Upload model

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,61 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen2.5-7B-Instruct
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: d1_science_mc_llm_10k
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# d1_science_mc_llm_10k
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/d1_science_mc_llm_10k dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 4e-05
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 16
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 128
+- total_eval_batch_size: 128
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 5.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.46.1
+- Pytorch 2.6.0a0+ecf3bae40a.nv25.01
+- Datasets 3.5.0
+- Tokenizers 0.20.3

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.992,
+    "total_flos": 7.862516792374067e+17,
+    "train_loss": 0.3952271275795423,
+    "train_runtime": 13928.0051,
+    "train_samples_per_second": 3.59,
+    "train_steps_per_second": 0.028
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "/p/data1/mmlaion/dcft/hub/models--Qwen--Qwen2.5-7B-Instruct/snapshots/a09a35458c702b33eeacc393d103063234e8bc28",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.1",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.46.1"
+}

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e67b2eb80abaf971d510863522f923cd7a5f81f7fa870814fe469d725cd3eba9
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac2b932f72239cb5a1510ccfd0ae854290103e6f42e0fe6c03a30c5727fff1df
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9462fe2587af48659764c57f9ddbcdab5e651d90ccef63a54403d2df7083e042
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e8ab3a4592a582eb81302ffbdfba3df06c1fb957d2bec94dab27e84215a0c7c
+size 1089994880

model.safetensors.index.json

@@ -0,0 +1,346 @@
+{
+  "metadata": {
+    "total_size": 15231233024
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00004-of-00004.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.norm.weight": "model-00003-of-00004.safetensors"
+  }
+}

runs/Apr27_23-26-55_jpbot-001-26.jupiter.internal/events.out.tfevents.1745789242.jpbot-001-26.jupiter.internal.2394476.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26648c00bd3e7223d0b1f53e75cb34465681c34e7d45bc921c8db8ecb66583b8
+size 87670

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

start_end.json

@@ -0,0 +1 @@
+{"start_time": "2025-04-27 23:26:55", "end_time": "2025-04-28 03:19:37"}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.992,
+    "total_flos": 7.862516792374067e+17,
+    "train_loss": 0.3952271275795423,
+    "train_runtime": 13928.0051,
+    "train_samples_per_second": 3.59,
+    "train_steps_per_second": 0.028
+}

trainer_log.jsonl

@@ -0,0 +1,391 @@
+{"current_steps": 1, "total_steps": 390, "loss": 0.9618, "lr": 1.0256410256410257e-06, "epoch": 0.0128, "percentage": 0.26, "elapsed_time": "0:00:43", "remaining_time": "4:44:42"}
+{"current_steps": 2, "total_steps": 390, "loss": 0.9669, "lr": 2.0512820512820513e-06, "epoch": 0.0256, "percentage": 0.51, "elapsed_time": "0:01:15", "remaining_time": "4:04:30"}
+{"current_steps": 3, "total_steps": 390, "loss": 1.021, "lr": 3.0769230769230774e-06, "epoch": 0.0384, "percentage": 0.77, "elapsed_time": "0:01:55", "remaining_time": "4:08:19"}
+{"current_steps": 4, "total_steps": 390, "loss": 0.9571, "lr": 4.102564102564103e-06, "epoch": 0.0512, "percentage": 1.03, "elapsed_time": "0:02:28", "remaining_time": "3:58:55"}
+{"current_steps": 5, "total_steps": 390, "loss": 0.8816, "lr": 5.128205128205128e-06, "epoch": 0.064, "percentage": 1.28, "elapsed_time": "0:03:03", "remaining_time": "3:55:07"}
+{"current_steps": 6, "total_steps": 390, "loss": 0.8964, "lr": 6.153846153846155e-06, "epoch": 0.0768, "percentage": 1.54, "elapsed_time": "0:03:39", "remaining_time": "3:53:42"}
+{"current_steps": 7, "total_steps": 390, "loss": 0.8921, "lr": 7.17948717948718e-06, "epoch": 0.0896, "percentage": 1.79, "elapsed_time": "0:04:09", "remaining_time": "3:47:55"}
+{"current_steps": 8, "total_steps": 390, "loss": 0.8709, "lr": 8.205128205128205e-06, "epoch": 0.1024, "percentage": 2.05, "elapsed_time": "0:04:51", "remaining_time": "3:52:05"}
+{"current_steps": 9, "total_steps": 390, "loss": 0.9305, "lr": 9.230769230769232e-06, "epoch": 0.1152, "percentage": 2.31, "elapsed_time": "0:05:24", "remaining_time": "3:49:00"}
+{"current_steps": 10, "total_steps": 390, "loss": 0.813, "lr": 1.0256410256410256e-05, "epoch": 0.128, "percentage": 2.56, "elapsed_time": "0:06:03", "remaining_time": "3:50:19"}
+{"current_steps": 11, "total_steps": 390, "loss": 0.8366, "lr": 1.1282051282051283e-05, "epoch": 0.1408, "percentage": 2.82, "elapsed_time": "0:06:39", "remaining_time": "3:49:27"}
+{"current_steps": 12, "total_steps": 390, "loss": 0.8205, "lr": 1.230769230769231e-05, "epoch": 0.1536, "percentage": 3.08, "elapsed_time": "0:07:19", "remaining_time": "3:50:31"}
+{"current_steps": 13, "total_steps": 390, "loss": 0.7825, "lr": 1.3333333333333333e-05, "epoch": 0.1664, "percentage": 3.33, "elapsed_time": "0:07:57", "remaining_time": "3:50:41"}
+{"current_steps": 14, "total_steps": 390, "loss": 0.7745, "lr": 1.435897435897436e-05, "epoch": 0.1792, "percentage": 3.59, "elapsed_time": "0:08:33", "remaining_time": "3:49:57"}
+{"current_steps": 15, "total_steps": 390, "loss": 0.7378, "lr": 1.5384615384615387e-05, "epoch": 0.192, "percentage": 3.85, "elapsed_time": "0:09:10", "remaining_time": "3:49:22"}
+{"current_steps": 16, "total_steps": 390, "loss": 0.7281, "lr": 1.641025641025641e-05, "epoch": 0.2048, "percentage": 4.1, "elapsed_time": "0:09:44", "remaining_time": "3:47:46"}
+{"current_steps": 17, "total_steps": 390, "loss": 0.7435, "lr": 1.7435897435897438e-05, "epoch": 0.2176, "percentage": 4.36, "elapsed_time": "0:10:16", "remaining_time": "3:45:26"}
+{"current_steps": 18, "total_steps": 390, "loss": 0.743, "lr": 1.8461538461538465e-05, "epoch": 0.2304, "percentage": 4.62, "elapsed_time": "0:10:52", "remaining_time": "3:44:50"}
+{"current_steps": 19, "total_steps": 390, "loss": 0.7486, "lr": 1.9487179487179488e-05, "epoch": 0.2432, "percentage": 4.87, "elapsed_time": "0:11:23", "remaining_time": "3:42:18"}
+{"current_steps": 20, "total_steps": 390, "loss": 0.7467, "lr": 2.0512820512820512e-05, "epoch": 0.256, "percentage": 5.13, "elapsed_time": "0:11:59", "remaining_time": "3:41:45"}
+{"current_steps": 21, "total_steps": 390, "loss": 0.6933, "lr": 2.153846153846154e-05, "epoch": 0.2688, "percentage": 5.38, "elapsed_time": "0:12:30", "remaining_time": "3:39:43"}
+{"current_steps": 22, "total_steps": 390, "loss": 0.7067, "lr": 2.2564102564102566e-05, "epoch": 0.2816, "percentage": 5.64, "elapsed_time": "0:13:02", "remaining_time": "3:38:11"}
+{"current_steps": 23, "total_steps": 390, "loss": 0.7046, "lr": 2.3589743589743593e-05, "epoch": 0.2944, "percentage": 5.9, "elapsed_time": "0:13:36", "remaining_time": "3:37:02"}
+{"current_steps": 24, "total_steps": 390, "loss": 0.7125, "lr": 2.461538461538462e-05, "epoch": 0.3072, "percentage": 6.15, "elapsed_time": "0:14:15", "remaining_time": "3:37:32"}
+{"current_steps": 25, "total_steps": 390, "loss": 0.6728, "lr": 2.5641025641025646e-05, "epoch": 0.32, "percentage": 6.41, "elapsed_time": "0:14:54", "remaining_time": "3:37:34"}
+{"current_steps": 26, "total_steps": 390, "loss": 0.7381, "lr": 2.6666666666666667e-05, "epoch": 0.3328, "percentage": 6.67, "elapsed_time": "0:15:27", "remaining_time": "3:36:27"}
+{"current_steps": 27, "total_steps": 390, "loss": 0.6553, "lr": 2.7692307692307694e-05, "epoch": 0.3456, "percentage": 6.92, "elapsed_time": "0:16:08", "remaining_time": "3:36:57"}
+{"current_steps": 28, "total_steps": 390, "loss": 0.6798, "lr": 2.871794871794872e-05, "epoch": 0.3584, "percentage": 7.18, "elapsed_time": "0:16:42", "remaining_time": "3:36:01"}
+{"current_steps": 29, "total_steps": 390, "loss": 0.6958, "lr": 2.9743589743589747e-05, "epoch": 0.3712, "percentage": 7.44, "elapsed_time": "0:17:19", "remaining_time": "3:35:35"}
+{"current_steps": 30, "total_steps": 390, "loss": 0.6497, "lr": 3.0769230769230774e-05, "epoch": 0.384, "percentage": 7.69, "elapsed_time": "0:17:57", "remaining_time": "3:35:34"}
+{"current_steps": 31, "total_steps": 390, "loss": 0.6844, "lr": 3.1794871794871795e-05, "epoch": 0.3968, "percentage": 7.95, "elapsed_time": "0:18:27", "remaining_time": "3:33:40"}
+{"current_steps": 32, "total_steps": 390, "loss": 0.6672, "lr": 3.282051282051282e-05, "epoch": 0.4096, "percentage": 8.21, "elapsed_time": "0:19:03", "remaining_time": "3:33:09"}
+{"current_steps": 33, "total_steps": 390, "loss": 0.6744, "lr": 3.384615384615385e-05, "epoch": 0.4224, "percentage": 8.46, "elapsed_time": "0:19:45", "remaining_time": "3:33:50"}
+{"current_steps": 34, "total_steps": 390, "loss": 0.6666, "lr": 3.4871794871794875e-05, "epoch": 0.4352, "percentage": 8.72, "elapsed_time": "0:20:17", "remaining_time": "3:32:28"}
+{"current_steps": 35, "total_steps": 390, "loss": 0.6642, "lr": 3.58974358974359e-05, "epoch": 0.448, "percentage": 8.97, "elapsed_time": "0:20:55", "remaining_time": "3:32:10"}
+{"current_steps": 36, "total_steps": 390, "loss": 0.6638, "lr": 3.692307692307693e-05, "epoch": 0.4608, "percentage": 9.23, "elapsed_time": "0:21:27", "remaining_time": "3:30:55"}
+{"current_steps": 37, "total_steps": 390, "loss": 0.6732, "lr": 3.794871794871795e-05, "epoch": 0.4736, "percentage": 9.49, "elapsed_time": "0:22:02", "remaining_time": "3:30:14"}
+{"current_steps": 38, "total_steps": 390, "loss": 0.6624, "lr": 3.8974358974358976e-05, "epoch": 0.4864, "percentage": 9.74, "elapsed_time": "0:22:39", "remaining_time": "3:29:54"}
+{"current_steps": 39, "total_steps": 390, "loss": 0.6962, "lr": 4e-05, "epoch": 0.4992, "percentage": 10.0, "elapsed_time": "0:23:08", "remaining_time": "3:28:17"}
+{"current_steps": 40, "total_steps": 390, "loss": 0.6245, "lr": 3.9999198907597046e-05, "epoch": 0.512, "percentage": 10.26, "elapsed_time": "0:23:48", "remaining_time": "3:28:17"}
+{"current_steps": 41, "total_steps": 390, "loss": 0.6225, "lr": 3.9996795694563096e-05, "epoch": 0.5248, "percentage": 10.51, "elapsed_time": "0:24:21", "remaining_time": "3:27:23"}
+{"current_steps": 42, "total_steps": 390, "loss": 0.6604, "lr": 3.999279055341771e-05, "epoch": 0.5376, "percentage": 10.77, "elapsed_time": "0:24:52", "remaining_time": "3:26:05"}
+{"current_steps": 43, "total_steps": 390, "loss": 0.6133, "lr": 3.998718380500971e-05, "epoch": 0.5504, "percentage": 11.03, "elapsed_time": "0:25:33", "remaining_time": "3:26:17"}
+{"current_steps": 44, "total_steps": 390, "loss": 0.6309, "lr": 3.997997589849145e-05, "epoch": 0.5632, "percentage": 11.28, "elapsed_time": "0:26:06", "remaining_time": "3:25:18"}
+{"current_steps": 45, "total_steps": 390, "loss": 0.6659, "lr": 3.9971167411282835e-05, "epoch": 0.576, "percentage": 11.54, "elapsed_time": "0:26:38", "remaining_time": "3:24:15"}
+{"current_steps": 46, "total_steps": 390, "loss": 0.6569, "lr": 3.99607590490251e-05, "epoch": 0.5888, "percentage": 11.79, "elapsed_time": "0:27:06", "remaining_time": "3:22:43"}
+{"current_steps": 47, "total_steps": 390, "loss": 0.6313, "lr": 3.9948751645524235e-05, "epoch": 0.6016, "percentage": 12.05, "elapsed_time": "0:27:43", "remaining_time": "3:22:20"}
+{"current_steps": 48, "total_steps": 390, "loss": 0.6326, "lr": 3.9935146162684206e-05, "epoch": 0.6144, "percentage": 12.31, "elapsed_time": "0:28:19", "remaining_time": "3:21:51"}
+{"current_steps": 49, "total_steps": 390, "loss": 0.6201, "lr": 3.9919943690429906e-05, "epoch": 0.6272, "percentage": 12.56, "elapsed_time": "0:28:58", "remaining_time": "3:21:41"}
+{"current_steps": 50, "total_steps": 390, "loss": 0.6329, "lr": 3.9903145446619837e-05, "epoch": 0.64, "percentage": 12.82, "elapsed_time": "0:29:37", "remaining_time": "3:21:24"}
+{"current_steps": 51, "total_steps": 390, "loss": 0.6422, "lr": 3.9884752776948564e-05, "epoch": 0.6528, "percentage": 13.08, "elapsed_time": "0:30:09", "remaining_time": "3:20:27"}
+{"current_steps": 52, "total_steps": 390, "loss": 0.6354, "lr": 3.9864767154838864e-05, "epoch": 0.6656, "percentage": 13.33, "elapsed_time": "0:30:47", "remaining_time": "3:20:09"}
+{"current_steps": 53, "total_steps": 390, "loss": 0.6667, "lr": 3.9843190181323744e-05, "epoch": 0.6784, "percentage": 13.59, "elapsed_time": "0:31:23", "remaining_time": "3:19:34"}
+{"current_steps": 54, "total_steps": 390, "loss": 0.6364, "lr": 3.982002358491817e-05, "epoch": 0.6912, "percentage": 13.85, "elapsed_time": "0:31:57", "remaining_time": "3:18:51"}
+{"current_steps": 55, "total_steps": 390, "loss": 0.6218, "lr": 3.979526922148058e-05, "epoch": 0.704, "percentage": 14.1, "elapsed_time": "0:32:36", "remaining_time": "3:18:33"}
+{"current_steps": 56, "total_steps": 390, "loss": 0.6401, "lr": 3.9768929074064206e-05, "epoch": 0.7168, "percentage": 14.36, "elapsed_time": "0:33:11", "remaining_time": "3:17:57"}
+{"current_steps": 57, "total_steps": 390, "loss": 0.6168, "lr": 3.9741005252758255e-05, "epoch": 0.7296, "percentage": 14.62, "elapsed_time": "0:33:47", "remaining_time": "3:17:23"}
+{"current_steps": 58, "total_steps": 390, "loss": 0.6423, "lr": 3.971149999451886e-05, "epoch": 0.7424, "percentage": 14.87, "elapsed_time": "0:34:23", "remaining_time": "3:16:51"}
+{"current_steps": 59, "total_steps": 390, "loss": 0.6644, "lr": 3.9680415662989806e-05, "epoch": 0.7552, "percentage": 15.13, "elapsed_time": "0:34:53", "remaining_time": "3:15:47"}
+{"current_steps": 60, "total_steps": 390, "loss": 0.6425, "lr": 3.9647754748313294e-05, "epoch": 0.768, "percentage": 15.38, "elapsed_time": "0:35:27", "remaining_time": "3:14:59"}
+{"current_steps": 61, "total_steps": 390, "loss": 0.6114, "lr": 3.96135198669304e-05, "epoch": 0.7808, "percentage": 15.64, "elapsed_time": "0:36:00", "remaining_time": "3:14:10"}
+{"current_steps": 62, "total_steps": 390, "loss": 0.6523, "lr": 3.957771376137144e-05, "epoch": 0.7936, "percentage": 15.9, "elapsed_time": "0:36:37", "remaining_time": "3:13:45"}
+{"current_steps": 63, "total_steps": 390, "loss": 0.6168, "lr": 3.954033930003634e-05, "epoch": 0.8064, "percentage": 16.15, "elapsed_time": "0:37:14", "remaining_time": "3:13:19"}
+{"current_steps": 64, "total_steps": 390, "loss": 0.6177, "lr": 3.9501399476964806e-05, "epoch": 0.8192, "percentage": 16.41, "elapsed_time": "0:37:59", "remaining_time": "3:13:32"}
+{"current_steps": 65, "total_steps": 390, "loss": 0.6338, "lr": 3.946089741159648e-05, "epoch": 0.832, "percentage": 16.67, "elapsed_time": "0:38:31", "remaining_time": "3:12:37"}
+{"current_steps": 66, "total_steps": 390, "loss": 0.6004, "lr": 3.9418836348521045e-05, "epoch": 0.8448, "percentage": 16.92, "elapsed_time": "0:39:12", "remaining_time": "3:12:27"}
+{"current_steps": 67, "total_steps": 390, "loss": 0.6324, "lr": 3.937521965721831e-05, "epoch": 0.8576, "percentage": 17.18, "elapsed_time": "0:39:45", "remaining_time": "3:11:41"}
+{"current_steps": 68, "total_steps": 390, "loss": 0.6246, "lr": 3.933005083178828e-05, "epoch": 0.8704, "percentage": 17.44, "elapsed_time": "0:40:19", "remaining_time": "3:10:57"}
+{"current_steps": 69, "total_steps": 390, "loss": 0.6219, "lr": 3.928333349067125e-05, "epoch": 0.8832, "percentage": 17.69, "elapsed_time": "0:41:04", "remaining_time": "3:11:03"}
+{"current_steps": 70, "total_steps": 390, "loss": 0.623, "lr": 3.923507137635792e-05, "epoch": 0.896, "percentage": 17.95, "elapsed_time": "0:41:47", "remaining_time": "3:11:02"}
+{"current_steps": 71, "total_steps": 390, "loss": 0.6269, "lr": 3.9185268355089606e-05, "epoch": 0.9088, "percentage": 18.21, "elapsed_time": "0:42:29", "remaining_time": "3:10:54"}
+{"current_steps": 72, "total_steps": 390, "loss": 0.6351, "lr": 3.913392841654851e-05, "epoch": 0.9216, "percentage": 18.46, "elapsed_time": "0:43:01", "remaining_time": "3:10:00"}
+{"current_steps": 73, "total_steps": 390, "loss": 0.6239, "lr": 3.9081055673538093e-05, "epoch": 0.9344, "percentage": 18.72, "elapsed_time": "0:43:34", "remaining_time": "3:09:13"}
+{"current_steps": 74, "total_steps": 390, "loss": 0.6573, "lr": 3.902665436165364e-05, "epoch": 0.9472, "percentage": 18.97, "elapsed_time": "0:44:06", "remaining_time": "3:08:20"}
+{"current_steps": 75, "total_steps": 390, "loss": 0.5857, "lr": 3.897072883894291e-05, "epoch": 0.96, "percentage": 19.23, "elapsed_time": "0:44:46", "remaining_time": "3:08:03"}
+{"current_steps": 76, "total_steps": 390, "loss": 0.6344, "lr": 3.8913283585557054e-05, "epoch": 0.9728, "percentage": 19.49, "elapsed_time": "0:45:19", "remaining_time": "3:07:14"}
+{"current_steps": 77, "total_steps": 390, "loss": 0.5906, "lr": 3.885432320339167e-05, "epoch": 0.9856, "percentage": 19.74, "elapsed_time": "0:45:53", "remaining_time": "3:06:31"}
+{"current_steps": 78, "total_steps": 390, "loss": 0.6364, "lr": 3.879385241571817e-05, "epoch": 0.9984, "percentage": 20.0, "elapsed_time": "0:46:25", "remaining_time": "3:05:43"}
+{"current_steps": 79, "total_steps": 390, "loss": 1.068, "lr": 3.873187606680543e-05, "epoch": 1.0112, "percentage": 20.26, "elapsed_time": "0:47:14", "remaining_time": "3:05:57"}
+{"current_steps": 80, "total_steps": 390, "loss": 0.5233, "lr": 3.866839912153168e-05, "epoch": 1.024, "percentage": 20.51, "elapsed_time": "0:47:44", "remaining_time": "3:05:00"}
+{"current_steps": 81, "total_steps": 390, "loss": 0.5753, "lr": 3.860342666498677e-05, "epoch": 1.0368, "percentage": 20.77, "elapsed_time": "0:48:19", "remaining_time": "3:04:19"}
+{"current_steps": 82, "total_steps": 390, "loss": 0.4947, "lr": 3.853696390206484e-05, "epoch": 1.0496, "percentage": 21.03, "elapsed_time": "0:48:52", "remaining_time": "3:03:36"}
+{"current_steps": 83, "total_steps": 390, "loss": 0.5649, "lr": 3.846901615704734e-05, "epoch": 1.0624, "percentage": 21.28, "elapsed_time": "0:49:28", "remaining_time": "3:02:58"}
+{"current_steps": 84, "total_steps": 390, "loss": 0.4996, "lr": 3.839958887317649e-05, "epoch": 1.0752, "percentage": 21.54, "elapsed_time": "0:50:00", "remaining_time": "3:02:11"}
+{"current_steps": 85, "total_steps": 390, "loss": 0.5333, "lr": 3.832868761221926e-05, "epoch": 1.088, "percentage": 21.79, "elapsed_time": "0:50:40", "remaining_time": "3:01:49"}
+{"current_steps": 86, "total_steps": 390, "loss": 0.4864, "lr": 3.825631805402182e-05, "epoch": 1.1008, "percentage": 22.05, "elapsed_time": "0:51:11", "remaining_time": "3:00:58"}
+{"current_steps": 87, "total_steps": 390, "loss": 0.5237, "lr": 3.818248599605448e-05, "epoch": 1.1136, "percentage": 22.31, "elapsed_time": "0:51:44", "remaining_time": "3:00:13"}
+{"current_steps": 88, "total_steps": 390, "loss": 0.5167, "lr": 3.810719735294731e-05, "epoch": 1.1264, "percentage": 22.56, "elapsed_time": "0:52:17", "remaining_time": "2:59:28"}
+{"current_steps": 89, "total_steps": 390, "loss": 0.4995, "lr": 3.8030458156016326e-05, "epoch": 1.1392, "percentage": 22.82, "elapsed_time": "0:52:58", "remaining_time": "2:59:08"}
+{"current_steps": 90, "total_steps": 390, "loss": 0.5106, "lr": 3.795227455278029e-05, "epoch": 1.152, "percentage": 23.08, "elapsed_time": "0:53:33", "remaining_time": "2:58:30"}
+{"current_steps": 91, "total_steps": 390, "loss": 0.5217, "lr": 3.787265280646825e-05, "epoch": 1.1648, "percentage": 23.33, "elapsed_time": "0:54:10", "remaining_time": "2:58:01"}
+{"current_steps": 92, "total_steps": 390, "loss": 0.5033, "lr": 3.7791599295517825e-05, "epoch": 1.1776, "percentage": 23.59, "elapsed_time": "0:54:47", "remaining_time": "2:57:30"}
+{"current_steps": 93, "total_steps": 390, "loss": 0.5028, "lr": 3.7709120513064196e-05, "epoch": 1.1904, "percentage": 23.85, "elapsed_time": "0:55:26", "remaining_time": "2:57:02"}
+{"current_steps": 94, "total_steps": 390, "loss": 0.4779, "lr": 3.762522306641998e-05, "epoch": 1.2032, "percentage": 24.1, "elapsed_time": "0:55:57", "remaining_time": "2:56:12"}
+{"current_steps": 95, "total_steps": 390, "loss": 0.5451, "lr": 3.7539913676545874e-05, "epoch": 1.216, "percentage": 24.36, "elapsed_time": "0:56:34", "remaining_time": "2:55:41"}
+{"current_steps": 96, "total_steps": 390, "loss": 0.5274, "lr": 3.745319917751229e-05, "epoch": 1.2288000000000001, "percentage": 24.62, "elapsed_time": "0:57:07", "remaining_time": "2:54:55"}
+{"current_steps": 97, "total_steps": 390, "loss": 0.539, "lr": 3.736508651595188e-05, "epoch": 1.2416, "percentage": 24.87, "elapsed_time": "0:57:43", "remaining_time": "2:54:21"}
+{"current_steps": 98, "total_steps": 390, "loss": 0.4945, "lr": 3.727558275050301e-05, "epoch": 1.2544, "percentage": 25.13, "elapsed_time": "0:58:23", "remaining_time": "2:53:58"}
+{"current_steps": 99, "total_steps": 390, "loss": 0.5272, "lr": 3.718469505124434e-05, "epoch": 1.2671999999999999, "percentage": 25.38, "elapsed_time": "0:59:02", "remaining_time": "2:53:33"}
+{"current_steps": 100, "total_steps": 390, "loss": 0.5066, "lr": 3.709243069912041e-05, "epoch": 1.28, "percentage": 25.64, "elapsed_time": "0:59:38", "remaining_time": "2:52:57"}
+{"current_steps": 101, "total_steps": 390, "loss": 0.5123, "lr": 3.699879708535838e-05, "epoch": 1.2928, "percentage": 25.9, "elapsed_time": "1:00:16", "remaining_time": "2:52:27"}
+{"current_steps": 102, "total_steps": 390, "loss": 0.5259, "lr": 3.69038017108759e-05, "epoch": 1.3056, "percentage": 26.15, "elapsed_time": "1:00:51", "remaining_time": "2:51:51"}
+{"current_steps": 103, "total_steps": 390, "loss": 0.473, "lr": 3.680745218568026e-05, "epoch": 1.3184, "percentage": 26.41, "elapsed_time": "1:01:23", "remaining_time": "2:51:04"}
+{"current_steps": 104, "total_steps": 390, "loss": 0.5262, "lr": 3.6709756228258735e-05, "epoch": 1.3312, "percentage": 26.67, "elapsed_time": "1:01:58", "remaining_time": "2:50:25"}
+{"current_steps": 105, "total_steps": 390, "loss": 0.4905, "lr": 3.6610721664960236e-05, "epoch": 1.3439999999999999, "percentage": 26.92, "elapsed_time": "1:02:32", "remaining_time": "2:49:46"}
+{"current_steps": 106, "total_steps": 390, "loss": 0.5479, "lr": 3.65103564293684e-05, "epoch": 1.3568, "percentage": 27.18, "elapsed_time": "1:03:10", "remaining_time": "2:49:14"}
+{"current_steps": 107, "total_steps": 390, "loss": 0.4948, "lr": 3.640866856166601e-05, "epoch": 1.3696, "percentage": 27.44, "elapsed_time": "1:03:47", "remaining_time": "2:48:44"}
+{"current_steps": 108, "total_steps": 390, "loss": 0.4785, "lr": 3.6305666207990886e-05, "epoch": 1.3824, "percentage": 27.69, "elapsed_time": "1:04:19", "remaining_time": "2:47:58"}
+{"current_steps": 109, "total_steps": 390, "loss": 0.4754, "lr": 3.6201357619783336e-05, "epoch": 1.3952, "percentage": 27.95, "elapsed_time": "1:04:52", "remaining_time": "2:47:15"}
+{"current_steps": 110, "total_steps": 390, "loss": 0.4869, "lr": 3.609575115312511e-05, "epoch": 1.408, "percentage": 28.21, "elapsed_time": "1:05:29", "remaining_time": "2:46:41"}
+{"current_steps": 111, "total_steps": 390, "loss": 0.4972, "lr": 3.598885526807003e-05, "epoch": 1.4208, "percentage": 28.46, "elapsed_time": "1:06:04", "remaining_time": "2:46:03"}
+{"current_steps": 112, "total_steps": 390, "loss": 0.4735, "lr": 3.5880678527966224e-05, "epoch": 1.4336, "percentage": 28.72, "elapsed_time": "1:06:41", "remaining_time": "2:45:32"}
+{"current_steps": 113, "total_steps": 390, "loss": 0.5055, "lr": 3.577122959877017e-05, "epoch": 1.4464000000000001, "percentage": 28.97, "elapsed_time": "1:07:18", "remaining_time": "2:45:00"}
+{"current_steps": 114, "total_steps": 390, "loss": 0.4923, "lr": 3.566051724835245e-05, "epoch": 1.4592, "percentage": 29.23, "elapsed_time": "1:07:55", "remaining_time": "2:44:26"}
+{"current_steps": 115, "total_steps": 390, "loss": 0.512, "lr": 3.554855034579532e-05, "epoch": 1.472, "percentage": 29.49, "elapsed_time": "1:08:31", "remaining_time": "2:43:51"}
+{"current_steps": 116, "total_steps": 390, "loss": 0.4981, "lr": 3.5435337860682304e-05, "epoch": 1.4848, "percentage": 29.74, "elapsed_time": "1:09:07", "remaining_time": "2:43:15"}
+{"current_steps": 117, "total_steps": 390, "loss": 0.4915, "lr": 3.532088886237956e-05, "epoch": 1.4976, "percentage": 30.0, "elapsed_time": "1:09:46", "remaining_time": "2:42:49"}
+{"current_steps": 118, "total_steps": 390, "loss": 0.4736, "lr": 3.520521251930941e-05, "epoch": 1.5104, "percentage": 30.26, "elapsed_time": "1:10:20", "remaining_time": "2:42:09"}
+{"current_steps": 119, "total_steps": 390, "loss": 0.5403, "lr": 3.5088318098215805e-05, "epoch": 1.5232, "percentage": 30.51, "elapsed_time": "1:10:54", "remaining_time": "2:41:29"}
+{"current_steps": 120, "total_steps": 390, "loss": 0.4694, "lr": 3.497021496342203e-05, "epoch": 1.536, "percentage": 30.77, "elapsed_time": "1:11:29", "remaining_time": "2:40:52"}
+{"current_steps": 121, "total_steps": 390, "loss": 0.4979, "lr": 3.485091257608047e-05, "epoch": 1.5488, "percentage": 31.03, "elapsed_time": "1:12:09", "remaining_time": "2:40:24"}
+{"current_steps": 122, "total_steps": 390, "loss": 0.5191, "lr": 3.473042049341474e-05, "epoch": 1.5615999999999999, "percentage": 31.28, "elapsed_time": "1:12:40", "remaining_time": "2:39:39"}
+{"current_steps": 123, "total_steps": 390, "loss": 0.4805, "lr": 3.4608748367954064e-05, "epoch": 1.5744, "percentage": 31.54, "elapsed_time": "1:13:11", "remaining_time": "2:38:53"}
+{"current_steps": 124, "total_steps": 390, "loss": 0.5199, "lr": 3.4485905946759965e-05, "epoch": 1.5872000000000002, "percentage": 31.79, "elapsed_time": "1:13:39", "remaining_time": "2:38:01"}
+{"current_steps": 125, "total_steps": 390, "loss": 0.5183, "lr": 3.4361903070645484e-05, "epoch": 1.6, "percentage": 32.05, "elapsed_time": "1:14:16", "remaining_time": "2:37:27"}
+{"current_steps": 126, "total_steps": 390, "loss": 0.5178, "lr": 3.423674967338681e-05, "epoch": 1.6128, "percentage": 32.31, "elapsed_time": "1:14:51", "remaining_time": "2:36:51"}
+{"current_steps": 127, "total_steps": 390, "loss": 0.4837, "lr": 3.411045578092754e-05, "epoch": 1.6256, "percentage": 32.56, "elapsed_time": "1:15:24", "remaining_time": "2:36:09"}
+{"current_steps": 128, "total_steps": 390, "loss": 0.5115, "lr": 3.398303151057543e-05, "epoch": 1.6383999999999999, "percentage": 32.82, "elapsed_time": "1:16:00", "remaining_time": "2:35:34"}
+{"current_steps": 129, "total_steps": 390, "loss": 0.5102, "lr": 3.385448707019199e-05, "epoch": 1.6512, "percentage": 33.08, "elapsed_time": "1:16:40", "remaining_time": "2:35:07"}
+{"current_steps": 130, "total_steps": 390, "loss": 0.5035, "lr": 3.372483275737468e-05, "epoch": 1.6640000000000001, "percentage": 33.33, "elapsed_time": "1:17:21", "remaining_time": "2:34:43"}
+{"current_steps": 131, "total_steps": 390, "loss": 0.4708, "lr": 3.359407895863199e-05, "epoch": 1.6768, "percentage": 33.59, "elapsed_time": "1:17:59", "remaining_time": "2:34:12"}
+{"current_steps": 132, "total_steps": 390, "loss": 0.5117, "lr": 3.34622361485514e-05, "epoch": 1.6896, "percentage": 33.85, "elapsed_time": "1:18:38", "remaining_time": "2:33:43"}
+{"current_steps": 133, "total_steps": 390, "loss": 0.5006, "lr": 3.332931488896029e-05, "epoch": 1.7024, "percentage": 34.1, "elapsed_time": "1:19:12", "remaining_time": "2:33:03"}
+{"current_steps": 134, "total_steps": 390, "loss": 0.4956, "lr": 3.319532582807977e-05, "epoch": 1.7151999999999998, "percentage": 34.36, "elapsed_time": "1:19:49", "remaining_time": "2:32:29"}
+{"current_steps": 135, "total_steps": 390, "loss": 0.5138, "lr": 3.30602796996717e-05, "epoch": 1.728, "percentage": 34.62, "elapsed_time": "1:20:22", "remaining_time": "2:31:49"}
+{"current_steps": 136, "total_steps": 390, "loss": 0.452, "lr": 3.2924187322178865e-05, "epoch": 1.7408000000000001, "percentage": 34.87, "elapsed_time": "1:21:00", "remaining_time": "2:31:16"}
+{"current_steps": 137, "total_steps": 390, "loss": 0.5132, "lr": 3.278705959785821e-05, "epoch": 1.7536, "percentage": 35.13, "elapsed_time": "1:21:35", "remaining_time": "2:30:40"}
+{"current_steps": 138, "total_steps": 390, "loss": 0.51, "lr": 3.2648907511907544e-05, "epoch": 1.7664, "percentage": 35.38, "elapsed_time": "1:22:15", "remaining_time": "2:30:11"}
+{"current_steps": 139, "total_steps": 390, "loss": 0.5055, "lr": 3.250974213158555e-05, "epoch": 1.7792, "percentage": 35.64, "elapsed_time": "1:22:48", "remaining_time": "2:29:31"}
+{"current_steps": 140, "total_steps": 390, "loss": 0.5128, "lr": 3.23695746053251e-05, "epoch": 1.792, "percentage": 35.9, "elapsed_time": "1:23:24", "remaining_time": "2:28:57"}
+{"current_steps": 141, "total_steps": 390, "loss": 0.4807, "lr": 3.222841616184025e-05, "epoch": 1.8048, "percentage": 36.15, "elapsed_time": "1:23:56", "remaining_time": "2:28:14"}
+{"current_steps": 142, "total_steps": 390, "loss": 0.4887, "lr": 3.208627810922665e-05, "epoch": 1.8176, "percentage": 36.41, "elapsed_time": "1:24:27", "remaining_time": "2:27:29"}
+{"current_steps": 143, "total_steps": 390, "loss": 0.5051, "lr": 3.194317183405573e-05, "epoch": 1.8304, "percentage": 36.67, "elapsed_time": "1:25:04", "remaining_time": "2:26:56"}
+{"current_steps": 144, "total_steps": 390, "loss": 0.5138, "lr": 3.1799108800462466e-05, "epoch": 1.8432, "percentage": 36.92, "elapsed_time": "1:25:38", "remaining_time": "2:26:18"}
+{"current_steps": 145, "total_steps": 390, "loss": 0.472, "lr": 3.1654100549227024e-05, "epoch": 1.8559999999999999, "percentage": 37.18, "elapsed_time": "1:26:19", "remaining_time": "2:25:51"}
+{"current_steps": 146, "total_steps": 390, "loss": 0.4954, "lr": 3.1508158696850275e-05, "epoch": 1.8688, "percentage": 37.44, "elapsed_time": "1:26:47", "remaining_time": "2:25:02"}
+{"current_steps": 147, "total_steps": 390, "loss": 0.4987, "lr": 3.136129493462312e-05, "epoch": 1.8816000000000002, "percentage": 37.69, "elapsed_time": "1:27:27", "remaining_time": "2:24:34"}
+{"current_steps": 148, "total_steps": 390, "loss": 0.4848, "lr": 3.121352102768998e-05, "epoch": 1.8944, "percentage": 37.95, "elapsed_time": "1:28:09", "remaining_time": "2:24:09"}
+{"current_steps": 149, "total_steps": 390, "loss": 0.4665, "lr": 3.106484881410628e-05, "epoch": 1.9072, "percentage": 38.21, "elapsed_time": "1:28:46", "remaining_time": "2:23:34"}
+{"current_steps": 150, "total_steps": 390, "loss": 0.4994, "lr": 3.091529020389009e-05, "epoch": 1.92, "percentage": 38.46, "elapsed_time": "1:29:24", "remaining_time": "2:23:03"}
+{"current_steps": 151, "total_steps": 390, "loss": 0.4927, "lr": 3.076485717806808e-05, "epoch": 1.9327999999999999, "percentage": 38.72, "elapsed_time": "1:29:54", "remaining_time": "2:22:17"}
+{"current_steps": 152, "total_steps": 390, "loss": 0.5066, "lr": 3.061356178771564e-05, "epoch": 1.9456, "percentage": 38.97, "elapsed_time": "1:30:31", "remaining_time": "2:21:45"}
+{"current_steps": 153, "total_steps": 390, "loss": 0.4687, "lr": 3.0461416152991555e-05, "epoch": 1.9584000000000001, "percentage": 39.23, "elapsed_time": "1:31:02", "remaining_time": "2:21:01"}
+{"current_steps": 154, "total_steps": 390, "loss": 0.5163, "lr": 3.0308432462167045e-05, "epoch": 1.9712, "percentage": 39.49, "elapsed_time": "1:31:39", "remaining_time": "2:20:28"}
+{"current_steps": 155, "total_steps": 390, "loss": 0.471, "lr": 3.015462297064936e-05, "epoch": 1.984, "percentage": 39.74, "elapsed_time": "1:32:18", "remaining_time": "2:19:57"}
+{"current_steps": 156, "total_steps": 390, "loss": 0.5087, "lr": 3.0000000000000004e-05, "epoch": 1.9968, "percentage": 40.0, "elapsed_time": "1:32:52", "remaining_time": "2:19:18"}
+{"current_steps": 157, "total_steps": 390, "loss": 0.8075, "lr": 2.98445759369477e-05, "epoch": 2.0096, "percentage": 40.26, "elapsed_time": "1:33:40", "remaining_time": "2:19:01"}
+{"current_steps": 158, "total_steps": 390, "loss": 0.4051, "lr": 2.9688363232396056e-05, "epoch": 2.0224, "percentage": 40.51, "elapsed_time": "1:34:10", "remaining_time": "2:18:17"}
+{"current_steps": 159, "total_steps": 390, "loss": 0.3823, "lr": 2.9531374400426158e-05, "epoch": 2.0352, "percentage": 40.77, "elapsed_time": "1:34:44", "remaining_time": "2:17:38"}
+{"current_steps": 160, "total_steps": 390, "loss": 0.3444, "lr": 2.9373622017294075e-05, "epoch": 2.048, "percentage": 41.03, "elapsed_time": "1:35:18", "remaining_time": "2:17:00"}
+{"current_steps": 161, "total_steps": 390, "loss": 0.3576, "lr": 2.9215118720423375e-05, "epoch": 2.0608, "percentage": 41.28, "elapsed_time": "1:35:58", "remaining_time": "2:16:30"}
+{"current_steps": 162, "total_steps": 390, "loss": 0.3585, "lr": 2.9055877207392752e-05, "epoch": 2.0736, "percentage": 41.54, "elapsed_time": "1:36:33", "remaining_time": "2:15:54"}
+{"current_steps": 163, "total_steps": 390, "loss": 0.3538, "lr": 2.8895910234918828e-05, "epoch": 2.0864, "percentage": 41.79, "elapsed_time": "1:37:11", "remaining_time": "2:15:20"}
+{"current_steps": 164, "total_steps": 390, "loss": 0.3908, "lr": 2.873523061783426e-05, "epoch": 2.0992, "percentage": 42.05, "elapsed_time": "1:37:43", "remaining_time": "2:14:40"}
+{"current_steps": 165, "total_steps": 390, "loss": 0.4022, "lr": 2.8573851228061084e-05, "epoch": 2.112, "percentage": 42.31, "elapsed_time": "1:38:20", "remaining_time": "2:14:05"}
+{"current_steps": 166, "total_steps": 390, "loss": 0.3443, "lr": 2.8411784993579633e-05, "epoch": 2.1248, "percentage": 42.56, "elapsed_time": "1:38:56", "remaining_time": "2:13:30"}
+{"current_steps": 167, "total_steps": 390, "loss": 0.3723, "lr": 2.8249044897392814e-05, "epoch": 2.1376, "percentage": 42.82, "elapsed_time": "1:39:34", "remaining_time": "2:12:57"}
+{"current_steps": 168, "total_steps": 390, "loss": 0.3498, "lr": 2.80856439764861e-05, "epoch": 2.1504, "percentage": 43.08, "elapsed_time": "1:40:14", "remaining_time": "2:12:27"}
+{"current_steps": 169, "total_steps": 390, "loss": 0.3497, "lr": 2.792159532078314e-05, "epoch": 2.1632, "percentage": 43.33, "elapsed_time": "1:40:47", "remaining_time": "2:11:48"}
+{"current_steps": 170, "total_steps": 390, "loss": 0.3677, "lr": 2.77569120720971e-05, "epoch": 2.176, "percentage": 43.59, "elapsed_time": "1:41:23", "remaining_time": "2:11:12"}
+{"current_steps": 171, "total_steps": 390, "loss": 0.3433, "lr": 2.7591607423077932e-05, "epoch": 2.1888, "percentage": 43.85, "elapsed_time": "1:41:48", "remaining_time": "2:10:23"}
+{"current_steps": 172, "total_steps": 390, "loss": 0.3971, "lr": 2.7425694616155474e-05, "epoch": 2.2016, "percentage": 44.1, "elapsed_time": "1:42:25", "remaining_time": "2:09:49"}
+{"current_steps": 173, "total_steps": 390, "loss": 0.3756, "lr": 2.7259186942478656e-05, "epoch": 2.2144, "percentage": 44.36, "elapsed_time": "1:43:01", "remaining_time": "2:09:13"}
+{"current_steps": 174, "total_steps": 390, "loss": 0.3403, "lr": 2.7092097740850712e-05, "epoch": 2.2272, "percentage": 44.62, "elapsed_time": "1:43:32", "remaining_time": "2:08:32"}
+{"current_steps": 175, "total_steps": 390, "loss": 0.3595, "lr": 2.692444039666066e-05, "epoch": 2.24, "percentage": 44.87, "elapsed_time": "1:44:05", "remaining_time": "2:07:53"}
+{"current_steps": 176, "total_steps": 390, "loss": 0.3314, "lr": 2.6756228340810946e-05, "epoch": 2.2528, "percentage": 45.13, "elapsed_time": "1:44:40", "remaining_time": "2:07:16"}
+{"current_steps": 177, "total_steps": 390, "loss": 0.3606, "lr": 2.6587475048641596e-05, "epoch": 2.2656, "percentage": 45.38, "elapsed_time": "1:45:12", "remaining_time": "2:06:36"}
+{"current_steps": 178, "total_steps": 390, "loss": 0.3799, "lr": 2.6418194038850634e-05, "epoch": 2.2784, "percentage": 45.64, "elapsed_time": "1:45:46", "remaining_time": "2:05:58"}
+{"current_steps": 179, "total_steps": 390, "loss": 0.3464, "lr": 2.624839887241115e-05, "epoch": 2.2912, "percentage": 45.9, "elapsed_time": "1:46:21", "remaining_time": "2:05:22"}
+{"current_steps": 180, "total_steps": 390, "loss": 0.366, "lr": 2.607810315148494e-05, "epoch": 2.304, "percentage": 46.15, "elapsed_time": "1:46:52", "remaining_time": "2:04:41"}
+{"current_steps": 181, "total_steps": 390, "loss": 0.3433, "lr": 2.5907320518332827e-05, "epoch": 2.3168, "percentage": 46.41, "elapsed_time": "1:47:27", "remaining_time": "2:04:05"}
+{"current_steps": 182, "total_steps": 390, "loss": 0.36, "lr": 2.5736064654221808e-05, "epoch": 2.3296, "percentage": 46.67, "elapsed_time": "1:48:00", "remaining_time": "2:03:26"}
+{"current_steps": 183, "total_steps": 390, "loss": 0.3337, "lr": 2.5564349278329056e-05, "epoch": 2.3424, "percentage": 46.92, "elapsed_time": "1:48:35", "remaining_time": "2:02:49"}
+{"current_steps": 184, "total_steps": 390, "loss": 0.3614, "lr": 2.539218814664288e-05, "epoch": 2.3552, "percentage": 47.18, "elapsed_time": "1:49:07", "remaining_time": "2:02:10"}
+{"current_steps": 185, "total_steps": 390, "loss": 0.364, "lr": 2.521959505086075e-05, "epoch": 2.368, "percentage": 47.44, "elapsed_time": "1:49:39", "remaining_time": "2:01:31"}
+{"current_steps": 186, "total_steps": 390, "loss": 0.3369, "lr": 2.5046583817284437e-05, "epoch": 2.3808, "percentage": 47.69, "elapsed_time": "1:50:11", "remaining_time": "2:00:51"}
+{"current_steps": 187, "total_steps": 390, "loss": 0.3397, "lr": 2.487316830571244e-05, "epoch": 2.3936, "percentage": 47.95, "elapsed_time": "1:50:47", "remaining_time": "2:00:15"}
+{"current_steps": 188, "total_steps": 390, "loss": 0.3609, "lr": 2.4699362408329646e-05, "epoch": 2.4064, "percentage": 48.21, "elapsed_time": "1:51:18", "remaining_time": "1:59:35"}
+{"current_steps": 189, "total_steps": 390, "loss": 0.3403, "lr": 2.4525180048594452e-05, "epoch": 2.4192, "percentage": 48.46, "elapsed_time": "1:51:51", "remaining_time": "1:58:58"}
+{"current_steps": 190, "total_steps": 390, "loss": 0.3559, "lr": 2.435063518012335e-05, "epoch": 2.432, "percentage": 48.72, "elapsed_time": "1:52:31", "remaining_time": "1:58:26"}
+{"current_steps": 191, "total_steps": 390, "loss": 0.3699, "lr": 2.4175741785573177e-05, "epoch": 2.4448, "percentage": 48.97, "elapsed_time": "1:53:05", "remaining_time": "1:57:49"}
+{"current_steps": 192, "total_steps": 390, "loss": 0.3389, "lr": 2.4000513875520892e-05, "epoch": 2.4576000000000002, "percentage": 49.23, "elapsed_time": "1:53:37", "remaining_time": "1:57:10"}
+{"current_steps": 193, "total_steps": 390, "loss": 0.3748, "lr": 2.3824965487341247e-05, "epoch": 2.4704, "percentage": 49.49, "elapsed_time": "1:54:14", "remaining_time": "1:56:36"}
+{"current_steps": 194, "total_steps": 390, "loss": 0.3649, "lr": 2.3649110684082258e-05, "epoch": 2.4832, "percentage": 49.74, "elapsed_time": "1:54:50", "remaining_time": "1:56:01"}
+{"current_steps": 195, "total_steps": 390, "loss": 0.3911, "lr": 2.3472963553338614e-05, "epoch": 2.496, "percentage": 50.0, "elapsed_time": "1:55:27", "remaining_time": "1:55:27"}
+{"current_steps": 196, "total_steps": 390, "loss": 0.3594, "lr": 2.3296538206123134e-05, "epoch": 2.5088, "percentage": 50.26, "elapsed_time": "1:56:05", "remaining_time": "1:54:54"}
+{"current_steps": 197, "total_steps": 390, "loss": 0.3528, "lr": 2.311984877573636e-05, "epoch": 2.5216, "percentage": 50.51, "elapsed_time": "1:56:39", "remaining_time": "1:54:17"}
+{"current_steps": 198, "total_steps": 390, "loss": 0.328, "lr": 2.2942909416634326e-05, "epoch": 2.5343999999999998, "percentage": 50.77, "elapsed_time": "1:57:14", "remaining_time": "1:53:41"}
+{"current_steps": 199, "total_steps": 390, "loss": 0.393, "lr": 2.2765734303294666e-05, "epoch": 2.5472, "percentage": 51.03, "elapsed_time": "1:57:54", "remaining_time": "1:53:09"}
+{"current_steps": 200, "total_steps": 390, "loss": 0.3705, "lr": 2.2588337629081107e-05, "epoch": 2.56, "percentage": 51.28, "elapsed_time": "1:58:37", "remaining_time": "1:52:41"}
+{"current_steps": 201, "total_steps": 390, "loss": 0.3333, "lr": 2.2410733605106462e-05, "epoch": 2.5728, "percentage": 51.54, "elapsed_time": "1:59:13", "remaining_time": "1:52:06"}
+{"current_steps": 202, "total_steps": 390, "loss": 0.3494, "lr": 2.2232936459094158e-05, "epoch": 2.5856, "percentage": 51.79, "elapsed_time": "1:59:49", "remaining_time": "1:51:31"}
+{"current_steps": 203, "total_steps": 390, "loss": 0.3405, "lr": 2.205496043423849e-05, "epoch": 2.5984, "percentage": 52.05, "elapsed_time": "2:00:26", "remaining_time": "1:50:56"}
+{"current_steps": 204, "total_steps": 390, "loss": 0.3888, "lr": 2.1876819788063586e-05, "epoch": 2.6112, "percentage": 52.31, "elapsed_time": "2:00:57", "remaining_time": "1:50:17"}
+{"current_steps": 205, "total_steps": 390, "loss": 0.3255, "lr": 2.16985287912813e-05, "epoch": 2.624, "percentage": 52.56, "elapsed_time": "2:01:29", "remaining_time": "1:49:37"}
+{"current_steps": 206, "total_steps": 390, "loss": 0.3865, "lr": 2.1520101726647922e-05, "epoch": 2.6368, "percentage": 52.82, "elapsed_time": "2:02:03", "remaining_time": "1:49:01"}
+{"current_steps": 207, "total_steps": 390, "loss": 0.3513, "lr": 2.1341552887820048e-05, "epoch": 2.6496, "percentage": 53.08, "elapsed_time": "2:02:37", "remaining_time": "1:48:24"}
+{"current_steps": 208, "total_steps": 390, "loss": 0.3609, "lr": 2.1162896578209517e-05, "epoch": 2.6624, "percentage": 53.33, "elapsed_time": "2:03:13", "remaining_time": "1:47:49"}
+{"current_steps": 209, "total_steps": 390, "loss": 0.3219, "lr": 2.0984147109837564e-05, "epoch": 2.6752000000000002, "percentage": 53.59, "elapsed_time": "2:03:53", "remaining_time": "1:47:18"}
+{"current_steps": 210, "total_steps": 390, "loss": 0.3744, "lr": 2.0805318802188307e-05, "epoch": 2.6879999999999997, "percentage": 53.85, "elapsed_time": "2:04:26", "remaining_time": "1:46:39"}
+{"current_steps": 211, "total_steps": 390, "loss": 0.3658, "lr": 2.0626425981061608e-05, "epoch": 2.7008, "percentage": 54.1, "elapsed_time": "2:05:05", "remaining_time": "1:46:06"}
+{"current_steps": 212, "total_steps": 390, "loss": 0.3405, "lr": 2.0447482977425465e-05, "epoch": 2.7136, "percentage": 54.36, "elapsed_time": "2:05:42", "remaining_time": "1:45:33"}
+{"current_steps": 213, "total_steps": 390, "loss": 0.368, "lr": 2.0268504126267952e-05, "epoch": 2.7264, "percentage": 54.62, "elapsed_time": "2:06:12", "remaining_time": "1:44:52"}
+{"current_steps": 214, "total_steps": 390, "loss": 0.3465, "lr": 2.008950376544887e-05, "epoch": 2.7392, "percentage": 54.87, "elapsed_time": "2:06:52", "remaining_time": "1:44:21"}
+{"current_steps": 215, "total_steps": 390, "loss": 0.3532, "lr": 1.9910496234551132e-05, "epoch": 2.752, "percentage": 55.13, "elapsed_time": "2:07:23", "remaining_time": "1:43:41"}
+{"current_steps": 216, "total_steps": 390, "loss": 0.353, "lr": 1.9731495873732055e-05, "epoch": 2.7648, "percentage": 55.38, "elapsed_time": "2:07:58", "remaining_time": "1:43:05"}
+{"current_steps": 217, "total_steps": 390, "loss": 0.3569, "lr": 1.9552517022574542e-05, "epoch": 2.7776, "percentage": 55.64, "elapsed_time": "2:08:36", "remaining_time": "1:42:31"}
+{"current_steps": 218, "total_steps": 390, "loss": 0.3568, "lr": 1.93735740189384e-05, "epoch": 2.7904, "percentage": 55.9, "elapsed_time": "2:09:15", "remaining_time": "1:41:59"}
+{"current_steps": 219, "total_steps": 390, "loss": 0.365, "lr": 1.9194681197811703e-05, "epoch": 2.8032, "percentage": 56.15, "elapsed_time": "2:09:54", "remaining_time": "1:41:26"}
+{"current_steps": 220, "total_steps": 390, "loss": 0.3427, "lr": 1.901585289016244e-05, "epoch": 2.816, "percentage": 56.41, "elapsed_time": "2:10:26", "remaining_time": "1:40:47"}
+{"current_steps": 221, "total_steps": 390, "loss": 0.3772, "lr": 1.8837103421790486e-05, "epoch": 2.8288, "percentage": 56.67, "elapsed_time": "2:11:04", "remaining_time": "1:40:14"}
+{"current_steps": 222, "total_steps": 390, "loss": 0.3534, "lr": 1.8658447112179952e-05, "epoch": 2.8416, "percentage": 56.92, "elapsed_time": "2:11:48", "remaining_time": "1:39:44"}
+{"current_steps": 223, "total_steps": 390, "loss": 0.369, "lr": 1.8479898273352084e-05, "epoch": 2.8544, "percentage": 57.18, "elapsed_time": "2:12:27", "remaining_time": "1:39:12"}
+{"current_steps": 224, "total_steps": 390, "loss": 0.3387, "lr": 1.83014712087187e-05, "epoch": 2.8672, "percentage": 57.44, "elapsed_time": "2:13:02", "remaining_time": "1:38:35"}
+{"current_steps": 225, "total_steps": 390, "loss": 0.3771, "lr": 1.8123180211936417e-05, "epoch": 2.88, "percentage": 57.69, "elapsed_time": "2:13:33", "remaining_time": "1:37:56"}
+{"current_steps": 226, "total_steps": 390, "loss": 0.3534, "lr": 1.794503956576152e-05, "epoch": 2.8928000000000003, "percentage": 57.95, "elapsed_time": "2:14:07", "remaining_time": "1:37:19"}
+{"current_steps": 227, "total_steps": 390, "loss": 0.3563, "lr": 1.776706354090585e-05, "epoch": 2.9055999999999997, "percentage": 58.21, "elapsed_time": "2:14:49", "remaining_time": "1:36:48"}
+{"current_steps": 228, "total_steps": 390, "loss": 0.376, "lr": 1.758926639489354e-05, "epoch": 2.9184, "percentage": 58.46, "elapsed_time": "2:15:27", "remaining_time": "1:36:14"}
+{"current_steps": 229, "total_steps": 390, "loss": 0.339, "lr": 1.7411662370918893e-05, "epoch": 2.9312, "percentage": 58.72, "elapsed_time": "2:15:59", "remaining_time": "1:35:36"}
+{"current_steps": 230, "total_steps": 390, "loss": 0.3758, "lr": 1.7234265696705344e-05, "epoch": 2.944, "percentage": 58.97, "elapsed_time": "2:16:43", "remaining_time": "1:35:06"}
+{"current_steps": 231, "total_steps": 390, "loss": 0.35, "lr": 1.7057090583365678e-05, "epoch": 2.9568, "percentage": 59.23, "elapsed_time": "2:17:14", "remaining_time": "1:34:28"}
+{"current_steps": 232, "total_steps": 390, "loss": 0.3553, "lr": 1.6880151224263646e-05, "epoch": 2.9696, "percentage": 59.49, "elapsed_time": "2:17:53", "remaining_time": "1:33:54"}
+{"current_steps": 233, "total_steps": 390, "loss": 0.3772, "lr": 1.6703461793876876e-05, "epoch": 2.9824, "percentage": 59.74, "elapsed_time": "2:18:23", "remaining_time": "1:33:15"}
+{"current_steps": 234, "total_steps": 390, "loss": 0.3486, "lr": 1.6527036446661396e-05, "epoch": 2.9952, "percentage": 60.0, "elapsed_time": "2:18:58", "remaining_time": "1:32:39"}
+{"current_steps": 235, "total_steps": 390, "loss": 0.6203, "lr": 1.635088931591775e-05, "epoch": 3.008, "percentage": 60.26, "elapsed_time": "2:19:55", "remaining_time": "1:32:17"}
+{"current_steps": 236, "total_steps": 390, "loss": 0.2424, "lr": 1.6175034512658753e-05, "epoch": 3.0208, "percentage": 60.51, "elapsed_time": "2:20:26", "remaining_time": "1:31:38"}
+{"current_steps": 237, "total_steps": 390, "loss": 0.2496, "lr": 1.5999486124479115e-05, "epoch": 3.0336, "percentage": 60.77, "elapsed_time": "2:21:01", "remaining_time": "1:31:02"}
+{"current_steps": 238, "total_steps": 390, "loss": 0.2474, "lr": 1.5824258214426833e-05, "epoch": 3.0464, "percentage": 61.03, "elapsed_time": "2:21:31", "remaining_time": "1:30:23"}
+{"current_steps": 239, "total_steps": 390, "loss": 0.2404, "lr": 1.5649364819876655e-05, "epoch": 3.0592, "percentage": 61.28, "elapsed_time": "2:22:12", "remaining_time": "1:29:50"}
+{"current_steps": 240, "total_steps": 390, "loss": 0.24, "lr": 1.547481995140556e-05, "epoch": 3.072, "percentage": 61.54, "elapsed_time": "2:22:50", "remaining_time": "1:29:16"}
+{"current_steps": 241, "total_steps": 390, "loss": 0.2274, "lr": 1.5300637591670357e-05, "epoch": 3.0848, "percentage": 61.79, "elapsed_time": "2:23:23", "remaining_time": "1:28:39"}
+{"current_steps": 242, "total_steps": 390, "loss": 0.26, "lr": 1.5126831694287564e-05, "epoch": 3.0976, "percentage": 62.05, "elapsed_time": "2:23:59", "remaining_time": "1:28:03"}
+{"current_steps": 243, "total_steps": 390, "loss": 0.2544, "lr": 1.4953416182715566e-05, "epoch": 3.1104, "percentage": 62.31, "elapsed_time": "2:24:30", "remaining_time": "1:27:25"}
+{"current_steps": 244, "total_steps": 390, "loss": 0.2206, "lr": 1.478040494913926e-05, "epoch": 3.1232, "percentage": 62.56, "elapsed_time": "2:25:00", "remaining_time": "1:26:46"}
+{"current_steps": 245, "total_steps": 390, "loss": 0.2417, "lr": 1.460781185335713e-05, "epoch": 3.136, "percentage": 62.82, "elapsed_time": "2:25:34", "remaining_time": "1:26:09"}
+{"current_steps": 246, "total_steps": 390, "loss": 0.2449, "lr": 1.443565072167095e-05, "epoch": 3.1488, "percentage": 63.08, "elapsed_time": "2:26:15", "remaining_time": "1:25:36"}
+{"current_steps": 247, "total_steps": 390, "loss": 0.2199, "lr": 1.4263935345778202e-05, "epoch": 3.1616, "percentage": 63.33, "elapsed_time": "2:26:44", "remaining_time": "1:24:57"}
+{"current_steps": 248, "total_steps": 390, "loss": 0.2351, "lr": 1.409267948166718e-05, "epoch": 3.1744, "percentage": 63.59, "elapsed_time": "2:27:24", "remaining_time": "1:24:24"}
+{"current_steps": 249, "total_steps": 390, "loss": 0.2556, "lr": 1.3921896848515064e-05, "epoch": 3.1872, "percentage": 63.85, "elapsed_time": "2:28:03", "remaining_time": "1:23:50"}
+{"current_steps": 250, "total_steps": 390, "loss": 0.2365, "lr": 1.3751601127588849e-05, "epoch": 3.2, "percentage": 64.1, "elapsed_time": "2:28:38", "remaining_time": "1:23:14"}
+{"current_steps": 251, "total_steps": 390, "loss": 0.245, "lr": 1.3581805961149371e-05, "epoch": 3.2128, "percentage": 64.36, "elapsed_time": "2:29:13", "remaining_time": "1:22:38"}
+{"current_steps": 252, "total_steps": 390, "loss": 0.2344, "lr": 1.341252495135841e-05, "epoch": 3.2256, "percentage": 64.62, "elapsed_time": "2:29:51", "remaining_time": "1:22:03"}
+{"current_steps": 253, "total_steps": 390, "loss": 0.2465, "lr": 1.324377165918906e-05, "epoch": 3.2384, "percentage": 64.87, "elapsed_time": "2:30:19", "remaining_time": "1:21:24"}
+{"current_steps": 254, "total_steps": 390, "loss": 0.2338, "lr": 1.3075559603339354e-05, "epoch": 3.2512, "percentage": 65.13, "elapsed_time": "2:30:59", "remaining_time": "1:20:50"}
+{"current_steps": 255, "total_steps": 390, "loss": 0.228, "lr": 1.2907902259149287e-05, "epoch": 3.2640000000000002, "percentage": 65.38, "elapsed_time": "2:31:40", "remaining_time": "1:20:17"}
+{"current_steps": 256, "total_steps": 390, "loss": 0.2532, "lr": 1.274081305752135e-05, "epoch": 3.2768, "percentage": 65.64, "elapsed_time": "2:32:17", "remaining_time": "1:19:42"}
+{"current_steps": 257, "total_steps": 390, "loss": 0.2086, "lr": 1.2574305383844528e-05, "epoch": 3.2896, "percentage": 65.9, "elapsed_time": "2:32:50", "remaining_time": "1:19:05"}
+{"current_steps": 258, "total_steps": 390, "loss": 0.2375, "lr": 1.2408392576922075e-05, "epoch": 3.3024, "percentage": 66.15, "elapsed_time": "2:33:27", "remaining_time": "1:18:30"}
+{"current_steps": 259, "total_steps": 390, "loss": 0.2216, "lr": 1.2243087927902905e-05, "epoch": 3.3152, "percentage": 66.41, "elapsed_time": "2:33:57", "remaining_time": "1:17:52"}
+{"current_steps": 260, "total_steps": 390, "loss": 0.2438, "lr": 1.2078404679216864e-05, "epoch": 3.328, "percentage": 66.67, "elapsed_time": "2:34:36", "remaining_time": "1:17:18"}
+{"current_steps": 261, "total_steps": 390, "loss": 0.2449, "lr": 1.1914356023513904e-05, "epoch": 3.3407999999999998, "percentage": 66.92, "elapsed_time": "2:35:12", "remaining_time": "1:16:42"}
+{"current_steps": 262, "total_steps": 390, "loss": 0.2298, "lr": 1.1750955102607193e-05, "epoch": 3.3536, "percentage": 67.18, "elapsed_time": "2:35:48", "remaining_time": "1:16:07"}
+{"current_steps": 263, "total_steps": 390, "loss": 0.2163, "lr": 1.1588215006420374e-05, "epoch": 3.3664, "percentage": 67.44, "elapsed_time": "2:36:21", "remaining_time": "1:15:30"}
+{"current_steps": 264, "total_steps": 390, "loss": 0.2189, "lr": 1.1426148771938915e-05, "epoch": 3.3792, "percentage": 67.69, "elapsed_time": "2:36:55", "remaining_time": "1:14:53"}
+{"current_steps": 265, "total_steps": 390, "loss": 0.2372, "lr": 1.1264769382165748e-05, "epoch": 3.392, "percentage": 67.95, "elapsed_time": "2:37:30", "remaining_time": "1:14:17"}
+{"current_steps": 266, "total_steps": 390, "loss": 0.2133, "lr": 1.110408976508118e-05, "epoch": 3.4048, "percentage": 68.21, "elapsed_time": "2:37:59", "remaining_time": "1:13:38"}
+{"current_steps": 267, "total_steps": 390, "loss": 0.2301, "lr": 1.094412279260726e-05, "epoch": 3.4176, "percentage": 68.46, "elapsed_time": "2:38:38", "remaining_time": "1:13:04"}
+{"current_steps": 268, "total_steps": 390, "loss": 0.2486, "lr": 1.0784881279576635e-05, "epoch": 3.4304, "percentage": 68.72, "elapsed_time": "2:39:15", "remaining_time": "1:12:29"}
+{"current_steps": 269, "total_steps": 390, "loss": 0.2141, "lr": 1.0626377982705929e-05, "epoch": 3.4432, "percentage": 68.97, "elapsed_time": "2:39:51", "remaining_time": "1:11:54"}
+{"current_steps": 270, "total_steps": 390, "loss": 0.2304, "lr": 1.0468625599573842e-05, "epoch": 3.456, "percentage": 69.23, "elapsed_time": "2:40:22", "remaining_time": "1:11:16"}
+{"current_steps": 271, "total_steps": 390, "loss": 0.2431, "lr": 1.0311636767603952e-05, "epoch": 3.4688, "percentage": 69.49, "elapsed_time": "2:40:54", "remaining_time": "1:10:39"}
+{"current_steps": 272, "total_steps": 390, "loss": 0.2231, "lr": 1.0155424063052306e-05, "epoch": 3.4816, "percentage": 69.74, "elapsed_time": "2:41:23", "remaining_time": "1:10:00"}
+{"current_steps": 273, "total_steps": 390, "loss": 0.219, "lr": 1.0000000000000006e-05, "epoch": 3.4944, "percentage": 70.0, "elapsed_time": "2:41:50", "remaining_time": "1:09:21"}
+{"current_steps": 274, "total_steps": 390, "loss": 0.2347, "lr": 9.84537702935065e-06, "epoch": 3.5072, "percentage": 70.26, "elapsed_time": "2:42:23", "remaining_time": "1:08:44"}
+{"current_steps": 275, "total_steps": 390, "loss": 0.2257, "lr": 9.691567537832964e-06, "epoch": 3.52, "percentage": 70.51, "elapsed_time": "2:43:02", "remaining_time": "1:08:10"}
+{"current_steps": 276, "total_steps": 390, "loss": 0.2367, "lr": 9.538583847008452e-06, "epoch": 3.5328, "percentage": 70.77, "elapsed_time": "2:43:31", "remaining_time": "1:07:32"}
+{"current_steps": 277, "total_steps": 390, "loss": 0.2415, "lr": 9.386438212284372e-06, "epoch": 3.5456, "percentage": 71.03, "elapsed_time": "2:44:11", "remaining_time": "1:06:58"}
+{"current_steps": 278, "total_steps": 390, "loss": 0.2333, "lr": 9.235142821931928e-06, "epoch": 3.5584, "percentage": 71.28, "elapsed_time": "2:44:46", "remaining_time": "1:06:23"}
+{"current_steps": 279, "total_steps": 390, "loss": 0.2235, "lr": 9.084709796109907e-06, "epoch": 3.5712, "percentage": 71.54, "elapsed_time": "2:45:10", "remaining_time": "1:05:42"}
+{"current_steps": 280, "total_steps": 390, "loss": 0.2573, "lr": 8.93515118589373e-06, "epoch": 3.584, "percentage": 71.79, "elapsed_time": "2:45:54", "remaining_time": "1:05:10"}
+{"current_steps": 281, "total_steps": 390, "loss": 0.2111, "lr": 8.786478972310023e-06, "epoch": 3.5968, "percentage": 72.05, "elapsed_time": "2:46:27", "remaining_time": "1:04:34"}
+{"current_steps": 282, "total_steps": 390, "loss": 0.2451, "lr": 8.638705065376887e-06, "epoch": 3.6096, "percentage": 72.31, "elapsed_time": "2:47:06", "remaining_time": "1:03:59"}
+{"current_steps": 283, "total_steps": 390, "loss": 0.2345, "lr": 8.491841303149728e-06, "epoch": 3.6224, "percentage": 72.56, "elapsed_time": "2:47:37", "remaining_time": "1:03:22"}
+{"current_steps": 284, "total_steps": 390, "loss": 0.2463, "lr": 8.345899450772975e-06, "epoch": 3.6352, "percentage": 72.82, "elapsed_time": "2:48:10", "remaining_time": "1:02:46"}
+{"current_steps": 285, "total_steps": 390, "loss": 0.2451, "lr": 8.200891199537549e-06, "epoch": 3.648, "percentage": 73.08, "elapsed_time": "2:48:43", "remaining_time": "1:02:09"}
+{"current_steps": 286, "total_steps": 390, "loss": 0.2292, "lr": 8.056828165944282e-06, "epoch": 3.6608, "percentage": 73.33, "elapsed_time": "2:49:20", "remaining_time": "1:01:34"}
+{"current_steps": 287, "total_steps": 390, "loss": 0.2499, "lr": 7.913721890773354e-06, "epoch": 3.6736, "percentage": 73.59, "elapsed_time": "2:50:01", "remaining_time": "1:01:01"}
+{"current_steps": 288, "total_steps": 390, "loss": 0.2284, "lr": 7.771583838159756e-06, "epoch": 3.6864, "percentage": 73.85, "elapsed_time": "2:50:29", "remaining_time": "1:00:22"}
+{"current_steps": 289, "total_steps": 390, "loss": 0.2477, "lr": 7.630425394674903e-06, "epoch": 3.6992000000000003, "percentage": 74.1, "elapsed_time": "2:51:02", "remaining_time": "0:59:46"}
+{"current_steps": 290, "total_steps": 390, "loss": 0.2256, "lr": 7.49025786841445e-06, "epoch": 3.7119999999999997, "percentage": 74.36, "elapsed_time": "2:51:32", "remaining_time": "0:59:09"}
+{"current_steps": 291, "total_steps": 390, "loss": 0.2525, "lr": 7.3510924880924575e-06, "epoch": 3.7248, "percentage": 74.62, "elapsed_time": "2:52:06", "remaining_time": "0:58:33"}
+{"current_steps": 292, "total_steps": 390, "loss": 0.2286, "lr": 7.212940402141808e-06, "epoch": 3.7376, "percentage": 74.87, "elapsed_time": "2:52:37", "remaining_time": "0:57:56"}
+{"current_steps": 293, "total_steps": 390, "loss": 0.2281, "lr": 7.075812677821145e-06, "epoch": 3.7504, "percentage": 75.13, "elapsed_time": "2:53:15", "remaining_time": "0:57:21"}
+{"current_steps": 294, "total_steps": 390, "loss": 0.2574, "lr": 6.939720300328303e-06, "epoch": 3.7632, "percentage": 75.38, "elapsed_time": "2:53:51", "remaining_time": "0:56:46"}
+{"current_steps": 295, "total_steps": 390, "loss": 0.2053, "lr": 6.8046741719202385e-06, "epoch": 3.776, "percentage": 75.64, "elapsed_time": "2:54:25", "remaining_time": "0:56:10"}
+{"current_steps": 296, "total_steps": 390, "loss": 0.2561, "lr": 6.67068511103971e-06, "epoch": 3.7888, "percentage": 75.9, "elapsed_time": "2:55:06", "remaining_time": "0:55:36"}
+{"current_steps": 297, "total_steps": 390, "loss": 0.2167, "lr": 6.537763851448593e-06, "epoch": 3.8016, "percentage": 76.15, "elapsed_time": "2:55:43", "remaining_time": "0:55:01"}
+{"current_steps": 298, "total_steps": 390, "loss": 0.2372, "lr": 6.4059210413680175e-06, "epoch": 3.8144, "percentage": 76.41, "elapsed_time": "2:56:22", "remaining_time": "0:54:27"}
+{"current_steps": 299, "total_steps": 390, "loss": 0.2545, "lr": 6.275167242625331e-06, "epoch": 3.8272, "percentage": 76.67, "elapsed_time": "2:57:02", "remaining_time": "0:53:52"}
+{"current_steps": 300, "total_steps": 390, "loss": 0.2171, "lr": 6.145512929808013e-06, "epoch": 3.84, "percentage": 76.92, "elapsed_time": "2:57:37", "remaining_time": "0:53:17"}
+{"current_steps": 301, "total_steps": 390, "loss": 0.2231, "lr": 6.016968489424572e-06, "epoch": 3.8528000000000002, "percentage": 77.18, "elapsed_time": "2:58:12", "remaining_time": "0:52:41"}
+{"current_steps": 302, "total_steps": 390, "loss": 0.2371, "lr": 5.889544219072465e-06, "epoch": 3.8656, "percentage": 77.44, "elapsed_time": "2:58:52", "remaining_time": "0:52:07"}
+{"current_steps": 303, "total_steps": 390, "loss": 0.2348, "lr": 5.7632503266131925e-06, "epoch": 3.8784, "percentage": 77.69, "elapsed_time": "2:59:28", "remaining_time": "0:51:31"}
+{"current_steps": 304, "total_steps": 390, "loss": 0.2048, "lr": 5.638096929354522e-06, "epoch": 3.8912, "percentage": 77.95, "elapsed_time": "3:00:05", "remaining_time": "0:50:56"}
+{"current_steps": 305, "total_steps": 390, "loss": 0.2562, "lr": 5.514094053240035e-06, "epoch": 3.904, "percentage": 78.21, "elapsed_time": "3:00:46", "remaining_time": "0:50:22"}
+{"current_steps": 306, "total_steps": 390, "loss": 0.225, "lr": 5.39125163204594e-06, "epoch": 3.9168, "percentage": 78.46, "elapsed_time": "3:01:18", "remaining_time": "0:49:46"}
+{"current_steps": 307, "total_steps": 390, "loss": 0.2279, "lr": 5.269579506585259e-06, "epoch": 3.9295999999999998, "percentage": 78.72, "elapsed_time": "3:01:55", "remaining_time": "0:49:11"}
+{"current_steps": 308, "total_steps": 390, "loss": 0.2219, "lr": 5.149087423919541e-06, "epoch": 3.9424, "percentage": 78.97, "elapsed_time": "3:02:34", "remaining_time": "0:48:36"}
+{"current_steps": 309, "total_steps": 390, "loss": 0.2354, "lr": 5.029785036577976e-06, "epoch": 3.9552, "percentage": 79.23, "elapsed_time": "3:03:13", "remaining_time": "0:48:01"}
+{"current_steps": 310, "total_steps": 390, "loss": 0.2422, "lr": 4.911681901784198e-06, "epoch": 3.968, "percentage": 79.49, "elapsed_time": "3:03:48", "remaining_time": "0:47:26"}
+{"current_steps": 311, "total_steps": 390, "loss": 0.2316, "lr": 4.794787480690597e-06, "epoch": 3.9808, "percentage": 79.74, "elapsed_time": "3:04:30", "remaining_time": "0:46:52"}
+{"current_steps": 312, "total_steps": 390, "loss": 0.2275, "lr": 4.679111137620442e-06, "epoch": 3.9936, "percentage": 80.0, "elapsed_time": "3:05:12", "remaining_time": "0:46:18"}
+{"current_steps": 313, "total_steps": 390, "loss": 0.3807, "lr": 4.5646621393177e-06, "epoch": 4.0064, "percentage": 80.26, "elapsed_time": "3:06:08", "remaining_time": "0:45:47"}
+{"current_steps": 314, "total_steps": 390, "loss": 0.183, "lr": 4.451449654204685e-06, "epoch": 4.0192, "percentage": 80.51, "elapsed_time": "3:06:43", "remaining_time": "0:45:11"}
+{"current_steps": 315, "total_steps": 390, "loss": 0.1775, "lr": 4.339482751647557e-06, "epoch": 4.032, "percentage": 80.77, "elapsed_time": "3:07:19", "remaining_time": "0:44:36"}
+{"current_steps": 316, "total_steps": 390, "loss": 0.1509, "lr": 4.228770401229824e-06, "epoch": 4.0448, "percentage": 81.03, "elapsed_time": "3:07:50", "remaining_time": "0:43:59"}
+{"current_steps": 317, "total_steps": 390, "loss": 0.1573, "lr": 4.119321472033779e-06, "epoch": 4.0576, "percentage": 81.28, "elapsed_time": "3:08:26", "remaining_time": "0:43:23"}
+{"current_steps": 318, "total_steps": 390, "loss": 0.1619, "lr": 4.011144731929981e-06, "epoch": 4.0704, "percentage": 81.54, "elapsed_time": "3:08:58", "remaining_time": "0:42:47"}
+{"current_steps": 319, "total_steps": 390, "loss": 0.1693, "lr": 3.904248846874894e-06, "epoch": 4.0832, "percentage": 81.79, "elapsed_time": "3:09:35", "remaining_time": "0:42:11"}
+{"current_steps": 320, "total_steps": 390, "loss": 0.1767, "lr": 3.7986423802166705e-06, "epoch": 4.096, "percentage": 82.05, "elapsed_time": "3:10:11", "remaining_time": "0:41:36"}
+{"current_steps": 321, "total_steps": 390, "loss": 0.1588, "lr": 3.694333792009115e-06, "epoch": 4.1088, "percentage": 82.31, "elapsed_time": "3:10:43", "remaining_time": "0:40:59"}
+{"current_steps": 322, "total_steps": 390, "loss": 0.1782, "lr": 3.5913314383339937e-06, "epoch": 4.1216, "percentage": 82.56, "elapsed_time": "3:11:19", "remaining_time": "0:40:24"}
+{"current_steps": 323, "total_steps": 390, "loss": 0.1661, "lr": 3.4896435706316e-06, "epoch": 4.1344, "percentage": 82.82, "elapsed_time": "3:11:51", "remaining_time": "0:39:47"}
+{"current_steps": 324, "total_steps": 390, "loss": 0.1424, "lr": 3.3892783350397675e-06, "epoch": 4.1472, "percentage": 83.08, "elapsed_time": "3:12:22", "remaining_time": "0:39:11"}
+{"current_steps": 325, "total_steps": 390, "loss": 0.1734, "lr": 3.290243771741275e-06, "epoch": 4.16, "percentage": 83.33, "elapsed_time": "3:12:53", "remaining_time": "0:38:34"}
+{"current_steps": 326, "total_steps": 390, "loss": 0.1608, "lr": 3.1925478143197418e-06, "epoch": 4.1728, "percentage": 83.59, "elapsed_time": "3:13:32", "remaining_time": "0:37:59"}
+{"current_steps": 327, "total_steps": 390, "loss": 0.158, "lr": 3.0961982891241083e-06, "epoch": 4.1856, "percentage": 83.85, "elapsed_time": "3:14:01", "remaining_time": "0:37:22"}
+{"current_steps": 328, "total_steps": 390, "loss": 0.1596, "lr": 3.001202914641628e-06, "epoch": 4.1984, "percentage": 84.1, "elapsed_time": "3:14:36", "remaining_time": "0:36:47"}
+{"current_steps": 329, "total_steps": 390, "loss": 0.1704, "lr": 2.907569300879596e-06, "epoch": 4.2112, "percentage": 84.36, "elapsed_time": "3:15:12", "remaining_time": "0:36:11"}
+{"current_steps": 330, "total_steps": 390, "loss": 0.1641, "lr": 2.815304948755664e-06, "epoch": 4.224, "percentage": 84.62, "elapsed_time": "3:15:49", "remaining_time": "0:35:36"}
+{"current_steps": 331, "total_steps": 390, "loss": 0.1561, "lr": 2.7244172494969978e-06, "epoch": 4.2368, "percentage": 84.87, "elapsed_time": "3:16:24", "remaining_time": "0:35:00"}
+{"current_steps": 332, "total_steps": 390, "loss": 0.1561, "lr": 2.6349134840481294e-06, "epoch": 4.2496, "percentage": 85.13, "elapsed_time": "3:16:57", "remaining_time": "0:34:24"}
+{"current_steps": 333, "total_steps": 390, "loss": 0.1765, "lr": 2.546800822487714e-06, "epoch": 4.2624, "percentage": 85.38, "elapsed_time": "3:17:38", "remaining_time": "0:33:49"}
+{"current_steps": 334, "total_steps": 390, "loss": 0.1731, "lr": 2.4600863234541338e-06, "epoch": 4.2752, "percentage": 85.64, "elapsed_time": "3:18:18", "remaining_time": "0:33:14"}
+{"current_steps": 335, "total_steps": 390, "loss": 0.1558, "lr": 2.374776933580025e-06, "epoch": 4.288, "percentage": 85.9, "elapsed_time": "3:18:56", "remaining_time": "0:32:39"}
+{"current_steps": 336, "total_steps": 390, "loss": 0.1557, "lr": 2.2908794869358044e-06, "epoch": 4.3008, "percentage": 86.15, "elapsed_time": "3:19:37", "remaining_time": "0:32:04"}
+{"current_steps": 337, "total_steps": 390, "loss": 0.1572, "lr": 2.2084007044821764e-06, "epoch": 4.3136, "percentage": 86.41, "elapsed_time": "3:20:11", "remaining_time": "0:31:29"}
+{"current_steps": 338, "total_steps": 390, "loss": 0.1557, "lr": 2.127347193531757e-06, "epoch": 4.3264, "percentage": 86.67, "elapsed_time": "3:20:52", "remaining_time": "0:30:54"}
+{"current_steps": 339, "total_steps": 390, "loss": 0.1514, "lr": 2.0477254472197237e-06, "epoch": 4.3392, "percentage": 86.92, "elapsed_time": "3:21:30", "remaining_time": "0:30:18"}
+{"current_steps": 340, "total_steps": 390, "loss": 0.1704, "lr": 1.96954184398368e-06, "epoch": 4.352, "percentage": 87.18, "elapsed_time": "3:22:04", "remaining_time": "0:29:43"}
+{"current_steps": 341, "total_steps": 390, "loss": 0.1469, "lr": 1.8928026470526917e-06, "epoch": 4.3648, "percentage": 87.44, "elapsed_time": "3:22:33", "remaining_time": "0:29:06"}
+{"current_steps": 342, "total_steps": 390, "loss": 0.1632, "lr": 1.817514003945524e-06, "epoch": 4.3776, "percentage": 87.69, "elapsed_time": "3:23:08", "remaining_time": "0:28:30"}
+{"current_steps": 343, "total_steps": 390, "loss": 0.1694, "lr": 1.743681945978184e-06, "epoch": 4.3904, "percentage": 87.95, "elapsed_time": "3:23:43", "remaining_time": "0:27:54"}
+{"current_steps": 344, "total_steps": 390, "loss": 0.1515, "lr": 1.6713123877807413e-06, "epoch": 4.4032, "percentage": 88.21, "elapsed_time": "3:24:10", "remaining_time": "0:27:18"}
+{"current_steps": 345, "total_steps": 390, "loss": 0.1549, "lr": 1.6004111268235156e-06, "epoch": 4.416, "percentage": 88.46, "elapsed_time": "3:24:47", "remaining_time": "0:26:42"}
+{"current_steps": 346, "total_steps": 390, "loss": 0.1691, "lr": 1.5309838429526714e-06, "epoch": 4.4288, "percentage": 88.72, "elapsed_time": "3:25:27", "remaining_time": "0:26:07"}
+{"current_steps": 347, "total_steps": 390, "loss": 0.16, "lr": 1.4630360979351644e-06, "epoch": 4.4416, "percentage": 88.97, "elapsed_time": "3:26:02", "remaining_time": "0:25:31"}
+{"current_steps": 348, "total_steps": 390, "loss": 0.1572, "lr": 1.396573335013236e-06, "epoch": 4.4544, "percentage": 89.23, "elapsed_time": "3:26:42", "remaining_time": "0:24:56"}
+{"current_steps": 349, "total_steps": 390, "loss": 0.1677, "lr": 1.3316008784683265e-06, "epoch": 4.4672, "percentage": 89.49, "elapsed_time": "3:27:18", "remaining_time": "0:24:21"}
+{"current_steps": 350, "total_steps": 390, "loss": 0.1656, "lr": 1.2681239331945695e-06, "epoch": 4.48, "percentage": 89.74, "elapsed_time": "3:27:53", "remaining_time": "0:23:45"}
+{"current_steps": 351, "total_steps": 390, "loss": 0.1575, "lr": 1.2061475842818337e-06, "epoch": 4.4928, "percentage": 90.0, "elapsed_time": "3:28:26", "remaining_time": "0:23:09"}
+{"current_steps": 352, "total_steps": 390, "loss": 0.1512, "lr": 1.1456767966083393e-06, "epoch": 4.5056, "percentage": 90.26, "elapsed_time": "3:29:02", "remaining_time": "0:22:34"}
+{"current_steps": 353, "total_steps": 390, "loss": 0.1719, "lr": 1.086716414442952e-06, "epoch": 4.5184, "percentage": 90.51, "elapsed_time": "3:29:31", "remaining_time": "0:21:57"}
+{"current_steps": 354, "total_steps": 390, "loss": 0.1525, "lr": 1.0292711610570904e-06, "epoch": 4.5312, "percentage": 90.77, "elapsed_time": "3:30:04", "remaining_time": "0:21:21"}
+{"current_steps": 355, "total_steps": 390, "loss": 0.163, "lr": 9.733456383463658e-07, "epoch": 4.5440000000000005, "percentage": 91.03, "elapsed_time": "3:30:36", "remaining_time": "0:20:45"}
+{"current_steps": 356, "total_steps": 390, "loss": 0.1621, "lr": 9.189443264619102e-07, "epoch": 4.5568, "percentage": 91.28, "elapsed_time": "3:31:07", "remaining_time": "0:20:09"}
+{"current_steps": 357, "total_steps": 390, "loss": 0.1517, "lr": 8.660715834514977e-07, "epoch": 4.5696, "percentage": 91.54, "elapsed_time": "3:31:43", "remaining_time": "0:19:34"}
+{"current_steps": 358, "total_steps": 390, "loss": 0.1579, "lr": 8.147316449103959e-07, "epoch": 4.5824, "percentage": 91.79, "elapsed_time": "3:32:19", "remaining_time": "0:18:58"}
+{"current_steps": 359, "total_steps": 390, "loss": 0.1589, "lr": 7.649286236420806e-07, "epoch": 4.5952, "percentage": 92.05, "elapsed_time": "3:32:58", "remaining_time": "0:18:23"}
+{"current_steps": 360, "total_steps": 390, "loss": 0.1516, "lr": 7.166665093287539e-07, "epoch": 4.608, "percentage": 92.31, "elapsed_time": "3:33:33", "remaining_time": "0:17:47"}
+{"current_steps": 361, "total_steps": 390, "loss": 0.1532, "lr": 6.69949168211721e-07, "epoch": 4.6208, "percentage": 92.56, "elapsed_time": "3:34:10", "remaining_time": "0:17:12"}
+{"current_steps": 362, "total_steps": 390, "loss": 0.1528, "lr": 6.247803427816945e-07, "epoch": 4.6336, "percentage": 92.82, "elapsed_time": "3:34:48", "remaining_time": "0:16:36"}
+{"current_steps": 363, "total_steps": 390, "loss": 0.1584, "lr": 5.811636514789598e-07, "epoch": 4.6464, "percentage": 93.08, "elapsed_time": "3:35:26", "remaining_time": "0:16:01"}
+{"current_steps": 364, "total_steps": 390, "loss": 0.1637, "lr": 5.391025884035239e-07, "epoch": 4.6592, "percentage": 93.33, "elapsed_time": "3:36:06", "remaining_time": "0:15:26"}
+{"current_steps": 365, "total_steps": 390, "loss": 0.154, "lr": 4.986005230351954e-07, "epoch": 4.672, "percentage": 93.59, "elapsed_time": "3:36:42", "remaining_time": "0:14:50"}
+{"current_steps": 366, "total_steps": 390, "loss": 0.1517, "lr": 4.5966069996365993e-07, "epoch": 4.6848, "percentage": 93.85, "elapsed_time": "3:37:17", "remaining_time": "0:14:14"}
+{"current_steps": 367, "total_steps": 390, "loss": 0.1613, "lr": 4.22286238628562e-07, "epoch": 4.6975999999999996, "percentage": 94.1, "elapsed_time": "3:38:00", "remaining_time": "0:13:39"}
+{"current_steps": 368, "total_steps": 390, "loss": 0.1563, "lr": 3.8648013306960664e-07, "epoch": 4.7104, "percentage": 94.36, "elapsed_time": "3:38:32", "remaining_time": "0:13:03"}
+{"current_steps": 369, "total_steps": 390, "loss": 0.1612, "lr": 3.522452516867048e-07, "epoch": 4.7232, "percentage": 94.62, "elapsed_time": "3:39:05", "remaining_time": "0:12:28"}
+{"current_steps": 370, "total_steps": 390, "loss": 0.168, "lr": 3.1958433701019697e-07, "epoch": 4.736, "percentage": 94.87, "elapsed_time": "3:39:46", "remaining_time": "0:11:52"}
+{"current_steps": 371, "total_steps": 390, "loss": 0.1472, "lr": 2.8850000548115155e-07, "epoch": 4.7488, "percentage": 95.13, "elapsed_time": "3:40:15", "remaining_time": "0:11:16"}
+{"current_steps": 372, "total_steps": 390, "loss": 0.1586, "lr": 2.5899474724174313e-07, "epoch": 4.7616, "percentage": 95.38, "elapsed_time": "3:40:51", "remaining_time": "0:10:41"}
+{"current_steps": 373, "total_steps": 390, "loss": 0.1578, "lr": 2.3107092593579905e-07, "epoch": 4.7744, "percentage": 95.64, "elapsed_time": "3:41:30", "remaining_time": "0:10:05"}
+{"current_steps": 374, "total_steps": 390, "loss": 0.1497, "lr": 2.0473077851942858e-07, "epoch": 4.7872, "percentage": 95.9, "elapsed_time": "3:42:02", "remaining_time": "0:09:29"}
+{"current_steps": 375, "total_steps": 390, "loss": 0.1584, "lr": 1.799764150818306e-07, "epoch": 4.8, "percentage": 96.15, "elapsed_time": "3:42:44", "remaining_time": "0:08:54"}
+{"current_steps": 376, "total_steps": 390, "loss": 0.1423, "lr": 1.5680981867625566e-07, "epoch": 4.8128, "percentage": 96.41, "elapsed_time": "3:43:25", "remaining_time": "0:08:19"}
+{"current_steps": 377, "total_steps": 390, "loss": 0.1533, "lr": 1.3523284516113955e-07, "epoch": 4.8256, "percentage": 96.67, "elapsed_time": "3:44:01", "remaining_time": "0:07:43"}
+{"current_steps": 378, "total_steps": 390, "loss": 0.1534, "lr": 1.1524722305144231e-07, "epoch": 4.8384, "percentage": 96.92, "elapsed_time": "3:44:37", "remaining_time": "0:07:07"}
+{"current_steps": 379, "total_steps": 390, "loss": 0.1624, "lr": 9.685455338016347e-08, "epoch": 4.8512, "percentage": 97.18, "elapsed_time": "3:45:12", "remaining_time": "0:06:32"}
+{"current_steps": 380, "total_steps": 390, "loss": 0.1641, "lr": 8.005630957010014e-08, "epoch": 4.864, "percentage": 97.44, "elapsed_time": "3:45:44", "remaining_time": "0:05:56"}
+{"current_steps": 381, "total_steps": 390, "loss": 0.1585, "lr": 6.485383731580142e-08, "epoch": 4.8768, "percentage": 97.69, "elapsed_time": "3:46:21", "remaining_time": "0:05:20"}
+{"current_steps": 382, "total_steps": 390, "loss": 0.1606, "lr": 5.1248354475768034e-08, "epoch": 4.8896, "percentage": 97.95, "elapsed_time": "3:46:54", "remaining_time": "0:04:45"}
+{"current_steps": 383, "total_steps": 390, "loss": 0.1486, "lr": 3.924095097489922e-08, "epoch": 4.9024, "percentage": 98.21, "elapsed_time": "3:47:29", "remaining_time": "0:04:09"}
+{"current_steps": 384, "total_steps": 390, "loss": 0.172, "lr": 2.8832588717164766e-08, "epoch": 4.9152000000000005, "percentage": 98.46, "elapsed_time": "3:48:07", "remaining_time": "0:03:33"}
+{"current_steps": 385, "total_steps": 390, "loss": 0.1655, "lr": 2.0024101508555604e-08, "epoch": 4.928, "percentage": 98.72, "elapsed_time": "3:48:47", "remaining_time": "0:02:58"}
+{"current_steps": 386, "total_steps": 390, "loss": 0.178, "lr": 1.281619499029274e-08, "epoch": 4.9408, "percentage": 98.97, "elapsed_time": "3:49:25", "remaining_time": "0:02:22"}
+{"current_steps": 387, "total_steps": 390, "loss": 0.1551, "lr": 7.209446582292501e-09, "epoch": 4.9536, "percentage": 99.23, "elapsed_time": "3:49:54", "remaining_time": "0:01:46"}
+{"current_steps": 388, "total_steps": 390, "loss": 0.1647, "lr": 3.2043054369057523e-09, "epoch": 4.9664, "percentage": 99.49, "elapsed_time": "3:50:25", "remaining_time": "0:01:11"}
+{"current_steps": 389, "total_steps": 390, "loss": 0.1696, "lr": 8.010924029533406e-10, "epoch": 4.9792, "percentage": 99.74, "elapsed_time": "3:51:03", "remaining_time": "0:00:35"}
+{"current_steps": 390, "total_steps": 390, "loss": 0.1485, "lr": 0.0, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "3:51:37", "remaining_time": "0:00:00"}
+{"current_steps": 390, "total_steps": 390, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "3:52:06", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2772 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.992,
+  "eval_steps": 500,
+  "global_step": 390,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0128,
+      "grad_norm": 5.839018847816655,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.9618,
+      "step": 1
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 6.044975028236581,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.9669,
+      "step": 2
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.219047217466672,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.021,
+      "step": 3
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 5.4838569963975985,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.9571,
+      "step": 4
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.047643058048521,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.8816,
+      "step": 5
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.2364546199415556,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.8964,
+      "step": 6
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.06808539264327,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.8921,
+      "step": 7
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 4.395847794140312,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.8709,
+      "step": 8
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 5.136863261262288,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9305,
+      "step": 9
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 4.504381262936195,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.813,
+      "step": 10
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 3.9344562843621285,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.8366,
+      "step": 11
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 3.14056368073828,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.8205,
+      "step": 12
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 2.2854369325531834,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7825,
+      "step": 13
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.55230766379796,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.7745,
+      "step": 14
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.1163531503614808,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.7378,
+      "step": 15
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.7545006007454378,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.7281,
+      "step": 16
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.4874110741930364,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.7435,
+      "step": 17
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.6189725301930469,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.743,
+      "step": 18
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.4046995785004956,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.7486,
+      "step": 19
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.9358645050027536,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.7467,
+      "step": 20
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.0721376654895611,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.6933,
+      "step": 21
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.1516256662003364,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.7067,
+      "step": 22
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.9223029393111457,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.7046,
+      "step": 23
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 1.1474983271538444,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.7125,
+      "step": 24
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8955100949974366,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.6728,
+      "step": 25
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 1.2251687166354326,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7381,
+      "step": 26
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.7816429553280688,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.6553,
+      "step": 27
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.9763360804698599,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.6798,
+      "step": 28
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.8354114194674345,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.6958,
+      "step": 29
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.6887105015637863,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.6497,
+      "step": 30
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.8779524514630032,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.6844,
+      "step": 31
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.7537997504917305,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.6672,
+      "step": 32
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.8385727211139176,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6744,
+      "step": 33
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.7602171774049509,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.6666,
+      "step": 34
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.7678255126965418,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.6642,
+      "step": 35
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.647081056305973,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.6638,
+      "step": 36
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.8379083769655634,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.6732,
+      "step": 37
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.9149239688006771,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.6624,
+      "step": 38
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.9772104074784475,
+      "learning_rate": 4e-05,
+      "loss": 0.6962,
+      "step": 39
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 1.1116947880572285,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.6245,
+      "step": 40
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.7227423304768713,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.6225,
+      "step": 41
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.9725040218297323,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.6604,
+      "step": 42
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.955643555394975,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.6133,
+      "step": 43
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.7567970803940156,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.6309,
+      "step": 44
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.7996454927598492,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.6659,
+      "step": 45
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.7366005891361118,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.6569,
+      "step": 46
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.7885369085723203,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.6313,
+      "step": 47
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.6751447378852333,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.6326,
+      "step": 48
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.8571310551925032,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.6201,
+      "step": 49
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7550362899021348,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.6329,
+      "step": 50
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.8280593896023649,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.6422,
+      "step": 51
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.8811902923700886,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6354,
+      "step": 52
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 1.1095721931863598,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.6667,
+      "step": 53
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.6445072194387576,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.6364,
+      "step": 54
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.9734320449566246,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.6218,
+      "step": 55
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.6247852399111969,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.6401,
+      "step": 56
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.6158328565284048,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.6168,
+      "step": 57
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.6280970629663458,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.6423,
+      "step": 58
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.876342261245329,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.6644,
+      "step": 59
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.6782228770929252,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.6425,
+      "step": 60
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.6213554461131593,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.6114,
+      "step": 61
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.8970158114530625,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.6523,
+      "step": 62
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.6928394207466625,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.6168,
+      "step": 63
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.6361855672357039,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.6177,
+      "step": 64
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.7475080798399086,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6338,
+      "step": 65
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.7201665058372186,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.6004,
+      "step": 66
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.6665201343517307,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.6324,
+      "step": 67
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.7733661737414401,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.6246,
+      "step": 68
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.6768671192755346,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.6219,
+      "step": 69
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.7097685593783307,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.623,
+      "step": 70
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.6536798448840015,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.6269,
+      "step": 71
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.6614424957485971,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.6351,
+      "step": 72
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.6977238316575177,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.6239,
+      "step": 73
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.7934882507910082,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.6573,
+      "step": 74
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6575515131086959,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.5857,
+      "step": 75
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.7723302740017406,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.6344,
+      "step": 76
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.6941667379950092,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.5906,
+      "step": 77
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.8021010313212673,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6364,
+      "step": 78
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 1.626971688308395,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 1.068,
+      "step": 79
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.9308988989110186,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.5233,
+      "step": 80
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.9488953391338323,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.5753,
+      "step": 81
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.7366798820619407,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.4947,
+      "step": 82
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.820021588310417,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.5649,
+      "step": 83
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.6449418020815564,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.4996,
+      "step": 84
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.9216006193216165,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.5333,
+      "step": 85
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.6725609871315669,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.4864,
+      "step": 86
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.9531128581935571,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.5237,
+      "step": 87
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.6641456185314702,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.5167,
+      "step": 88
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.8823519514010172,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.4995,
+      "step": 89
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.5548207275448888,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.5106,
+      "step": 90
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.8288761814178127,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5217,
+      "step": 91
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.5061675211682982,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.5033,
+      "step": 92
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.7190426537289591,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.5028,
+      "step": 93
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.6128890949411272,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.4779,
+      "step": 94
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.62309405530766,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.5451,
+      "step": 95
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.6334625248101861,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.5274,
+      "step": 96
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.5653360526439104,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.539,
+      "step": 97
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.6224327796221331,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.4945,
+      "step": 98
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.7081659483273709,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.5272,
+      "step": 99
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5845719830086954,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.5066,
+      "step": 100
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.6882406051818343,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.5123,
+      "step": 101
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.6051334323383424,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.5259,
+      "step": 102
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.5879222709126353,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.473,
+      "step": 103
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.5592006776867556,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5262,
+      "step": 104
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.5242844070496,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.4905,
+      "step": 105
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.5941812532382632,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.5479,
+      "step": 106
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.5387721208985127,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.4948,
+      "step": 107
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.5854840592046221,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.4785,
+      "step": 108
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.522507074880471,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.4754,
+      "step": 109
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.5625209088620615,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.4869,
+      "step": 110
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.43969509643876364,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.4972,
+      "step": 111
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.5619860822540647,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.4735,
+      "step": 112
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.4421063815613501,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.5055,
+      "step": 113
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.5369412032192468,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.4923,
+      "step": 114
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.48910911632729354,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.512,
+      "step": 115
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.5136909704915422,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.4981,
+      "step": 116
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.48631677940560963,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4915,
+      "step": 117
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.4840473244318828,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.4736,
+      "step": 118
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.5064523309799703,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.5403,
+      "step": 119
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.589977688148516,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.4694,
+      "step": 120
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.45092131277939423,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.4979,
+      "step": 121
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.5694759979218573,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.5191,
+      "step": 122
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.5065392756850746,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.4805,
+      "step": 123
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.5466527945164157,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.5199,
+      "step": 124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7191169252596367,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.5183,
+      "step": 125
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.519742080744315,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.5178,
+      "step": 126
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.738483275841055,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.4837,
+      "step": 127
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.5192697289694931,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.5115,
+      "step": 128
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.7555019608159453,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.5102,
+      "step": 129
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.525986636066217,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5035,
+      "step": 130
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.6237338996072936,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.4708,
+      "step": 131
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.5347925953394986,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.5117,
+      "step": 132
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.47804164439498226,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.5006,
+      "step": 133
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.5097745208171317,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.4956,
+      "step": 134
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.472002785073484,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.5138,
+      "step": 135
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.40000804145024127,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.452,
+      "step": 136
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.452939045182372,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.5132,
+      "step": 137
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.4492202701835303,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.51,
+      "step": 138
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.5217146312190998,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.5055,
+      "step": 139
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.45917544834665075,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.5128,
+      "step": 140
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.48608619608415743,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.4807,
+      "step": 141
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.4153980294330535,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.4887,
+      "step": 142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.4635350891032439,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5051,
+      "step": 143
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.5452295259370951,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.5138,
+      "step": 144
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.45699342832042916,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.472,
+      "step": 145
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.5082490486526375,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.4954,
+      "step": 146
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.6384690552060961,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.4987,
+      "step": 147
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.4371303247604508,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.4848,
+      "step": 148
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.41511391912882867,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.4665,
+      "step": 149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.4249372081891107,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.4994,
+      "step": 150
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.44138936645964416,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.4927,
+      "step": 151
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.43910408608963586,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.5066,
+      "step": 152
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.43088729689416577,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.4687,
+      "step": 153
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.4278640166834245,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.5163,
+      "step": 154
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.4742909576609465,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.471,
+      "step": 155
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.4587718963346401,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5087,
+      "step": 156
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 0.9966628173120351,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.8075,
+      "step": 157
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.6863182224101623,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.4051,
+      "step": 158
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 2.0900892617168663,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.3823,
+      "step": 159
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.7960735194376195,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.3444,
+      "step": 160
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 0.655937861814063,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.3576,
+      "step": 161
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.5911864199285094,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.3585,
+      "step": 162
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.6532713894850588,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.3538,
+      "step": 163
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.6592758774119294,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.3908,
+      "step": 164
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.6444342767269957,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.4022,
+      "step": 165
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.6495217170344411,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.3443,
+      "step": 166
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.5590722491328468,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.3723,
+      "step": 167
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.5648244135782439,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.3498,
+      "step": 168
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.4820229598242042,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3497,
+      "step": 169
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.5561337750606294,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.3677,
+      "step": 170
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.5455054065895131,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.3433,
+      "step": 171
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.5916420617982933,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.3971,
+      "step": 172
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.6508258578592079,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.3756,
+      "step": 173
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.47618701005958114,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.3403,
+      "step": 174
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5593921807214691,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.3595,
+      "step": 175
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.4198057083178645,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.3314,
+      "step": 176
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.5110303368655525,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.3606,
+      "step": 177
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.5011779517833854,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.3799,
+      "step": 178
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.5246653684584236,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.3464,
+      "step": 179
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.4855556982779856,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.366,
+      "step": 180
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.5305125098314,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.3433,
+      "step": 181
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.4970558826142017,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.36,
+      "step": 182
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.499133817248476,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.3337,
+      "step": 183
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.4862098543776659,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.3614,
+      "step": 184
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.49489877938395055,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.364,
+      "step": 185
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.47745055778943635,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.3369,
+      "step": 186
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.4518599993209865,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.3397,
+      "step": 187
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.49190428712388085,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.3609,
+      "step": 188
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.42695016108008466,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.3403,
+      "step": 189
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.48399579137589493,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.3559,
+      "step": 190
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.43535579894833265,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.3699,
+      "step": 191
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.42236836330089367,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.3389,
+      "step": 192
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.4254052985368455,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.3748,
+      "step": 193
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.4124026511474054,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.3649,
+      "step": 194
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.4358743463877152,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3911,
+      "step": 195
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.3787171361989274,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.3594,
+      "step": 196
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.402330181025512,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.3528,
+      "step": 197
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.3742935305188853,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.328,
+      "step": 198
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.41761460975640535,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.393,
+      "step": 199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3555221908620572,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.3705,
+      "step": 200
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.4632160156775803,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.3333,
+      "step": 201
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.3854892191886437,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.3494,
+      "step": 202
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.3637067298468189,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.3405,
+      "step": 203
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.40494042381970147,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.3888,
+      "step": 204
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.3755956967947316,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.3255,
+      "step": 205
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.4304305110638947,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.3865,
+      "step": 206
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.4062101646966682,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.3513,
+      "step": 207
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.4000767842561131,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3609,
+      "step": 208
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.3945734513213183,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.3219,
+      "step": 209
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.42058336216796155,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.3744,
+      "step": 210
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.4132723550394805,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.3658,
+      "step": 211
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.39649934238958967,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.3405,
+      "step": 212
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.4208572250100482,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.368,
+      "step": 213
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.3978701788145883,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.3465,
+      "step": 214
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.4001369867888692,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.3532,
+      "step": 215
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.4034806906167456,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.353,
+      "step": 216
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.4307848928295819,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.3569,
+      "step": 217
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.3999064020664724,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.3568,
+      "step": 218
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.4090043678053143,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.365,
+      "step": 219
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.42532758658903885,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.3427,
+      "step": 220
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.39419334493438113,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3772,
+      "step": 221
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.4016749191362796,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.3534,
+      "step": 222
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.40847223940384836,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.369,
+      "step": 223
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.4001323176748629,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.3387,
+      "step": 224
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4248079947067711,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.3771,
+      "step": 225
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.3945282895021062,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.3534,
+      "step": 226
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.3497014179956054,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.3563,
+      "step": 227
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.3986345065122307,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.376,
+      "step": 228
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.370859934109157,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.339,
+      "step": 229
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.3747895589693441,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.3758,
+      "step": 230
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.3685623113569053,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.35,
+      "step": 231
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.36104623960562526,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.3553,
+      "step": 232
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.40586050975133925,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.3772,
+      "step": 233
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.41959521563610697,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3486,
+      "step": 234
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 1.1794707358197836,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.6203,
+      "step": 235
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.6207295094733497,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.2424,
+      "step": 236
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.5155664642605005,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.2496,
+      "step": 237
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.841858506357091,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.2474,
+      "step": 238
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.6482412245876622,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.2404,
+      "step": 239
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.4828258709239939,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.24,
+      "step": 240
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.5795204158855689,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.2274,
+      "step": 241
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.5853029607020683,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.26,
+      "step": 242
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.4828643470497328,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.2544,
+      "step": 243
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.4329500194495083,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.2206,
+      "step": 244
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.5165455521545157,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.2417,
+      "step": 245
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.5275383950137368,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.2449,
+      "step": 246
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.39340503477267014,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2199,
+      "step": 247
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.4317866921063053,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.2351,
+      "step": 248
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.4443169086208162,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.2556,
+      "step": 249
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.4005458806639983,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.2365,
+      "step": 250
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.4181384968270446,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.245,
+      "step": 251
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.37836401384344864,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.2344,
+      "step": 252
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.3713195178763742,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.2465,
+      "step": 253
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.36043895430414663,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.2338,
+      "step": 254
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.34938091694810997,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.228,
+      "step": 255
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.3625745323291279,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.2532,
+      "step": 256
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.32852523370479425,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.2086,
+      "step": 257
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.3579171159966551,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.2375,
+      "step": 258
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.33573195229266234,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.2216,
+      "step": 259
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.3678272441344075,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2438,
+      "step": 260
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.33596560147428767,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.2449,
+      "step": 261
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.35108960586375987,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.2298,
+      "step": 262
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.3505624560113376,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.2163,
+      "step": 263
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.3504767757115922,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.2189,
+      "step": 264
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.34430891718788276,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.2372,
+      "step": 265
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.33778027498109425,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.2133,
+      "step": 266
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.33636617129539925,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.2301,
+      "step": 267
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.3504754683249517,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.2486,
+      "step": 268
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.34795897812025656,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.2141,
+      "step": 269
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.36580061553563076,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.2304,
+      "step": 270
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.3701212255172201,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.2431,
+      "step": 271
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.3405446548694439,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.2231,
+      "step": 272
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.3553166437308296,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.219,
+      "step": 273
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.37282631650361,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.2347,
+      "step": 274
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.32702491987456384,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.2257,
+      "step": 275
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.34275469394870306,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.2367,
+      "step": 276
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.35933691031419707,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.2415,
+      "step": 277
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.3571317614246153,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.2333,
+      "step": 278
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.34359614641158914,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.2235,
+      "step": 279
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.3521601969192374,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.2573,
+      "step": 280
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.3283557605845052,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.2111,
+      "step": 281
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.3608233567186079,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.2451,
+      "step": 282
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.33885700019167236,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.2345,
+      "step": 283
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.3558837243949352,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.2463,
+      "step": 284
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.3417347616378375,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.2451,
+      "step": 285
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.3265691533149346,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2292,
+      "step": 286
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.35543632739891745,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.2499,
+      "step": 287
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.32721794505380586,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.2284,
+      "step": 288
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.3473173577929293,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.2477,
+      "step": 289
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.3267429658875211,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.2256,
+      "step": 290
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.3308565691014178,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.2525,
+      "step": 291
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.32775713637571846,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.2286,
+      "step": 292
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.3230110993671641,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.2281,
+      "step": 293
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.34319449892245124,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.2574,
+      "step": 294
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.3052306427404134,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.2053,
+      "step": 295
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.3389886313628059,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.2561,
+      "step": 296
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.3096027527849823,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.2167,
+      "step": 297
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.3320520175361268,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.2372,
+      "step": 298
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.3406340780226723,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2545,
+      "step": 299
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.2988000824396719,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.2171,
+      "step": 300
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.30931376951446365,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.2231,
+      "step": 301
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.33622255017983976,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.2371,
+      "step": 302
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.32360161834108225,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.2348,
+      "step": 303
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.30292782196099155,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.2048,
+      "step": 304
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.3090810070307293,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.2562,
+      "step": 305
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.29309856186061733,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.225,
+      "step": 306
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.31956453050611927,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.2279,
+      "step": 307
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.3112087941940966,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.2219,
+      "step": 308
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.3153817831789216,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.2354,
+      "step": 309
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.31372150880773386,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.2422,
+      "step": 310
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.32510096573711394,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.2316,
+      "step": 311
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.29906942590577396,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2275,
+      "step": 312
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 1.2139953461603725,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.3807,
+      "step": 313
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.5237410389498884,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.183,
+      "step": 314
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.4427029532721314,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.1775,
+      "step": 315
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.34230370359902346,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.1509,
+      "step": 316
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.29601757434559484,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.1573,
+      "step": 317
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.3814257012353634,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.1619,
+      "step": 318
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.5041931343111281,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.1693,
+      "step": 319
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.5823550587186476,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.1767,
+      "step": 320
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.4625808142431024,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.1588,
+      "step": 321
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.37116127475914895,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.1782,
+      "step": 322
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.3294610298238965,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.1661,
+      "step": 323
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.300193661172271,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.1424,
+      "step": 324
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.4038054890315199,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1734,
+      "step": 325
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.3513663152569111,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.1608,
+      "step": 326
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.39787351845147634,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.158,
+      "step": 327
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.33655397686316324,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.1596,
+      "step": 328
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.318527802100109,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.1704,
+      "step": 329
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.2974865136489592,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.1641,
+      "step": 330
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.3016788088406415,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.1561,
+      "step": 331
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.32849297882511297,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.1561,
+      "step": 332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.33069183157158233,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.1765,
+      "step": 333
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.3540783313652625,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.1731,
+      "step": 334
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.3086978594001584,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.1558,
+      "step": 335
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.29331458561932094,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.1557,
+      "step": 336
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.3085758487601604,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.1572,
+      "step": 337
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.2906868443213144,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1557,
+      "step": 338
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.2970918857148492,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.1514,
+      "step": 339
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.3173460879970613,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.1704,
+      "step": 340
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.2961358934681771,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.1469,
+      "step": 341
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.3130008308479644,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.1632,
+      "step": 342
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.30139540096728357,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.1694,
+      "step": 343
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.30497921291683516,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.1515,
+      "step": 344
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.2962093457026487,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.1549,
+      "step": 345
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.29145197933072053,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.1691,
+      "step": 346
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.2805215505614425,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.16,
+      "step": 347
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.2716802696786645,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.1572,
+      "step": 348
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.29116806810849954,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.1677,
+      "step": 349
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.30403120086929486,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.1656,
+      "step": 350
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.2888776980210721,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1575,
+      "step": 351
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.28436489114232044,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.1512,
+      "step": 352
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.2945919286625033,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.1719,
+      "step": 353
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.2915528020141999,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.1525,
+      "step": 354
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.29895984099909056,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.163,
+      "step": 355
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.30792150165168203,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.1621,
+      "step": 356
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.2909817524180524,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.1517,
+      "step": 357
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.2950182202862721,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.1579,
+      "step": 358
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.2786025132207038,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.1589,
+      "step": 359
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.28027422258603657,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.1516,
+      "step": 360
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.2751820247938203,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.1532,
+      "step": 361
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.2734147517610915,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.1528,
+      "step": 362
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.28075210904691406,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.1584,
+      "step": 363
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.2887601716240021,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1637,
+      "step": 364
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.277925460841827,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.154,
+      "step": 365
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.2844367789603032,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.1517,
+      "step": 366
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.28931672628419347,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.1613,
+      "step": 367
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.2860034658051635,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.1563,
+      "step": 368
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.3089159326788425,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.1612,
+      "step": 369
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.29457815167214546,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.168,
+      "step": 370
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.27208637514557227,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.1472,
+      "step": 371
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.27590131884272984,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.1586,
+      "step": 372
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.28602629295970033,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.1578,
+      "step": 373
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.28428689397441836,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.1497,
+      "step": 374
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.2836450607351416,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.1584,
+      "step": 375
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.2631863358532374,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.1423,
+      "step": 376
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.2746300466010014,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1533,
+      "step": 377
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.2916563865381194,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.1534,
+      "step": 378
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.2801924376684902,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.1624,
+      "step": 379
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.3066114462534377,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.1641,
+      "step": 380
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.28936538978957893,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.1585,
+      "step": 381
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.29197481999400926,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.1606,
+      "step": 382
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.27877119220344143,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.1486,
+      "step": 383
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.29040742011215104,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.172,
+      "step": 384
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.27366147923175643,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.1655,
+      "step": 385
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.2928012431572296,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.178,
+      "step": 386
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.28735601580788767,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.1551,
+      "step": 387
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.2946376966049083,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.1647,
+      "step": 388
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.2925440423275787,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.1696,
+      "step": 389
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.2681299045611345,
+      "learning_rate": 0.0,
+      "loss": 0.1485,
+      "step": 390
+    },
+    {
+      "epoch": 4.992,
+      "step": 390,
+      "total_flos": 7.862516792374067e+17,
+      "train_loss": 0.3952271275795423,
+      "train_runtime": 13928.0051,
+      "train_samples_per_second": 3.59,
+      "train_steps_per_second": 0.028
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 390,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.862516792374067e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3a92a830b2992e48944ac103901980dec21af842486e707372b36997a61fe7d
+size 7352