ "learning_rate": 0.0027600000000000003,
+ "loss": 1.2255799770355225,
+ "step": 93
+ },
+ {
+ "epoch": 5.875,
+ "grad_norm": 0.06040462478995323,
+ "learning_rate": 0.0027900000000000004,
+ "loss": 1.136023759841919,
+ "step": 94
+ },
+ {
+ "epoch": 5.9375,
+ "grad_norm": 0.09845872968435287,
+ "learning_rate": 0.00282,
+ "loss": 1.1807364225387573,
+ "step": 95
+ },
+ {
+ "epoch": 6.0,
+ "grad_norm": 0.09086572378873825,
+ "learning_rate": 0.00285,
+ "loss": 1.2171341180801392,
+ "step": 96
+ },
+ {
+ "epoch": 6.0,
+ "eval_loss": 0.6200892329216003,
+ "eval_runtime": 30.1252,
+ "eval_samples_per_second": 3.386,
+ "eval_steps_per_second": 0.432,
+ "step": 96
+ },
+ {
+ "epoch": 6.0625,
+ "grad_norm": 0.11787034571170807,
+ "learning_rate": 0.0028799999999999997,
+ "loss": 1.242072582244873,
+ "step": 97
+ },
+ {
+ "epoch": 6.125,
+ "grad_norm": 0.20093177258968353,
+ "learning_rate": 0.00291,
+ "loss": 1.2056899070739746,
+ "step": 98
+ },
+ {
+ "epoch": 6.1875,
+ "grad_norm": 0.03221331909298897,
+ "learning_rate": 0.00294,
+ "loss": 1.1774635314941406,
+ "step": 99
+ },
+ {
+ "epoch": 6.25,
+ "grad_norm": 0.06112038344144821,
+ "learning_rate": 0.00297,
+ "loss": 1.1831618547439575,
+ "step": 100
+ },
+ {
+ "epoch": 6.3125,
+ "grad_norm": 0.05896962434053421,
+ "learning_rate": 0.003,
+ "loss": 1.219610571861267,
+ "step": 101
+ },
+ {
+ "epoch": 6.375,
+ "grad_norm": 0.4068014323711395,
+ "learning_rate": 0.0029999850445531283,
+ "loss": 1.231442928314209,
+ "step": 102
+ },
+ {
+ "epoch": 6.4375,
+ "grad_norm": 3.013302803039551,
+ "learning_rate": 0.0029999401785137437,
+ "loss": 1.2512195110321045,
+ "step": 103
+ },
+ {
+ "epoch": 6.5,
+ "grad_norm": 0.07003403455018997,
+ "learning_rate": 0.0029998654027855406,
+ "loss": 1.1912157535552979,
+ "step": 104
+ },
+ {
+ "epoch": 6.5625,
+ "grad_norm": 1.1927955150604248,
+ "learning_rate": 0.002999760718874652,
+ "loss": 1.1158745288848877,
+ "step": 105
+ },
+ {
+ "epoch": 6.625,
+ "grad_norm": 0.11684458702802658,
+ "learning_rate": 0.0029996261288896225,
+ "loss": 1.1835001707077026,
+ "step": 106
+ },
+ {
+ "epoch": 6.6875,
+ "grad_norm": 0.07812155783176422,
+ "learning_rate": 0.0029994616355413666,
+ "loss": 1.1712734699249268,
+ "step": 107
+ },
+ {
+ "epoch": 6.75,
+ "grad_norm": 0.19128946959972382,
+ "learning_rate": 0.0029992672421431115,
+ "loss": 1.0906574726104736,
+ "step": 108
+ },
+ {
+ "epoch": 6.8125,
+ "grad_norm": 0.033625371754169464,
+ "learning_rate": 0.002999042952610332,
+ "loss": 1.115525245666504,
+ "step": 109
+ },
+ {
+ "epoch": 6.875,
+ "grad_norm": 0.2518763244152069,
+ "learning_rate": 0.0029987887714606723,
+ "loss": 1.1433348655700684,
+ "step": 110
+ },
+ {
+ "epoch": 6.9375,
+ "grad_norm": 0.037159573286771774,
+ "learning_rate": 0.002998504703813852,
+ "loss": 1.1867601871490479,
+ "step": 111
+ },
+ {
+ "epoch": 7.0,
+ "grad_norm": 0.04564391076564789,
+ "learning_rate": 0.0029981907553915674,
+ "loss": 1.1302828788757324,
+ "step": 112
+ },
+ {
+ "epoch": 7.0,
+ "eval_loss": 0.5897142887115479,
+ "eval_runtime": 31.3671,
+ "eval_samples_per_second": 3.252,
+ "eval_steps_per_second": 0.414,
+ "step": 112
+ },
+ {
+ "epoch": 7.0625,
+ "grad_norm": 0.05205775797367096,
+ "learning_rate": 0.0029978469325173717,
+ "loss": 1.182741641998291,
+ "step": 113
+ },
+ {
+ "epoch": 7.125,
+ "grad_norm": 0.12080052495002747,
+ "learning_rate": 0.002997473242116551,
+ "loss": 1.1457600593566895,
+ "step": 114
+ },
+ {
+ "epoch": 7.1875,
+ "grad_norm": 0.05600045621395111,
+ "learning_rate": 0.0029970696917159833,
+ "loss": 1.1101720333099365,
+ "step": 115
+ },
+ {
+ "epoch": 7.25,
+ "grad_norm": 0.0507950522005558,
+ "learning_rate": 0.0029966362894439874,
+ "loss": 1.1785833835601807,
+ "step": 116
+ },
+ {
+ "epoch": 7.3125,
+ "grad_norm": 0.6396550536155701,
+ "learning_rate": 0.002996173044030159,
+ "loss": 1.1538987159729004,
+ "step": 117
+ },
+ {
+ "epoch": 7.375,
+ "grad_norm": 0.06849977374076843,
+ "learning_rate": 0.0029956799648051934,
+ "loss": 1.1528857946395874,
+ "step": 118
+ },
+ {
+ "epoch": 7.4375,
+ "grad_norm": 0.0685896947979927,
+ "learning_rate": 0.002995157061700702,
+ "loss": 1.1049470901489258,
+ "step": 119
+ },
+ {
+ "epoch": 7.5,
+ "grad_norm": 1.027722716331482,
+ "learning_rate": 0.002994604345249006,
+ "loss": 1.1308050155639648,
+ "step": 120
+ },
+ {
+ "epoch": 7.5625,
+ "grad_norm": 0.21406109631061554,
+ "learning_rate": 0.00299402182658293,
+ "loss": 1.0960861444473267,
+ "step": 121
+ },
+ {
+ "epoch": 7.625,
+ "grad_norm": 0.0503656342625618,
+ "learning_rate": 0.002993409517435574,
+ "loss": 1.1096141338348389,
+ "step": 122
+ },
+ {
+ "epoch": 7.6875,
+ "grad_norm": 0.04800846800208092,
+ "learning_rate": 0.002992767430140078,
+ "loss": 1.1386680603027344,
+ "step": 123
+ },
+ {
+ "epoch": 7.75,
+ "grad_norm": 0.13091202080249786,
+ "learning_rate": 0.002992095577629376,
+ "loss": 1.1129441261291504,
+ "step": 124
+ },
+ {
+ "epoch": 7.8125,
+ "grad_norm": 0.04949796572327614,
+ "learning_rate": 0.0029913939734359306,
+ "loss": 1.106520175933838,
+ "step": 125
+ },
+ {
+ "epoch": 7.875,
+ "grad_norm": 0.07686582207679749,
+ "learning_rate": 0.0029906626316914655,
+ "loss": 1.1572234630584717,
+ "step": 126
+ },
+ {
+ "epoch": 7.9375,
+ "grad_norm": 0.053587451577186584,
+ "learning_rate": 0.0029899015671266773,
+ "loss": 1.125717282295227,
+ "step": 127
+ },
+ {
+ "epoch": 8.0,
+ "grad_norm": 0.08031050115823746,
+ "learning_rate": 0.0029891107950709406,
+ "loss": 1.153022050857544,
+ "step": 128
+ },
+ {
+ "epoch": 8.0,
+ "eval_loss": 0.5910838842391968,
+ "eval_runtime": 30.7073,
+ "eval_samples_per_second": 3.322,
+ "eval_steps_per_second": 0.423,
+ "step": 128
+ },
+ {
+ "epoch": 8.0625,
+ "grad_norm": 0.3520950675010681,
+ "learning_rate": 0.0029882903314519997,
+ "loss": 1.1582988500595093,
+ "step": 129
+ },
+ {
+ "epoch": 8.125,
+ "grad_norm": 0.07153584063053131,
+ "learning_rate": 0.002987440192795645,
+ "loss": 1.1344678401947021,
+ "step": 130
+ },
+ {
+ "epoch": 8.1875,
+ "grad_norm": 0.06336807459592819,
+ "learning_rate": 0.0029865603962253833,
+ "loss": 1.120319128036499,
+ "step": 131
+ },
+ {
+ "epoch": 8.25,
+ "grad_norm": 0.11782127618789673,
+ "learning_rate": 0.0029856509594620912,
+ "loss": 1.1245379447937012,
+ "step": 132
+ },
+ {
+ "epoch": 8.3125,
+ "grad_norm": 0.2592264413833618,
+ "learning_rate": 0.0029847119008236576,
+ "loss": 1.1454271078109741,
+ "step": 133
+ },
+ {
+ "epoch": 8.375,
+ "grad_norm": 0.21255695819854736,
+ "learning_rate": 0.0029837432392246185,
+ "loss": 1.231805443763733,
+ "step": 134
+ },
+ {
+ "epoch": 8.4375,
+ "grad_norm": 15.057058334350586,
+ "learning_rate": 0.0029827449941757698,
+ "loss": 1.1380956172943115,
+ "step": 135
+ },
+ {
+ "epoch": 8.5,
+ "grad_norm": 0.17857126891613007,
+ "learning_rate": 0.00298171718578378,
+ "loss": 1.1401441097259521,
+ "step": 136
+ },
+ {
+ "epoch": 8.5625,
+ "grad_norm": 0.18281924724578857,
+ "learning_rate": 0.002980659834750781,
+ "loss": 1.171494722366333,
+ "step": 137
+ },
+ {
+ "epoch": 8.625,
+ "grad_norm": 0.10490825027227402,
+ "learning_rate": 0.0029795729623739555,
+ "loss": 1.1622605323791504,
+ "step": 138
+ },
+ {
+ "epoch": 8.6875,
+ "grad_norm": 0.344013512134552,
+ "learning_rate": 0.002978456590545104,
+ "loss": 1.1702141761779785,
+ "step": 139
+ },
+ {
+ "epoch": 8.75,
+ "grad_norm": 0.05058368295431137,
+ "learning_rate": 0.0029773107417502055,
+ "loss": 1.2006211280822754,
+ "step": 140
+ },
+ {
+ "epoch": 8.8125,
+ "grad_norm": 0.09775760769844055,
+ "learning_rate": 0.002976135439068965,
+ "loss": 1.1407063007354736,
+ "step": 141
+ },
+ {
+ "epoch": 8.875,
+ "grad_norm": 0.0785030648112297,
+ "learning_rate": 0.0029749307061743475,
+ "loss": 1.1778209209442139,
+ "step": 142
+ },
+ {
+ "epoch": 8.9375,
+ "grad_norm": 0.053056441247463226,
+ "learning_rate": 0.0029736965673321032,
+ "loss": 1.1302082538604736,
+ "step": 143
+ },
+ {
+ "epoch": 9.0,
+ "grad_norm": 0.4020994305610657,
+ "learning_rate": 0.0029724330474002743,
+ "loss": 1.1655206680297852,
+ "step": 144
+ },
+ {
+ "epoch": 9.0,
+ "eval_loss": 0.6083483099937439,
+ "eval_runtime": 30.0453,
+ "eval_samples_per_second": 3.395,
+ "eval_steps_per_second": 0.433,
+ "step": 144
+ },
+ {
+ "epoch": 9.0625,
+ "grad_norm": 0.04755420982837677,
+ "learning_rate": 0.002971140171828702,
+ "loss": 1.1900575160980225,
+ "step": 145
+ },
+ {
+ "epoch": 9.125,
+ "grad_norm": 0.2549609839916229,
+ "learning_rate": 0.002969817966658505,
+ "loss": 1.1891400814056396,
+ "step": 146
+ },
+ {
+ "epoch": 9.1875,
+ "grad_norm": 0.11641769111156464,
+ "learning_rate": 0.0029684664585215618,
+ "loss": 1.1930484771728516,
+ "step": 147
+ },
+ {
+ "epoch": 9.25,
+ "grad_norm": 0.13052672147750854,
+ "learning_rate": 0.00296708567463997,
+ "loss": 1.1905772686004639,
+ "step": 148
+ },
+ {
+ "epoch": 9.3125,
+ "grad_norm": 0.15126636624336243,
+ "learning_rate": 0.0029656756428255017,
+ "loss": 1.1417806148529053,
+ "step": 149
+ },
+ {
+ "epoch": 9.375,
+ "grad_norm": 0.08826905488967896,
+ "learning_rate": 0.00296423639147904,
+ "loss": 1.1802985668182373,
+ "step": 150
+ },
+ {
+ "epoch": 9.4375,
+ "grad_norm": 0.01381592359393835,
+ "learning_rate": 0.002962767949590008,
+ "loss": 1.1525968313217163,
+ "step": 151
+ },
+ {
+ "epoch": 9.5,
+ "grad_norm": 0.04648139327764511,
+ "learning_rate": 0.0029612703467357866,
+ "loss": 1.169856309890747,
+ "step": 152
+ },
+ {
+ "epoch": 9.5625,
+ "grad_norm": 0.014019235037267208,
+ "learning_rate": 0.0029597436130811153,
+ "loss": 1.1442453861236572,
+ "step": 153
+ },
+ {
+ "epoch": 9.625,
+ "grad_norm": 0.03701096400618553,
+ "learning_rate": 0.0029581877793774885,
+ "loss": 1.112680435180664,
+ "step": 154
+ },
+ {
+ "epoch": 9.6875,
+ "grad_norm": 0.030421797186136246,
+ "learning_rate": 0.0029566028769625334,
+ "loss": 1.223894715309143,
+ "step": 155
+ },
+ {
+ "epoch": 9.75,
+ "grad_norm": 0.12336083501577377,
+ "learning_rate": 0.00295498893775938,
+ "loss": 1.220346212387085,
+ "step": 156
+ },
+ {
+ "epoch": 9.8125,
+ "grad_norm": 0.038585737347602844,
+ "learning_rate": 0.0029533459942760174,
+ "loss": 1.2078583240509033,
+ "step": 157
+ },
+ {
+ "epoch": 9.875,
+ "grad_norm": 0.4653286039829254,
+ "learning_rate": 0.002951674079604639,
+ "loss": 1.2035815715789795,
+ "step": 158
+ },
+ {
+ "epoch": 9.9375,
+ "grad_norm": 0.020640619099140167,
+ "learning_rate": 0.0029499732274209777,
+ "loss": 1.1313148736953735,
+ "step": 159
+ },
+ {
+ "epoch": 10.0,
+ "grad_norm": 0.04939777031540871,
+ "learning_rate": 0.002948243471983625,
+ "loss": 1.1253995895385742,
+ "step": 160
+ },
+ {
+ "epoch": 10.0,
+ "eval_loss": 0.604421079158783,
+ "eval_runtime": 30.1556,
+ "eval_samples_per_second": 3.382,
+ "eval_steps_per_second": 0.431,
+ "step": 160
+ },
+ {
+ "epoch": 10.0625,
+ "grad_norm": 0.05968557298183441,
+ "learning_rate": 0.0029464848481333423,
+ "loss": 1.1324307918548584,
+ "step": 161
+ },
+ {
+ "epoch": 10.125,
+ "grad_norm": 1.0001988410949707,
+ "learning_rate": 0.0029446973912923597,
+ "loss": 1.126110315322876,
+ "step": 162
+ },
+ {
+ "epoch": 10.1875,
+ "grad_norm": 0.2294934093952179,
+ "learning_rate": 0.002942881137463661,
+ "loss": 1.1873570680618286,
+ "step": 163
+ },
+ {
+ "epoch": 10.25,
+ "grad_norm": 0.06396346539258957,
+ "learning_rate": 0.002941036123230261,
+ "loss": 1.1288526058197021,
+ "step": 164
+ },
+ {
+ "epoch": 10.3125,
+ "grad_norm": 0.3410445749759674,
+ "learning_rate": 0.0029391623857544643,
+ "loss": 1.1219967603683472,
+ "step": 165
+ },
+ {
+ "epoch": 10.375,
+ "grad_norm": 0.09721948206424713,
+ "learning_rate": 0.002937259962777122,
+ "loss": 1.141520619392395,
+ "step": 166
+ },
+ {
+ "epoch": 10.4375,
+ "grad_norm": 0.033714450895786285,
+ "learning_rate": 0.002935328892616868,
+ "loss": 1.1983453035354614,
+ "step": 167
+ },
+ {
+ "epoch": 10.5,
+ "grad_norm": 0.028446367010474205,
+ "learning_rate": 0.0029333692141693484,
+ "loss": 1.153167724609375,
+ "step": 168
+ },
+ {
+ "epoch": 10.5625,
+ "grad_norm": 0.04298163205385208,
+ "learning_rate": 0.0029313809669064377,
+ "loss": 1.120814561843872,
+ "step": 169
+ },
+ {
+ "epoch": 10.625,
+ "grad_norm": 0.08190953731536865,
+ "learning_rate": 0.002929364190875444,
+ "loss": 1.1164218187332153,
+ "step": 170
+ },
+ {
+ "epoch": 10.6875,
+ "grad_norm": 0.0602508969604969,
+ "learning_rate": 0.0029273189266983035,
+ "loss": 1.136858344078064,
+ "step": 171
+ },
+ {
+ "epoch": 10.75,
+ "grad_norm": 0.07118846476078033,
+ "learning_rate": 0.0029252452155707582,
+ "loss": 1.1356765031814575,
+ "step": 172
+ },
+ {
+ "epoch": 10.8125,
+ "grad_norm": 0.0445970743894577,
+ "learning_rate": 0.002923143099261531,
+ "loss": 1.121457815170288,
+ "step": 173
+ },
+ {
+ "epoch": 10.875,
+ "grad_norm": 0.07745611667633057,
+ "learning_rate": 0.0029210126201114836,
+ "loss": 1.0903944969177246,
+ "step": 174
+ },
+ {
+ "epoch": 10.9375,
+ "grad_norm": 0.08945920318365097,
+ "learning_rate": 0.0029188538210327595,
+ "loss": 1.1194133758544922,
+ "step": 175
+ },
+ {
+ "epoch": 11.0,
+ "grad_norm": 0.07996643334627151,
+ "learning_rate": 0.002916666745507926,
+ "loss": 1.1248202323913574,
+ "step": 176
+ },
+ {
+ "epoch": 11.0,
+ "eval_loss": 0.5789638757705688,
+ "eval_runtime": 30.3538,
+ "eval_samples_per_second": 3.36,
+ "eval_steps_per_second": 0.428,
+ "step": 176
+ },
+ {
+ "epoch": 11.0625,
+ "grad_norm": 0.10533642023801804,
+ "learning_rate": 0.002914451437589092,
+ "loss": 1.1199195384979248,
+ "step": 177
+ },
+ {
+ "epoch": 11.125,
+ "grad_norm": 0.044989872723817825,
+ "learning_rate": 0.002912207941897025,
+ "loss": 1.1084158420562744,
+ "step": 178
+ },
+ {
+ "epoch": 11.1875,
+ "grad_norm": 0.04824615642428398,
+ "learning_rate": 0.0029099363036202515,
+ "loss": 1.0993731021881104,
+ "step": 179
+ },
+ {
+ "epoch": 11.25,
+ "grad_norm": 0.03144732117652893,
+ "learning_rate": 0.0029076365685141463,
+ "loss": 1.1078145503997803,
+ "step": 180
+ },
+ {
+ "epoch": 11.3125,
+ "grad_norm": 0.03728745877742767,
+ "learning_rate": 0.0029053087829000097,
+ "loss": 1.0995632410049438,
+ "step": 181
+ },
+ {
+ "epoch": 11.375,
+ "grad_norm": 0.02090257592499256,
+ "learning_rate": 0.002902952993664138,
+ "loss": 1.0722652673721313,
+ "step": 182
+ },
+ {
+ "epoch": 11.4375,
+ "grad_norm": 0.020907679572701454,
+ "learning_rate": 0.0029005692482568747,
+ "loss": 1.0840482711791992,
+ "step": 183
+ },
+ {
+ "epoch": 11.5,
+ "grad_norm": 0.023914694786071777,
+ "learning_rate": 0.002898157594691658,
+ "loss": 1.1289193630218506,
+ "step": 184
+ },
+ {
+ "epoch": 11.5625,
+ "grad_norm": 0.01937722973525524,
+ "learning_rate": 0.0028957180815440534,
+ "loss": 1.1244895458221436,
+ "step": 185
+ },
+ {
+ "epoch": 11.625,
+ "grad_norm": 0.02047949656844139,
+ "learning_rate": 0.002893250757950773,
+ "loss": 1.0527830123901367,
+ "step": 186
+ },
+ {
+ "epoch": 11.6875,
+ "grad_norm": 0.04898907244205475,
+ "learning_rate": 0.002890755673608688,
+ "loss": 1.1234995126724243,
+ "step": 187
+ },
+ {
+ "epoch": 11.75,
+ "grad_norm": 0.024841446429491043,
+ "learning_rate": 0.0028882328787738273,
+ "loss": 1.138411045074463,
+ "step": 188
+ },
+ {
+ "epoch": 11.8125,
+ "grad_norm": 0.02533135376870632,
+ "learning_rate": 0.002885682424260365,
+ "loss": 1.1021623611450195,
+ "step": 189
+ },
+ {
+ "epoch": 11.875,
+ "grad_norm": 0.04309404641389847,
+ "learning_rate": 0.0028831043614395966,
+ "loss": 1.0767666101455688,
+ "step": 190
+ },
+ {
+ "epoch": 11.9375,
+ "grad_norm": 0.029682395979762077,
+ "learning_rate": 0.0028804987422389045,
+ "loss": 1.1310434341430664,
+ "step": 191
+ },
+ {
+ "epoch": 12.0,
+ "grad_norm": 0.021236268803477287,
+ "learning_rate": 0.002877865619140712,
+ "loss": 1.0969300270080566,
+ "step": 192
+ },
+ {
+ "epoch": 12.0,
+ "eval_loss": 0.5664089322090149,
+ "eval_runtime": 30.1046,
+ "eval_samples_per_second": 3.388,
+ "eval_steps_per_second": 0.432,
+ "step": 192
+ },
+ {
+ "epoch": 12.0625,
+ "grad_norm": 0.09511730819940567,
+ "learning_rate": 0.0028752050451814276,
+ "loss": 1.0866225957870483,
+ "step": 193
+ },
+ {
+ "epoch": 12.125,
+ "grad_norm": 0.1858353465795517,
+ "learning_rate": 0.0028725170739503734,
+ "loss": 1.1082183122634888,
+ "step": 194
+ },
+ {
+ "epoch": 12.1875,
+ "grad_norm": 0.018049633130431175,
+ "learning_rate": 0.00286980175958871,
+ "loss": 1.0569005012512207,
+ "step": 195
+ },
+ {
+ "epoch": 12.25,
+ "grad_norm": 0.06167157366871834,
+ "learning_rate": 0.002867059156788342,
+ "loss": 1.1418097019195557,
+ "step": 196
+ },
+ {
+ "epoch": 12.3125,
+ "grad_norm": 0.06766022741794586,
+ "learning_rate": 0.0028642893207908187,
+ "loss": 1.0995159149169922,
+ "step": 197
+ },
+ {
+ "epoch": 12.375,
+ "grad_norm": 0.03883042931556702,
+ "learning_rate": 0.0028614923073862213,
+ "loss": 1.0847246646881104,
+ "step": 198
+ },
+ {
+ "epoch": 12.4375,
+ "grad_norm": 0.05537786707282066,
+ "learning_rate": 0.0028586681729120393,
+ "loss": 1.1187801361083984,
+ "step": 199
+ },
+ {
+ "epoch": 12.5,
+ "grad_norm": 0.0653286948800087,
+ "learning_rate": 0.002855816974252033,
+ "loss": 1.0798084735870361,
+ "step": 200
+ },
+ {
+ "epoch": 12.5625,
+ "grad_norm": 0.04398065432906151,
+ "learning_rate": 0.002852938768835092,
+ "loss": 1.0946357250213623,
+ "step": 201
+ },
+ {
+ "epoch": 12.625,
+ "grad_norm": 0.05614284798502922,
+ "learning_rate": 0.0028500336146340772,
+ "loss": 1.1144163608551025,
+ "step": 202
+ },
+ {
+ "epoch": 12.6875,
+ "grad_norm": 0.035650063306093216,
+ "learning_rate": 0.0028471015701646485,
+ "loss": 1.0421898365020752,
+ "step": 203
+ },
+ {
+ "epoch": 12.75,
+ "grad_norm": 0.02544066496193409,
+ "learning_rate": 0.002844142694484094,
+ "loss": 1.0994564294815063,
+ "step": 204
+ },
+ {
+ "epoch": 12.8125,
+ "grad_norm": 0.07384507358074188,
+ "learning_rate": 0.002841157047190135,
+ "loss": 1.0665926933288574,
+ "step": 205
+ },
+ {
+ "epoch": 12.875,
+ "grad_norm": 0.04875814542174339,
+ "learning_rate": 0.002838144688419726,
+ "loss": 1.0665624141693115,
+ "step": 206
+ },
+ {
+ "epoch": 12.9375,
+ "grad_norm": 0.05739036574959755,
+ "learning_rate": 0.002835105678847847,
+ "loss": 1.1085939407348633,
+ "step": 207
+ },
+ {
+ "epoch": 13.0,
+ "grad_norm": 0.03088603913784027,
+ "learning_rate": 0.002832040079686276,
+ "loss": 1.1357779502868652,
+ "step": 208
+ },
+ {
+ "epoch": 13.0,
+ "eval_loss": 0.5675534009933472,
+ "eval_runtime": 31.0754,
+ "eval_samples_per_second": 3.282,
+ "eval_steps_per_second": 0.418,
+ "step": 208
+ },
+ {
+ "epoch": 13.0625,
+ "grad_norm": 0.052636317908763885,
+ "learning_rate": 0.0028289479526823614,
+ "loss": 1.0684641599655151,
+ "step": 209
+ },
+ {
+ "epoch": 13.125,
+ "grad_norm": 0.26156774163246155,
+ "learning_rate": 0.0028258293601177743,
+ "loss": 1.1112414598464966,
+ "step": 210
+ },
+ {
+ "epoch": 13.1875,
+ "grad_norm": 0.013046488165855408,
+ "learning_rate": 0.0028226843648072562,
+ "loss": 1.0339481830596924,
+ "step": 211
+ },
+ {
+ "epoch": 13.25,
+ "grad_norm": 0.029812589287757874,
+ "learning_rate": 0.002819513030097352,
+ "loss": 1.0734567642211914,
+ "step": 212
+ },
+ {
+ "epoch": 13.3125,
+ "grad_norm": 0.03711036592721939,
+ "learning_rate": 0.0028163154198651372,
+ "loss": 1.071413516998291,
+ "step": 213
+ },
+ {
+ "epoch": 13.375,
+ "grad_norm": 0.09543009102344513,
+ "learning_rate": 0.002813091598516928,
+ "loss": 1.1175503730773926,
+ "step": 214
+ },
+ {
+ "epoch": 13.4375,
+ "grad_norm": 0.02742145210504532,
+ "learning_rate": 0.0028098416309869857,
+ "loss": 1.0401530265808105,
+ "step": 215
+ },
+ {
+ "epoch": 13.5,
+ "grad_norm": 0.05347360298037529,
+ "learning_rate": 0.002806565582736208,
+ "loss": 1.053802728652954,
+ "step": 216
+ },
+ {
+ "epoch": 13.5625,
+ "grad_norm": 0.023394430056214333,
+ "learning_rate": 0.002803263519750811,
+ "loss": 1.0774943828582764,
+ "step": 217
+ },
+ {
+ "epoch": 13.625,
+ "grad_norm": 0.03274613991379738,
+ "learning_rate": 0.0027999355085409996,
+ "loss": 1.1332664489746094,
+ "step": 218
+ },
+ {
+ "epoch": 13.6875,
+ "grad_norm": 0.029956744983792305,
+ "learning_rate": 0.00279658161613963,
+ "loss": 1.1192328929901123,
+ "step": 219
+ },
+ {
+ "epoch": 13.75,
+ "grad_norm": 0.027843108400702477,
+ "learning_rate": 0.002793201910100856,
+ "loss": 1.0953505039215088,
+ "step": 220
+ },
+ {
+ "epoch": 13.8125,
+ "grad_norm": 0.3722478151321411,
+ "learning_rate": 0.0027897964584987713,
+ "loss": 1.1045243740081787,
+ "step": 221
+ },
+ {
+ "epoch": 13.875,
+ "grad_norm": 0.05480271577835083,
+ "learning_rate": 0.0027863653299260356,
+ "loss": 1.0857124328613281,
+ "step": 222
+ },
+ {
+ "epoch": 13.9375,
+ "grad_norm": 0.05806222930550575,
+ "learning_rate": 0.0027829085934924958,
+ "loss": 1.092329502105713,
+ "step": 223
+ },
+ {
+ "epoch": 14.0,
+ "grad_norm": 0.07507319748401642,
+ "learning_rate": 0.0027794263188237922,
+ "loss": 1.0946941375732422,
+ "step": 224
+ },
+ {
+ "epoch": 14.0,
+ "eval_loss": 0.5624361038208008,
+ "eval_runtime": 30.3035,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 224
+ },
+ {
+ "epoch": 14.0625,
+ "grad_norm": 0.08125249296426773,
+ "learning_rate": 0.0027759185760599576,
+ "loss": 1.082047939300537,
+ "step": 225
+ },
+ {
+ "epoch": 14.125,
+ "grad_norm": 0.08838590234518051,
+ "learning_rate": 0.002772385435854002,
+ "loss": 1.0747947692871094,
+ "step": 226
+ },
+ {
+ "epoch": 14.1875,
+ "grad_norm": 0.18887385725975037,
+ "learning_rate": 0.0027688269693704926,
+ "loss": 1.0601060390472412,
+ "step": 227
+ },
+ {
+ "epoch": 14.25,
+ "grad_norm": 0.4315606653690338,
+ "learning_rate": 0.002765243248284118,
+ "loss": 1.167302131652832,
+ "step": 228
+ },
+ {
+ "epoch": 14.3125,
+ "grad_norm": 0.07665648311376572,
+ "learning_rate": 0.002761634344778245,
+ "loss": 1.130738615989685,
+ "step": 229
+ },
+ {
+ "epoch": 14.375,
+ "grad_norm": 0.051125217229127884,
+ "learning_rate": 0.002758000331543466,
+ "loss": 1.0479257106781006,
+ "step": 230
+ },
+ {
+ "epoch": 14.4375,
+ "grad_norm": 0.0658797100186348,
+ "learning_rate": 0.0027543412817761345,
+ "loss": 1.0913889408111572,
+ "step": 231
+ },
+ {
+ "epoch": 14.5,
+ "grad_norm": 0.1936631202697754,
+ "learning_rate": 0.0027506572691768888,
+ "loss": 1.0582077503204346,
+ "step": 232
+ },
+ {
+ "epoch": 14.5625,
+ "grad_norm": 2.202739953994751,
+ "learning_rate": 0.0027469483679491693,
+ "loss": 1.0946757793426514,
+ "step": 233
+ },
+ {
+ "epoch": 14.625,
+ "grad_norm": 0.07856857776641846,
+ "learning_rate": 0.002743214652797724,
+ "loss": 1.074622631072998,
+ "step": 234
+ },
+ {
+ "epoch": 14.6875,
+ "grad_norm": 0.04288165271282196,
+ "learning_rate": 0.002739456198927104,
+ "loss": 1.0788205862045288,
+ "step": 235
+ },
+ {
+ "epoch": 14.75,
+ "grad_norm": 0.08169898390769958,
+ "learning_rate": 0.0027356730820401465,
+ "loss": 1.0906422138214111,
+ "step": 236
+ },
+ {
+ "epoch": 14.8125,
+ "grad_norm": 0.04536404460668564,
+ "learning_rate": 0.002731865378336452,
+ "loss": 1.0542757511138916,
+ "step": 237
+ },
+ {
+ "epoch": 14.875,
+ "grad_norm": 0.0246584415435791,
+ "learning_rate": 0.0027280331645108498,
+ "loss": 1.0401986837387085,
+ "step": 238
+ },
+ {
+ "epoch": 14.9375,
+ "grad_norm": 0.03399847820401192,
+ "learning_rate": 0.0027241765177518523,
+ "loss": 1.109243392944336,
+ "step": 239
+ },
+ {
+ "epoch": 15.0,
+ "grad_norm": 0.12445052713155746,
+ "learning_rate": 0.0027202955157401,
+ "loss": 1.0952987670898438,
+ "step": 240
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 800,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 50,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": true,
+ "should_log": false,
+ "should_save": false,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3217062038077440.0,
+ "train_batch_size": 8,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/runs/test-model/checkpoint-240/training_args.bin b/runs/test-model/checkpoint-240/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1c106baba014e8cee356170e7acaf136a1e2efb8
--- /dev/null
+++ b/runs/test-model/checkpoint-240/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
+size 5265
diff --git a/runs/test-model/checkpoint-320/chat_template.jinja b/runs/test-model/checkpoint-320/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..699ff8df401fe4788525e9c1f9b86a99eadd6230
--- /dev/null
+++ b/runs/test-model/checkpoint-320/chat_template.jinja
@@ -0,0 +1,85 @@
+{%- 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 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.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set content = message.content %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in message.content %}
+ {%- set content = message.content.split('')[-1].lstrip('\n') %}
+ {%- set reasoning_content = message.content.split('')[0].rstrip('\n').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' }}
+ {{- message.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/runs/test-model/checkpoint-320/config.json b/runs/test-model/checkpoint-320/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..44f7b99e220689bde520b89c11fdd83d1b5348de
--- /dev/null
+++ b/runs/test-model/checkpoint-320/config.json
@@ -0,0 +1,32 @@
+{
+ "architectures": [
+ "LlamaForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "float32",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 512,
+ "initializer_range": 0.02,
+ "intermediate_size": 1536,
+ "max_position_embeddings": 2048,
+ "mlp_bias": false,
+ "model_type": "llama",
+ "num_attention_heads": 4,
+ "num_hidden_layers": 20,
+ "num_key_value_heads": 4,
+ "pad_token_id": 151645,
+ "pretraining_tp": 1,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 10000.0,
+ "rope_type": "default"
+ },
+ "tie_word_embeddings": true,
+ "transformers_version": "5.5.0",
+ "use_cache": false,
+ "vocab_size": 151671
+}
diff --git a/runs/test-model/checkpoint-320/generation_config.json b/runs/test-model/checkpoint-320/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f962c4bfc66159cdeb7ba836cbbd79365e9304f3
--- /dev/null
+++ b/runs/test-model/checkpoint-320/generation_config.json
@@ -0,0 +1,11 @@
+{
+ "_from_model_config": true,
+ "eos_token_id": [
+ 151645
+ ],
+ "output_attentions": false,
+ "output_hidden_states": false,
+ "pad_token_id": 151645,
+ "transformers_version": "5.5.0",
+ "use_cache": true
+}
diff --git a/runs/test-model/checkpoint-320/model.safetensors b/runs/test-model/checkpoint-320/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d661247f149903be270bcf378f42946e7f168b7c
--- /dev/null
+++ b/runs/test-model/checkpoint-320/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2aa2e118c526b9d0af1ed0fd002df979c49d3ce6f2a967821e998acdca70de76
+size 583356232
diff --git a/runs/test-model/checkpoint-320/optimizer.pt b/runs/test-model/checkpoint-320/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ed696c9b4fcdeefa138ad61d857123007ffcbefa
--- /dev/null
+++ b/runs/test-model/checkpoint-320/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a3a7d82ce002ce9c6932773a9943a10e26e5f2a8e81b38b0041ad823cf8705a
+size 1166825803
diff --git a/runs/test-model/checkpoint-320/rng_state.pth b/runs/test-model/checkpoint-320/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb38b0a47e00c4d6bcc13949fbbe074c0294f3bc
--- /dev/null
+++ b/runs/test-model/checkpoint-320/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92797ce4fcd09363a266ca38b4355b9ce342ee666fccb4450bc247f277b9c363
+size 14709
diff --git a/runs/test-model/checkpoint-320/scheduler.pt b/runs/test-model/checkpoint-320/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c42289ffba99934c18c221ca21db5c93500f3b51
--- /dev/null
+++ b/runs/test-model/checkpoint-320/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d37018ddedbaa9aa250e8eac93c964b8bfb9fa3b839f8ed59513490593702820
+size 1465
diff --git a/runs/test-model/checkpoint-320/tokenizer.json b/runs/test-model/checkpoint-320/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..b83d93986de8ecfc4343943be1c86e9532682c15
--- /dev/null
+++ b/runs/test-model/checkpoint-320/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
+size 11424079
diff --git a/runs/test-model/checkpoint-320/tokenizer_config.json b/runs/test-model/checkpoint-320/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1d6bb391bb8fbc446de576db5220ec91c98a150
--- /dev/null
+++ b/runs/test-model/checkpoint-320/tokenizer_config.json
@@ -0,0 +1,18 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|l2r_pred|>",
+ "<|r2l_pred|>"
+ ],
+ "is_local": false,
+ "model_max_length": 131072,
+ "pad_token": "<|im_end|>",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/runs/test-model/checkpoint-320/trainer_state.json b/runs/test-model/checkpoint-320/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d4c2aa09c3c20df2b78bdd086ae4a81bcb8a84d
--- /dev/null
+++ b/runs/test-model/checkpoint-320/trainer_state.json
@@ -0,0 +1,2426 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 20.0,
+ "eval_steps": 500,
+ "global_step": 320,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0625,
+ "grad_norm": 0.4655202031135559,
+ "learning_rate": 0.0,
+ "loss": 2.00244402885437,
+ "step": 1
+ },
+ {
+ "epoch": 0.125,
+ "grad_norm": 0.4847956895828247,
+ "learning_rate": 3e-05,
+ "loss": 2.008800506591797,
+ "step": 2
+ },
+ {
+ "epoch": 0.1875,
+ "grad_norm": 0.4177771806716919,
+ "learning_rate": 6e-05,
+ "loss": 1.9685651063919067,
+ "step": 3
+ },
+ {
+ "epoch": 0.25,
+ "grad_norm": 0.3550562858581543,
+ "learning_rate": 8.999999999999999e-05,
+ "loss": 1.9168976545333862,
+ "step": 4
+ },
+ {
+ "epoch": 0.3125,
+ "grad_norm": 0.23293137550354004,
+ "learning_rate": 0.00012,
+ "loss": 1.8660516738891602,
+ "step": 5
+ },
+ {
+ "epoch": 0.375,
+ "grad_norm": 0.13890701532363892,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.8209998607635498,
+ "step": 6
+ },
+ {
+ "epoch": 0.4375,
+ "grad_norm": 0.08890970796346664,
+ "learning_rate": 0.00017999999999999998,
+ "loss": 1.806854248046875,
+ "step": 7
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.061611492186784744,
+ "learning_rate": 0.00021000000000000004,
+ "loss": 1.793172836303711,
+ "step": 8
+ },
+ {
+ "epoch": 0.5625,
+ "grad_norm": 0.05786389857530594,
+ "learning_rate": 0.00024,
+ "loss": 1.7882877588272095,
+ "step": 9
+ },
+ {
+ "epoch": 0.625,
+ "grad_norm": 0.03947344794869423,
+ "learning_rate": 0.00027,
+ "loss": 1.7696826457977295,
+ "step": 10
+ },
+ {
+ "epoch": 0.6875,
+ "grad_norm": 0.029783710837364197,
+ "learning_rate": 0.00030000000000000003,
+ "loss": 1.7767844200134277,
+ "step": 11
+ },
+ {
+ "epoch": 0.75,
+ "grad_norm": 0.025729672983288765,
+ "learning_rate": 0.00033,
+ "loss": 1.7827998399734497,
+ "step": 12
+ },
+ {
+ "epoch": 0.8125,
+ "grad_norm": 0.020592188462615013,
+ "learning_rate": 0.00035999999999999997,
+ "loss": 1.766276478767395,
+ "step": 13
+ },
+ {
+ "epoch": 0.875,
+ "grad_norm": 0.01877395436167717,
+ "learning_rate": 0.00039000000000000005,
+ "loss": 1.7644124031066895,
+ "step": 14
+ },
+ {
+ "epoch": 0.9375,
+ "grad_norm": 0.015196867287158966,
+ "learning_rate": 0.00042000000000000007,
+ "loss": 1.771270990371704,
+ "step": 15
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.014262701384723186,
+ "learning_rate": 0.00045,
+ "loss": 1.755500316619873,
+ "step": 16
+ },
+ {
+ "epoch": 1.0,
+ "eval_loss": 0.8736820816993713,
+ "eval_runtime": 30.2743,
+ "eval_samples_per_second": 3.369,
+ "eval_steps_per_second": 0.429,
+ "step": 16
+ },
+ {
+ "epoch": 1.0625,
+ "grad_norm": 0.01042763702571392,
+ "learning_rate": 0.00048,
+ "loss": 1.7556875944137573,
+ "step": 17
+ },
+ {
+ "epoch": 1.125,
+ "grad_norm": 0.009123577736318111,
+ "learning_rate": 0.00051,
+ "loss": 1.753890037536621,
+ "step": 18
+ },
+ {
+ "epoch": 1.1875,
+ "grad_norm": 0.009028254076838493,
+ "learning_rate": 0.00054,
+ "loss": 1.7515414953231812,
+ "step": 19
+ },
+ {
+ "epoch": 1.25,
+ "grad_norm": 0.009630529209971428,
+ "learning_rate": 0.00057,
+ "loss": 1.7465155124664307,
+ "step": 20
+ },
+ {
+ "epoch": 1.3125,
+ "grad_norm": 0.008235524408519268,
+ "learning_rate": 0.0006000000000000001,
+ "loss": 1.7533702850341797,
+ "step": 21
+ },
+ {
+ "epoch": 1.375,
+ "grad_norm": 0.009377758949995041,
+ "learning_rate": 0.00063,
+ "loss": 1.738031029701233,
+ "step": 22
+ },
+ {
+ "epoch": 1.4375,
+ "grad_norm": 0.008636261336505413,
+ "learning_rate": 0.00066,
+ "loss": 1.735835313796997,
+ "step": 23
+ },
+ {
+ "epoch": 1.5,
+ "grad_norm": 0.008001700975000858,
+ "learning_rate": 0.0006900000000000001,
+ "loss": 1.7505712509155273,
+ "step": 24
+ },
+ {
+ "epoch": 1.5625,
+ "grad_norm": 0.007517208810895681,
+ "learning_rate": 0.0007199999999999999,
+ "loss": 1.7330820560455322,
+ "step": 25
+ },
+ {
+ "epoch": 1.625,
+ "grad_norm": 0.00952192209661007,
+ "learning_rate": 0.00075,
+ "loss": 1.7318685054779053,
+ "step": 26
+ },
+ {
+ "epoch": 1.6875,
+ "grad_norm": 0.006763546261936426,
+ "learning_rate": 0.0007800000000000001,
+ "loss": 1.7329007387161255,
+ "step": 27
+ },
+ {
+ "epoch": 1.75,
+ "grad_norm": 0.006364013999700546,
+ "learning_rate": 0.0008100000000000001,
+ "loss": 1.7285535335540771,
+ "step": 28
+ },
+ {
+ "epoch": 1.8125,
+ "grad_norm": 0.007770394906401634,
+ "learning_rate": 0.0008400000000000001,
+ "loss": 1.7320656776428223,
+ "step": 29
+ },
+ {
+ "epoch": 1.875,
+ "grad_norm": 0.006639651022851467,
+ "learning_rate": 0.00087,
+ "loss": 1.7304966449737549,
+ "step": 30
+ },
+ {
+ "epoch": 1.9375,
+ "grad_norm": 0.008190952241420746,
+ "learning_rate": 0.0009,
+ "loss": 1.7399260997772217,
+ "step": 31
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 0.008188036270439625,
+ "learning_rate": 0.00093,
+ "loss": 1.7375996112823486,
+ "step": 32
+ },
+ {
+ "epoch": 2.0,
+ "eval_loss": 0.860374927520752,
+ "eval_runtime": 30.3049,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 32
+ },
+ {
+ "epoch": 2.0625,
+ "grad_norm": 0.005768604576587677,
+ "learning_rate": 0.00096,
+ "loss": 1.7309706211090088,
+ "step": 33
+ },
+ {
+ "epoch": 2.125,
+ "grad_norm": 0.006822533905506134,
+ "learning_rate": 0.00099,
+ "loss": 1.7321696281433105,
+ "step": 34
+ },
+ {
+ "epoch": 2.1875,
+ "grad_norm": 0.009974546730518341,
+ "learning_rate": 0.00102,
+ "loss": 1.7290048599243164,
+ "step": 35
+ },
+ {
+ "epoch": 2.25,
+ "grad_norm": 0.00856814906001091,
+ "learning_rate": 0.00105,
+ "loss": 1.7231603860855103,
+ "step": 36
+ },
+ {
+ "epoch": 2.3125,
+ "grad_norm": 0.011694240383803844,
+ "learning_rate": 0.00108,
+ "loss": 1.728348731994629,
+ "step": 37
+ },
+ {
+ "epoch": 2.375,
+ "grad_norm": 0.005738088861107826,
+ "learning_rate": 0.00111,
+ "loss": 1.7308874130249023,
+ "step": 38
+ },
+ {
+ "epoch": 2.4375,
+ "grad_norm": 0.010038000531494617,
+ "learning_rate": 0.00114,
+ "loss": 1.7276837825775146,
+ "step": 39
+ },
+ {
+ "epoch": 2.5,
+ "grad_norm": 0.01586989127099514,
+ "learning_rate": 0.00117,
+ "loss": 1.7368175983428955,
+ "step": 40
+ },
+ {
+ "epoch": 2.5625,
+ "grad_norm": 0.16198667883872986,
+ "learning_rate": 0.0012000000000000001,
+ "loss": 1.730246663093567,
+ "step": 41
+ },
+ {
+ "epoch": 2.625,
+ "grad_norm": 0.01653525047004223,
+ "learning_rate": 0.00123,
+ "loss": 1.7403852939605713,
+ "step": 42
+ },
+ {
+ "epoch": 2.6875,
+ "grad_norm": 0.01241418905556202,
+ "learning_rate": 0.00126,
+ "loss": 1.7306722402572632,
+ "step": 43
+ },
+ {
+ "epoch": 2.75,
+ "grad_norm": 0.022560030221939087,
+ "learning_rate": 0.00129,
+ "loss": 1.7223260402679443,
+ "step": 44
+ },
+ {
+ "epoch": 2.8125,
+ "grad_norm": 0.019568368792533875,
+ "learning_rate": 0.00132,
+ "loss": 1.7304983139038086,
+ "step": 45
+ },
+ {
+ "epoch": 2.875,
+ "grad_norm": 0.016755228862166405,
+ "learning_rate": 0.00135,
+ "loss": 1.7223992347717285,
+ "step": 46
+ },
+ {
+ "epoch": 2.9375,
+ "grad_norm": 0.02981281839311123,
+ "learning_rate": 0.0013800000000000002,
+ "loss": 1.7094981670379639,
+ "step": 47
+ },
+ {
+ "epoch": 3.0,
+ "grad_norm": 0.02506650611758232,
+ "learning_rate": 0.00141,
+ "loss": 1.7054555416107178,
+ "step": 48
+ },
+ {
+ "epoch": 3.0,
+ "eval_loss": 0.8416222333908081,
+ "eval_runtime": 30.3027,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 48
+ },
+ {
+ "epoch": 3.0625,
+ "grad_norm": 0.02260257676243782,
+ "learning_rate": 0.0014399999999999999,
+ "loss": 1.6888771057128906,
+ "step": 49
+ },
+ {
+ "epoch": 3.125,
+ "grad_norm": 0.02620658650994301,
+ "learning_rate": 0.00147,
+ "loss": 1.6639578342437744,
+ "step": 50
+ },
+ {
+ "epoch": 3.1875,
+ "grad_norm": 0.047354575246572495,
+ "learning_rate": 0.0015,
+ "loss": 1.6736350059509277,
+ "step": 51
+ },
+ {
+ "epoch": 3.25,
+ "grad_norm": 0.04594860598444939,
+ "learning_rate": 0.0015300000000000001,
+ "loss": 1.6675360202789307,
+ "step": 52
+ },
+ {
+ "epoch": 3.3125,
+ "grad_norm": 0.08131621032953262,
+ "learning_rate": 0.0015600000000000002,
+ "loss": 1.6574077606201172,
+ "step": 53
+ },
+ {
+ "epoch": 3.375,
+ "grad_norm": 0.028789900243282318,
+ "learning_rate": 0.00159,
+ "loss": 1.6299148797988892,
+ "step": 54
+ },
+ {
+ "epoch": 3.4375,
+ "grad_norm": 0.2623404860496521,
+ "learning_rate": 0.0016200000000000001,
+ "loss": 1.6315665245056152,
+ "step": 55
+ },
+ {
+ "epoch": 3.5,
+ "grad_norm": 0.056828465312719345,
+ "learning_rate": 0.0016500000000000002,
+ "loss": 1.6279325485229492,
+ "step": 56
+ },
+ {
+ "epoch": 3.5625,
+ "grad_norm": 0.0549125075340271,
+ "learning_rate": 0.0016800000000000003,
+ "loss": 1.6549303531646729,
+ "step": 57
+ },
+ {
+ "epoch": 3.625,
+ "grad_norm": 0.07287958264350891,
+ "learning_rate": 0.00171,
+ "loss": 1.606963872909546,
+ "step": 58
+ },
+ {
+ "epoch": 3.6875,
+ "grad_norm": 0.07426343113183975,
+ "learning_rate": 0.00174,
+ "loss": 1.6250534057617188,
+ "step": 59
+ },
+ {
+ "epoch": 3.75,
+ "grad_norm": 0.1364986002445221,
+ "learning_rate": 0.0017699999999999999,
+ "loss": 1.5964961051940918,
+ "step": 60
+ },
+ {
+ "epoch": 3.8125,
+ "grad_norm": 0.15007159113883972,
+ "learning_rate": 0.0018,
+ "loss": 1.6019010543823242,
+ "step": 61
+ },
+ {
+ "epoch": 3.875,
+ "grad_norm": 0.08447311073541641,
+ "learning_rate": 0.00183,
+ "loss": 1.5763543844223022,
+ "step": 62
+ },
+ {
+ "epoch": 3.9375,
+ "grad_norm": 0.1681874394416809,
+ "learning_rate": 0.00186,
+ "loss": 1.5769858360290527,
+ "step": 63
+ },
+ {
+ "epoch": 4.0,
+ "grad_norm": 0.09677319973707199,
+ "learning_rate": 0.00189,
+ "loss": 1.534682273864746,
+ "step": 64
+ },
+ {
+ "epoch": 4.0,
+ "eval_loss": 0.7682511806488037,
+ "eval_runtime": 30.4228,
+ "eval_samples_per_second": 3.353,
+ "eval_steps_per_second": 0.427,
+ "step": 64
+ },
+ {
+ "epoch": 4.0625,
+ "grad_norm": 0.42635980248451233,
+ "learning_rate": 0.00192,
+ "loss": 1.5509562492370605,
+ "step": 65
+ },
+ {
+ "epoch": 4.125,
+ "grad_norm": 0.8467249274253845,
+ "learning_rate": 0.0019500000000000001,
+ "loss": 1.5001531839370728,
+ "step": 66
+ },
+ {
+ "epoch": 4.1875,
+ "grad_norm": 0.22631719708442688,
+ "learning_rate": 0.00198,
+ "loss": 1.4554829597473145,
+ "step": 67
+ },
+ {
+ "epoch": 4.25,
+ "grad_norm": 0.05312464386224747,
+ "learning_rate": 0.00201,
+ "loss": 1.4732775688171387,
+ "step": 68
+ },
+ {
+ "epoch": 4.3125,
+ "grad_norm": 0.050442785024642944,
+ "learning_rate": 0.00204,
+ "loss": 1.4176194667816162,
+ "step": 69
+ },
+ {
+ "epoch": 4.375,
+ "grad_norm": 0.18886858224868774,
+ "learning_rate": 0.00207,
+ "loss": 1.4325997829437256,
+ "step": 70
+ },
+ {
+ "epoch": 4.4375,
+ "grad_norm": 0.05258706212043762,
+ "learning_rate": 0.0021,
+ "loss": 1.39802086353302,
+ "step": 71
+ },
+ {
+ "epoch": 4.5,
+ "grad_norm": 0.07482657581567764,
+ "learning_rate": 0.00213,
+ "loss": 1.4000320434570312,
+ "step": 72
+ },
+ {
+ "epoch": 4.5625,
+ "grad_norm": 0.4853312075138092,
+ "learning_rate": 0.00216,
+ "loss": 1.4199151992797852,
+ "step": 73
+ },
+ {
+ "epoch": 4.625,
+ "grad_norm": 0.48888301849365234,
+ "learning_rate": 0.00219,
+ "loss": 1.3049702644348145,
+ "step": 74
+ },
+ {
+ "epoch": 4.6875,
+ "grad_norm": 0.284348726272583,
+ "learning_rate": 0.00222,
+ "loss": 1.3884068727493286,
+ "step": 75
+ },
+ {
+ "epoch": 4.75,
+ "grad_norm": 0.20248942077159882,
+ "learning_rate": 0.0022500000000000003,
+ "loss": 1.3502120971679688,
+ "step": 76
+ },
+ {
+ "epoch": 4.8125,
+ "grad_norm": 0.2714093327522278,
+ "learning_rate": 0.00228,
+ "loss": 1.3195693492889404,
+ "step": 77
+ },
+ {
+ "epoch": 4.875,
+ "grad_norm": 0.2544061243534088,
+ "learning_rate": 0.00231,
+ "loss": 1.3194304704666138,
+ "step": 78
+ },
+ {
+ "epoch": 4.9375,
+ "grad_norm": 0.18475523591041565,
+ "learning_rate": 0.00234,
+ "loss": 1.3006361722946167,
+ "step": 79
+ },
+ {
+ "epoch": 5.0,
+ "grad_norm": 0.6583278775215149,
+ "learning_rate": 0.00237,
+ "loss": 1.2868517637252808,
+ "step": 80
+ },
+ {
+ "epoch": 5.0,
+ "eval_loss": 0.6618342399597168,
+ "eval_runtime": 30.1233,
+ "eval_samples_per_second": 3.386,
+ "eval_steps_per_second": 0.432,
+ "step": 80
+ },
+ {
+ "epoch": 5.0625,
+ "grad_norm": 0.04499061033129692,
+ "learning_rate": 0.0024000000000000002,
+ "loss": 1.25992751121521,
+ "step": 81
+ },
+ {
+ "epoch": 5.125,
+ "grad_norm": 0.05609860643744469,
+ "learning_rate": 0.0024300000000000003,
+ "loss": 1.2764487266540527,
+ "step": 82
+ },
+ {
+ "epoch": 5.1875,
+ "grad_norm": 0.03665148839354515,
+ "learning_rate": 0.00246,
+ "loss": 1.2908766269683838,
+ "step": 83
+ },
+ {
+ "epoch": 5.25,
+ "grad_norm": 0.17949481308460236,
+ "learning_rate": 0.00249,
+ "loss": 1.253080129623413,
+ "step": 84
+ },
+ {
+ "epoch": 5.3125,
+ "grad_norm": 0.14031393826007843,
+ "learning_rate": 0.00252,
+ "loss": 1.2648870944976807,
+ "step": 85
+ },
+ {
+ "epoch": 5.375,
+ "grad_norm": 0.08193210512399673,
+ "learning_rate": 0.00255,
+ "loss": 1.296234130859375,
+ "step": 86
+ },
+ {
+ "epoch": 5.4375,
+ "grad_norm": 0.056413356214761734,
+ "learning_rate": 0.00258,
+ "loss": 1.2784960269927979,
+ "step": 87
+ },
+ {
+ "epoch": 5.5,
+ "grad_norm": 0.09004318714141846,
+ "learning_rate": 0.00261,
+ "loss": 1.2054948806762695,
+ "step": 88
+ },
+ {
+ "epoch": 5.5625,
+ "grad_norm": 0.044756047427654266,
+ "learning_rate": 0.00264,
+ "loss": 1.2707666158676147,
+ "step": 89
+ },
+ {
+ "epoch": 5.625,
+ "grad_norm": 0.049932755529880524,
+ "learning_rate": 0.00267,
+ "loss": 1.2369787693023682,
+ "step": 90
+ },
+ {
+ "epoch": 5.6875,
+ "grad_norm": 0.09624820947647095,
+ "learning_rate": 0.0027,
+ "loss": 1.2419278621673584,
+ "step": 91
+ },
+ {
+ "epoch": 5.75,
+ "grad_norm": 0.17648059129714966,
+ "learning_rate": 0.0027300000000000002,
+ "loss": 1.2992033958435059,
+ "step": 92
+ },
+ {
+ "epoch": 5.8125,
+ "grad_norm": 0.07027823477983475,
+ "learning_rate": 0.0027600000000000003,
+ "loss": 1.2255799770355225,
+ "step": 93
+ },
+ {
+ "epoch": 5.875,
+ "grad_norm": 0.06040462478995323,
+ "learning_rate": 0.0027900000000000004,
+ "loss": 1.136023759841919,
+ "step": 94
+ },
+ {
+ "epoch": 5.9375,
+ "grad_norm": 0.09845872968435287,
+ "learning_rate": 0.00282,
+ "loss": 1.1807364225387573,
+ "step": 95
+ },
+ {
+ "epoch": 6.0,
+ "grad_norm": 0.09086572378873825,
+ "learning_rate": 0.00285,
+ "loss": 1.2171341180801392,
+ "step": 96
+ },
+ {
+ "epoch": 6.0,
+ "eval_loss": 0.6200892329216003,
+ "eval_runtime": 30.1252,
+ "eval_samples_per_second": 3.386,
+ "eval_steps_per_second": 0.432,
+ "step": 96
+ },
+ {
+ "epoch": 6.0625,
+ "grad_norm": 0.11787034571170807,
+ "learning_rate": 0.0028799999999999997,
+ "loss": 1.242072582244873,
+ "step": 97
+ },
+ {
+ "epoch": 6.125,
+ "grad_norm": 0.20093177258968353,
+ "learning_rate": 0.00291,
+ "loss": 1.2056899070739746,
+ "step": 98
+ },
+ {
+ "epoch": 6.1875,
+ "grad_norm": 0.03221331909298897,
+ "learning_rate": 0.00294,
+ "loss": 1.1774635314941406,
+ "step": 99
+ },
+ {
+ "epoch": 6.25,
+ "grad_norm": 0.06112038344144821,
+ "learning_rate": 0.00297,
+ "loss": 1.1831618547439575,
+ "step": 100
+ },
+ {
+ "epoch": 6.3125,
+ "grad_norm": 0.05896962434053421,
+ "learning_rate": 0.003,
+ "loss": 1.219610571861267,
+ "step": 101
+ },
+ {
+ "epoch": 6.375,
+ "grad_norm": 0.4068014323711395,
+ "learning_rate": 0.0029999850445531283,
+ "loss": 1.231442928314209,
+ "step": 102
+ },
+ {
+ "epoch": 6.4375,
+ "grad_norm": 3.013302803039551,
+ "learning_rate": 0.0029999401785137437,
+ "loss": 1.2512195110321045,
+ "step": 103
+ },
+ {
+ "epoch": 6.5,
+ "grad_norm": 0.07003403455018997,
+ "learning_rate": 0.0029998654027855406,
+ "loss": 1.1912157535552979,
+ "step": 104
+ },
+ {
+ "epoch": 6.5625,
+ "grad_norm": 1.1927955150604248,
+ "learning_rate": 0.002999760718874652,
+ "loss": 1.1158745288848877,
+ "step": 105
+ },
+ {
+ "epoch": 6.625,
+ "grad_norm": 0.11684458702802658,
+ "learning_rate": 0.0029996261288896225,
+ "loss": 1.1835001707077026,
+ "step": 106
+ },
+ {
+ "epoch": 6.6875,
+ "grad_norm": 0.07812155783176422,
+ "learning_rate": 0.0029994616355413666,
+ "loss": 1.1712734699249268,
+ "step": 107
+ },
+ {
+ "epoch": 6.75,
+ "grad_norm": 0.19128946959972382,
+ "learning_rate": 0.0029992672421431115,
+ "loss": 1.0906574726104736,
+ "step": 108
+ },
+ {
+ "epoch": 6.8125,
+ "grad_norm": 0.033625371754169464,
+ "learning_rate": 0.002999042952610332,
+ "loss": 1.115525245666504,
+ "step": 109
+ },
+ {
+ "epoch": 6.875,
+ "grad_norm": 0.2518763244152069,
+ "learning_rate": 0.0029987887714606723,
+ "loss": 1.1433348655700684,
+ "step": 110
+ },
+ {
+ "epoch": 6.9375,
+ "grad_norm": 0.037159573286771774,
+ "learning_rate": 0.002998504703813852,
+ "loss": 1.1867601871490479,
+ "step": 111
+ },
+ {
+ "epoch": 7.0,
+ "grad_norm": 0.04564391076564789,
+ "learning_rate": 0.0029981907553915674,
+ "loss": 1.1302828788757324,
+ "step": 112
+ },
+ {
+ "epoch": 7.0,
+ "eval_loss": 0.5897142887115479,
+ "eval_runtime": 31.3671,
+ "eval_samples_per_second": 3.252,
+ "eval_steps_per_second": 0.414,
+ "step": 112
+ },
+ {
+ "epoch": 7.0625,
+ "grad_norm": 0.05205775797367096,
+ "learning_rate": 0.0029978469325173717,
+ "loss": 1.182741641998291,
+ "step": 113
+ },
+ {
+ "epoch": 7.125,
+ "grad_norm": 0.12080052495002747,
+ "learning_rate": 0.002997473242116551,
+ "loss": 1.1457600593566895,
+ "step": 114
+ },
+ {
+ "epoch": 7.1875,
+ "grad_norm": 0.05600045621395111,
+ "learning_rate": 0.0029970696917159833,
+ "loss": 1.1101720333099365,
+ "step": 115
+ },
+ {
+ "epoch": 7.25,
+ "grad_norm": 0.0507950522005558,
+ "learning_rate": 0.0029966362894439874,
+ "loss": 1.1785833835601807,
+ "step": 116
+ },
+ {
+ "epoch": 7.3125,
+ "grad_norm": 0.6396550536155701,
+ "learning_rate": 0.002996173044030159,
+ "loss": 1.1538987159729004,
+ "step": 117
+ },
+ {
+ "epoch": 7.375,
+ "grad_norm": 0.06849977374076843,
+ "learning_rate": 0.0029956799648051934,
+ "loss": 1.1528857946395874,
+ "step": 118
+ },
+ {
+ "epoch": 7.4375,
+ "grad_norm": 0.0685896947979927,
+ "learning_rate": 0.002995157061700702,
+ "loss": 1.1049470901489258,
+ "step": 119
+ },
+ {
+ "epoch": 7.5,
+ "grad_norm": 1.027722716331482,
+ "learning_rate": 0.002994604345249006,
+ "loss": 1.1308050155639648,
+ "step": 120
+ },
+ {
+ "epoch": 7.5625,
+ "grad_norm": 0.21406109631061554,
+ "learning_rate": 0.00299402182658293,
+ "loss": 1.0960861444473267,
+ "step": 121
+ },
+ {
+ "epoch": 7.625,
+ "grad_norm": 0.0503656342625618,
+ "learning_rate": 0.002993409517435574,
+ "loss": 1.1096141338348389,
+ "step": 122
+ },
+ {
+ "epoch": 7.6875,
+ "grad_norm": 0.04800846800208092,
+ "learning_rate": 0.002992767430140078,
+ "loss": 1.1386680603027344,
+ "step": 123
+ },
+ {
+ "epoch": 7.75,
+ "grad_norm": 0.13091202080249786,
+ "learning_rate": 0.002992095577629376,
+ "loss": 1.1129441261291504,
+ "step": 124
+ },
+ {
+ "epoch": 7.8125,
+ "grad_norm": 0.04949796572327614,
+ "learning_rate": 0.0029913939734359306,
+ "loss": 1.106520175933838,
+ "step": 125
+ },
+ {
+ "epoch": 7.875,
+ "grad_norm": 0.07686582207679749,
+ "learning_rate": 0.0029906626316914655,
+ "loss": 1.1572234630584717,
+ "step": 126
+ },
+ {
+ "epoch": 7.9375,
+ "grad_norm": 0.053587451577186584,
+ "learning_rate": 0.0029899015671266773,
+ "loss": 1.125717282295227,
+ "step": 127
+ },
+ {
+ "epoch": 8.0,
+ "grad_norm": 0.08031050115823746,
+ "learning_rate": 0.0029891107950709406,
+ "loss": 1.153022050857544,
+ "step": 128
+ },
+ {
+ "epoch": 8.0,
+ "eval_loss": 0.5910838842391968,
+ "eval_runtime": 30.7073,
+ "eval_samples_per_second": 3.322,
+ "eval_steps_per_second": 0.423,
+ "step": 128
+ },
+ {
+ "epoch": 8.0625,
+ "grad_norm": 0.3520950675010681,
+ "learning_rate": 0.0029882903314519997,
+ "loss": 1.1582988500595093,
+ "step": 129
+ },
+ {
+ "epoch": 8.125,
+ "grad_norm": 0.07153584063053131,
+ "learning_rate": 0.002987440192795645,
+ "loss": 1.1344678401947021,
+ "step": 130
+ },
+ {
+ "epoch": 8.1875,
+ "grad_norm": 0.06336807459592819,
+ "learning_rate": 0.0029865603962253833,
+ "loss": 1.120319128036499,
+ "step": 131
+ },
+ {
+ "epoch": 8.25,
+ "grad_norm": 0.11782127618789673,
+ "learning_rate": 0.0029856509594620912,
+ "loss": 1.1245379447937012,
+ "step": 132
+ },
+ {
+ "epoch": 8.3125,
+ "grad_norm": 0.2592264413833618,
+ "learning_rate": 0.0029847119008236576,
+ "loss": 1.1454271078109741,
+ "step": 133
+ },
+ {
+ "epoch": 8.375,
+ "grad_norm": 0.21255695819854736,
+ "learning_rate": 0.0029837432392246185,
+ "loss": 1.231805443763733,
+ "step": 134
+ },
+ {
+ "epoch": 8.4375,
+ "grad_norm": 15.057058334350586,
+ "learning_rate": 0.0029827449941757698,
+ "loss": 1.1380956172943115,
+ "step": 135
+ },
+ {
+ "epoch": 8.5,
+ "grad_norm": 0.17857126891613007,
+ "learning_rate": 0.00298171718578378,
+ "loss": 1.1401441097259521,
+ "step": 136
+ },
+ {
+ "epoch": 8.5625,
+ "grad_norm": 0.18281924724578857,
+ "learning_rate": 0.002980659834750781,
+ "loss": 1.171494722366333,
+ "step": 137
+ },
+ {
+ "epoch": 8.625,
+ "grad_norm": 0.10490825027227402,
+ "learning_rate": 0.0029795729623739555,
+ "loss": 1.1622605323791504,
+ "step": 138
+ },
+ {
+ "epoch": 8.6875,
+ "grad_norm": 0.344013512134552,
+ "learning_rate": 0.002978456590545104,
+ "loss": 1.1702141761779785,
+ "step": 139
+ },
+ {
+ "epoch": 8.75,
+ "grad_norm": 0.05058368295431137,
+ "learning_rate": 0.0029773107417502055,
+ "loss": 1.2006211280822754,
+ "step": 140
+ },
+ {
+ "epoch": 8.8125,
+ "grad_norm": 0.09775760769844055,
+ "learning_rate": 0.002976135439068965,
+ "loss": 1.1407063007354736,
+ "step": 141
+ },
+ {
+ "epoch": 8.875,
+ "grad_norm": 0.0785030648112297,
+ "learning_rate": 0.0029749307061743475,
+ "loss": 1.1778209209442139,
+ "step": 142
+ },
+ {
+ "epoch": 8.9375,
+ "grad_norm": 0.053056441247463226,
+ "learning_rate": 0.0029736965673321032,
+ "loss": 1.1302082538604736,
+ "step": 143
+ },
+ {
+ "epoch": 9.0,
+ "grad_norm": 0.4020994305610657,
+ "learning_rate": 0.0029724330474002743,
+ "loss": 1.1655206680297852,
+ "step": 144
+ },
+ {
+ "epoch": 9.0,
+ "eval_loss": 0.6083483099937439,
+ "eval_runtime": 30.0453,
+ "eval_samples_per_second": 3.395,
+ "eval_steps_per_second": 0.433,
+ "step": 144
+ },
+ {
+ "epoch": 9.0625,
+ "grad_norm": 0.04755420982837677,
+ "learning_rate": 0.002971140171828702,
+ "loss": 1.1900575160980225,
+ "step": 145
+ },
+ {
+ "epoch": 9.125,
+ "grad_norm": 0.2549609839916229,
+ "learning_rate": 0.002969817966658505,
+ "loss": 1.1891400814056396,
+ "step": 146
+ },
+ {
+ "epoch": 9.1875,
+ "grad_norm": 0.11641769111156464,
+ "learning_rate": 0.0029684664585215618,
+ "loss": 1.1930484771728516,
+ "step": 147
+ },
+ {
+ "epoch": 9.25,
+ "grad_norm": 0.13052672147750854,
+ "learning_rate": 0.00296708567463997,
+ "loss": 1.1905772686004639,
+ "step": 148
+ },
+ {
+ "epoch": 9.3125,
+ "grad_norm": 0.15126636624336243,
+ "learning_rate": 0.0029656756428255017,
+ "loss": 1.1417806148529053,
+ "step": 149
+ },
+ {
+ "epoch": 9.375,
+ "grad_norm": 0.08826905488967896,
+ "learning_rate": 0.00296423639147904,
+ "loss": 1.1802985668182373,
+ "step": 150
+ },
+ {
+ "epoch": 9.4375,
+ "grad_norm": 0.01381592359393835,
+ "learning_rate": 0.002962767949590008,
+ "loss": 1.1525968313217163,
+ "step": 151
+ },
+ {
+ "epoch": 9.5,
+ "grad_norm": 0.04648139327764511,
+ "learning_rate": 0.0029612703467357866,
+ "loss": 1.169856309890747,
+ "step": 152
+ },
+ {
+ "epoch": 9.5625,
+ "grad_norm": 0.014019235037267208,
+ "learning_rate": 0.0029597436130811153,
+ "loss": 1.1442453861236572,
+ "step": 153
+ },
+ {
+ "epoch": 9.625,
+ "grad_norm": 0.03701096400618553,
+ "learning_rate": 0.0029581877793774885,
+ "loss": 1.112680435180664,
+ "step": 154
+ },
+ {
+ "epoch": 9.6875,
+ "grad_norm": 0.030421797186136246,
+ "learning_rate": 0.0029566028769625334,
+ "loss": 1.223894715309143,
+ "step": 155
+ },
+ {
+ "epoch": 9.75,
+ "grad_norm": 0.12336083501577377,
+ "learning_rate": 0.00295498893775938,
+ "loss": 1.220346212387085,
+ "step": 156
+ },
+ {
+ "epoch": 9.8125,
+ "grad_norm": 0.038585737347602844,
+ "learning_rate": 0.0029533459942760174,
+ "loss": 1.2078583240509033,
+ "step": 157
+ },
+ {
+ "epoch": 9.875,
+ "grad_norm": 0.4653286039829254,
+ "learning_rate": 0.002951674079604639,
+ "loss": 1.2035815715789795,
+ "step": 158
+ },
+ {
+ "epoch": 9.9375,
+ "grad_norm": 0.020640619099140167,
+ "learning_rate": 0.0029499732274209777,
+ "loss": 1.1313148736953735,
+ "step": 159
+ },
+ {
+ "epoch": 10.0,
+ "grad_norm": 0.04939777031540871,
+ "learning_rate": 0.002948243471983625,
+ "loss": 1.1253995895385742,
+ "step": 160
+ },
+ {
+ "epoch": 10.0,
+ "eval_loss": 0.604421079158783,
+ "eval_runtime": 30.1556,
+ "eval_samples_per_second": 3.382,
+ "eval_steps_per_second": 0.431,
+ "step": 160
+ },
+ {
+ "epoch": 10.0625,
+ "grad_norm": 0.05968557298183441,
+ "learning_rate": 0.0029464848481333423,
+ "loss": 1.1324307918548584,
+ "step": 161
+ },
+ {
+ "epoch": 10.125,
+ "grad_norm": 1.0001988410949707,
+ "learning_rate": 0.0029446973912923597,
+ "loss": 1.126110315322876,
+ "step": 162
+ },
+ {
+ "epoch": 10.1875,
+ "grad_norm": 0.2294934093952179,
+ "learning_rate": 0.002942881137463661,
+ "loss": 1.1873570680618286,
+ "step": 163
+ },
+ {
+ "epoch": 10.25,
+ "grad_norm": 0.06396346539258957,
+ "learning_rate": 0.002941036123230261,
+ "loss": 1.1288526058197021,
+ "step": 164
+ },
+ {
+ "epoch": 10.3125,
+ "grad_norm": 0.3410445749759674,
+ "learning_rate": 0.0029391623857544643,
+ "loss": 1.1219967603683472,
+ "step": 165
+ },
+ {
+ "epoch": 10.375,
+ "grad_norm": 0.09721948206424713,
+ "learning_rate": 0.002937259962777122,
+ "loss": 1.141520619392395,
+ "step": 166
+ },
+ {
+ "epoch": 10.4375,
+ "grad_norm": 0.033714450895786285,
+ "learning_rate": 0.002935328892616868,
+ "loss": 1.1983453035354614,
+ "step": 167
+ },
+ {
+ "epoch": 10.5,
+ "grad_norm": 0.028446367010474205,
+ "learning_rate": 0.0029333692141693484,
+ "loss": 1.153167724609375,
+ "step": 168
+ },
+ {
+ "epoch": 10.5625,
+ "grad_norm": 0.04298163205385208,
+ "learning_rate": 0.0029313809669064377,
+ "loss": 1.120814561843872,
+ "step": 169
+ },
+ {
+ "epoch": 10.625,
+ "grad_norm": 0.08190953731536865,
+ "learning_rate": 0.002929364190875444,
+ "loss": 1.1164218187332153,
+ "step": 170
+ },
+ {
+ "epoch": 10.6875,
+ "grad_norm": 0.0602508969604969,
+ "learning_rate": 0.0029273189266983035,
+ "loss": 1.136858344078064,
+ "step": 171
+ },
+ {
+ "epoch": 10.75,
+ "grad_norm": 0.07118846476078033,
+ "learning_rate": 0.0029252452155707582,
+ "loss": 1.1356765031814575,
+ "step": 172
+ },
+ {
+ "epoch": 10.8125,
+ "grad_norm": 0.0445970743894577,
+ "learning_rate": 0.002923143099261531,
+ "loss": 1.121457815170288,
+ "step": 173
+ },
+ {
+ "epoch": 10.875,
+ "grad_norm": 0.07745611667633057,
+ "learning_rate": 0.0029210126201114836,
+ "loss": 1.0903944969177246,
+ "step": 174
+ },
+ {
+ "epoch": 10.9375,
+ "grad_norm": 0.08945920318365097,
+ "learning_rate": 0.0029188538210327595,
+ "loss": 1.1194133758544922,
+ "step": 175
+ },
+ {
+ "epoch": 11.0,
+ "grad_norm": 0.07996643334627151,
+ "learning_rate": 0.002916666745507926,
+ "loss": 1.1248202323913574,
+ "step": 176
+ },
+ {
+ "epoch": 11.0,
+ "eval_loss": 0.5789638757705688,
+ "eval_runtime": 30.3538,
+ "eval_samples_per_second": 3.36,
+ "eval_steps_per_second": 0.428,
+ "step": 176
+ },
+ {
+ "epoch": 11.0625,
+ "grad_norm": 0.10533642023801804,
+ "learning_rate": 0.002914451437589092,
+ "loss": 1.1199195384979248,
+ "step": 177
+ },
+ {
+ "epoch": 11.125,
+ "grad_norm": 0.044989872723817825,
+ "learning_rate": 0.002912207941897025,
+ "loss": 1.1084158420562744,
+ "step": 178
+ },
+ {
+ "epoch": 11.1875,
+ "grad_norm": 0.04824615642428398,
+ "learning_rate": 0.0029099363036202515,
+ "loss": 1.0993731021881104,
+ "step": 179
+ },
+ {
+ "epoch": 11.25,
+ "grad_norm": 0.03144732117652893,
+ "learning_rate": 0.0029076365685141463,
+ "loss": 1.1078145503997803,
+ "step": 180
+ },
+ {
+ "epoch": 11.3125,
+ "grad_norm": 0.03728745877742767,
+ "learning_rate": 0.0029053087829000097,
+ "loss": 1.0995632410049438,
+ "step": 181
+ },
+ {
+ "epoch": 11.375,
+ "grad_norm": 0.02090257592499256,
+ "learning_rate": 0.002902952993664138,
+ "loss": 1.0722652673721313,
+ "step": 182
+ },
+ {
+ "epoch": 11.4375,
+ "grad_norm": 0.020907679572701454,
+ "learning_rate": 0.0029005692482568747,
+ "loss": 1.0840482711791992,
+ "step": 183
+ },
+ {
+ "epoch": 11.5,
+ "grad_norm": 0.023914694786071777,
+ "learning_rate": 0.002898157594691658,
+ "loss": 1.1289193630218506,
+ "step": 184
+ },
+ {
+ "epoch": 11.5625,
+ "grad_norm": 0.01937722973525524,
+ "learning_rate": 0.0028957180815440534,
+ "loss": 1.1244895458221436,
+ "step": 185
+ },
+ {
+ "epoch": 11.625,
+ "grad_norm": 0.02047949656844139,
+ "learning_rate": 0.002893250757950773,
+ "loss": 1.0527830123901367,
+ "step": 186
+ },
+ {
+ "epoch": 11.6875,
+ "grad_norm": 0.04898907244205475,
+ "learning_rate": 0.002890755673608688,
+ "loss": 1.1234995126724243,
+ "step": 187
+ },
+ {
+ "epoch": 11.75,
+ "grad_norm": 0.024841446429491043,
+ "learning_rate": 0.0028882328787738273,
+ "loss": 1.138411045074463,
+ "step": 188
+ },
+ {
+ "epoch": 11.8125,
+ "grad_norm": 0.02533135376870632,
+ "learning_rate": 0.002885682424260365,
+ "loss": 1.1021623611450195,
+ "step": 189
+ },
+ {
+ "epoch": 11.875,
+ "grad_norm": 0.04309404641389847,
+ "learning_rate": 0.0028831043614395966,
+ "loss": 1.0767666101455688,
+ "step": 190
+ },
+ {
+ "epoch": 11.9375,
+ "grad_norm": 0.029682395979762077,
+ "learning_rate": 0.0028804987422389045,
+ "loss": 1.1310434341430664,
+ "step": 191
+ },
+ {
+ "epoch": 12.0,
+ "grad_norm": 0.021236268803477287,
+ "learning_rate": 0.002877865619140712,
+ "loss": 1.0969300270080566,
+ "step": 192
+ },
+ {
+ "epoch": 12.0,
+ "eval_loss": 0.5664089322090149,
+ "eval_runtime": 30.1046,
+ "eval_samples_per_second": 3.388,
+ "eval_steps_per_second": 0.432,
+ "step": 192
+ },
+ {
+ "epoch": 12.0625,
+ "grad_norm": 0.09511730819940567,
+ "learning_rate": 0.0028752050451814276,
+ "loss": 1.0866225957870483,
+ "step": 193
+ },
+ {
+ "epoch": 12.125,
+ "grad_norm": 0.1858353465795517,
+ "learning_rate": 0.0028725170739503734,
+ "loss": 1.1082183122634888,
+ "step": 194
+ },
+ {
+ "epoch": 12.1875,
+ "grad_norm": 0.018049633130431175,
+ "learning_rate": 0.00286980175958871,
+ "loss": 1.0569005012512207,
+ "step": 195
+ },
+ {
+ "epoch": 12.25,
+ "grad_norm": 0.06167157366871834,
+ "learning_rate": 0.002867059156788342,
+ "loss": 1.1418097019195557,
+ "step": 196
+ },
+ {
+ "epoch": 12.3125,
+ "grad_norm": 0.06766022741794586,
+ "learning_rate": 0.0028642893207908187,
+ "loss": 1.0995159149169922,
+ "step": 197
+ },
+ {
+ "epoch": 12.375,
+ "grad_norm": 0.03883042931556702,
+ "learning_rate": 0.0028614923073862213,
+ "loss": 1.0847246646881104,
+ "step": 198
+ },
+ {
+ "epoch": 12.4375,
+ "grad_norm": 0.05537786707282066,
+ "learning_rate": 0.0028586681729120393,
+ "loss": 1.1187801361083984,
+ "step": 199
+ },
+ {
+ "epoch": 12.5,
+ "grad_norm": 0.0653286948800087,
+ "learning_rate": 0.002855816974252033,
+ "loss": 1.0798084735870361,
+ "step": 200
+ },
+ {
+ "epoch": 12.5625,
+ "grad_norm": 0.04398065432906151,
+ "learning_rate": 0.002852938768835092,
+ "loss": 1.0946357250213623,
+ "step": 201
+ },
+ {
+ "epoch": 12.625,
+ "grad_norm": 0.05614284798502922,
+ "learning_rate": 0.0028500336146340772,
+ "loss": 1.1144163608551025,
+ "step": 202
+ },
+ {
+ "epoch": 12.6875,
+ "grad_norm": 0.035650063306093216,
+ "learning_rate": 0.0028471015701646485,
+ "loss": 1.0421898365020752,
+ "step": 203
+ },
+ {
+ "epoch": 12.75,
+ "grad_norm": 0.02544066496193409,
+ "learning_rate": 0.002844142694484094,
+ "loss": 1.0994564294815063,
+ "step": 204
+ },
+ {
+ "epoch": 12.8125,
+ "grad_norm": 0.07384507358074188,
+ "learning_rate": 0.002841157047190135,
+ "loss": 1.0665926933288574,
+ "step": 205
+ },
+ {
+ "epoch": 12.875,
+ "grad_norm": 0.04875814542174339,
+ "learning_rate": 0.002838144688419726,
+ "loss": 1.0665624141693115,
+ "step": 206
+ },
+ {
+ "epoch": 12.9375,
+ "grad_norm": 0.05739036574959755,
+ "learning_rate": 0.002835105678847847,
+ "loss": 1.1085939407348633,
+ "step": 207
+ },
+ {
+ "epoch": 13.0,
+ "grad_norm": 0.03088603913784027,
+ "learning_rate": 0.002832040079686276,
+ "loss": 1.1357779502868652,
+ "step": 208
+ },
+ {
+ "epoch": 13.0,
+ "eval_loss": 0.5675534009933472,
+ "eval_runtime": 31.0754,
+ "eval_samples_per_second": 3.282,
+ "eval_steps_per_second": 0.418,
+ "step": 208
+ },
+ {
+ "epoch": 13.0625,
+ "grad_norm": 0.052636317908763885,
+ "learning_rate": 0.0028289479526823614,
+ "loss": 1.0684641599655151,
+ "step": 209
+ },
+ {
+ "epoch": 13.125,
+ "grad_norm": 0.26156774163246155,
+ "learning_rate": 0.0028258293601177743,
+ "loss": 1.1112414598464966,
+ "step": 210
+ },
+ {
+ "epoch": 13.1875,
+ "grad_norm": 0.013046488165855408,
+ "learning_rate": 0.0028226843648072562,
+ "loss": 1.0339481830596924,
+ "step": 211
+ },
+ {
+ "epoch": 13.25,
+ "grad_norm": 0.029812589287757874,
+ "learning_rate": 0.002819513030097352,
+ "loss": 1.0734567642211914,
+ "step": 212
+ },
+ {
+ "epoch": 13.3125,
+ "grad_norm": 0.03711036592721939,
+ "learning_rate": 0.0028163154198651372,
+ "loss": 1.071413516998291,
+ "step": 213
+ },
+ {
+ "epoch": 13.375,
+ "grad_norm": 0.09543009102344513,
+ "learning_rate": 0.002813091598516928,
+ "loss": 1.1175503730773926,
+ "step": 214
+ },
+ {
+ "epoch": 13.4375,
+ "grad_norm": 0.02742145210504532,
+ "learning_rate": 0.0028098416309869857,
+ "loss": 1.0401530265808105,
+ "step": 215
+ },
+ {
+ "epoch": 13.5,
+ "grad_norm": 0.05347360298037529,
+ "learning_rate": 0.002806565582736208,
+ "loss": 1.053802728652954,
+ "step": 216
+ },
+ {
+ "epoch": 13.5625,
+ "grad_norm": 0.023394430056214333,
+ "learning_rate": 0.002803263519750811,
+ "loss": 1.0774943828582764,
+ "step": 217
+ },
+ {
+ "epoch": 13.625,
+ "grad_norm": 0.03274613991379738,
+ "learning_rate": 0.0027999355085409996,
+ "loss": 1.1332664489746094,
+ "step": 218
+ },
+ {
+ "epoch": 13.6875,
+ "grad_norm": 0.029956744983792305,
+ "learning_rate": 0.00279658161613963,
+ "loss": 1.1192328929901123,
+ "step": 219
+ },
+ {
+ "epoch": 13.75,
+ "grad_norm": 0.027843108400702477,
+ "learning_rate": 0.002793201910100856,
+ "loss": 1.0953505039215088,
+ "step": 220
+ },
+ {
+ "epoch": 13.8125,
+ "grad_norm": 0.3722478151321411,
+ "learning_rate": 0.0027897964584987713,
+ "loss": 1.1045243740081787,
+ "step": 221
+ },
+ {
+ "epoch": 13.875,
+ "grad_norm": 0.05480271577835083,
+ "learning_rate": 0.0027863653299260356,
+ "loss": 1.0857124328613281,
+ "step": 222
+ },
+ {
+ "epoch": 13.9375,
+ "grad_norm": 0.05806222930550575,
+ "learning_rate": 0.0027829085934924958,
+ "loss": 1.092329502105713,
+ "step": 223
+ },
+ {
+ "epoch": 14.0,
+ "grad_norm": 0.07507319748401642,
+ "learning_rate": 0.0027794263188237922,
+ "loss": 1.0946941375732422,
+ "step": 224
+ },
+ {
+ "epoch": 14.0,
+ "eval_loss": 0.5624361038208008,
+ "eval_runtime": 30.3035,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 224
+ },
+ {
+ "epoch": 14.0625,
+ "grad_norm": 0.08125249296426773,
+ "learning_rate": 0.0027759185760599576,
+ "loss": 1.082047939300537,
+ "step": 225
+ },
+ {
+ "epoch": 14.125,
+ "grad_norm": 0.08838590234518051,
+ "learning_rate": 0.002772385435854002,
+ "loss": 1.0747947692871094,
+ "step": 226
+ },
+ {
+ "epoch": 14.1875,
+ "grad_norm": 0.18887385725975037,
+ "learning_rate": 0.0027688269693704926,
+ "loss": 1.0601060390472412,
+ "step": 227
+ },
+ {
+ "epoch": 14.25,
+ "grad_norm": 0.4315606653690338,
+ "learning_rate": 0.002765243248284118,
+ "loss": 1.167302131652832,
+ "step": 228
+ },
+ {
+ "epoch": 14.3125,
+ "grad_norm": 0.07665648311376572,
+ "learning_rate": 0.002761634344778245,
+ "loss": 1.130738615989685,
+ "step": 229
+ },
+ {
+ "epoch": 14.375,
+ "grad_norm": 0.051125217229127884,
+ "learning_rate": 0.002758000331543466,
+ "loss": 1.0479257106781006,
+ "step": 230
+ },
+ {
+ "epoch": 14.4375,
+ "grad_norm": 0.0658797100186348,
+ "learning_rate": 0.0027543412817761345,
+ "loss": 1.0913889408111572,
+ "step": 231
+ },
+ {
+ "epoch": 14.5,
+ "grad_norm": 0.1936631202697754,
+ "learning_rate": 0.0027506572691768888,
+ "loss": 1.0582077503204346,
+ "step": 232
+ },
+ {
+ "epoch": 14.5625,
+ "grad_norm": 2.202739953994751,
+ "learning_rate": 0.0027469483679491693,
+ "loss": 1.0946757793426514,
+ "step": 233
+ },
+ {
+ "epoch": 14.625,
+ "grad_norm": 0.07856857776641846,
+ "learning_rate": 0.002743214652797724,
+ "loss": 1.074622631072998,
+ "step": 234
+ },
+ {
+ "epoch": 14.6875,
+ "grad_norm": 0.04288165271282196,
+ "learning_rate": 0.002739456198927104,
+ "loss": 1.0788205862045288,
+ "step": 235
+ },
+ {
+ "epoch": 14.75,
+ "grad_norm": 0.08169898390769958,
+ "learning_rate": 0.0027356730820401465,
+ "loss": 1.0906422138214111,
+ "step": 236
+ },
+ {
+ "epoch": 14.8125,
+ "grad_norm": 0.04536404460668564,
+ "learning_rate": 0.002731865378336452,
+ "loss": 1.0542757511138916,
+ "step": 237
+ },
+ {
+ "epoch": 14.875,
+ "grad_norm": 0.0246584415435791,
+ "learning_rate": 0.0027280331645108498,
+ "loss": 1.0401986837387085,
+ "step": 238
+ },
+ {
+ "epoch": 14.9375,
+ "grad_norm": 0.03399847820401192,
+ "learning_rate": 0.0027241765177518523,
+ "loss": 1.109243392944336,
+ "step": 239
+ },
+ {
+ "epoch": 15.0,
+ "grad_norm": 0.12445052713155746,
+ "learning_rate": 0.0027202955157401,
+ "loss": 1.0952987670898438,
+ "step": 240
+ },
+ {
+ "epoch": 15.0,
+ "eval_loss": 0.5670443177223206,
+ "eval_runtime": 30.7619,
+ "eval_samples_per_second": 3.316,
+ "eval_steps_per_second": 0.423,
+ "step": 240
+ },
+ {
+ "epoch": 15.0625,
+ "grad_norm": 0.029460057616233826,
+ "learning_rate": 0.002716390236646797,
+ "loss": 1.0749149322509766,
+ "step": 241
+ },
+ {
+ "epoch": 15.125,
+ "grad_norm": 0.09484875947237015,
+ "learning_rate": 0.0027124607591321377,
+ "loss": 1.1196999549865723,
+ "step": 242
+ },
+ {
+ "epoch": 15.1875,
+ "grad_norm": 0.46802467107772827,
+ "learning_rate": 0.002708507162343721,
+ "loss": 1.0824053287506104,
+ "step": 243
+ },
+ {
+ "epoch": 15.25,
+ "grad_norm": 0.02694086730480194,
+ "learning_rate": 0.002704529525914957,
+ "loss": 1.015854835510254,
+ "step": 244
+ },
+ {
+ "epoch": 15.3125,
+ "grad_norm": 0.0534524992108345,
+ "learning_rate": 0.0027005279299634616,
+ "loss": 1.1357049942016602,
+ "step": 245
+ },
+ {
+ "epoch": 15.375,
+ "grad_norm": 0.17367368936538696,
+ "learning_rate": 0.002696502455089445,
+ "loss": 1.0634117126464844,
+ "step": 246
+ },
+ {
+ "epoch": 15.4375,
+ "grad_norm": 0.044344011694192886,
+ "learning_rate": 0.002692453182374086,
+ "loss": 1.1020934581756592,
+ "step": 247
+ },
+ {
+ "epoch": 15.5,
+ "grad_norm": 0.2549358606338501,
+ "learning_rate": 0.0026883801933779003,
+ "loss": 1.0967700481414795,
+ "step": 248
+ },
+ {
+ "epoch": 15.5625,
+ "grad_norm": 0.09740083664655685,
+ "learning_rate": 0.0026842835701390982,
+ "loss": 1.0662806034088135,
+ "step": 249
+ },
+ {
+ "epoch": 15.625,
+ "grad_norm": 0.11962032318115234,
+ "learning_rate": 0.0026801633951719288,
+ "loss": 1.0632579326629639,
+ "step": 250
+ },
+ {
+ "epoch": 15.6875,
+ "grad_norm": 0.11223616451025009,
+ "learning_rate": 0.0026760197514650247,
+ "loss": 1.089836835861206,
+ "step": 251
+ },
+ {
+ "epoch": 15.75,
+ "grad_norm": 0.2540510892868042,
+ "learning_rate": 0.002671852722479721,
+ "loss": 1.088876485824585,
+ "step": 252
+ },
+ {
+ "epoch": 15.8125,
+ "grad_norm": 0.08267078548669815,
+ "learning_rate": 0.0026676623921483844,
+ "loss": 1.077115535736084,
+ "step": 253
+ },
+ {
+ "epoch": 15.875,
+ "grad_norm": 0.4830060601234436,
+ "learning_rate": 0.0026634488448727144,
+ "loss": 1.067319393157959,
+ "step": 254
+ },
+ {
+ "epoch": 15.9375,
+ "grad_norm": 0.06782206147909164,
+ "learning_rate": 0.0026592121655220472,
+ "loss": 1.105837106704712,
+ "step": 255
+ },
+ {
+ "epoch": 16.0,
+ "grad_norm": 0.0953136682510376,
+ "learning_rate": 0.0026549524394316464,
+ "loss": 1.1199421882629395,
+ "step": 256
+ },
+ {
+ "epoch": 16.0,
+ "eval_loss": 0.5653528571128845,
+ "eval_runtime": 30.0549,
+ "eval_samples_per_second": 3.394,
+ "eval_steps_per_second": 0.433,
+ "step": 256
+ },
+ {
+ "epoch": 16.0625,
+ "grad_norm": 9.851863861083984,
+ "learning_rate": 0.0026506697524009836,
+ "loss": 1.1010339260101318,
+ "step": 257
+ },
+ {
+ "epoch": 16.125,
+ "grad_norm": 0.10633312165737152,
+ "learning_rate": 0.002646364190692009,
+ "loss": 1.059409737586975,
+ "step": 258
+ },
+ {
+ "epoch": 16.1875,
+ "grad_norm": 0.155309796333313,
+ "learning_rate": 0.0026420358410274153,
+ "loss": 1.0807991027832031,
+ "step": 259
+ },
+ {
+ "epoch": 16.25,
+ "grad_norm": 0.05278189107775688,
+ "learning_rate": 0.0026376847905888903,
+ "loss": 1.0858765840530396,
+ "step": 260
+ },
+ {
+ "epoch": 16.3125,
+ "grad_norm": 0.05539517104625702,
+ "learning_rate": 0.0026333111270153624,
+ "loss": 1.0652291774749756,
+ "step": 261
+ },
+ {
+ "epoch": 16.375,
+ "grad_norm": 0.5698870420455933,
+ "learning_rate": 0.0026289149384012326,
+ "loss": 1.1399967670440674,
+ "step": 262
+ },
+ {
+ "epoch": 16.4375,
+ "grad_norm": 0.0724017322063446,
+ "learning_rate": 0.002624496313294602,
+ "loss": 1.1068029403686523,
+ "step": 263
+ },
+ {
+ "epoch": 16.5,
+ "grad_norm": 0.08893640339374542,
+ "learning_rate": 0.0026200553406954892,
+ "loss": 1.1509923934936523,
+ "step": 264
+ },
+ {
+ "epoch": 16.5625,
+ "grad_norm": 0.29909902811050415,
+ "learning_rate": 0.0026155921100540344,
+ "loss": 1.0652649402618408,
+ "step": 265
+ },
+ {
+ "epoch": 16.625,
+ "grad_norm": 0.09181151539087296,
+ "learning_rate": 0.0026111067112687012,
+ "loss": 1.1081140041351318,
+ "step": 266
+ },
+ {
+ "epoch": 16.6875,
+ "grad_norm": 0.15625476837158203,
+ "learning_rate": 0.0026065992346844626,
+ "loss": 1.1298624277114868,
+ "step": 267
+ },
+ {
+ "epoch": 16.75,
+ "grad_norm": 0.05208205059170723,
+ "learning_rate": 0.002602069771090985,
+ "loss": 1.1198302507400513,
+ "step": 268
+ },
+ {
+ "epoch": 16.8125,
+ "grad_norm": 0.25109273195266724,
+ "learning_rate": 0.002597518411720796,
+ "loss": 1.0928356647491455,
+ "step": 269
+ },
+ {
+ "epoch": 16.875,
+ "grad_norm": 0.026749474927783012,
+ "learning_rate": 0.0025929452482474497,
+ "loss": 1.0576367378234863,
+ "step": 270
+ },
+ {
+ "epoch": 16.9375,
+ "grad_norm": 0.11184096336364746,
+ "learning_rate": 0.0025883503727836764,
+ "loss": 1.1186749935150146,
+ "step": 271
+ },
+ {
+ "epoch": 17.0,
+ "grad_norm": 0.06893426924943924,
+ "learning_rate": 0.0025837338778795325,
+ "loss": 1.1332192420959473,
+ "step": 272
+ },
+ {
+ "epoch": 17.0,
+ "eval_loss": 0.5755453705787659,
+ "eval_runtime": 30.0862,
+ "eval_samples_per_second": 3.39,
+ "eval_steps_per_second": 0.432,
+ "step": 272
+ },
+ {
+ "epoch": 17.0625,
+ "grad_norm": 0.28368332982063293,
+ "learning_rate": 0.0025790958565205316,
+ "loss": 1.1554040908813477,
+ "step": 273
+ },
+ {
+ "epoch": 17.125,
+ "grad_norm": 0.13374564051628113,
+ "learning_rate": 0.0025744364021257743,
+ "loss": 1.119795322418213,
+ "step": 274
+ },
+ {
+ "epoch": 17.1875,
+ "grad_norm": 0.46172675490379333,
+ "learning_rate": 0.0025697556085460646,
+ "loss": 1.1097784042358398,
+ "step": 275
+ },
+ {
+ "epoch": 17.25,
+ "grad_norm": 0.05883564427495003,
+ "learning_rate": 0.002565053570062023,
+ "loss": 1.1001802682876587,
+ "step": 276
+ },
+ {
+ "epoch": 17.3125,
+ "grad_norm": 0.038139838725328445,
+ "learning_rate": 0.002560330381382183,
+ "loss": 1.0831687450408936,
+ "step": 277
+ },
+ {
+ "epoch": 17.375,
+ "grad_norm": 0.029911940917372704,
+ "learning_rate": 0.0025555861376410868,
+ "loss": 1.1179802417755127,
+ "step": 278
+ },
+ {
+ "epoch": 17.4375,
+ "grad_norm": 0.038241755217313766,
+ "learning_rate": 0.002550820934397368,
+ "loss": 1.128745675086975,
+ "step": 279
+ },
+ {
+ "epoch": 17.5,
+ "grad_norm": 0.11490178853273392,
+ "learning_rate": 0.0025460348676318257,
+ "loss": 1.1357793807983398,
+ "step": 280
+ },
+ {
+ "epoch": 17.5625,
+ "grad_norm": 0.09644758701324463,
+ "learning_rate": 0.0025412280337454942,
+ "loss": 1.0953055620193481,
+ "step": 281
+ },
+ {
+ "epoch": 17.625,
+ "grad_norm": 0.0858362540602684,
+ "learning_rate": 0.0025364005295576975,
+ "loss": 1.0640878677368164,
+ "step": 282
+ },
+ {
+ "epoch": 17.6875,
+ "grad_norm": 0.04771006479859352,
+ "learning_rate": 0.002531552452304103,
+ "loss": 1.0404937267303467,
+ "step": 283
+ },
+ {
+ "epoch": 17.75,
+ "grad_norm": 0.06784195452928543,
+ "learning_rate": 0.0025266838996347595,
+ "loss": 1.1567411422729492,
+ "step": 284
+ },
+ {
+ "epoch": 17.8125,
+ "grad_norm": 0.07531201839447021,
+ "learning_rate": 0.002521794969612133,
+ "loss": 1.0844026803970337,
+ "step": 285
+ },
+ {
+ "epoch": 17.875,
+ "grad_norm": 0.06453386694192886,
+ "learning_rate": 0.002516885760709131,
+ "loss": 1.1036568880081177,
+ "step": 286
+ },
+ {
+ "epoch": 17.9375,
+ "grad_norm": 0.022719847038388252,
+ "learning_rate": 0.002511956371807116,
+ "loss": 1.1051126718521118,
+ "step": 287
+ },
+ {
+ "epoch": 18.0,
+ "grad_norm": 0.03833983093500137,
+ "learning_rate": 0.0025070069021939194,
+ "loss": 1.1458027362823486,
+ "step": 288
+ },
+ {
+ "epoch": 18.0,
+ "eval_loss": 0.5735718011856079,
+ "eval_runtime": 30.3576,
+ "eval_samples_per_second": 3.36,
+ "eval_steps_per_second": 0.428,
+ "step": 288
+ },
+ {
+ "epoch": 18.0625,
+ "grad_norm": 0.03980008512735367,
+ "learning_rate": 0.0025020374515618373,
+ "loss": 1.0746936798095703,
+ "step": 289
+ },
+ {
+ "epoch": 18.125,
+ "grad_norm": 0.015201105736196041,
+ "learning_rate": 0.002497048120005623,
+ "loss": 1.0497227907180786,
+ "step": 290
+ },
+ {
+ "epoch": 18.1875,
+ "grad_norm": 0.08461925387382507,
+ "learning_rate": 0.0024920390080204733,
+ "loss": 1.1033759117126465,
+ "step": 291
+ },
+ {
+ "epoch": 18.25,
+ "grad_norm": 0.02152029238641262,
+ "learning_rate": 0.0024870102165000015,
+ "loss": 1.1139872074127197,
+ "step": 292
+ },
+ {
+ "epoch": 18.3125,
+ "grad_norm": 0.030964918434619904,
+ "learning_rate": 0.0024819618467342066,
+ "loss": 1.0686633586883545,
+ "step": 293
+ },
+ {
+ "epoch": 18.375,
+ "grad_norm": 0.03928976133465767,
+ "learning_rate": 0.0024768940004074335,
+ "loss": 1.1165332794189453,
+ "step": 294
+ },
+ {
+ "epoch": 18.4375,
+ "grad_norm": 0.013071025721728802,
+ "learning_rate": 0.002471806779596324,
+ "loss": 1.0793699026107788,
+ "step": 295
+ },
+ {
+ "epoch": 18.5,
+ "grad_norm": 0.017155099660158157,
+ "learning_rate": 0.0024667002867677606,
+ "loss": 1.1448696851730347,
+ "step": 296
+ },
+ {
+ "epoch": 18.5625,
+ "grad_norm": 0.025132035836577415,
+ "learning_rate": 0.0024615746247768043,
+ "loss": 1.1179168224334717,
+ "step": 297
+ },
+ {
+ "epoch": 18.625,
+ "grad_norm": 0.030385689809918404,
+ "learning_rate": 0.0024564298968646203,
+ "loss": 1.0827066898345947,
+ "step": 298
+ },
+ {
+ "epoch": 18.6875,
+ "grad_norm": 0.08415888249874115,
+ "learning_rate": 0.0024512662066563996,
+ "loss": 1.1512041091918945,
+ "step": 299
+ },
+ {
+ "epoch": 18.75,
+ "grad_norm": 0.0354602225124836,
+ "learning_rate": 0.0024460836581592744,
+ "loss": 1.1055684089660645,
+ "step": 300
+ },
+ {
+ "epoch": 18.8125,
+ "grad_norm": 0.061975862830877304,
+ "learning_rate": 0.0024408823557602196,
+ "loss": 1.1121817827224731,
+ "step": 301
+ },
+ {
+ "epoch": 18.875,
+ "grad_norm": 0.12300977855920792,
+ "learning_rate": 0.0024356624042239498,
+ "loss": 1.1116411685943604,
+ "step": 302
+ },
+ {
+ "epoch": 18.9375,
+ "grad_norm": 0.024768438190221786,
+ "learning_rate": 0.002430423908690813,
+ "loss": 1.1031725406646729,
+ "step": 303
+ },
+ {
+ "epoch": 19.0,
+ "grad_norm": 0.20697493851184845,
+ "learning_rate": 0.0024251669746746703,
+ "loss": 1.1268653869628906,
+ "step": 304
+ },
+ {
+ "epoch": 19.0,
+ "eval_loss": 0.5697882175445557,
+ "eval_runtime": 30.3468,
+ "eval_samples_per_second": 3.361,
+ "eval_steps_per_second": 0.428,
+ "step": 304
+ },
+ {
+ "epoch": 19.0625,
+ "grad_norm": 0.04372715577483177,
+ "learning_rate": 0.0024198917080607706,
+ "loss": 1.0879970788955688,
+ "step": 305
+ },
+ {
+ "epoch": 19.125,
+ "grad_norm": 0.3106955885887146,
+ "learning_rate": 0.0024145982151036176,
+ "loss": 1.0981900691986084,
+ "step": 306
+ },
+ {
+ "epoch": 19.1875,
+ "grad_norm": 0.04271978512406349,
+ "learning_rate": 0.0024092866024248323,
+ "loss": 1.067917823791504,
+ "step": 307
+ },
+ {
+ "epoch": 19.25,
+ "grad_norm": 0.03746578097343445,
+ "learning_rate": 0.0024039569770110012,
+ "loss": 1.168632984161377,
+ "step": 308
+ },
+ {
+ "epoch": 19.3125,
+ "grad_norm": 0.02499796636402607,
+ "learning_rate": 0.0023986094462115255,
+ "loss": 1.1304916143417358,
+ "step": 309
+ },
+ {
+ "epoch": 19.375,
+ "grad_norm": 0.031224632635712624,
+ "learning_rate": 0.002393244117736455,
+ "loss": 1.1047842502593994,
+ "step": 310
+ },
+ {
+ "epoch": 19.4375,
+ "grad_norm": 0.03269527107477188,
+ "learning_rate": 0.0023878610996543224,
+ "loss": 1.085580825805664,
+ "step": 311
+ },
+ {
+ "epoch": 19.5,
+ "grad_norm": 0.027562979608774185,
+ "learning_rate": 0.002382460500389964,
+ "loss": 1.1179475784301758,
+ "step": 312
+ },
+ {
+ "epoch": 19.5625,
+ "grad_norm": 0.05654411390423775,
+ "learning_rate": 0.0023770424287223365,
+ "loss": 1.1345707178115845,
+ "step": 313
+ },
+ {
+ "epoch": 19.625,
+ "grad_norm": 0.23459751904010773,
+ "learning_rate": 0.002371606993782326,
+ "loss": 1.0679315328598022,
+ "step": 314
+ },
+ {
+ "epoch": 19.6875,
+ "grad_norm": 0.09804576635360718,
+ "learning_rate": 0.0023661543050505496,
+ "loss": 1.0888817310333252,
+ "step": 315
+ },
+ {
+ "epoch": 19.75,
+ "grad_norm": 0.09066950529813766,
+ "learning_rate": 0.002360684472355151,
+ "loss": 1.094346046447754,
+ "step": 316
+ },
+ {
+ "epoch": 19.8125,
+ "grad_norm": 0.043143078684806824,
+ "learning_rate": 0.0023551976058695887,
+ "loss": 1.081190586090088,
+ "step": 317
+ },
+ {
+ "epoch": 19.875,
+ "grad_norm": 0.03301508352160454,
+ "learning_rate": 0.002349693816110414,
+ "loss": 1.0623350143432617,
+ "step": 318
+ },
+ {
+ "epoch": 19.9375,
+ "grad_norm": 0.022927623242139816,
+ "learning_rate": 0.0023441732139350487,
+ "loss": 1.0692837238311768,
+ "step": 319
+ },
+ {
+ "epoch": 20.0,
+ "grad_norm": 0.03282017260789871,
+ "learning_rate": 0.002338635910539551,
+ "loss": 1.128983736038208,
+ "step": 320
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 800,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 50,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": true,
+ "should_log": false,
+ "should_save": false,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4289416050769920.0,
+ "train_batch_size": 8,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/runs/test-model/checkpoint-320/training_args.bin b/runs/test-model/checkpoint-320/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1c106baba014e8cee356170e7acaf136a1e2efb8
--- /dev/null
+++ b/runs/test-model/checkpoint-320/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
+size 5265
diff --git a/runs/test-model/checkpoint-80/chat_template.jinja b/runs/test-model/checkpoint-80/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..699ff8df401fe4788525e9c1f9b86a99eadd6230
--- /dev/null
+++ b/runs/test-model/checkpoint-80/chat_template.jinja
@@ -0,0 +1,85 @@
+{%- 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 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.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set content = message.content %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in message.content %}
+ {%- set content = message.content.split('')[-1].lstrip('\n') %}
+ {%- set reasoning_content = message.content.split('')[0].rstrip('\n').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' }}
+ {{- message.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/runs/test-model/checkpoint-80/config.json b/runs/test-model/checkpoint-80/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..44f7b99e220689bde520b89c11fdd83d1b5348de
--- /dev/null
+++ b/runs/test-model/checkpoint-80/config.json
@@ -0,0 +1,32 @@
+{
+ "architectures": [
+ "LlamaForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "float32",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 512,
+ "initializer_range": 0.02,
+ "intermediate_size": 1536,
+ "max_position_embeddings": 2048,
+ "mlp_bias": false,
+ "model_type": "llama",
+ "num_attention_heads": 4,
+ "num_hidden_layers": 20,
+ "num_key_value_heads": 4,
+ "pad_token_id": 151645,
+ "pretraining_tp": 1,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 10000.0,
+ "rope_type": "default"
+ },
+ "tie_word_embeddings": true,
+ "transformers_version": "5.5.0",
+ "use_cache": false,
+ "vocab_size": 151671
+}
diff --git a/runs/test-model/checkpoint-80/generation_config.json b/runs/test-model/checkpoint-80/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f962c4bfc66159cdeb7ba836cbbd79365e9304f3
--- /dev/null
+++ b/runs/test-model/checkpoint-80/generation_config.json
@@ -0,0 +1,11 @@
+{
+ "_from_model_config": true,
+ "eos_token_id": [
+ 151645
+ ],
+ "output_attentions": false,
+ "output_hidden_states": false,
+ "pad_token_id": 151645,
+ "transformers_version": "5.5.0",
+ "use_cache": true
+}
diff --git a/runs/test-model/checkpoint-80/model.safetensors b/runs/test-model/checkpoint-80/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ad679b82964ed7242e732e59c330df9035270fad
--- /dev/null
+++ b/runs/test-model/checkpoint-80/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7010337ddd1e2cfefbabe44c6658ae857e41e281477422edf569bd56c444c48e
+size 583356232
diff --git a/runs/test-model/checkpoint-80/optimizer.pt b/runs/test-model/checkpoint-80/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d205cb255ce79f68d99cb50748ddb461cc028344
--- /dev/null
+++ b/runs/test-model/checkpoint-80/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:126c8397d533d83915415be6d3ef122ab4eb3cf3fcf8a1ab4ac2f9e67ad45199
+size 1166825803
diff --git a/runs/test-model/checkpoint-80/rng_state.pth b/runs/test-model/checkpoint-80/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fa41c03a2e286e603e3dfdd6d959bd35b23fff1b
--- /dev/null
+++ b/runs/test-model/checkpoint-80/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecd2f0f0059662cc1490f507d1442b5fd1833fca9ed58923f6b3fa726ed8982a
+size 14709
diff --git a/runs/test-model/checkpoint-80/scheduler.pt b/runs/test-model/checkpoint-80/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d5c607f0dde4aaa6678625fa2236b360b15f4c12
--- /dev/null
+++ b/runs/test-model/checkpoint-80/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3521e8394c8134bca28f9ee35ceeaee5c470453d616f391f758dbecf1b0f1906
+size 1465
diff --git a/runs/test-model/checkpoint-80/tokenizer.json b/runs/test-model/checkpoint-80/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..b83d93986de8ecfc4343943be1c86e9532682c15
--- /dev/null
+++ b/runs/test-model/checkpoint-80/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
+size 11424079
diff --git a/runs/test-model/checkpoint-80/tokenizer_config.json b/runs/test-model/checkpoint-80/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1d6bb391bb8fbc446de576db5220ec91c98a150
--- /dev/null
+++ b/runs/test-model/checkpoint-80/tokenizer_config.json
@@ -0,0 +1,18 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|l2r_pred|>",
+ "<|r2l_pred|>"
+ ],
+ "is_local": false,
+ "model_max_length": 131072,
+ "pad_token": "<|im_end|>",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/runs/test-model/checkpoint-80/trainer_state.json b/runs/test-model/checkpoint-80/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..0168ec3f05e6212db819f981b115a7e757bda11e
--- /dev/null
+++ b/runs/test-model/checkpoint-80/trainer_state.json
@@ -0,0 +1,626 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 5.0,
+ "eval_steps": 500,
+ "global_step": 80,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0625,
+ "grad_norm": 0.4655202031135559,
+ "learning_rate": 0.0,
+ "loss": 2.00244402885437,
+ "step": 1
+ },
+ {
+ "epoch": 0.125,
+ "grad_norm": 0.4847956895828247,
+ "learning_rate": 3e-05,
+ "loss": 2.008800506591797,
+ "step": 2
+ },
+ {
+ "epoch": 0.1875,
+ "grad_norm": 0.4177771806716919,
+ "learning_rate": 6e-05,
+ "loss": 1.9685651063919067,
+ "step": 3
+ },
+ {
+ "epoch": 0.25,
+ "grad_norm": 0.3550562858581543,
+ "learning_rate": 8.999999999999999e-05,
+ "loss": 1.9168976545333862,
+ "step": 4
+ },
+ {
+ "epoch": 0.3125,
+ "grad_norm": 0.23293137550354004,
+ "learning_rate": 0.00012,
+ "loss": 1.8660516738891602,
+ "step": 5
+ },
+ {
+ "epoch": 0.375,
+ "grad_norm": 0.13890701532363892,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.8209998607635498,
+ "step": 6
+ },
+ {
+ "epoch": 0.4375,
+ "grad_norm": 0.08890970796346664,
+ "learning_rate": 0.00017999999999999998,
+ "loss": 1.806854248046875,
+ "step": 7
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.061611492186784744,
+ "learning_rate": 0.00021000000000000004,
+ "loss": 1.793172836303711,
+ "step": 8
+ },
+ {
+ "epoch": 0.5625,
+ "grad_norm": 0.05786389857530594,
+ "learning_rate": 0.00024,
+ "loss": 1.7882877588272095,
+ "step": 9
+ },
+ {
+ "epoch": 0.625,
+ "grad_norm": 0.03947344794869423,
+ "learning_rate": 0.00027,
+ "loss": 1.7696826457977295,
+ "step": 10
+ },
+ {
+ "epoch": 0.6875,
+ "grad_norm": 0.029783710837364197,
+ "learning_rate": 0.00030000000000000003,
+ "loss": 1.7767844200134277,
+ "step": 11
+ },
+ {
+ "epoch": 0.75,
+ "grad_norm": 0.025729672983288765,
+ "learning_rate": 0.00033,
+ "loss": 1.7827998399734497,
+ "step": 12
+ },
+ {
+ "epoch": 0.8125,
+ "grad_norm": 0.020592188462615013,
+ "learning_rate": 0.00035999999999999997,
+ "loss": 1.766276478767395,
+ "step": 13
+ },
+ {
+ "epoch": 0.875,
+ "grad_norm": 0.01877395436167717,
+ "learning_rate": 0.00039000000000000005,
+ "loss": 1.7644124031066895,
+ "step": 14
+ },
+ {
+ "epoch": 0.9375,
+ "grad_norm": 0.015196867287158966,
+ "learning_rate": 0.00042000000000000007,
+ "loss": 1.771270990371704,
+ "step": 15
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.014262701384723186,
+ "learning_rate": 0.00045,
+ "loss": 1.755500316619873,
+ "step": 16
+ },
+ {
+ "epoch": 1.0,
+ "eval_loss": 0.8736820816993713,
+ "eval_runtime": 30.2743,
+ "eval_samples_per_second": 3.369,
+ "eval_steps_per_second": 0.429,
+ "step": 16
+ },
+ {
+ "epoch": 1.0625,
+ "grad_norm": 0.01042763702571392,
+ "learning_rate": 0.00048,
+ "loss": 1.7556875944137573,
+ "step": 17
+ },
+ {
+ "epoch": 1.125,
+ "grad_norm": 0.009123577736318111,
+ "learning_rate": 0.00051,
+ "loss": 1.753890037536621,
+ "step": 18
+ },
+ {
+ "epoch": 1.1875,
+ "grad_norm": 0.009028254076838493,
+ "learning_rate": 0.00054,
+ "loss": 1.7515414953231812,
+ "step": 19
+ },
+ {
+ "epoch": 1.25,
+ "grad_norm": 0.009630529209971428,
+ "learning_rate": 0.00057,
+ "loss": 1.7465155124664307,
+ "step": 20
+ },
+ {
+ "epoch": 1.3125,
+ "grad_norm": 0.008235524408519268,
+ "learning_rate": 0.0006000000000000001,
+ "loss": 1.7533702850341797,
+ "step": 21
+ },
+ {
+ "epoch": 1.375,
+ "grad_norm": 0.009377758949995041,
+ "learning_rate": 0.00063,
+ "loss": 1.738031029701233,
+ "step": 22
+ },
+ {
+ "epoch": 1.4375,
+ "grad_norm": 0.008636261336505413,
+ "learning_rate": 0.00066,
+ "loss": 1.735835313796997,
+ "step": 23
+ },
+ {
+ "epoch": 1.5,
+ "grad_norm": 0.008001700975000858,
+ "learning_rate": 0.0006900000000000001,
+ "loss": 1.7505712509155273,
+ "step": 24
+ },
+ {
+ "epoch": 1.5625,
+ "grad_norm": 0.007517208810895681,
+ "learning_rate": 0.0007199999999999999,
+ "loss": 1.7330820560455322,
+ "step": 25
+ },
+ {
+ "epoch": 1.625,
+ "grad_norm": 0.00952192209661007,
+ "learning_rate": 0.00075,
+ "loss": 1.7318685054779053,
+ "step": 26
+ },
+ {
+ "epoch": 1.6875,
+ "grad_norm": 0.006763546261936426,
+ "learning_rate": 0.0007800000000000001,
+ "loss": 1.7329007387161255,
+ "step": 27
+ },
+ {
+ "epoch": 1.75,
+ "grad_norm": 0.006364013999700546,
+ "learning_rate": 0.0008100000000000001,
+ "loss": 1.7285535335540771,
+ "step": 28
+ },
+ {
+ "epoch": 1.8125,
+ "grad_norm": 0.007770394906401634,
+ "learning_rate": 0.0008400000000000001,
+ "loss": 1.7320656776428223,
+ "step": 29
+ },
+ {
+ "epoch": 1.875,
+ "grad_norm": 0.006639651022851467,
+ "learning_rate": 0.00087,
+ "loss": 1.7304966449737549,
+ "step": 30
+ },
+ {
+ "epoch": 1.9375,
+ "grad_norm": 0.008190952241420746,
+ "learning_rate": 0.0009,
+ "loss": 1.7399260997772217,
+ "step": 31
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 0.008188036270439625,
+ "learning_rate": 0.00093,
+ "loss": 1.7375996112823486,
+ "step": 32
+ },
+ {
+ "epoch": 2.0,
+ "eval_loss": 0.860374927520752,
+ "eval_runtime": 30.3049,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 32
+ },
+ {
+ "epoch": 2.0625,
+ "grad_norm": 0.005768604576587677,
+ "learning_rate": 0.00096,
+ "loss": 1.7309706211090088,
+ "step": 33
+ },
+ {
+ "epoch": 2.125,
+ "grad_norm": 0.006822533905506134,
+ "learning_rate": 0.00099,
+ "loss": 1.7321696281433105,
+ "step": 34
+ },
+ {
+ "epoch": 2.1875,
+ "grad_norm": 0.009974546730518341,
+ "learning_rate": 0.00102,
+ "loss": 1.7290048599243164,
+ "step": 35
+ },
+ {
+ "epoch": 2.25,
+ "grad_norm": 0.00856814906001091,
+ "learning_rate": 0.00105,
+ "loss": 1.7231603860855103,
+ "step": 36
+ },
+ {
+ "epoch": 2.3125,
+ "grad_norm": 0.011694240383803844,
+ "learning_rate": 0.00108,
+ "loss": 1.728348731994629,
+ "step": 37
+ },
+ {
+ "epoch": 2.375,
+ "grad_norm": 0.005738088861107826,
+ "learning_rate": 0.00111,
+ "loss": 1.7308874130249023,
+ "step": 38
+ },
+ {
+ "epoch": 2.4375,
+ "grad_norm": 0.010038000531494617,
+ "learning_rate": 0.00114,
+ "loss": 1.7276837825775146,
+ "step": 39
+ },
+ {
+ "epoch": 2.5,
+ "grad_norm": 0.01586989127099514,
+ "learning_rate": 0.00117,
+ "loss": 1.7368175983428955,
+ "step": 40
+ },
+ {
+ "epoch": 2.5625,
+ "grad_norm": 0.16198667883872986,
+ "learning_rate": 0.0012000000000000001,
+ "loss": 1.730246663093567,
+ "step": 41
+ },
+ {
+ "epoch": 2.625,
+ "grad_norm": 0.01653525047004223,
+ "learning_rate": 0.00123,
+ "loss": 1.7403852939605713,
+ "step": 42
+ },
+ {
+ "epoch": 2.6875,
+ "grad_norm": 0.01241418905556202,
+ "learning_rate": 0.00126,
+ "loss": 1.7306722402572632,
+ "step": 43
+ },
+ {
+ "epoch": 2.75,
+ "grad_norm": 0.022560030221939087,
+ "learning_rate": 0.00129,
+ "loss": 1.7223260402679443,
+ "step": 44
+ },
+ {
+ "epoch": 2.8125,
+ "grad_norm": 0.019568368792533875,
+ "learning_rate": 0.00132,
+ "loss": 1.7304983139038086,
+ "step": 45
+ },
+ {
+ "epoch": 2.875,
+ "grad_norm": 0.016755228862166405,
+ "learning_rate": 0.00135,
+ "loss": 1.7223992347717285,
+ "step": 46
+ },
+ {
+ "epoch": 2.9375,
+ "grad_norm": 0.02981281839311123,
+ "learning_rate": 0.0013800000000000002,
+ "loss": 1.7094981670379639,
+ "step": 47
+ },
+ {
+ "epoch": 3.0,
+ "grad_norm": 0.02506650611758232,
+ "learning_rate": 0.00141,
+ "loss": 1.7054555416107178,
+ "step": 48
+ },
+ {
+ "epoch": 3.0,
+ "eval_loss": 0.8416222333908081,
+ "eval_runtime": 30.3027,
+ "eval_samples_per_second": 3.366,
+ "eval_steps_per_second": 0.429,
+ "step": 48
+ },
+ {
+ "epoch": 3.0625,
+ "grad_norm": 0.02260257676243782,
+ "learning_rate": 0.0014399999999999999,
+ "loss": 1.6888771057128906,
+ "step": 49
+ },
+ {
+ "epoch": 3.125,
+ "grad_norm": 0.02620658650994301,
+ "learning_rate": 0.00147,
+ "loss": 1.6639578342437744,
+ "step": 50
+ },
+ {
+ "epoch": 3.1875,
+ "grad_norm": 0.047354575246572495,
+ "learning_rate": 0.0015,
+ "loss": 1.6736350059509277,
+ "step": 51
+ },
+ {
+ "epoch": 3.25,
+ "grad_norm": 0.04594860598444939,
+ "learning_rate": 0.0015300000000000001,
+ "loss": 1.6675360202789307,
+ "step": 52
+ },
+ {
+ "epoch": 3.3125,
+ "grad_norm": 0.08131621032953262,
+ "learning_rate": 0.0015600000000000002,
+ "loss": 1.6574077606201172,
+ "step": 53
+ },
+ {
+ "epoch": 3.375,
+ "grad_norm": 0.028789900243282318,
+ "learning_rate": 0.00159,
+ "loss": 1.6299148797988892,
+ "step": 54
+ },
+ {
+ "epoch": 3.4375,
+ "grad_norm": 0.2623404860496521,
+ "learning_rate": 0.0016200000000000001,
+ "loss": 1.6315665245056152,
+ "step": 55
+ },
+ {
+ "epoch": 3.5,
+ "grad_norm": 0.056828465312719345,
+ "learning_rate": 0.0016500000000000002,
+ "loss": 1.6279325485229492,
+ "step": 56
+ },
+ {
+ "epoch": 3.5625,
+ "grad_norm": 0.0549125075340271,
+ "learning_rate": 0.0016800000000000003,
+ "loss": 1.6549303531646729,
+ "step": 57
+ },
+ {
+ "epoch": 3.625,
+ "grad_norm": 0.07287958264350891,
+ "learning_rate": 0.00171,
+ "loss": 1.606963872909546,
+ "step": 58
+ },
+ {
+ "epoch": 3.6875,
+ "grad_norm": 0.07426343113183975,
+ "learning_rate": 0.00174,
+ "loss": 1.6250534057617188,
+ "step": 59
+ },
+ {
+ "epoch": 3.75,
+ "grad_norm": 0.1364986002445221,
+ "learning_rate": 0.0017699999999999999,
+ "loss": 1.5964961051940918,
+ "step": 60
+ },
+ {
+ "epoch": 3.8125,
+ "grad_norm": 0.15007159113883972,
+ "learning_rate": 0.0018,
+ "loss": 1.6019010543823242,
+ "step": 61
+ },
+ {
+ "epoch": 3.875,
+ "grad_norm": 0.08447311073541641,
+ "learning_rate": 0.00183,
+ "loss": 1.5763543844223022,
+ "step": 62
+ },
+ {
+ "epoch": 3.9375,
+ "grad_norm": 0.1681874394416809,
+ "learning_rate": 0.00186,
+ "loss": 1.5769858360290527,
+ "step": 63
+ },
+ {
+ "epoch": 4.0,
+ "grad_norm": 0.09677319973707199,
+ "learning_rate": 0.00189,
+ "loss": 1.534682273864746,
+ "step": 64
+ },
+ {
+ "epoch": 4.0,
+ "eval_loss": 0.7682511806488037,
+ "eval_runtime": 30.4228,
+ "eval_samples_per_second": 3.353,
+ "eval_steps_per_second": 0.427,
+ "step": 64
+ },
+ {
+ "epoch": 4.0625,
+ "grad_norm": 0.42635980248451233,
+ "learning_rate": 0.00192,
+ "loss": 1.5509562492370605,
+ "step": 65
+ },
+ {
+ "epoch": 4.125,
+ "grad_norm": 0.8467249274253845,
+ "learning_rate": 0.0019500000000000001,
+ "loss": 1.5001531839370728,
+ "step": 66
+ },
+ {
+ "epoch": 4.1875,
+ "grad_norm": 0.22631719708442688,
+ "learning_rate": 0.00198,
+ "loss": 1.4554829597473145,
+ "step": 67
+ },
+ {
+ "epoch": 4.25,
+ "grad_norm": 0.05312464386224747,
+ "learning_rate": 0.00201,
+ "loss": 1.4732775688171387,
+ "step": 68
+ },
+ {
+ "epoch": 4.3125,
+ "grad_norm": 0.050442785024642944,
+ "learning_rate": 0.00204,
+ "loss": 1.4176194667816162,
+ "step": 69
+ },
+ {
+ "epoch": 4.375,
+ "grad_norm": 0.18886858224868774,
+ "learning_rate": 0.00207,
+ "loss": 1.4325997829437256,
+ "step": 70
+ },
+ {
+ "epoch": 4.4375,
+ "grad_norm": 0.05258706212043762,
+ "learning_rate": 0.0021,
+ "loss": 1.39802086353302,
+ "step": 71
+ },
+ {
+ "epoch": 4.5,
+ "grad_norm": 0.07482657581567764,
+ "learning_rate": 0.00213,
+ "loss": 1.4000320434570312,
+ "step": 72
+ },
+ {
+ "epoch": 4.5625,
+ "grad_norm": 0.4853312075138092,
+ "learning_rate": 0.00216,
+ "loss": 1.4199151992797852,
+ "step": 73
+ },
+ {
+ "epoch": 4.625,
+ "grad_norm": 0.48888301849365234,
+ "learning_rate": 0.00219,
+ "loss": 1.3049702644348145,
+ "step": 74
+ },
+ {
+ "epoch": 4.6875,
+ "grad_norm": 0.284348726272583,
+ "learning_rate": 0.00222,
+ "loss": 1.3884068727493286,
+ "step": 75
+ },
+ {
+ "epoch": 4.75,
+ "grad_norm": 0.20248942077159882,
+ "learning_rate": 0.0022500000000000003,
+ "loss": 1.3502120971679688,
+ "step": 76
+ },
+ {
+ "epoch": 4.8125,
+ "grad_norm": 0.2714093327522278,
+ "learning_rate": 0.00228,
+ "loss": 1.3195693492889404,
+ "step": 77
+ },
+ {
+ "epoch": 4.875,
+ "grad_norm": 0.2544061243534088,
+ "learning_rate": 0.00231,
+ "loss": 1.3194304704666138,
+ "step": 78
+ },
+ {
+ "epoch": 4.9375,
+ "grad_norm": 0.18475523591041565,
+ "learning_rate": 0.00234,
+ "loss": 1.3006361722946167,
+ "step": 79
+ },
+ {
+ "epoch": 5.0,
+ "grad_norm": 0.6583278775215149,
+ "learning_rate": 0.00237,
+ "loss": 1.2868517637252808,
+ "step": 80
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 800,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 50,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": true,
+ "should_log": false,
+ "should_save": false,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 1072354012692480.0,
+ "train_batch_size": 8,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/runs/test-model/checkpoint-80/training_args.bin b/runs/test-model/checkpoint-80/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1c106baba014e8cee356170e7acaf136a1e2efb8
--- /dev/null
+++ b/runs/test-model/checkpoint-80/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
+size 5265