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: c1_math_0d_4s_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. -->
+
+# c1_math_0d_4s_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/c1_math_0d_4s_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": 1.8773457441390592e+18,
+    "train_loss": 0.3770592579474816,
+    "train_runtime": 19616.6723,
+    "train_samples_per_second": 2.549,
+    "train_steps_per_second": 0.02
+}

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 sha256:90ce0667da84368e284cc15904f975bc7057d149161b374a362c803b355ce097
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0deb9627b71a6d3cd75160b0f7460b4355592cad5886a091e356f46b494284b
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fd192334fb81eb0d33b3d42a294d12168add271cb82837b313666f4e3318c56
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9416328786a524bf140f7d8d220163ccdd41b932cf0b6cdb1cac69f1094affdb
+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_00-18-50_jpbot-001-15.jupiter.internal/events.out.tfevents.1745705984.jpbot-001-15.jupiter.internal.1552287.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb1ba3c8fae455c8817be933a11d8380f82e692bb0c7bc6d3b7fd3e6641b7c63
+size 87654

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 00:18:50", "end_time": "2025-04-27 05:46:48"}

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": 1.8773457441390592e+18,
+    "train_loss": 0.3770592579474816,
+    "train_runtime": 19616.6723,
+    "train_samples_per_second": 2.549,
+    "train_steps_per_second": 0.02
+}

trainer_log.jsonl

@@ -0,0 +1,391 @@
+{"current_steps": 1, "total_steps": 390, "loss": 0.8108, "lr": 1.0256410256410257e-06, "epoch": 0.0128, "percentage": 0.26, "elapsed_time": "0:00:58", "remaining_time": "6:22:10"}
+{"current_steps": 2, "total_steps": 390, "loss": 0.7847, "lr": 2.0512820512820513e-06, "epoch": 0.0256, "percentage": 0.51, "elapsed_time": "0:01:49", "remaining_time": "5:52:29"}
+{"current_steps": 3, "total_steps": 390, "loss": 0.8192, "lr": 3.0769230769230774e-06, "epoch": 0.0384, "percentage": 0.77, "elapsed_time": "0:02:39", "remaining_time": "5:43:26"}
+{"current_steps": 4, "total_steps": 390, "loss": 0.7829, "lr": 4.102564102564103e-06, "epoch": 0.0512, "percentage": 1.03, "elapsed_time": "0:03:33", "remaining_time": "5:42:57"}
+{"current_steps": 5, "total_steps": 390, "loss": 0.7687, "lr": 5.128205128205128e-06, "epoch": 0.064, "percentage": 1.28, "elapsed_time": "0:04:24", "remaining_time": "5:39:18"}
+{"current_steps": 6, "total_steps": 390, "loss": 0.7336, "lr": 6.153846153846155e-06, "epoch": 0.0768, "percentage": 1.54, "elapsed_time": "0:05:13", "remaining_time": "5:34:40"}
+{"current_steps": 7, "total_steps": 390, "loss": 0.7165, "lr": 7.17948717948718e-06, "epoch": 0.0896, "percentage": 1.79, "elapsed_time": "0:06:06", "remaining_time": "5:33:48"}
+{"current_steps": 8, "total_steps": 390, "loss": 0.7287, "lr": 8.205128205128205e-06, "epoch": 0.1024, "percentage": 2.05, "elapsed_time": "0:06:55", "remaining_time": "5:30:51"}
+{"current_steps": 9, "total_steps": 390, "loss": 0.7252, "lr": 9.230769230769232e-06, "epoch": 0.1152, "percentage": 2.31, "elapsed_time": "0:07:44", "remaining_time": "5:27:44"}
+{"current_steps": 10, "total_steps": 390, "loss": 0.6916, "lr": 1.0256410256410256e-05, "epoch": 0.128, "percentage": 2.56, "elapsed_time": "0:08:37", "remaining_time": "5:27:57"}
+{"current_steps": 11, "total_steps": 390, "loss": 0.6714, "lr": 1.1282051282051283e-05, "epoch": 0.1408, "percentage": 2.82, "elapsed_time": "0:09:28", "remaining_time": "5:26:32"}
+{"current_steps": 12, "total_steps": 390, "loss": 0.6509, "lr": 1.230769230769231e-05, "epoch": 0.1536, "percentage": 3.08, "elapsed_time": "0:10:17", "remaining_time": "5:24:06"}
+{"current_steps": 13, "total_steps": 390, "loss": 0.6462, "lr": 1.3333333333333333e-05, "epoch": 0.1664, "percentage": 3.33, "elapsed_time": "0:11:06", "remaining_time": "5:21:55"}
+{"current_steps": 14, "total_steps": 390, "loss": 0.6075, "lr": 1.435897435897436e-05, "epoch": 0.1792, "percentage": 3.59, "elapsed_time": "0:11:55", "remaining_time": "5:20:09"}
+{"current_steps": 15, "total_steps": 390, "loss": 0.6198, "lr": 1.5384615384615387e-05, "epoch": 0.192, "percentage": 3.85, "elapsed_time": "0:12:44", "remaining_time": "5:18:41"}
+{"current_steps": 16, "total_steps": 390, "loss": 0.5743, "lr": 1.641025641025641e-05, "epoch": 0.2048, "percentage": 4.1, "elapsed_time": "0:13:35", "remaining_time": "5:17:30"}
+{"current_steps": 17, "total_steps": 390, "loss": 0.5916, "lr": 1.7435897435897438e-05, "epoch": 0.2176, "percentage": 4.36, "elapsed_time": "0:14:24", "remaining_time": "5:16:18"}
+{"current_steps": 18, "total_steps": 390, "loss": 0.5856, "lr": 1.8461538461538465e-05, "epoch": 0.2304, "percentage": 4.62, "elapsed_time": "0:15:15", "remaining_time": "5:15:13"}
+{"current_steps": 19, "total_steps": 390, "loss": 0.5385, "lr": 1.9487179487179488e-05, "epoch": 0.2432, "percentage": 4.87, "elapsed_time": "0:16:06", "remaining_time": "5:14:24"}
+{"current_steps": 20, "total_steps": 390, "loss": 0.568, "lr": 2.0512820512820512e-05, "epoch": 0.256, "percentage": 5.13, "elapsed_time": "0:16:54", "remaining_time": "5:12:57"}
+{"current_steps": 21, "total_steps": 390, "loss": 0.5631, "lr": 2.153846153846154e-05, "epoch": 0.2688, "percentage": 5.38, "elapsed_time": "0:17:47", "remaining_time": "5:12:45"}
+{"current_steps": 22, "total_steps": 390, "loss": 0.5431, "lr": 2.2564102564102566e-05, "epoch": 0.2816, "percentage": 5.64, "elapsed_time": "0:18:38", "remaining_time": "5:11:44"}
+{"current_steps": 23, "total_steps": 390, "loss": 0.5403, "lr": 2.3589743589743593e-05, "epoch": 0.2944, "percentage": 5.9, "elapsed_time": "0:19:27", "remaining_time": "5:10:30"}
+{"current_steps": 24, "total_steps": 390, "loss": 0.5346, "lr": 2.461538461538462e-05, "epoch": 0.3072, "percentage": 6.15, "elapsed_time": "0:20:18", "remaining_time": "5:09:47"}
+{"current_steps": 25, "total_steps": 390, "loss": 0.5391, "lr": 2.5641025641025646e-05, "epoch": 0.32, "percentage": 6.41, "elapsed_time": "0:21:08", "remaining_time": "5:08:46"}
+{"current_steps": 26, "total_steps": 390, "loss": 0.5183, "lr": 2.6666666666666667e-05, "epoch": 0.3328, "percentage": 6.67, "elapsed_time": "0:21:56", "remaining_time": "5:07:17"}
+{"current_steps": 27, "total_steps": 390, "loss": 0.5211, "lr": 2.7692307692307694e-05, "epoch": 0.3456, "percentage": 6.92, "elapsed_time": "0:22:43", "remaining_time": "5:05:35"}
+{"current_steps": 28, "total_steps": 390, "loss": 0.5116, "lr": 2.871794871794872e-05, "epoch": 0.3584, "percentage": 7.18, "elapsed_time": "0:23:33", "remaining_time": "5:04:30"}
+{"current_steps": 29, "total_steps": 390, "loss": 0.5252, "lr": 2.9743589743589747e-05, "epoch": 0.3712, "percentage": 7.44, "elapsed_time": "0:24:22", "remaining_time": "5:03:21"}
+{"current_steps": 30, "total_steps": 390, "loss": 0.5117, "lr": 3.0769230769230774e-05, "epoch": 0.384, "percentage": 7.69, "elapsed_time": "0:25:08", "remaining_time": "5:01:39"}
+{"current_steps": 31, "total_steps": 390, "loss": 0.5165, "lr": 3.1794871794871795e-05, "epoch": 0.3968, "percentage": 7.95, "elapsed_time": "0:25:58", "remaining_time": "5:00:43"}
+{"current_steps": 32, "total_steps": 390, "loss": 0.5302, "lr": 3.282051282051282e-05, "epoch": 0.4096, "percentage": 8.21, "elapsed_time": "0:26:47", "remaining_time": "4:59:42"}
+{"current_steps": 33, "total_steps": 390, "loss": 0.5131, "lr": 3.384615384615385e-05, "epoch": 0.4224, "percentage": 8.46, "elapsed_time": "0:27:39", "remaining_time": "4:59:17"}
+{"current_steps": 34, "total_steps": 390, "loss": 0.4912, "lr": 3.4871794871794875e-05, "epoch": 0.4352, "percentage": 8.72, "elapsed_time": "0:28:29", "remaining_time": "4:58:16"}
+{"current_steps": 35, "total_steps": 390, "loss": 0.5064, "lr": 3.58974358974359e-05, "epoch": 0.448, "percentage": 8.97, "elapsed_time": "0:29:18", "remaining_time": "4:57:12"}
+{"current_steps": 36, "total_steps": 390, "loss": 0.5085, "lr": 3.692307692307693e-05, "epoch": 0.4608, "percentage": 9.23, "elapsed_time": "0:30:05", "remaining_time": "4:55:51"}
+{"current_steps": 37, "total_steps": 390, "loss": 0.5093, "lr": 3.794871794871795e-05, "epoch": 0.4736, "percentage": 9.49, "elapsed_time": "0:30:54", "remaining_time": "4:54:51"}
+{"current_steps": 38, "total_steps": 390, "loss": 0.5151, "lr": 3.8974358974358976e-05, "epoch": 0.4864, "percentage": 9.74, "elapsed_time": "0:31:45", "remaining_time": "4:54:12"}
+{"current_steps": 39, "total_steps": 390, "loss": 0.4846, "lr": 4e-05, "epoch": 0.4992, "percentage": 10.0, "elapsed_time": "0:32:36", "remaining_time": "4:53:26"}
+{"current_steps": 40, "total_steps": 390, "loss": 0.4902, "lr": 3.9999198907597046e-05, "epoch": 0.512, "percentage": 10.26, "elapsed_time": "0:33:27", "remaining_time": "4:52:44"}
+{"current_steps": 41, "total_steps": 390, "loss": 0.499, "lr": 3.9996795694563096e-05, "epoch": 0.5248, "percentage": 10.51, "elapsed_time": "0:34:16", "remaining_time": "4:51:44"}
+{"current_steps": 42, "total_steps": 390, "loss": 0.4961, "lr": 3.999279055341771e-05, "epoch": 0.5376, "percentage": 10.77, "elapsed_time": "0:35:08", "remaining_time": "4:51:10"}
+{"current_steps": 43, "total_steps": 390, "loss": 0.5058, "lr": 3.998718380500971e-05, "epoch": 0.5504, "percentage": 11.03, "elapsed_time": "0:35:57", "remaining_time": "4:50:13"}
+{"current_steps": 44, "total_steps": 390, "loss": 0.4806, "lr": 3.997997589849145e-05, "epoch": 0.5632, "percentage": 11.28, "elapsed_time": "0:36:47", "remaining_time": "4:49:16"}
+{"current_steps": 45, "total_steps": 390, "loss": 0.5032, "lr": 3.9971167411282835e-05, "epoch": 0.576, "percentage": 11.54, "elapsed_time": "0:37:35", "remaining_time": "4:48:14"}
+{"current_steps": 46, "total_steps": 390, "loss": 0.5021, "lr": 3.99607590490251e-05, "epoch": 0.5888, "percentage": 11.79, "elapsed_time": "0:38:26", "remaining_time": "4:47:24"}
+{"current_steps": 47, "total_steps": 390, "loss": 0.4857, "lr": 3.9948751645524235e-05, "epoch": 0.6016, "percentage": 12.05, "elapsed_time": "0:39:57", "remaining_time": "4:51:36"}
+{"current_steps": 48, "total_steps": 390, "loss": 0.4747, "lr": 3.9935146162684206e-05, "epoch": 0.6144, "percentage": 12.31, "elapsed_time": "0:40:44", "remaining_time": "4:50:17"}
+{"current_steps": 49, "total_steps": 390, "loss": 0.4726, "lr": 3.9919943690429906e-05, "epoch": 0.6272, "percentage": 12.56, "elapsed_time": "0:41:30", "remaining_time": "4:48:54"}
+{"current_steps": 50, "total_steps": 390, "loss": 0.4884, "lr": 3.9903145446619837e-05, "epoch": 0.64, "percentage": 12.82, "elapsed_time": "0:42:19", "remaining_time": "4:47:49"}
+{"current_steps": 51, "total_steps": 390, "loss": 0.4906, "lr": 3.9884752776948564e-05, "epoch": 0.6528, "percentage": 13.08, "elapsed_time": "0:43:11", "remaining_time": "4:47:03"}
+{"current_steps": 52, "total_steps": 390, "loss": 0.4821, "lr": 3.9864767154838864e-05, "epoch": 0.6656, "percentage": 13.33, "elapsed_time": "0:44:02", "remaining_time": "4:46:14"}
+{"current_steps": 53, "total_steps": 390, "loss": 0.4823, "lr": 3.9843190181323744e-05, "epoch": 0.6784, "percentage": 13.59, "elapsed_time": "0:44:51", "remaining_time": "4:45:12"}
+{"current_steps": 54, "total_steps": 390, "loss": 0.4719, "lr": 3.982002358491817e-05, "epoch": 0.6912, "percentage": 13.85, "elapsed_time": "0:45:40", "remaining_time": "4:44:11"}
+{"current_steps": 55, "total_steps": 390, "loss": 0.4665, "lr": 3.979526922148058e-05, "epoch": 0.704, "percentage": 14.1, "elapsed_time": "0:46:26", "remaining_time": "4:42:51"}
+{"current_steps": 56, "total_steps": 390, "loss": 0.4743, "lr": 3.9768929074064206e-05, "epoch": 0.7168, "percentage": 14.36, "elapsed_time": "0:47:18", "remaining_time": "4:42:07"}
+{"current_steps": 57, "total_steps": 390, "loss": 0.4723, "lr": 3.9741005252758255e-05, "epoch": 0.7296, "percentage": 14.62, "elapsed_time": "0:48:07", "remaining_time": "4:41:08"}
+{"current_steps": 58, "total_steps": 390, "loss": 0.4726, "lr": 3.971149999451886e-05, "epoch": 0.7424, "percentage": 14.87, "elapsed_time": "0:48:56", "remaining_time": "4:40:08"}
+{"current_steps": 59, "total_steps": 390, "loss": 0.4671, "lr": 3.9680415662989806e-05, "epoch": 0.7552, "percentage": 15.13, "elapsed_time": "0:49:47", "remaining_time": "4:39:19"}
+{"current_steps": 60, "total_steps": 390, "loss": 0.466, "lr": 3.9647754748313294e-05, "epoch": 0.768, "percentage": 15.38, "elapsed_time": "0:50:37", "remaining_time": "4:38:27"}
+{"current_steps": 61, "total_steps": 390, "loss": 0.4741, "lr": 3.96135198669304e-05, "epoch": 0.7808, "percentage": 15.64, "elapsed_time": "0:51:28", "remaining_time": "4:37:37"}
+{"current_steps": 62, "total_steps": 390, "loss": 0.4601, "lr": 3.957771376137144e-05, "epoch": 0.7936, "percentage": 15.9, "elapsed_time": "0:52:16", "remaining_time": "4:36:30"}
+{"current_steps": 63, "total_steps": 390, "loss": 0.4706, "lr": 3.954033930003634e-05, "epoch": 0.8064, "percentage": 16.15, "elapsed_time": "0:53:06", "remaining_time": "4:35:41"}
+{"current_steps": 64, "total_steps": 390, "loss": 0.467, "lr": 3.9501399476964806e-05, "epoch": 0.8192, "percentage": 16.41, "elapsed_time": "0:53:54", "remaining_time": "4:34:36"}
+{"current_steps": 65, "total_steps": 390, "loss": 0.4695, "lr": 3.946089741159648e-05, "epoch": 0.832, "percentage": 16.67, "elapsed_time": "0:54:45", "remaining_time": "4:33:49"}
+{"current_steps": 66, "total_steps": 390, "loss": 0.4748, "lr": 3.9418836348521045e-05, "epoch": 0.8448, "percentage": 16.92, "elapsed_time": "0:55:37", "remaining_time": "4:33:03"}
+{"current_steps": 67, "total_steps": 390, "loss": 0.4537, "lr": 3.937521965721831e-05, "epoch": 0.8576, "percentage": 17.18, "elapsed_time": "0:56:29", "remaining_time": "4:32:19"}
+{"current_steps": 68, "total_steps": 390, "loss": 0.4586, "lr": 3.933005083178828e-05, "epoch": 0.8704, "percentage": 17.44, "elapsed_time": "0:57:17", "remaining_time": "4:31:18"}
+{"current_steps": 69, "total_steps": 390, "loss": 0.4724, "lr": 3.928333349067125e-05, "epoch": 0.8832, "percentage": 17.69, "elapsed_time": "0:58:07", "remaining_time": "4:30:25"}
+{"current_steps": 70, "total_steps": 390, "loss": 0.4674, "lr": 3.923507137635792e-05, "epoch": 0.896, "percentage": 17.95, "elapsed_time": "0:58:58", "remaining_time": "4:29:36"}
+{"current_steps": 71, "total_steps": 390, "loss": 0.4508, "lr": 3.9185268355089606e-05, "epoch": 0.9088, "percentage": 18.21, "elapsed_time": "0:59:46", "remaining_time": "4:28:34"}
+{"current_steps": 72, "total_steps": 390, "loss": 0.4655, "lr": 3.913392841654851e-05, "epoch": 0.9216, "percentage": 18.46, "elapsed_time": "1:00:39", "remaining_time": "4:27:52"}
+{"current_steps": 73, "total_steps": 390, "loss": 0.4661, "lr": 3.9081055673538093e-05, "epoch": 0.9344, "percentage": 18.72, "elapsed_time": "1:01:28", "remaining_time": "4:26:58"}
+{"current_steps": 74, "total_steps": 390, "loss": 0.4803, "lr": 3.902665436165364e-05, "epoch": 0.9472, "percentage": 18.97, "elapsed_time": "1:02:19", "remaining_time": "4:26:10"}
+{"current_steps": 75, "total_steps": 390, "loss": 0.4639, "lr": 3.897072883894291e-05, "epoch": 0.96, "percentage": 19.23, "elapsed_time": "1:03:11", "remaining_time": "4:25:25"}
+{"current_steps": 76, "total_steps": 390, "loss": 0.4708, "lr": 3.8913283585557054e-05, "epoch": 0.9728, "percentage": 19.49, "elapsed_time": "1:04:00", "remaining_time": "4:24:28"}
+{"current_steps": 77, "total_steps": 390, "loss": 0.4524, "lr": 3.885432320339167e-05, "epoch": 0.9856, "percentage": 19.74, "elapsed_time": "1:04:52", "remaining_time": "4:23:41"}
+{"current_steps": 78, "total_steps": 390, "loss": 0.4643, "lr": 3.879385241571817e-05, "epoch": 0.9984, "percentage": 20.0, "elapsed_time": "1:05:43", "remaining_time": "4:22:53"}
+{"current_steps": 79, "total_steps": 390, "loss": 0.8012, "lr": 3.873187606680543e-05, "epoch": 1.0112, "percentage": 20.26, "elapsed_time": "1:06:59", "remaining_time": "4:23:42"}
+{"current_steps": 80, "total_steps": 390, "loss": 0.4398, "lr": 3.866839912153168e-05, "epoch": 1.024, "percentage": 20.51, "elapsed_time": "1:07:50", "remaining_time": "4:22:52"}
+{"current_steps": 81, "total_steps": 390, "loss": 0.4121, "lr": 3.860342666498677e-05, "epoch": 1.0368, "percentage": 20.77, "elapsed_time": "1:08:39", "remaining_time": "4:21:53"}
+{"current_steps": 82, "total_steps": 390, "loss": 0.4285, "lr": 3.853696390206484e-05, "epoch": 1.0496, "percentage": 21.03, "elapsed_time": "1:09:29", "remaining_time": "4:20:59"}
+{"current_steps": 83, "total_steps": 390, "loss": 0.4184, "lr": 3.846901615704734e-05, "epoch": 1.0624, "percentage": 21.28, "elapsed_time": "1:10:19", "remaining_time": "4:20:08"}
+{"current_steps": 84, "total_steps": 390, "loss": 0.4219, "lr": 3.839958887317649e-05, "epoch": 1.0752, "percentage": 21.54, "elapsed_time": "1:11:08", "remaining_time": "4:19:08"}
+{"current_steps": 85, "total_steps": 390, "loss": 0.4266, "lr": 3.832868761221926e-05, "epoch": 1.088, "percentage": 21.79, "elapsed_time": "1:11:58", "remaining_time": "4:18:17"}
+{"current_steps": 86, "total_steps": 390, "loss": 0.4306, "lr": 3.825631805402182e-05, "epoch": 1.1008, "percentage": 22.05, "elapsed_time": "1:12:46", "remaining_time": "4:17:14"}
+{"current_steps": 87, "total_steps": 390, "loss": 0.4363, "lr": 3.818248599605448e-05, "epoch": 1.1136, "percentage": 22.31, "elapsed_time": "1:13:37", "remaining_time": "4:16:25"}
+{"current_steps": 88, "total_steps": 390, "loss": 0.4198, "lr": 3.810719735294731e-05, "epoch": 1.1264, "percentage": 22.56, "elapsed_time": "1:14:25", "remaining_time": "4:15:25"}
+{"current_steps": 89, "total_steps": 390, "loss": 0.427, "lr": 3.8030458156016326e-05, "epoch": 1.1392, "percentage": 22.82, "elapsed_time": "1:15:11", "remaining_time": "4:14:19"}
+{"current_steps": 90, "total_steps": 390, "loss": 0.434, "lr": 3.795227455278029e-05, "epoch": 1.152, "percentage": 23.08, "elapsed_time": "1:16:03", "remaining_time": "4:13:30"}
+{"current_steps": 91, "total_steps": 390, "loss": 0.4081, "lr": 3.787265280646825e-05, "epoch": 1.1648, "percentage": 23.33, "elapsed_time": "1:16:54", "remaining_time": "4:12:42"}
+{"current_steps": 92, "total_steps": 390, "loss": 0.4277, "lr": 3.7791599295517825e-05, "epoch": 1.1776, "percentage": 23.59, "elapsed_time": "1:17:44", "remaining_time": "4:11:50"}
+{"current_steps": 93, "total_steps": 390, "loss": 0.4176, "lr": 3.7709120513064196e-05, "epoch": 1.1904, "percentage": 23.85, "elapsed_time": "1:18:35", "remaining_time": "4:10:59"}
+{"current_steps": 94, "total_steps": 390, "loss": 0.4153, "lr": 3.762522306641998e-05, "epoch": 1.2032, "percentage": 24.1, "elapsed_time": "1:19:23", "remaining_time": "4:10:01"}
+{"current_steps": 95, "total_steps": 390, "loss": 0.4172, "lr": 3.7539913676545874e-05, "epoch": 1.216, "percentage": 24.36, "elapsed_time": "1:20:11", "remaining_time": "4:09:01"}
+{"current_steps": 96, "total_steps": 390, "loss": 0.4102, "lr": 3.745319917751229e-05, "epoch": 1.2288000000000001, "percentage": 24.62, "elapsed_time": "1:21:03", "remaining_time": "4:08:13"}
+{"current_steps": 97, "total_steps": 390, "loss": 0.4161, "lr": 3.736508651595188e-05, "epoch": 1.2416, "percentage": 24.87, "elapsed_time": "1:21:54", "remaining_time": "4:07:23"}
+{"current_steps": 98, "total_steps": 390, "loss": 0.4197, "lr": 3.727558275050301e-05, "epoch": 1.2544, "percentage": 25.13, "elapsed_time": "1:22:42", "remaining_time": "4:06:26"}
+{"current_steps": 99, "total_steps": 390, "loss": 0.4048, "lr": 3.718469505124434e-05, "epoch": 1.2671999999999999, "percentage": 25.38, "elapsed_time": "1:23:32", "remaining_time": "4:05:33"}
+{"current_steps": 100, "total_steps": 390, "loss": 0.419, "lr": 3.709243069912041e-05, "epoch": 1.28, "percentage": 25.64, "elapsed_time": "1:24:19", "remaining_time": "4:04:32"}
+{"current_steps": 101, "total_steps": 390, "loss": 0.4417, "lr": 3.699879708535838e-05, "epoch": 1.2928, "percentage": 25.9, "elapsed_time": "1:25:08", "remaining_time": "4:03:37"}
+{"current_steps": 102, "total_steps": 390, "loss": 0.4215, "lr": 3.69038017108759e-05, "epoch": 1.3056, "percentage": 26.15, "elapsed_time": "1:25:58", "remaining_time": "4:02:43"}
+{"current_steps": 103, "total_steps": 390, "loss": 0.4201, "lr": 3.680745218568026e-05, "epoch": 1.3184, "percentage": 26.41, "elapsed_time": "1:26:47", "remaining_time": "4:01:51"}
+{"current_steps": 104, "total_steps": 390, "loss": 0.4347, "lr": 3.6709756228258735e-05, "epoch": 1.3312, "percentage": 26.67, "elapsed_time": "1:27:37", "remaining_time": "4:00:57"}
+{"current_steps": 105, "total_steps": 390, "loss": 0.4206, "lr": 3.6610721664960236e-05, "epoch": 1.3439999999999999, "percentage": 26.92, "elapsed_time": "1:28:28", "remaining_time": "4:00:09"}
+{"current_steps": 106, "total_steps": 390, "loss": 0.3881, "lr": 3.65103564293684e-05, "epoch": 1.3568, "percentage": 27.18, "elapsed_time": "1:29:19", "remaining_time": "3:59:19"}
+{"current_steps": 107, "total_steps": 390, "loss": 0.4118, "lr": 3.640866856166601e-05, "epoch": 1.3696, "percentage": 27.44, "elapsed_time": "1:30:09", "remaining_time": "3:58:28"}
+{"current_steps": 108, "total_steps": 390, "loss": 0.416, "lr": 3.6305666207990886e-05, "epoch": 1.3824, "percentage": 27.69, "elapsed_time": "1:30:59", "remaining_time": "3:57:35"}
+{"current_steps": 109, "total_steps": 390, "loss": 0.4135, "lr": 3.6201357619783336e-05, "epoch": 1.3952, "percentage": 27.95, "elapsed_time": "1:31:48", "remaining_time": "3:56:42"}
+{"current_steps": 110, "total_steps": 390, "loss": 0.417, "lr": 3.609575115312511e-05, "epoch": 1.408, "percentage": 28.21, "elapsed_time": "1:32:37", "remaining_time": "3:55:46"}
+{"current_steps": 111, "total_steps": 390, "loss": 0.4111, "lr": 3.598885526807003e-05, "epoch": 1.4208, "percentage": 28.46, "elapsed_time": "1:33:27", "remaining_time": "3:54:54"}
+{"current_steps": 112, "total_steps": 390, "loss": 0.4036, "lr": 3.5880678527966224e-05, "epoch": 1.4336, "percentage": 28.72, "elapsed_time": "1:34:18", "remaining_time": "3:54:05"}
+{"current_steps": 113, "total_steps": 390, "loss": 0.4207, "lr": 3.577122959877017e-05, "epoch": 1.4464000000000001, "percentage": 28.97, "elapsed_time": "1:35:07", "remaining_time": "3:53:11"}
+{"current_steps": 114, "total_steps": 390, "loss": 0.4089, "lr": 3.566051724835245e-05, "epoch": 1.4592, "percentage": 29.23, "elapsed_time": "1:35:58", "remaining_time": "3:52:22"}
+{"current_steps": 115, "total_steps": 390, "loss": 0.4164, "lr": 3.554855034579532e-05, "epoch": 1.472, "percentage": 29.49, "elapsed_time": "1:36:47", "remaining_time": "3:51:28"}
+{"current_steps": 116, "total_steps": 390, "loss": 0.3981, "lr": 3.5435337860682304e-05, "epoch": 1.4848, "percentage": 29.74, "elapsed_time": "1:37:34", "remaining_time": "3:50:28"}
+{"current_steps": 117, "total_steps": 390, "loss": 0.4196, "lr": 3.532088886237956e-05, "epoch": 1.4976, "percentage": 30.0, "elapsed_time": "1:38:26", "remaining_time": "3:49:40"}
+{"current_steps": 118, "total_steps": 390, "loss": 0.4107, "lr": 3.520521251930941e-05, "epoch": 1.5104, "percentage": 30.26, "elapsed_time": "1:39:16", "remaining_time": "3:48:50"}
+{"current_steps": 119, "total_steps": 390, "loss": 0.4122, "lr": 3.5088318098215805e-05, "epoch": 1.5232, "percentage": 30.51, "elapsed_time": "1:40:05", "remaining_time": "3:47:56"}
+{"current_steps": 120, "total_steps": 390, "loss": 0.4073, "lr": 3.497021496342203e-05, "epoch": 1.536, "percentage": 30.77, "elapsed_time": "1:40:54", "remaining_time": "3:47:02"}
+{"current_steps": 121, "total_steps": 390, "loss": 0.4252, "lr": 3.485091257608047e-05, "epoch": 1.5488, "percentage": 31.03, "elapsed_time": "1:41:45", "remaining_time": "3:46:13"}
+{"current_steps": 122, "total_steps": 390, "loss": 0.4122, "lr": 3.473042049341474e-05, "epoch": 1.5615999999999999, "percentage": 31.28, "elapsed_time": "1:42:36", "remaining_time": "3:45:23"}
+{"current_steps": 123, "total_steps": 390, "loss": 0.416, "lr": 3.4608748367954064e-05, "epoch": 1.5744, "percentage": 31.54, "elapsed_time": "1:43:24", "remaining_time": "3:44:28"}
+{"current_steps": 124, "total_steps": 390, "loss": 0.4205, "lr": 3.4485905946759965e-05, "epoch": 1.5872000000000002, "percentage": 31.79, "elapsed_time": "1:44:16", "remaining_time": "3:43:41"}
+{"current_steps": 125, "total_steps": 390, "loss": 0.4185, "lr": 3.4361903070645484e-05, "epoch": 1.6, "percentage": 32.05, "elapsed_time": "1:45:08", "remaining_time": "3:42:54"}
+{"current_steps": 126, "total_steps": 390, "loss": 0.4103, "lr": 3.423674967338681e-05, "epoch": 1.6128, "percentage": 32.31, "elapsed_time": "1:45:57", "remaining_time": "3:42:00"}
+{"current_steps": 127, "total_steps": 390, "loss": 0.4205, "lr": 3.411045578092754e-05, "epoch": 1.6256, "percentage": 32.56, "elapsed_time": "1:46:47", "remaining_time": "3:41:09"}
+{"current_steps": 128, "total_steps": 390, "loss": 0.4256, "lr": 3.398303151057543e-05, "epoch": 1.6383999999999999, "percentage": 32.82, "elapsed_time": "1:47:36", "remaining_time": "3:40:14"}
+{"current_steps": 129, "total_steps": 390, "loss": 0.4168, "lr": 3.385448707019199e-05, "epoch": 1.6512, "percentage": 33.08, "elapsed_time": "1:48:25", "remaining_time": "3:39:22"}
+{"current_steps": 130, "total_steps": 390, "loss": 0.411, "lr": 3.372483275737468e-05, "epoch": 1.6640000000000001, "percentage": 33.33, "elapsed_time": "1:49:15", "remaining_time": "3:38:30"}
+{"current_steps": 131, "total_steps": 390, "loss": 0.4166, "lr": 3.359407895863199e-05, "epoch": 1.6768, "percentage": 33.59, "elapsed_time": "1:50:05", "remaining_time": "3:37:39"}
+{"current_steps": 132, "total_steps": 390, "loss": 0.4185, "lr": 3.34622361485514e-05, "epoch": 1.6896, "percentage": 33.85, "elapsed_time": "1:50:56", "remaining_time": "3:36:49"}
+{"current_steps": 133, "total_steps": 390, "loss": 0.4139, "lr": 3.332931488896029e-05, "epoch": 1.7024, "percentage": 34.1, "elapsed_time": "1:51:44", "remaining_time": "3:35:54"}
+{"current_steps": 134, "total_steps": 390, "loss": 0.4159, "lr": 3.319532582807977e-05, "epoch": 1.7151999999999998, "percentage": 34.36, "elapsed_time": "1:52:34", "remaining_time": "3:35:03"}
+{"current_steps": 135, "total_steps": 390, "loss": 0.4196, "lr": 3.30602796996717e-05, "epoch": 1.728, "percentage": 34.62, "elapsed_time": "1:53:24", "remaining_time": "3:34:13"}
+{"current_steps": 136, "total_steps": 390, "loss": 0.3985, "lr": 3.2924187322178865e-05, "epoch": 1.7408000000000001, "percentage": 34.87, "elapsed_time": "1:54:15", "remaining_time": "3:33:22"}
+{"current_steps": 137, "total_steps": 390, "loss": 0.4073, "lr": 3.278705959785821e-05, "epoch": 1.7536, "percentage": 35.13, "elapsed_time": "1:55:02", "remaining_time": "3:32:27"}
+{"current_steps": 138, "total_steps": 390, "loss": 0.4201, "lr": 3.2648907511907544e-05, "epoch": 1.7664, "percentage": 35.38, "elapsed_time": "1:55:54", "remaining_time": "3:31:39"}
+{"current_steps": 139, "total_steps": 390, "loss": 0.4087, "lr": 3.250974213158555e-05, "epoch": 1.7792, "percentage": 35.64, "elapsed_time": "1:56:46", "remaining_time": "3:30:51"}
+{"current_steps": 140, "total_steps": 390, "loss": 0.4147, "lr": 3.23695746053251e-05, "epoch": 1.792, "percentage": 35.9, "elapsed_time": "1:57:37", "remaining_time": "3:30:01"}
+{"current_steps": 141, "total_steps": 390, "loss": 0.4004, "lr": 3.222841616184025e-05, "epoch": 1.8048, "percentage": 36.15, "elapsed_time": "1:58:27", "remaining_time": "3:29:11"}
+{"current_steps": 142, "total_steps": 390, "loss": 0.4126, "lr": 3.208627810922665e-05, "epoch": 1.8176, "percentage": 36.41, "elapsed_time": "1:59:17", "remaining_time": "3:28:20"}
+{"current_steps": 143, "total_steps": 390, "loss": 0.4007, "lr": 3.194317183405573e-05, "epoch": 1.8304, "percentage": 36.67, "elapsed_time": "2:00:08", "remaining_time": "3:27:31"}
+{"current_steps": 144, "total_steps": 390, "loss": 0.4087, "lr": 3.1799108800462466e-05, "epoch": 1.8432, "percentage": 36.92, "elapsed_time": "2:00:57", "remaining_time": "3:26:38"}
+{"current_steps": 145, "total_steps": 390, "loss": 0.419, "lr": 3.1654100549227024e-05, "epoch": 1.8559999999999999, "percentage": 37.18, "elapsed_time": "2:01:47", "remaining_time": "3:25:46"}
+{"current_steps": 146, "total_steps": 390, "loss": 0.4138, "lr": 3.1508158696850275e-05, "epoch": 1.8688, "percentage": 37.44, "elapsed_time": "2:02:37", "remaining_time": "3:24:55"}
+{"current_steps": 147, "total_steps": 390, "loss": 0.4081, "lr": 3.136129493462312e-05, "epoch": 1.8816000000000002, "percentage": 37.69, "elapsed_time": "2:03:29", "remaining_time": "3:24:07"}
+{"current_steps": 148, "total_steps": 390, "loss": 0.4009, "lr": 3.121352102768998e-05, "epoch": 1.8944, "percentage": 37.95, "elapsed_time": "2:04:20", "remaining_time": "3:23:19"}
+{"current_steps": 149, "total_steps": 390, "loss": 0.4258, "lr": 3.106484881410628e-05, "epoch": 1.9072, "percentage": 38.21, "elapsed_time": "2:05:12", "remaining_time": "3:22:30"}
+{"current_steps": 150, "total_steps": 390, "loss": 0.4203, "lr": 3.091529020389009e-05, "epoch": 1.92, "percentage": 38.46, "elapsed_time": "2:06:01", "remaining_time": "3:21:38"}
+{"current_steps": 151, "total_steps": 390, "loss": 0.4006, "lr": 3.076485717806808e-05, "epoch": 1.9327999999999999, "percentage": 38.72, "elapsed_time": "2:06:46", "remaining_time": "3:20:39"}
+{"current_steps": 152, "total_steps": 390, "loss": 0.4094, "lr": 3.061356178771564e-05, "epoch": 1.9456, "percentage": 38.97, "elapsed_time": "2:07:38", "remaining_time": "3:19:51"}
+{"current_steps": 153, "total_steps": 390, "loss": 0.4213, "lr": 3.0461416152991555e-05, "epoch": 1.9584000000000001, "percentage": 39.23, "elapsed_time": "2:08:27", "remaining_time": "3:18:59"}
+{"current_steps": 154, "total_steps": 390, "loss": 0.4199, "lr": 3.0308432462167045e-05, "epoch": 1.9712, "percentage": 39.49, "elapsed_time": "2:09:18", "remaining_time": "3:18:10"}
+{"current_steps": 155, "total_steps": 390, "loss": 0.3961, "lr": 3.015462297064936e-05, "epoch": 1.984, "percentage": 39.74, "elapsed_time": "2:10:09", "remaining_time": "3:17:19"}
+{"current_steps": 156, "total_steps": 390, "loss": 0.4055, "lr": 3.0000000000000004e-05, "epoch": 1.9968, "percentage": 40.0, "elapsed_time": "2:10:57", "remaining_time": "3:16:26"}
+{"current_steps": 157, "total_steps": 390, "loss": 0.7441, "lr": 2.98445759369477e-05, "epoch": 2.0096, "percentage": 40.26, "elapsed_time": "2:12:06", "remaining_time": "3:16:04"}
+{"current_steps": 158, "total_steps": 390, "loss": 0.3507, "lr": 2.9688363232396056e-05, "epoch": 2.0224, "percentage": 40.51, "elapsed_time": "2:12:55", "remaining_time": "3:15:10"}
+{"current_steps": 159, "total_steps": 390, "loss": 0.3736, "lr": 2.9531374400426158e-05, "epoch": 2.0352, "percentage": 40.77, "elapsed_time": "2:13:46", "remaining_time": "3:14:21"}
+{"current_steps": 160, "total_steps": 390, "loss": 0.3587, "lr": 2.9373622017294075e-05, "epoch": 2.048, "percentage": 41.03, "elapsed_time": "2:14:34", "remaining_time": "3:13:27"}
+{"current_steps": 161, "total_steps": 390, "loss": 0.3623, "lr": 2.9215118720423375e-05, "epoch": 2.0608, "percentage": 41.28, "elapsed_time": "2:15:23", "remaining_time": "3:12:33"}
+{"current_steps": 162, "total_steps": 390, "loss": 0.3659, "lr": 2.9055877207392752e-05, "epoch": 2.0736, "percentage": 41.54, "elapsed_time": "2:16:12", "remaining_time": "3:11:42"}
+{"current_steps": 163, "total_steps": 390, "loss": 0.3476, "lr": 2.8895910234918828e-05, "epoch": 2.0864, "percentage": 41.79, "elapsed_time": "2:17:01", "remaining_time": "3:10:49"}
+{"current_steps": 164, "total_steps": 390, "loss": 0.3553, "lr": 2.873523061783426e-05, "epoch": 2.0992, "percentage": 42.05, "elapsed_time": "2:17:51", "remaining_time": "3:09:59"}
+{"current_steps": 165, "total_steps": 390, "loss": 0.36, "lr": 2.8573851228061084e-05, "epoch": 2.112, "percentage": 42.31, "elapsed_time": "2:18:40", "remaining_time": "3:09:06"}
+{"current_steps": 166, "total_steps": 390, "loss": 0.3481, "lr": 2.8411784993579633e-05, "epoch": 2.1248, "percentage": 42.56, "elapsed_time": "2:19:23", "remaining_time": "3:08:05"}
+{"current_steps": 167, "total_steps": 390, "loss": 0.3645, "lr": 2.8249044897392814e-05, "epoch": 2.1376, "percentage": 42.82, "elapsed_time": "2:20:13", "remaining_time": "3:07:15"}
+{"current_steps": 168, "total_steps": 390, "loss": 0.3635, "lr": 2.80856439764861e-05, "epoch": 2.1504, "percentage": 43.08, "elapsed_time": "2:21:06", "remaining_time": "3:06:27"}
+{"current_steps": 169, "total_steps": 390, "loss": 0.3542, "lr": 2.792159532078314e-05, "epoch": 2.1632, "percentage": 43.33, "elapsed_time": "2:21:55", "remaining_time": "3:05:35"}
+{"current_steps": 170, "total_steps": 390, "loss": 0.3556, "lr": 2.77569120720971e-05, "epoch": 2.176, "percentage": 43.59, "elapsed_time": "2:22:44", "remaining_time": "3:04:44"}
+{"current_steps": 171, "total_steps": 390, "loss": 0.3647, "lr": 2.7591607423077932e-05, "epoch": 2.1888, "percentage": 43.85, "elapsed_time": "2:23:36", "remaining_time": "3:03:54"}
+{"current_steps": 172, "total_steps": 390, "loss": 0.3554, "lr": 2.7425694616155474e-05, "epoch": 2.2016, "percentage": 44.1, "elapsed_time": "2:24:28", "remaining_time": "3:03:06"}
+{"current_steps": 173, "total_steps": 390, "loss": 0.3532, "lr": 2.7259186942478656e-05, "epoch": 2.2144, "percentage": 44.36, "elapsed_time": "2:25:18", "remaining_time": "3:02:15"}
+{"current_steps": 174, "total_steps": 390, "loss": 0.3544, "lr": 2.7092097740850712e-05, "epoch": 2.2272, "percentage": 44.62, "elapsed_time": "2:26:07", "remaining_time": "3:01:24"}
+{"current_steps": 175, "total_steps": 390, "loss": 0.3455, "lr": 2.692444039666066e-05, "epoch": 2.24, "percentage": 44.87, "elapsed_time": "2:27:00", "remaining_time": "3:00:36"}
+{"current_steps": 176, "total_steps": 390, "loss": 0.3621, "lr": 2.6756228340810946e-05, "epoch": 2.2528, "percentage": 45.13, "elapsed_time": "2:27:51", "remaining_time": "2:59:47"}
+{"current_steps": 177, "total_steps": 390, "loss": 0.3419, "lr": 2.6587475048641596e-05, "epoch": 2.2656, "percentage": 45.38, "elapsed_time": "2:28:41", "remaining_time": "2:58:56"}
+{"current_steps": 178, "total_steps": 390, "loss": 0.3524, "lr": 2.6418194038850634e-05, "epoch": 2.2784, "percentage": 45.64, "elapsed_time": "2:29:31", "remaining_time": "2:58:05"}
+{"current_steps": 179, "total_steps": 390, "loss": 0.3493, "lr": 2.624839887241115e-05, "epoch": 2.2912, "percentage": 45.9, "elapsed_time": "2:30:22", "remaining_time": "2:57:15"}
+{"current_steps": 180, "total_steps": 390, "loss": 0.3587, "lr": 2.607810315148494e-05, "epoch": 2.304, "percentage": 46.15, "elapsed_time": "2:31:11", "remaining_time": "2:56:23"}
+{"current_steps": 181, "total_steps": 390, "loss": 0.35, "lr": 2.5907320518332827e-05, "epoch": 2.3168, "percentage": 46.41, "elapsed_time": "2:32:01", "remaining_time": "2:55:33"}
+{"current_steps": 182, "total_steps": 390, "loss": 0.3582, "lr": 2.5736064654221808e-05, "epoch": 2.3296, "percentage": 46.67, "elapsed_time": "2:32:50", "remaining_time": "2:54:40"}
+{"current_steps": 183, "total_steps": 390, "loss": 0.3559, "lr": 2.5564349278329056e-05, "epoch": 2.3424, "percentage": 46.92, "elapsed_time": "2:33:39", "remaining_time": "2:53:49"}
+{"current_steps": 184, "total_steps": 390, "loss": 0.3568, "lr": 2.539218814664288e-05, "epoch": 2.3552, "percentage": 47.18, "elapsed_time": "2:34:29", "remaining_time": "2:52:57"}
+{"current_steps": 185, "total_steps": 390, "loss": 0.3532, "lr": 2.521959505086075e-05, "epoch": 2.368, "percentage": 47.44, "elapsed_time": "2:35:20", "remaining_time": "2:52:08"}
+{"current_steps": 186, "total_steps": 390, "loss": 0.3458, "lr": 2.5046583817284437e-05, "epoch": 2.3808, "percentage": 47.69, "elapsed_time": "2:36:11", "remaining_time": "2:51:17"}
+{"current_steps": 187, "total_steps": 390, "loss": 0.3421, "lr": 2.487316830571244e-05, "epoch": 2.3936, "percentage": 47.95, "elapsed_time": "2:37:01", "remaining_time": "2:50:27"}
+{"current_steps": 188, "total_steps": 390, "loss": 0.3453, "lr": 2.4699362408329646e-05, "epoch": 2.4064, "percentage": 48.21, "elapsed_time": "2:37:50", "remaining_time": "2:49:35"}
+{"current_steps": 189, "total_steps": 390, "loss": 0.3432, "lr": 2.4525180048594452e-05, "epoch": 2.4192, "percentage": 48.46, "elapsed_time": "2:38:37", "remaining_time": "2:48:42"}
+{"current_steps": 190, "total_steps": 390, "loss": 0.3548, "lr": 2.435063518012335e-05, "epoch": 2.432, "percentage": 48.72, "elapsed_time": "2:39:26", "remaining_time": "2:47:50"}
+{"current_steps": 191, "total_steps": 390, "loss": 0.3613, "lr": 2.4175741785573177e-05, "epoch": 2.4448, "percentage": 48.97, "elapsed_time": "2:40:18", "remaining_time": "2:47:01"}
+{"current_steps": 192, "total_steps": 390, "loss": 0.3417, "lr": 2.4000513875520892e-05, "epoch": 2.4576000000000002, "percentage": 49.23, "elapsed_time": "2:41:07", "remaining_time": "2:46:09"}
+{"current_steps": 193, "total_steps": 390, "loss": 0.3581, "lr": 2.3824965487341247e-05, "epoch": 2.4704, "percentage": 49.49, "elapsed_time": "2:41:59", "remaining_time": "2:45:21"}
+{"current_steps": 194, "total_steps": 390, "loss": 0.3474, "lr": 2.3649110684082258e-05, "epoch": 2.4832, "percentage": 49.74, "elapsed_time": "2:42:49", "remaining_time": "2:44:30"}
+{"current_steps": 195, "total_steps": 390, "loss": 0.3442, "lr": 2.3472963553338614e-05, "epoch": 2.496, "percentage": 50.0, "elapsed_time": "2:43:41", "remaining_time": "2:43:41"}
+{"current_steps": 196, "total_steps": 390, "loss": 0.3683, "lr": 2.3296538206123134e-05, "epoch": 2.5088, "percentage": 50.26, "elapsed_time": "2:44:33", "remaining_time": "2:42:52"}
+{"current_steps": 197, "total_steps": 390, "loss": 0.3477, "lr": 2.311984877573636e-05, "epoch": 2.5216, "percentage": 50.51, "elapsed_time": "2:45:24", "remaining_time": "2:42:03"}
+{"current_steps": 198, "total_steps": 390, "loss": 0.3595, "lr": 2.2942909416634326e-05, "epoch": 2.5343999999999998, "percentage": 50.77, "elapsed_time": "2:46:13", "remaining_time": "2:41:11"}
+{"current_steps": 199, "total_steps": 390, "loss": 0.3473, "lr": 2.2765734303294666e-05, "epoch": 2.5472, "percentage": 51.03, "elapsed_time": "2:47:03", "remaining_time": "2:40:20"}
+{"current_steps": 200, "total_steps": 390, "loss": 0.3521, "lr": 2.2588337629081107e-05, "epoch": 2.56, "percentage": 51.28, "elapsed_time": "2:47:54", "remaining_time": "2:39:30"}
+{"current_steps": 201, "total_steps": 390, "loss": 0.3543, "lr": 2.2410733605106462e-05, "epoch": 2.5728, "percentage": 51.54, "elapsed_time": "2:48:43", "remaining_time": "2:38:39"}
+{"current_steps": 202, "total_steps": 390, "loss": 0.3533, "lr": 2.2232936459094158e-05, "epoch": 2.5856, "percentage": 51.79, "elapsed_time": "2:49:33", "remaining_time": "2:37:47"}
+{"current_steps": 203, "total_steps": 390, "loss": 0.3575, "lr": 2.205496043423849e-05, "epoch": 2.5984, "percentage": 52.05, "elapsed_time": "2:50:24", "remaining_time": "2:36:58"}
+{"current_steps": 204, "total_steps": 390, "loss": 0.346, "lr": 2.1876819788063586e-05, "epoch": 2.6112, "percentage": 52.31, "elapsed_time": "2:51:15", "remaining_time": "2:36:08"}
+{"current_steps": 205, "total_steps": 390, "loss": 0.3671, "lr": 2.16985287912813e-05, "epoch": 2.624, "percentage": 52.56, "elapsed_time": "2:52:05", "remaining_time": "2:35:18"}
+{"current_steps": 206, "total_steps": 390, "loss": 0.3424, "lr": 2.1520101726647922e-05, "epoch": 2.6368, "percentage": 52.82, "elapsed_time": "2:52:53", "remaining_time": "2:34:25"}
+{"current_steps": 207, "total_steps": 390, "loss": 0.3526, "lr": 2.1341552887820048e-05, "epoch": 2.6496, "percentage": 53.08, "elapsed_time": "2:53:44", "remaining_time": "2:33:35"}
+{"current_steps": 208, "total_steps": 390, "loss": 0.3441, "lr": 2.1162896578209517e-05, "epoch": 2.6624, "percentage": 53.33, "elapsed_time": "2:54:32", "remaining_time": "2:32:43"}
+{"current_steps": 209, "total_steps": 390, "loss": 0.3598, "lr": 2.0984147109837564e-05, "epoch": 2.6752000000000002, "percentage": 53.59, "elapsed_time": "2:55:23", "remaining_time": "2:31:53"}
+{"current_steps": 210, "total_steps": 390, "loss": 0.3595, "lr": 2.0805318802188307e-05, "epoch": 2.6879999999999997, "percentage": 53.85, "elapsed_time": "2:56:13", "remaining_time": "2:31:02"}
+{"current_steps": 211, "total_steps": 390, "loss": 0.3514, "lr": 2.0626425981061608e-05, "epoch": 2.7008, "percentage": 54.1, "elapsed_time": "2:57:04", "remaining_time": "2:30:12"}
+{"current_steps": 212, "total_steps": 390, "loss": 0.3578, "lr": 2.0447482977425465e-05, "epoch": 2.7136, "percentage": 54.36, "elapsed_time": "2:57:55", "remaining_time": "2:29:23"}
+{"current_steps": 213, "total_steps": 390, "loss": 0.3658, "lr": 2.0268504126267952e-05, "epoch": 2.7264, "percentage": 54.62, "elapsed_time": "2:58:46", "remaining_time": "2:28:33"}
+{"current_steps": 214, "total_steps": 390, "loss": 0.3475, "lr": 2.008950376544887e-05, "epoch": 2.7392, "percentage": 54.87, "elapsed_time": "2:59:37", "remaining_time": "2:27:43"}
+{"current_steps": 215, "total_steps": 390, "loss": 0.3509, "lr": 1.9910496234551132e-05, "epoch": 2.752, "percentage": 55.13, "elapsed_time": "3:00:26", "remaining_time": "2:26:51"}
+{"current_steps": 216, "total_steps": 390, "loss": 0.3541, "lr": 1.9731495873732055e-05, "epoch": 2.7648, "percentage": 55.38, "elapsed_time": "3:01:18", "remaining_time": "2:26:03"}
+{"current_steps": 217, "total_steps": 390, "loss": 0.3602, "lr": 1.9552517022574542e-05, "epoch": 2.7776, "percentage": 55.64, "elapsed_time": "3:02:06", "remaining_time": "2:25:10"}
+{"current_steps": 218, "total_steps": 390, "loss": 0.3433, "lr": 1.93735740189384e-05, "epoch": 2.7904, "percentage": 55.9, "elapsed_time": "3:02:56", "remaining_time": "2:24:20"}
+{"current_steps": 219, "total_steps": 390, "loss": 0.3547, "lr": 1.9194681197811703e-05, "epoch": 2.8032, "percentage": 56.15, "elapsed_time": "3:03:46", "remaining_time": "2:23:29"}
+{"current_steps": 220, "total_steps": 390, "loss": 0.3678, "lr": 1.901585289016244e-05, "epoch": 2.816, "percentage": 56.41, "elapsed_time": "3:04:36", "remaining_time": "2:22:38"}
+{"current_steps": 221, "total_steps": 390, "loss": 0.3523, "lr": 1.8837103421790486e-05, "epoch": 2.8288, "percentage": 56.67, "elapsed_time": "3:05:25", "remaining_time": "2:21:48"}
+{"current_steps": 222, "total_steps": 390, "loss": 0.3449, "lr": 1.8658447112179952e-05, "epoch": 2.8416, "percentage": 56.92, "elapsed_time": "3:06:13", "remaining_time": "2:20:55"}
+{"current_steps": 223, "total_steps": 390, "loss": 0.3631, "lr": 1.8479898273352084e-05, "epoch": 2.8544, "percentage": 57.18, "elapsed_time": "3:07:02", "remaining_time": "2:20:04"}
+{"current_steps": 224, "total_steps": 390, "loss": 0.3526, "lr": 1.83014712087187e-05, "epoch": 2.8672, "percentage": 57.44, "elapsed_time": "3:07:55", "remaining_time": "2:19:15"}
+{"current_steps": 225, "total_steps": 390, "loss": 0.3428, "lr": 1.8123180211936417e-05, "epoch": 2.88, "percentage": 57.69, "elapsed_time": "3:08:44", "remaining_time": "2:18:24"}
+{"current_steps": 226, "total_steps": 390, "loss": 0.345, "lr": 1.794503956576152e-05, "epoch": 2.8928000000000003, "percentage": 57.95, "elapsed_time": "3:09:33", "remaining_time": "2:17:33"}
+{"current_steps": 227, "total_steps": 390, "loss": 0.3487, "lr": 1.776706354090585e-05, "epoch": 2.9055999999999997, "percentage": 58.21, "elapsed_time": "3:10:23", "remaining_time": "2:16:42"}
+{"current_steps": 228, "total_steps": 390, "loss": 0.356, "lr": 1.758926639489354e-05, "epoch": 2.9184, "percentage": 58.46, "elapsed_time": "3:11:13", "remaining_time": "2:15:52"}
+{"current_steps": 229, "total_steps": 390, "loss": 0.3578, "lr": 1.7411662370918893e-05, "epoch": 2.9312, "percentage": 58.72, "elapsed_time": "3:12:01", "remaining_time": "2:15:00"}
+{"current_steps": 230, "total_steps": 390, "loss": 0.3619, "lr": 1.7234265696705344e-05, "epoch": 2.944, "percentage": 58.97, "elapsed_time": "3:12:49", "remaining_time": "2:14:08"}
+{"current_steps": 231, "total_steps": 390, "loss": 0.3587, "lr": 1.7057090583365678e-05, "epoch": 2.9568, "percentage": 59.23, "elapsed_time": "3:13:36", "remaining_time": "2:13:15"}
+{"current_steps": 232, "total_steps": 390, "loss": 0.3447, "lr": 1.6880151224263646e-05, "epoch": 2.9696, "percentage": 59.49, "elapsed_time": "3:14:24", "remaining_time": "2:12:23"}
+{"current_steps": 233, "total_steps": 390, "loss": 0.359, "lr": 1.6703461793876876e-05, "epoch": 2.9824, "percentage": 59.74, "elapsed_time": "3:15:15", "remaining_time": "2:11:33"}
+{"current_steps": 234, "total_steps": 390, "loss": 0.3487, "lr": 1.6527036446661396e-05, "epoch": 2.9952, "percentage": 60.0, "elapsed_time": "3:16:06", "remaining_time": "2:10:44"}
+{"current_steps": 235, "total_steps": 390, "loss": 0.6241, "lr": 1.635088931591775e-05, "epoch": 3.008, "percentage": 60.26, "elapsed_time": "3:17:07", "remaining_time": "2:10:00"}
+{"current_steps": 236, "total_steps": 390, "loss": 0.3034, "lr": 1.6175034512658753e-05, "epoch": 3.0208, "percentage": 60.51, "elapsed_time": "3:17:57", "remaining_time": "2:09:10"}
+{"current_steps": 237, "total_steps": 390, "loss": 0.3032, "lr": 1.5999486124479115e-05, "epoch": 3.0336, "percentage": 60.77, "elapsed_time": "3:18:49", "remaining_time": "2:08:21"}
+{"current_steps": 238, "total_steps": 390, "loss": 0.3061, "lr": 1.5824258214426833e-05, "epoch": 3.0464, "percentage": 61.03, "elapsed_time": "3:19:40", "remaining_time": "2:07:31"}
+{"current_steps": 239, "total_steps": 390, "loss": 0.2937, "lr": 1.5649364819876655e-05, "epoch": 3.0592, "percentage": 61.28, "elapsed_time": "3:20:27", "remaining_time": "2:06:38"}
+{"current_steps": 240, "total_steps": 390, "loss": 0.3138, "lr": 1.547481995140556e-05, "epoch": 3.072, "percentage": 61.54, "elapsed_time": "3:21:17", "remaining_time": "2:05:48"}
+{"current_steps": 241, "total_steps": 390, "loss": 0.308, "lr": 1.5300637591670357e-05, "epoch": 3.0848, "percentage": 61.79, "elapsed_time": "3:22:07", "remaining_time": "2:04:57"}
+{"current_steps": 242, "total_steps": 390, "loss": 0.3076, "lr": 1.5126831694287564e-05, "epoch": 3.0976, "percentage": 62.05, "elapsed_time": "3:22:56", "remaining_time": "2:04:06"}
+{"current_steps": 243, "total_steps": 390, "loss": 0.3071, "lr": 1.4953416182715566e-05, "epoch": 3.1104, "percentage": 62.31, "elapsed_time": "3:23:44", "remaining_time": "2:03:15"}
+{"current_steps": 244, "total_steps": 390, "loss": 0.3056, "lr": 1.478040494913926e-05, "epoch": 3.1232, "percentage": 62.56, "elapsed_time": "3:24:33", "remaining_time": "2:02:23"}
+{"current_steps": 245, "total_steps": 390, "loss": 0.3047, "lr": 1.460781185335713e-05, "epoch": 3.136, "percentage": 62.82, "elapsed_time": "3:25:23", "remaining_time": "2:01:33"}
+{"current_steps": 246, "total_steps": 390, "loss": 0.2925, "lr": 1.443565072167095e-05, "epoch": 3.1488, "percentage": 63.08, "elapsed_time": "3:26:13", "remaining_time": "2:00:43"}
+{"current_steps": 247, "total_steps": 390, "loss": 0.2965, "lr": 1.4263935345778202e-05, "epoch": 3.1616, "percentage": 63.33, "elapsed_time": "3:27:03", "remaining_time": "1:59:52"}
+{"current_steps": 248, "total_steps": 390, "loss": 0.3127, "lr": 1.409267948166718e-05, "epoch": 3.1744, "percentage": 63.59, "elapsed_time": "3:27:54", "remaining_time": "1:59:02"}
+{"current_steps": 249, "total_steps": 390, "loss": 0.3035, "lr": 1.3921896848515064e-05, "epoch": 3.1872, "percentage": 63.85, "elapsed_time": "3:28:42", "remaining_time": "1:58:10"}
+{"current_steps": 250, "total_steps": 390, "loss": 0.299, "lr": 1.3751601127588849e-05, "epoch": 3.2, "percentage": 64.1, "elapsed_time": "3:29:30", "remaining_time": "1:57:19"}
+{"current_steps": 251, "total_steps": 390, "loss": 0.2928, "lr": 1.3581805961149371e-05, "epoch": 3.2128, "percentage": 64.36, "elapsed_time": "3:30:21", "remaining_time": "1:56:29"}
+{"current_steps": 252, "total_steps": 390, "loss": 0.291, "lr": 1.341252495135841e-05, "epoch": 3.2256, "percentage": 64.62, "elapsed_time": "3:31:12", "remaining_time": "1:55:39"}
+{"current_steps": 253, "total_steps": 390, "loss": 0.2888, "lr": 1.324377165918906e-05, "epoch": 3.2384, "percentage": 64.87, "elapsed_time": "3:32:00", "remaining_time": "1:54:48"}
+{"current_steps": 254, "total_steps": 390, "loss": 0.2953, "lr": 1.3075559603339354e-05, "epoch": 3.2512, "percentage": 65.13, "elapsed_time": "3:32:49", "remaining_time": "1:53:57"}
+{"current_steps": 255, "total_steps": 390, "loss": 0.2912, "lr": 1.2907902259149287e-05, "epoch": 3.2640000000000002, "percentage": 65.38, "elapsed_time": "3:33:38", "remaining_time": "1:53:06"}
+{"current_steps": 256, "total_steps": 390, "loss": 0.3111, "lr": 1.274081305752135e-05, "epoch": 3.2768, "percentage": 65.64, "elapsed_time": "3:34:29", "remaining_time": "1:52:16"}
+{"current_steps": 257, "total_steps": 390, "loss": 0.2899, "lr": 1.2574305383844528e-05, "epoch": 3.2896, "percentage": 65.9, "elapsed_time": "3:35:23", "remaining_time": "1:51:28"}
+{"current_steps": 258, "total_steps": 390, "loss": 0.2916, "lr": 1.2408392576922075e-05, "epoch": 3.3024, "percentage": 66.15, "elapsed_time": "3:36:12", "remaining_time": "1:50:37"}
+{"current_steps": 259, "total_steps": 390, "loss": 0.305, "lr": 1.2243087927902905e-05, "epoch": 3.3152, "percentage": 66.41, "elapsed_time": "3:37:01", "remaining_time": "1:49:46"}
+{"current_steps": 260, "total_steps": 390, "loss": 0.3039, "lr": 1.2078404679216864e-05, "epoch": 3.328, "percentage": 66.67, "elapsed_time": "3:37:49", "remaining_time": "1:48:54"}
+{"current_steps": 261, "total_steps": 390, "loss": 0.2878, "lr": 1.1914356023513904e-05, "epoch": 3.3407999999999998, "percentage": 66.92, "elapsed_time": "3:38:38", "remaining_time": "1:48:03"}
+{"current_steps": 262, "total_steps": 390, "loss": 0.2949, "lr": 1.1750955102607193e-05, "epoch": 3.3536, "percentage": 67.18, "elapsed_time": "3:39:26", "remaining_time": "1:47:12"}
+{"current_steps": 263, "total_steps": 390, "loss": 0.2993, "lr": 1.1588215006420374e-05, "epoch": 3.3664, "percentage": 67.44, "elapsed_time": "3:40:15", "remaining_time": "1:46:21"}
+{"current_steps": 264, "total_steps": 390, "loss": 0.2944, "lr": 1.1426148771938915e-05, "epoch": 3.3792, "percentage": 67.69, "elapsed_time": "3:41:05", "remaining_time": "1:45:31"}
+{"current_steps": 265, "total_steps": 390, "loss": 0.2955, "lr": 1.1264769382165748e-05, "epoch": 3.392, "percentage": 67.95, "elapsed_time": "3:41:55", "remaining_time": "1:44:40"}
+{"current_steps": 266, "total_steps": 390, "loss": 0.288, "lr": 1.110408976508118e-05, "epoch": 3.4048, "percentage": 68.21, "elapsed_time": "3:42:46", "remaining_time": "1:43:51"}
+{"current_steps": 267, "total_steps": 390, "loss": 0.2904, "lr": 1.094412279260726e-05, "epoch": 3.4176, "percentage": 68.46, "elapsed_time": "3:43:36", "remaining_time": "1:43:00"}
+{"current_steps": 268, "total_steps": 390, "loss": 0.2989, "lr": 1.0784881279576635e-05, "epoch": 3.4304, "percentage": 68.72, "elapsed_time": "3:44:28", "remaining_time": "1:42:10"}
+{"current_steps": 269, "total_steps": 390, "loss": 0.2975, "lr": 1.0626377982705929e-05, "epoch": 3.4432, "percentage": 68.97, "elapsed_time": "3:45:14", "remaining_time": "1:41:19"}
+{"current_steps": 270, "total_steps": 390, "loss": 0.3064, "lr": 1.0468625599573842e-05, "epoch": 3.456, "percentage": 69.23, "elapsed_time": "3:46:04", "remaining_time": "1:40:28"}
+{"current_steps": 271, "total_steps": 390, "loss": 0.2862, "lr": 1.0311636767603952e-05, "epoch": 3.4688, "percentage": 69.49, "elapsed_time": "3:46:54", "remaining_time": "1:39:38"}
+{"current_steps": 272, "total_steps": 390, "loss": 0.3092, "lr": 1.0155424063052306e-05, "epoch": 3.4816, "percentage": 69.74, "elapsed_time": "3:47:44", "remaining_time": "1:38:47"}
+{"current_steps": 273, "total_steps": 390, "loss": 0.2971, "lr": 1.0000000000000006e-05, "epoch": 3.4944, "percentage": 70.0, "elapsed_time": "3:48:35", "remaining_time": "1:37:57"}
+{"current_steps": 274, "total_steps": 390, "loss": 0.2984, "lr": 9.84537702935065e-06, "epoch": 3.5072, "percentage": 70.26, "elapsed_time": "3:49:26", "remaining_time": "1:37:07"}
+{"current_steps": 275, "total_steps": 390, "loss": 0.2961, "lr": 9.691567537832964e-06, "epoch": 3.52, "percentage": 70.51, "elapsed_time": "3:50:17", "remaining_time": "1:36:18"}
+{"current_steps": 276, "total_steps": 390, "loss": 0.3033, "lr": 9.538583847008452e-06, "epoch": 3.5328, "percentage": 70.77, "elapsed_time": "3:51:06", "remaining_time": "1:35:27"}
+{"current_steps": 277, "total_steps": 390, "loss": 0.3039, "lr": 9.386438212284372e-06, "epoch": 3.5456, "percentage": 71.03, "elapsed_time": "3:51:53", "remaining_time": "1:34:35"}
+{"current_steps": 278, "total_steps": 390, "loss": 0.2943, "lr": 9.235142821931928e-06, "epoch": 3.5584, "percentage": 71.28, "elapsed_time": "3:52:40", "remaining_time": "1:33:44"}
+{"current_steps": 279, "total_steps": 390, "loss": 0.2919, "lr": 9.084709796109907e-06, "epoch": 3.5712, "percentage": 71.54, "elapsed_time": "3:53:30", "remaining_time": "1:32:54"}
+{"current_steps": 280, "total_steps": 390, "loss": 0.2997, "lr": 8.93515118589373e-06, "epoch": 3.584, "percentage": 71.79, "elapsed_time": "3:54:20", "remaining_time": "1:32:03"}
+{"current_steps": 281, "total_steps": 390, "loss": 0.3002, "lr": 8.786478972310023e-06, "epoch": 3.5968, "percentage": 72.05, "elapsed_time": "3:55:11", "remaining_time": "1:31:14"}
+{"current_steps": 282, "total_steps": 390, "loss": 0.2921, "lr": 8.638705065376887e-06, "epoch": 3.6096, "percentage": 72.31, "elapsed_time": "3:56:02", "remaining_time": "1:30:23"}
+{"current_steps": 283, "total_steps": 390, "loss": 0.311, "lr": 8.491841303149728e-06, "epoch": 3.6224, "percentage": 72.56, "elapsed_time": "3:56:53", "remaining_time": "1:29:33"}
+{"current_steps": 284, "total_steps": 390, "loss": 0.2919, "lr": 8.345899450772975e-06, "epoch": 3.6352, "percentage": 72.82, "elapsed_time": "3:57:43", "remaining_time": "1:28:43"}
+{"current_steps": 285, "total_steps": 390, "loss": 0.2841, "lr": 8.200891199537549e-06, "epoch": 3.648, "percentage": 73.08, "elapsed_time": "3:58:31", "remaining_time": "1:27:52"}
+{"current_steps": 286, "total_steps": 390, "loss": 0.3014, "lr": 8.056828165944282e-06, "epoch": 3.6608, "percentage": 73.33, "elapsed_time": "3:59:22", "remaining_time": "1:27:02"}
+{"current_steps": 287, "total_steps": 390, "loss": 0.3019, "lr": 7.913721890773354e-06, "epoch": 3.6736, "percentage": 73.59, "elapsed_time": "4:00:14", "remaining_time": "1:26:12"}
+{"current_steps": 288, "total_steps": 390, "loss": 0.2976, "lr": 7.771583838159756e-06, "epoch": 3.6864, "percentage": 73.85, "elapsed_time": "4:01:04", "remaining_time": "1:25:22"}
+{"current_steps": 289, "total_steps": 390, "loss": 0.2989, "lr": 7.630425394674903e-06, "epoch": 3.6992000000000003, "percentage": 74.1, "elapsed_time": "4:01:54", "remaining_time": "1:24:32"}
+{"current_steps": 290, "total_steps": 390, "loss": 0.2923, "lr": 7.49025786841445e-06, "epoch": 3.7119999999999997, "percentage": 74.36, "elapsed_time": "4:02:44", "remaining_time": "1:23:42"}
+{"current_steps": 291, "total_steps": 390, "loss": 0.2969, "lr": 7.3510924880924575e-06, "epoch": 3.7248, "percentage": 74.62, "elapsed_time": "4:03:36", "remaining_time": "1:22:52"}
+{"current_steps": 292, "total_steps": 390, "loss": 0.3061, "lr": 7.212940402141808e-06, "epoch": 3.7376, "percentage": 74.87, "elapsed_time": "4:04:28", "remaining_time": "1:22:02"}
+{"current_steps": 293, "total_steps": 390, "loss": 0.2823, "lr": 7.075812677821145e-06, "epoch": 3.7504, "percentage": 75.13, "elapsed_time": "4:05:16", "remaining_time": "1:21:12"}
+{"current_steps": 294, "total_steps": 390, "loss": 0.2997, "lr": 6.939720300328303e-06, "epoch": 3.7632, "percentage": 75.38, "elapsed_time": "4:06:06", "remaining_time": "1:20:21"}
+{"current_steps": 295, "total_steps": 390, "loss": 0.2959, "lr": 6.8046741719202385e-06, "epoch": 3.776, "percentage": 75.64, "elapsed_time": "4:06:57", "remaining_time": "1:19:31"}
+{"current_steps": 296, "total_steps": 390, "loss": 0.2979, "lr": 6.67068511103971e-06, "epoch": 3.7888, "percentage": 75.9, "elapsed_time": "4:07:46", "remaining_time": "1:18:41"}
+{"current_steps": 297, "total_steps": 390, "loss": 0.2926, "lr": 6.537763851448593e-06, "epoch": 3.8016, "percentage": 76.15, "elapsed_time": "4:08:36", "remaining_time": "1:17:50"}
+{"current_steps": 298, "total_steps": 390, "loss": 0.2913, "lr": 6.4059210413680175e-06, "epoch": 3.8144, "percentage": 76.41, "elapsed_time": "4:09:29", "remaining_time": "1:17:01"}
+{"current_steps": 299, "total_steps": 390, "loss": 0.3006, "lr": 6.275167242625331e-06, "epoch": 3.8272, "percentage": 76.67, "elapsed_time": "4:10:18", "remaining_time": "1:16:10"}
+{"current_steps": 300, "total_steps": 390, "loss": 0.2923, "lr": 6.145512929808013e-06, "epoch": 3.84, "percentage": 76.92, "elapsed_time": "4:11:05", "remaining_time": "1:15:19"}
+{"current_steps": 301, "total_steps": 390, "loss": 0.2915, "lr": 6.016968489424572e-06, "epoch": 3.8528000000000002, "percentage": 77.18, "elapsed_time": "4:11:53", "remaining_time": "1:14:28"}
+{"current_steps": 302, "total_steps": 390, "loss": 0.2866, "lr": 5.889544219072465e-06, "epoch": 3.8656, "percentage": 77.44, "elapsed_time": "4:12:43", "remaining_time": "1:13:38"}
+{"current_steps": 303, "total_steps": 390, "loss": 0.3032, "lr": 5.7632503266131925e-06, "epoch": 3.8784, "percentage": 77.69, "elapsed_time": "4:13:33", "remaining_time": "1:12:48"}
+{"current_steps": 304, "total_steps": 390, "loss": 0.2909, "lr": 5.638096929354522e-06, "epoch": 3.8912, "percentage": 77.95, "elapsed_time": "4:14:23", "remaining_time": "1:11:57"}
+{"current_steps": 305, "total_steps": 390, "loss": 0.3084, "lr": 5.514094053240035e-06, "epoch": 3.904, "percentage": 78.21, "elapsed_time": "4:15:14", "remaining_time": "1:11:07"}
+{"current_steps": 306, "total_steps": 390, "loss": 0.2905, "lr": 5.39125163204594e-06, "epoch": 3.9168, "percentage": 78.46, "elapsed_time": "4:16:03", "remaining_time": "1:10:17"}
+{"current_steps": 307, "total_steps": 390, "loss": 0.2937, "lr": 5.269579506585259e-06, "epoch": 3.9295999999999998, "percentage": 78.72, "elapsed_time": "4:16:52", "remaining_time": "1:09:27"}
+{"current_steps": 308, "total_steps": 390, "loss": 0.2989, "lr": 5.149087423919541e-06, "epoch": 3.9424, "percentage": 78.97, "elapsed_time": "4:17:41", "remaining_time": "1:08:36"}
+{"current_steps": 309, "total_steps": 390, "loss": 0.3126, "lr": 5.029785036577976e-06, "epoch": 3.9552, "percentage": 79.23, "elapsed_time": "4:18:32", "remaining_time": "1:07:46"}
+{"current_steps": 310, "total_steps": 390, "loss": 0.2934, "lr": 4.911681901784198e-06, "epoch": 3.968, "percentage": 79.49, "elapsed_time": "4:19:22", "remaining_time": "1:06:56"}
+{"current_steps": 311, "total_steps": 390, "loss": 0.2905, "lr": 4.794787480690597e-06, "epoch": 3.9808, "percentage": 79.74, "elapsed_time": "4:20:12", "remaining_time": "1:06:05"}
+{"current_steps": 312, "total_steps": 390, "loss": 0.3044, "lr": 4.679111137620442e-06, "epoch": 3.9936, "percentage": 80.0, "elapsed_time": "4:21:01", "remaining_time": "1:05:15"}
+{"current_steps": 313, "total_steps": 390, "loss": 0.5472, "lr": 4.5646621393177e-06, "epoch": 4.0064, "percentage": 80.26, "elapsed_time": "4:22:32", "remaining_time": "1:04:35"}
+{"current_steps": 314, "total_steps": 390, "loss": 0.2704, "lr": 4.451449654204685e-06, "epoch": 4.0192, "percentage": 80.51, "elapsed_time": "4:23:20", "remaining_time": "1:03:44"}
+{"current_steps": 315, "total_steps": 390, "loss": 0.2549, "lr": 4.339482751647557e-06, "epoch": 4.032, "percentage": 80.77, "elapsed_time": "4:24:12", "remaining_time": "1:02:54"}
+{"current_steps": 316, "total_steps": 390, "loss": 0.2621, "lr": 4.228770401229824e-06, "epoch": 4.0448, "percentage": 81.03, "elapsed_time": "4:25:03", "remaining_time": "1:02:04"}
+{"current_steps": 317, "total_steps": 390, "loss": 0.2838, "lr": 4.119321472033779e-06, "epoch": 4.0576, "percentage": 81.28, "elapsed_time": "4:25:53", "remaining_time": "1:01:13"}
+{"current_steps": 318, "total_steps": 390, "loss": 0.2622, "lr": 4.011144731929981e-06, "epoch": 4.0704, "percentage": 81.54, "elapsed_time": "4:26:42", "remaining_time": "1:00:23"}
+{"current_steps": 319, "total_steps": 390, "loss": 0.2482, "lr": 3.904248846874894e-06, "epoch": 4.0832, "percentage": 81.79, "elapsed_time": "4:27:33", "remaining_time": "0:59:33"}
+{"current_steps": 320, "total_steps": 390, "loss": 0.2637, "lr": 3.7986423802166705e-06, "epoch": 4.096, "percentage": 82.05, "elapsed_time": "4:28:26", "remaining_time": "0:58:43"}
+{"current_steps": 321, "total_steps": 390, "loss": 0.2583, "lr": 3.694333792009115e-06, "epoch": 4.1088, "percentage": 82.31, "elapsed_time": "4:29:15", "remaining_time": "0:57:52"}
+{"current_steps": 322, "total_steps": 390, "loss": 0.2636, "lr": 3.5913314383339937e-06, "epoch": 4.1216, "percentage": 82.56, "elapsed_time": "4:30:05", "remaining_time": "0:57:02"}
+{"current_steps": 323, "total_steps": 390, "loss": 0.2677, "lr": 3.4896435706316e-06, "epoch": 4.1344, "percentage": 82.82, "elapsed_time": "4:30:55", "remaining_time": "0:56:11"}
+{"current_steps": 324, "total_steps": 390, "loss": 0.2611, "lr": 3.3892783350397675e-06, "epoch": 4.1472, "percentage": 83.08, "elapsed_time": "4:31:46", "remaining_time": "0:55:21"}
+{"current_steps": 325, "total_steps": 390, "loss": 0.2585, "lr": 3.290243771741275e-06, "epoch": 4.16, "percentage": 83.33, "elapsed_time": "4:32:37", "remaining_time": "0:54:31"}
+{"current_steps": 326, "total_steps": 390, "loss": 0.259, "lr": 3.1925478143197418e-06, "epoch": 4.1728, "percentage": 83.59, "elapsed_time": "4:33:30", "remaining_time": "0:53:41"}
+{"current_steps": 327, "total_steps": 390, "loss": 0.2514, "lr": 3.0961982891241083e-06, "epoch": 4.1856, "percentage": 83.85, "elapsed_time": "4:34:18", "remaining_time": "0:52:50"}
+{"current_steps": 328, "total_steps": 390, "loss": 0.2717, "lr": 3.001202914641628e-06, "epoch": 4.1984, "percentage": 84.1, "elapsed_time": "4:35:10", "remaining_time": "0:52:00"}
+{"current_steps": 329, "total_steps": 390, "loss": 0.2568, "lr": 2.907569300879596e-06, "epoch": 4.2112, "percentage": 84.36, "elapsed_time": "4:36:01", "remaining_time": "0:51:10"}
+{"current_steps": 330, "total_steps": 390, "loss": 0.2593, "lr": 2.815304948755664e-06, "epoch": 4.224, "percentage": 84.62, "elapsed_time": "4:36:49", "remaining_time": "0:50:19"}
+{"current_steps": 331, "total_steps": 390, "loss": 0.2672, "lr": 2.7244172494969978e-06, "epoch": 4.2368, "percentage": 84.87, "elapsed_time": "4:37:39", "remaining_time": "0:49:29"}
+{"current_steps": 332, "total_steps": 390, "loss": 0.2672, "lr": 2.6349134840481294e-06, "epoch": 4.2496, "percentage": 85.13, "elapsed_time": "4:38:28", "remaining_time": "0:48:39"}
+{"current_steps": 333, "total_steps": 390, "loss": 0.2577, "lr": 2.546800822487714e-06, "epoch": 4.2624, "percentage": 85.38, "elapsed_time": "4:39:18", "remaining_time": "0:47:48"}
+{"current_steps": 334, "total_steps": 390, "loss": 0.2608, "lr": 2.4600863234541338e-06, "epoch": 4.2752, "percentage": 85.64, "elapsed_time": "4:40:09", "remaining_time": "0:46:58"}
+{"current_steps": 335, "total_steps": 390, "loss": 0.2523, "lr": 2.374776933580025e-06, "epoch": 4.288, "percentage": 85.9, "elapsed_time": "4:40:58", "remaining_time": "0:46:07"}
+{"current_steps": 336, "total_steps": 390, "loss": 0.2545, "lr": 2.2908794869358044e-06, "epoch": 4.3008, "percentage": 86.15, "elapsed_time": "4:41:45", "remaining_time": "0:45:16"}
+{"current_steps": 337, "total_steps": 390, "loss": 0.2589, "lr": 2.2084007044821764e-06, "epoch": 4.3136, "percentage": 86.41, "elapsed_time": "4:42:36", "remaining_time": "0:44:26"}
+{"current_steps": 338, "total_steps": 390, "loss": 0.264, "lr": 2.127347193531757e-06, "epoch": 4.3264, "percentage": 86.67, "elapsed_time": "4:43:26", "remaining_time": "0:43:36"}
+{"current_steps": 339, "total_steps": 390, "loss": 0.275, "lr": 2.0477254472197237e-06, "epoch": 4.3392, "percentage": 86.92, "elapsed_time": "4:44:15", "remaining_time": "0:42:45"}
+{"current_steps": 340, "total_steps": 390, "loss": 0.2668, "lr": 1.96954184398368e-06, "epoch": 4.352, "percentage": 87.18, "elapsed_time": "4:45:07", "remaining_time": "0:41:55"}
+{"current_steps": 341, "total_steps": 390, "loss": 0.2626, "lr": 1.8928026470526917e-06, "epoch": 4.3648, "percentage": 87.44, "elapsed_time": "4:45:53", "remaining_time": "0:41:04"}
+{"current_steps": 342, "total_steps": 390, "loss": 0.2539, "lr": 1.817514003945524e-06, "epoch": 4.3776, "percentage": 87.69, "elapsed_time": "4:46:45", "remaining_time": "0:40:14"}
+{"current_steps": 343, "total_steps": 390, "loss": 0.2683, "lr": 1.743681945978184e-06, "epoch": 4.3904, "percentage": 87.95, "elapsed_time": "4:47:34", "remaining_time": "0:39:24"}
+{"current_steps": 344, "total_steps": 390, "loss": 0.2629, "lr": 1.6713123877807413e-06, "epoch": 4.4032, "percentage": 88.21, "elapsed_time": "4:48:25", "remaining_time": "0:38:34"}
+{"current_steps": 345, "total_steps": 390, "loss": 0.2574, "lr": 1.6004111268235156e-06, "epoch": 4.416, "percentage": 88.46, "elapsed_time": "4:49:13", "remaining_time": "0:37:43"}
+{"current_steps": 346, "total_steps": 390, "loss": 0.2583, "lr": 1.5309838429526714e-06, "epoch": 4.4288, "percentage": 88.72, "elapsed_time": "4:50:00", "remaining_time": "0:36:52"}
+{"current_steps": 347, "total_steps": 390, "loss": 0.2538, "lr": 1.4630360979351644e-06, "epoch": 4.4416, "percentage": 88.97, "elapsed_time": "4:50:49", "remaining_time": "0:36:02"}
+{"current_steps": 348, "total_steps": 390, "loss": 0.2573, "lr": 1.396573335013236e-06, "epoch": 4.4544, "percentage": 89.23, "elapsed_time": "4:51:39", "remaining_time": "0:35:11"}
+{"current_steps": 349, "total_steps": 390, "loss": 0.2636, "lr": 1.3316008784683265e-06, "epoch": 4.4672, "percentage": 89.49, "elapsed_time": "4:52:28", "remaining_time": "0:34:21"}
+{"current_steps": 350, "total_steps": 390, "loss": 0.2564, "lr": 1.2681239331945695e-06, "epoch": 4.48, "percentage": 89.74, "elapsed_time": "4:53:18", "remaining_time": "0:33:31"}
+{"current_steps": 351, "total_steps": 390, "loss": 0.2658, "lr": 1.2061475842818337e-06, "epoch": 4.4928, "percentage": 90.0, "elapsed_time": "4:54:07", "remaining_time": "0:32:40"}
+{"current_steps": 352, "total_steps": 390, "loss": 0.2642, "lr": 1.1456767966083393e-06, "epoch": 4.5056, "percentage": 90.26, "elapsed_time": "4:54:59", "remaining_time": "0:31:50"}
+{"current_steps": 353, "total_steps": 390, "loss": 0.2585, "lr": 1.086716414442952e-06, "epoch": 4.5184, "percentage": 90.51, "elapsed_time": "4:55:51", "remaining_time": "0:31:00"}
+{"current_steps": 354, "total_steps": 390, "loss": 0.2676, "lr": 1.0292711610570904e-06, "epoch": 4.5312, "percentage": 90.77, "elapsed_time": "4:56:41", "remaining_time": "0:30:10"}
+{"current_steps": 355, "total_steps": 390, "loss": 0.2707, "lr": 9.733456383463658e-07, "epoch": 4.5440000000000005, "percentage": 91.03, "elapsed_time": "4:57:31", "remaining_time": "0:29:19"}
+{"current_steps": 356, "total_steps": 390, "loss": 0.2622, "lr": 9.189443264619102e-07, "epoch": 4.5568, "percentage": 91.28, "elapsed_time": "4:58:23", "remaining_time": "0:28:29"}
+{"current_steps": 357, "total_steps": 390, "loss": 0.262, "lr": 8.660715834514977e-07, "epoch": 4.5696, "percentage": 91.54, "elapsed_time": "4:59:13", "remaining_time": "0:27:39"}
+{"current_steps": 358, "total_steps": 390, "loss": 0.2547, "lr": 8.147316449103959e-07, "epoch": 4.5824, "percentage": 91.79, "elapsed_time": "5:00:00", "remaining_time": "0:26:49"}
+{"current_steps": 359, "total_steps": 390, "loss": 0.2623, "lr": 7.649286236420806e-07, "epoch": 4.5952, "percentage": 92.05, "elapsed_time": "5:00:50", "remaining_time": "0:25:58"}
+{"current_steps": 360, "total_steps": 390, "loss": 0.2694, "lr": 7.166665093287539e-07, "epoch": 4.608, "percentage": 92.31, "elapsed_time": "5:01:43", "remaining_time": "0:25:08"}
+{"current_steps": 361, "total_steps": 390, "loss": 0.2575, "lr": 6.69949168211721e-07, "epoch": 4.6208, "percentage": 92.56, "elapsed_time": "5:02:32", "remaining_time": "0:24:18"}
+{"current_steps": 362, "total_steps": 390, "loss": 0.2459, "lr": 6.247803427816945e-07, "epoch": 4.6336, "percentage": 92.82, "elapsed_time": "5:03:21", "remaining_time": "0:23:27"}
+{"current_steps": 363, "total_steps": 390, "loss": 0.2716, "lr": 5.811636514789598e-07, "epoch": 4.6464, "percentage": 93.08, "elapsed_time": "5:04:12", "remaining_time": "0:22:37"}
+{"current_steps": 364, "total_steps": 390, "loss": 0.2534, "lr": 5.391025884035239e-07, "epoch": 4.6592, "percentage": 93.33, "elapsed_time": "5:05:01", "remaining_time": "0:21:47"}
+{"current_steps": 365, "total_steps": 390, "loss": 0.2586, "lr": 4.986005230351954e-07, "epoch": 4.672, "percentage": 93.59, "elapsed_time": "5:05:51", "remaining_time": "0:20:56"}
+{"current_steps": 366, "total_steps": 390, "loss": 0.2645, "lr": 4.5966069996365993e-07, "epoch": 4.6848, "percentage": 93.85, "elapsed_time": "5:06:41", "remaining_time": "0:20:06"}
+{"current_steps": 367, "total_steps": 390, "loss": 0.2569, "lr": 4.22286238628562e-07, "epoch": 4.6975999999999996, "percentage": 94.1, "elapsed_time": "5:07:31", "remaining_time": "0:19:16"}
+{"current_steps": 368, "total_steps": 390, "loss": 0.2643, "lr": 3.8648013306960664e-07, "epoch": 4.7104, "percentage": 94.36, "elapsed_time": "5:08:22", "remaining_time": "0:18:26"}
+{"current_steps": 369, "total_steps": 390, "loss": 0.2623, "lr": 3.522452516867048e-07, "epoch": 4.7232, "percentage": 94.62, "elapsed_time": "5:09:10", "remaining_time": "0:17:35"}
+{"current_steps": 370, "total_steps": 390, "loss": 0.2613, "lr": 3.1958433701019697e-07, "epoch": 4.736, "percentage": 94.87, "elapsed_time": "5:10:03", "remaining_time": "0:16:45"}
+{"current_steps": 371, "total_steps": 390, "loss": 0.2589, "lr": 2.8850000548115155e-07, "epoch": 4.7488, "percentage": 95.13, "elapsed_time": "5:10:51", "remaining_time": "0:15:55"}
+{"current_steps": 372, "total_steps": 390, "loss": 0.2632, "lr": 2.5899474724174313e-07, "epoch": 4.7616, "percentage": 95.38, "elapsed_time": "5:11:43", "remaining_time": "0:15:04"}
+{"current_steps": 373, "total_steps": 390, "loss": 0.2478, "lr": 2.3107092593579905e-07, "epoch": 4.7744, "percentage": 95.64, "elapsed_time": "5:12:30", "remaining_time": "0:14:14"}
+{"current_steps": 374, "total_steps": 390, "loss": 0.2516, "lr": 2.0473077851942858e-07, "epoch": 4.7872, "percentage": 95.9, "elapsed_time": "5:13:20", "remaining_time": "0:13:24"}
+{"current_steps": 375, "total_steps": 390, "loss": 0.2676, "lr": 1.799764150818306e-07, "epoch": 4.8, "percentage": 96.15, "elapsed_time": "5:14:08", "remaining_time": "0:12:33"}
+{"current_steps": 376, "total_steps": 390, "loss": 0.2605, "lr": 1.5680981867625566e-07, "epoch": 4.8128, "percentage": 96.41, "elapsed_time": "5:14:56", "remaining_time": "0:11:43"}
+{"current_steps": 377, "total_steps": 390, "loss": 0.2718, "lr": 1.3523284516113955e-07, "epoch": 4.8256, "percentage": 96.67, "elapsed_time": "5:15:48", "remaining_time": "0:10:53"}
+{"current_steps": 378, "total_steps": 390, "loss": 0.2706, "lr": 1.1524722305144231e-07, "epoch": 4.8384, "percentage": 96.92, "elapsed_time": "5:16:38", "remaining_time": "0:10:03"}
+{"current_steps": 379, "total_steps": 390, "loss": 0.2622, "lr": 9.685455338016347e-08, "epoch": 4.8512, "percentage": 97.18, "elapsed_time": "5:17:28", "remaining_time": "0:09:12"}
+{"current_steps": 380, "total_steps": 390, "loss": 0.2532, "lr": 8.005630957010014e-08, "epoch": 4.864, "percentage": 97.44, "elapsed_time": "5:18:17", "remaining_time": "0:08:22"}
+{"current_steps": 381, "total_steps": 390, "loss": 0.2543, "lr": 6.485383731580142e-08, "epoch": 4.8768, "percentage": 97.69, "elapsed_time": "5:19:08", "remaining_time": "0:07:32"}
+{"current_steps": 382, "total_steps": 390, "loss": 0.254, "lr": 5.1248354475768034e-08, "epoch": 4.8896, "percentage": 97.95, "elapsed_time": "5:19:57", "remaining_time": "0:06:42"}
+{"current_steps": 383, "total_steps": 390, "loss": 0.2678, "lr": 3.924095097489922e-08, "epoch": 4.9024, "percentage": 98.21, "elapsed_time": "5:20:43", "remaining_time": "0:05:51"}
+{"current_steps": 384, "total_steps": 390, "loss": 0.2528, "lr": 2.8832588717164766e-08, "epoch": 4.9152000000000005, "percentage": 98.46, "elapsed_time": "5:21:29", "remaining_time": "0:05:01"}
+{"current_steps": 385, "total_steps": 390, "loss": 0.2568, "lr": 2.0024101508555604e-08, "epoch": 4.928, "percentage": 98.72, "elapsed_time": "5:22:19", "remaining_time": "0:04:11"}
+{"current_steps": 386, "total_steps": 390, "loss": 0.2625, "lr": 1.281619499029274e-08, "epoch": 4.9408, "percentage": 98.97, "elapsed_time": "5:23:10", "remaining_time": "0:03:20"}
+{"current_steps": 387, "total_steps": 390, "loss": 0.2552, "lr": 7.209446582292501e-09, "epoch": 4.9536, "percentage": 99.23, "elapsed_time": "5:23:56", "remaining_time": "0:02:30"}
+{"current_steps": 388, "total_steps": 390, "loss": 0.2553, "lr": 3.2043054369057523e-09, "epoch": 4.9664, "percentage": 99.49, "elapsed_time": "5:24:45", "remaining_time": "0:01:40"}
+{"current_steps": 389, "total_steps": 390, "loss": 0.2546, "lr": 8.010924029533406e-10, "epoch": 4.9792, "percentage": 99.74, "elapsed_time": "5:25:35", "remaining_time": "0:00:50"}
+{"current_steps": 390, "total_steps": 390, "loss": 0.268, "lr": 0.0, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "5:26:26", "remaining_time": "0:00:00"}
+{"current_steps": 390, "total_steps": 390, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "5:26:55", "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": 6.233467267119553,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.8108,
+      "step": 1
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 5.904196326876289,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.7847,
+      "step": 2
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.23851753064735,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.8192,
+      "step": 3
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 5.415128171631446,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.7829,
+      "step": 4
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.04380559994215,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.7687,
+      "step": 5
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.2587637623659984,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.7336,
+      "step": 6
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.046066080232856,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.7165,
+      "step": 7
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 4.069511156688382,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.7287,
+      "step": 8
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 4.2611185393960005,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.7252,
+      "step": 9
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 4.260307749363621,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.6916,
+      "step": 10
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 3.5860528236126874,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.6714,
+      "step": 11
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 2.0408678402576386,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.6509,
+      "step": 12
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.7405635844975456,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6462,
+      "step": 13
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.089378409802929,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.6075,
+      "step": 14
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.5136568318392694,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.6198,
+      "step": 15
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.1263674006123536,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.5743,
+      "step": 16
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.0303264886226746,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.5916,
+      "step": 17
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.9294082822832469,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.5856,
+      "step": 18
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.947170872331032,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.5385,
+      "step": 19
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.8518591702550199,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.568,
+      "step": 20
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.8725288325536481,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.5631,
+      "step": 21
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.0592322309688051,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.5431,
+      "step": 22
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.7594966772209739,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.5403,
+      "step": 23
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 1.0147100967231533,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.5346,
+      "step": 24
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8092853991109913,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.5391,
+      "step": 25
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.7480207959415263,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5183,
+      "step": 26
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.6725540446498556,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.5211,
+      "step": 27
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.752524063992389,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.5116,
+      "step": 28
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.8714792547605933,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.5252,
+      "step": 29
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.587686864084582,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.5117,
+      "step": 30
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.7359163335855232,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.5165,
+      "step": 31
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.683074670763304,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.5302,
+      "step": 32
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.7899607579501703,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.5131,
+      "step": 33
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.7203007550703167,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.4912,
+      "step": 34
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.0497807505193113,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.5064,
+      "step": 35
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 1.1723249764639616,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.5085,
+      "step": 36
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.8733264133015174,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.5093,
+      "step": 37
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 1.2145299440192203,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.5151,
+      "step": 38
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.8295361766624695,
+      "learning_rate": 4e-05,
+      "loss": 0.4846,
+      "step": 39
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.8655831204630615,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.4902,
+      "step": 40
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.7831772024924701,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.499,
+      "step": 41
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 1.084114214652019,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.4961,
+      "step": 42
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.9639426639425154,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.5058,
+      "step": 43
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.6394336440537755,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.4806,
+      "step": 44
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.7819635417728371,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.5032,
+      "step": 45
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.7955295880993887,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.5021,
+      "step": 46
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.6671700101502559,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.4857,
+      "step": 47
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.5514514988706577,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.4747,
+      "step": 48
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.6556176731195519,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.4726,
+      "step": 49
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6240651998663801,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.4884,
+      "step": 50
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.5551386581113544,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.4906,
+      "step": 51
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.5904946318868938,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.4821,
+      "step": 52
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.538919779997332,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.4823,
+      "step": 53
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.5333470534988501,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.4719,
+      "step": 54
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.4850010563636148,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.4665,
+      "step": 55
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.44539126503300186,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.4743,
+      "step": 56
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.5032408043020593,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.4723,
+      "step": 57
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.46974810278602536,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.4726,
+      "step": 58
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.4608092351714356,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.4671,
+      "step": 59
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.45516143398106607,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.466,
+      "step": 60
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.4744623675372948,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.4741,
+      "step": 61
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.43856795287418615,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.4601,
+      "step": 62
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.3761545034457208,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.4706,
+      "step": 63
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.41810263359022676,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.467,
+      "step": 64
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.40583434815588343,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.4695,
+      "step": 65
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.3914490774493594,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.4748,
+      "step": 66
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.389130816543567,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.4537,
+      "step": 67
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.4581445439882817,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.4586,
+      "step": 68
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.5126383880525731,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.4724,
+      "step": 69
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.5181747207489492,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.4674,
+      "step": 70
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.4295166893762927,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.4508,
+      "step": 71
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.4868403059192456,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.4655,
+      "step": 72
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.5989861771380917,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.4661,
+      "step": 73
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.5656347544570441,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.4803,
+      "step": 74
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46441753698075355,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.4639,
+      "step": 75
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.510393828467784,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.4708,
+      "step": 76
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.48942431176760787,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.4524,
+      "step": 77
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.4968874160759365,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4643,
+      "step": 78
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 0.8861501634736185,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 0.8012,
+      "step": 79
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 1.231652143097992,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.4398,
+      "step": 80
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.6712625292151638,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.4121,
+      "step": 81
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 1.1356345596864996,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.4285,
+      "step": 82
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.6662175711028467,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.4184,
+      "step": 83
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.8930266388840874,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.4219,
+      "step": 84
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.5713183757359342,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.4266,
+      "step": 85
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.6446930790537315,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.4306,
+      "step": 86
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.5449251995753388,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.4363,
+      "step": 87
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.49840540333271255,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.4198,
+      "step": 88
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.636524520280604,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.427,
+      "step": 89
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.5763577377012762,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.434,
+      "step": 90
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.5518205698347075,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4081,
+      "step": 91
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.5685776152816718,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.4277,
+      "step": 92
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.6757526217251246,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.4176,
+      "step": 93
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.536729337543259,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.4153,
+      "step": 94
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.42544799934531624,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.4172,
+      "step": 95
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.42010888741086944,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.4102,
+      "step": 96
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.43678400955846786,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.4161,
+      "step": 97
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.5082670907410142,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.4197,
+      "step": 98
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.357137321779293,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.4048,
+      "step": 99
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.40817714129576205,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.419,
+      "step": 100
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.47645094237883096,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.4417,
+      "step": 101
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.39257867572254557,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.4215,
+      "step": 102
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.4206670690315583,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.4201,
+      "step": 103
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.3865293036754881,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.4347,
+      "step": 104
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.39367421741174213,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.4206,
+      "step": 105
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.4656673614605536,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.3881,
+      "step": 106
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.35335946805147883,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.4118,
+      "step": 107
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.4628742760548775,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.416,
+      "step": 108
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.39568236935918416,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.4135,
+      "step": 109
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.5466276296066325,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.417,
+      "step": 110
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.5090028901174583,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.4111,
+      "step": 111
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.48855888787545754,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.4036,
+      "step": 112
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.45101035460617356,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.4207,
+      "step": 113
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.5737035543815443,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.4089,
+      "step": 114
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.39146919766674915,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.4164,
+      "step": 115
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.3630072800119681,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.3981,
+      "step": 116
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.4547222351106011,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4196,
+      "step": 117
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.37216593927043445,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.4107,
+      "step": 118
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.4050307805559563,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.4122,
+      "step": 119
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.38269243213429416,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.4073,
+      "step": 120
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.36260299781848226,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.4252,
+      "step": 121
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.4438690588669624,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.4122,
+      "step": 122
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.41544202097631083,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.416,
+      "step": 123
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.641936060850495,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.4205,
+      "step": 124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5487268480206667,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.4185,
+      "step": 125
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.3876577101511856,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.4103,
+      "step": 126
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.42818180937132194,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.4205,
+      "step": 127
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.2680370619157418,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.4256,
+      "step": 128
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.39573261758619904,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.4168,
+      "step": 129
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.3543678918298844,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.411,
+      "step": 130
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.3717876519061496,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.4166,
+      "step": 131
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.41594670318211513,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.4185,
+      "step": 132
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.39133110888303313,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.4139,
+      "step": 133
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.43836287131270635,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.4159,
+      "step": 134
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.3757188844435322,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.4196,
+      "step": 135
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.34971023678161894,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.3985,
+      "step": 136
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.3512090580148879,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.4073,
+      "step": 137
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.34629713258117534,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.4201,
+      "step": 138
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.46977338380502903,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.4087,
+      "step": 139
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.40101157445187413,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.4147,
+      "step": 140
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.47691626196033376,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.4004,
+      "step": 141
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.3435761452724062,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.4126,
+      "step": 142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.3645128134974153,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4007,
+      "step": 143
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.35491499023320583,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.4087,
+      "step": 144
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.3734869903646821,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.419,
+      "step": 145
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.30082037154900365,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.4138,
+      "step": 146
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.4590029162532056,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.4081,
+      "step": 147
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.29447660882935095,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.4009,
+      "step": 148
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.48573583205938736,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.4258,
+      "step": 149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3631988369037435,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.4203,
+      "step": 150
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.3332268388037149,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.4006,
+      "step": 151
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.35727991875553283,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.4094,
+      "step": 152
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.3559631224614753,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.4213,
+      "step": 153
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.37804917682629285,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.4199,
+      "step": 154
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.37950577893585047,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.3961,
+      "step": 155
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.32608515028658464,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4055,
+      "step": 156
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 0.7669011198805715,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.7441,
+      "step": 157
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.6652887388218975,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.3507,
+      "step": 158
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.7385081279732356,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.3736,
+      "step": 159
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.7320172922150582,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.3587,
+      "step": 160
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 0.49807265418424446,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.3623,
+      "step": 161
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.48163598394041085,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.3659,
+      "step": 162
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.48731905853762586,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.3476,
+      "step": 163
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.42827838103728594,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.3553,
+      "step": 164
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.34579240618250967,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.36,
+      "step": 165
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.38008557988183295,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.3481,
+      "step": 166
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.3376533317787802,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.3645,
+      "step": 167
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.348488923971927,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.3635,
+      "step": 168
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.3599933955892379,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3542,
+      "step": 169
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.333690464884174,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.3556,
+      "step": 170
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.3350844064785929,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.3647,
+      "step": 171
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.37148148955239746,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.3554,
+      "step": 172
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.3280010701875189,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.3532,
+      "step": 173
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.332491142609001,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.3544,
+      "step": 174
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2881390860119507,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.3455,
+      "step": 175
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.32116573700586765,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.3621,
+      "step": 176
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.30659625060436463,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.3419,
+      "step": 177
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.30519135259940783,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.3524,
+      "step": 178
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.2980472772642158,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.3493,
+      "step": 179
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.32253352509389444,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.3587,
+      "step": 180
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.27321091859832713,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.35,
+      "step": 181
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.2985443161344337,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3582,
+      "step": 182
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.323527392124704,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.3559,
+      "step": 183
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.2948657008807223,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.3568,
+      "step": 184
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.2991813048805828,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.3532,
+      "step": 185
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.2783668607947485,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.3458,
+      "step": 186
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.28042445346441125,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.3421,
+      "step": 187
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.3471962042525829,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.3453,
+      "step": 188
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.2823227915730697,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.3432,
+      "step": 189
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.3530570165142917,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.3548,
+      "step": 190
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.3056427441596844,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.3613,
+      "step": 191
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.3086741723859052,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.3417,
+      "step": 192
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.31428784464812703,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.3581,
+      "step": 193
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.2859896069564287,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.3474,
+      "step": 194
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.27059096851688597,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3442,
+      "step": 195
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.3326884769262199,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.3683,
+      "step": 196
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.29505923783406746,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.3477,
+      "step": 197
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.32073943104242425,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.3595,
+      "step": 198
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.2840222270592631,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.3473,
+      "step": 199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3418714380147271,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.3521,
+      "step": 200
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.2797908282100119,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.3543,
+      "step": 201
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.36420961265518004,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.3533,
+      "step": 202
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.314568754851869,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.3575,
+      "step": 203
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.3342111687945849,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.346,
+      "step": 204
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.3064734258766498,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.3671,
+      "step": 205
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.29854230408342164,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.3424,
+      "step": 206
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.2827054055912993,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.3526,
+      "step": 207
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.28275958845846305,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3441,
+      "step": 208
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.312064605803253,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.3598,
+      "step": 209
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.28452488535848225,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.3595,
+      "step": 210
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.2744945291582551,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.3514,
+      "step": 211
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.2851263975583895,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.3578,
+      "step": 212
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.24934062786332797,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.3658,
+      "step": 213
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.32531564637396143,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.3475,
+      "step": 214
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.24195021899743993,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.3509,
+      "step": 215
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.33103816433572836,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.3541,
+      "step": 216
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.2269071264266118,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.3602,
+      "step": 217
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.29798422862653967,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.3433,
+      "step": 218
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.25616892838087085,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.3547,
+      "step": 219
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.2569631346633302,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.3678,
+      "step": 220
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.32778475734128903,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3523,
+      "step": 221
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.2369726285253331,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.3449,
+      "step": 222
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.2810512494548086,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.3631,
+      "step": 223
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.24745825973149008,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.3526,
+      "step": 224
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2586808546029739,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.3428,
+      "step": 225
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.23360130898404668,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.345,
+      "step": 226
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.240900500337055,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.3487,
+      "step": 227
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.23747073767970822,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.356,
+      "step": 228
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.2201875141981078,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.3578,
+      "step": 229
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.23449074731345043,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.3619,
+      "step": 230
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.23953740851380273,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.3587,
+      "step": 231
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.20232912490254967,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.3447,
+      "step": 232
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.25646246387237664,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.359,
+      "step": 233
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.22729675252261325,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3487,
+      "step": 234
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 0.6215353715874271,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.6241,
+      "step": 235
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.3696909872798802,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.3034,
+      "step": 236
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.5375575773965741,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.3032,
+      "step": 237
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.38516960983552195,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.3061,
+      "step": 238
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.4742991871870546,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.2937,
+      "step": 239
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.4135969453525089,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.3138,
+      "step": 240
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.42172855815907073,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.308,
+      "step": 241
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.3771894608374879,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.3076,
+      "step": 242
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.4163462974642452,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.3071,
+      "step": 243
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.3704830350749835,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.3056,
+      "step": 244
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.36115694049544883,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.3047,
+      "step": 245
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.3358650426314368,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.2925,
+      "step": 246
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.3744792395559403,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2965,
+      "step": 247
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.3361916360656167,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.3127,
+      "step": 248
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.3091988714486053,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.3035,
+      "step": 249
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3277119477509228,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.299,
+      "step": 250
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.2873917613244393,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.2928,
+      "step": 251
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.28358641106288984,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.291,
+      "step": 252
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.27138723040994367,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.2888,
+      "step": 253
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.2753898544881945,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.2953,
+      "step": 254
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.2647672143999282,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.2912,
+      "step": 255
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.2879829562601937,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.3111,
+      "step": 256
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.2685828458948742,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.2899,
+      "step": 257
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.2618038733902564,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.2916,
+      "step": 258
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.22753375801417833,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.305,
+      "step": 259
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.2930244026853143,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3039,
+      "step": 260
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.22977159381404028,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.2878,
+      "step": 261
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.24270939761725097,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.2949,
+      "step": 262
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.24447085983026712,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.2993,
+      "step": 263
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.23152423854583742,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.2944,
+      "step": 264
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.22985233893166343,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.2955,
+      "step": 265
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.23372953797080312,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.288,
+      "step": 266
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.21690130609713476,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.2904,
+      "step": 267
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.2246463430594481,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.2989,
+      "step": 268
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.21332209794451956,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.2975,
+      "step": 269
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.24970279498059386,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.3064,
+      "step": 270
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.21261729326140266,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.2862,
+      "step": 271
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.22206764450175856,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.3092,
+      "step": 272
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.24047571831325013,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2971,
+      "step": 273
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.21849137758843346,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.2984,
+      "step": 274
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.2229834366517268,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.2961,
+      "step": 275
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.22968975608492073,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.3033,
+      "step": 276
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.22151889834022945,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.3039,
+      "step": 277
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.2128895075658792,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.2943,
+      "step": 278
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.21164754242909364,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.2919,
+      "step": 279
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.20884461582582384,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.2997,
+      "step": 280
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.21541996842202368,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.3002,
+      "step": 281
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.20772455990874986,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.2921,
+      "step": 282
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.21297651398483633,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.311,
+      "step": 283
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.2190112768874125,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.2919,
+      "step": 284
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.21673928683300617,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.2841,
+      "step": 285
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.21487488035766797,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.3014,
+      "step": 286
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.21578163765129071,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.3019,
+      "step": 287
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.224427277921113,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.2976,
+      "step": 288
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.22042655864613764,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.2989,
+      "step": 289
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.20374940342407785,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.2923,
+      "step": 290
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.2114696586781579,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.2969,
+      "step": 291
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.20549511788083236,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.3061,
+      "step": 292
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.19979906515975762,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.2823,
+      "step": 293
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.19229453116296086,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.2997,
+      "step": 294
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.20712539472061123,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.2959,
+      "step": 295
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.2015468139758988,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.2979,
+      "step": 296
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.20097170185132207,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.2926,
+      "step": 297
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.19852055662210327,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.2913,
+      "step": 298
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.18906994406437755,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3006,
+      "step": 299
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.2021153809962757,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.2923,
+      "step": 300
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.1999899974233003,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.2915,
+      "step": 301
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.19630347526748365,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.2866,
+      "step": 302
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.1920809980534528,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.3032,
+      "step": 303
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.1901195320725009,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.2909,
+      "step": 304
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.21032225123095197,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.3084,
+      "step": 305
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.1868686917904852,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.2905,
+      "step": 306
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.18502790597756422,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.2937,
+      "step": 307
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.1928945785679494,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.2989,
+      "step": 308
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.19560615341315463,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.3126,
+      "step": 309
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.18840138055047603,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.2934,
+      "step": 310
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.19588915112265848,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.2905,
+      "step": 311
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.19730007784376413,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3044,
+      "step": 312
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 0.572586285609381,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.5472,
+      "step": 313
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.2949768631584251,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.2704,
+      "step": 314
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.2574596792309237,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.2549,
+      "step": 315
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.21477368558781973,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.2621,
+      "step": 316
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.2832750553273873,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.2838,
+      "step": 317
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.31022611032250574,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.2622,
+      "step": 318
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.2628883778841394,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.2482,
+      "step": 319
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.25960167507018217,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.2637,
+      "step": 320
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.23079097623497014,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.2583,
+      "step": 321
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.23734069583525863,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.2636,
+      "step": 322
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.24231979125793324,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.2677,
+      "step": 323
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.222245031459651,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.2611,
+      "step": 324
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.20312084213998194,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2585,
+      "step": 325
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.197752485913063,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.259,
+      "step": 326
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.2026654928471488,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.2514,
+      "step": 327
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.2226046433033049,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.2717,
+      "step": 328
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.19738358880511364,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.2568,
+      "step": 329
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.20518193085549377,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.2593,
+      "step": 330
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.2010784131419505,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.2672,
+      "step": 331
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.2040581583418926,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.2672,
+      "step": 332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.2051153804496881,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.2577,
+      "step": 333
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.19036151835844245,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.2608,
+      "step": 334
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.1878364416043157,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.2523,
+      "step": 335
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.19031269408103865,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.2545,
+      "step": 336
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.18967066362788515,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.2589,
+      "step": 337
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.18796101285613187,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.264,
+      "step": 338
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.18728077492824852,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.275,
+      "step": 339
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.1920736215287993,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.2668,
+      "step": 340
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.18710726206783124,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.2626,
+      "step": 341
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.18070835403401606,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.2539,
+      "step": 342
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.18040022199428507,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.2683,
+      "step": 343
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.18024941762685995,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.2629,
+      "step": 344
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.17679299731428164,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.2574,
+      "step": 345
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.17535840405973557,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.2583,
+      "step": 346
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.17209557575763554,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.2538,
+      "step": 347
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.17269823781217156,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.2573,
+      "step": 348
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.17569630903609573,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.2636,
+      "step": 349
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.17222091170740247,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.2564,
+      "step": 350
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.17986994282527735,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2658,
+      "step": 351
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.17749306159579448,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.2642,
+      "step": 352
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.17369779306753141,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.2585,
+      "step": 353
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.1762932829589361,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.2676,
+      "step": 354
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.17826711105619464,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.2707,
+      "step": 355
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.17362813852251957,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.2622,
+      "step": 356
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.17946540544791956,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.262,
+      "step": 357
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.17382409916390584,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.2547,
+      "step": 358
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.17625896763947824,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.2623,
+      "step": 359
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.1699245760301564,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.2694,
+      "step": 360
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.16895611272526004,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.2575,
+      "step": 361
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.1693241796677865,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.2459,
+      "step": 362
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.17362449206981356,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.2716,
+      "step": 363
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.17025045249936588,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.2534,
+      "step": 364
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.16875385878273286,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.2586,
+      "step": 365
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.17389976665994425,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.2645,
+      "step": 366
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.17023072304351422,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.2569,
+      "step": 367
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.17186663723394052,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.2643,
+      "step": 368
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.1680634909255728,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.2623,
+      "step": 369
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.17120199576002865,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.2613,
+      "step": 370
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.16412195443224825,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.2589,
+      "step": 371
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.17379655511927136,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.2632,
+      "step": 372
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.1656832409615132,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.2478,
+      "step": 373
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.16951075627616047,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.2516,
+      "step": 374
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.17537557692366856,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.2676,
+      "step": 375
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.17374885403540272,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.2605,
+      "step": 376
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.16888552813586052,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2718,
+      "step": 377
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.17354897645874484,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.2706,
+      "step": 378
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.1700437193077696,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.2622,
+      "step": 379
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.16925795308526098,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.2532,
+      "step": 380
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.17169956156014887,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.2543,
+      "step": 381
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.16991884248389066,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.254,
+      "step": 382
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.17543705217318195,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.2678,
+      "step": 383
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.17309881142676284,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.2528,
+      "step": 384
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.16669028171466665,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.2568,
+      "step": 385
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.16908323842167242,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.2625,
+      "step": 386
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.16654126250710174,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.2552,
+      "step": 387
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.16486951649258705,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.2553,
+      "step": 388
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.16746724674913815,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.2546,
+      "step": 389
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.17103013097653133,
+      "learning_rate": 0.0,
+      "loss": 0.268,
+      "step": 390
+    },
+    {
+      "epoch": 4.992,
+      "step": 390,
+      "total_flos": 1.8773457441390592e+18,
+      "train_loss": 0.3770592579474816,
+      "train_runtime": 19616.6723,
+      "train_samples_per_second": 2.549,
+      "train_steps_per_second": 0.02
+    }
+  ],
+  "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": 1.8773457441390592e+18,
+  "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 sha256:51820805243d613b44a07ac793c7b5b52f0bc131ac862cbc9a64cf2a5a9e9504
+size 7352