Upload model

.gitattributes

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

README.md

@@ -0,0 +1,61 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen2.5-7B-Instruct
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: d1_science_all_10k
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# d1_science_all_10k
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/d1_science_all_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": 8.005281734805422e+17,
+    "train_loss": 0.40785385312942357,
+    "train_runtime": 14479.4896,
+    "train_samples_per_second": 3.453,
+    "train_steps_per_second": 0.027
+}

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 sha256:6b9f3c1959fdde5cebf736c1b5b3a9135284a10e6d471da185ff9f58e8cd401f
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f690c49dc588f9496421d0fc08708cd3afa5dbc8ce4afc3ab80eb5c12d6644ae
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:933091a420944f60415a560be36895f630e58255a42c43445c4ef6a9877de11a
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a3b7c27e17cd4e2f6037cd9221a4ed95bf925f16a1f25a5fae161ee539c512e
+size 1089994880

model.safetensors.index.json

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

runs/Apr27_23-26-23_jpbot-001-21.jupiter.internal/events.out.tfevents.1745789210.jpbot-001-21.jupiter.internal.2074777.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20899bd85e48ba72878ab7f3399177ab6ab0f036a1719f447af5e91ceeedeb06
+size 87658

special_tokens_map.json

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

start_end.json

@@ -0,0 +1 @@
+{"start_time": "2025-04-27 23:26:23", "end_time": "2025-04-28 03:28:16"}

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": 8.005281734805422e+17,
+    "train_loss": 0.40785385312942357,
+    "train_runtime": 14479.4896,
+    "train_samples_per_second": 3.453,
+    "train_steps_per_second": 0.027
+}

trainer_log.jsonl

@@ -0,0 +1,391 @@
+{"current_steps": 1, "total_steps": 390, "loss": 0.9596, "lr": 1.0256410256410257e-06, "epoch": 0.0128, "percentage": 0.26, "elapsed_time": "0:00:43", "remaining_time": "4:39:26"}
+{"current_steps": 2, "total_steps": 390, "loss": 1.0162, "lr": 2.0512820512820513e-06, "epoch": 0.0256, "percentage": 0.51, "elapsed_time": "0:01:20", "remaining_time": "4:19:45"}
+{"current_steps": 3, "total_steps": 390, "loss": 1.0178, "lr": 3.0769230769230774e-06, "epoch": 0.0384, "percentage": 0.77, "elapsed_time": "0:01:53", "remaining_time": "4:04:33"}
+{"current_steps": 4, "total_steps": 390, "loss": 0.9835, "lr": 4.102564102564103e-06, "epoch": 0.0512, "percentage": 1.03, "elapsed_time": "0:02:25", "remaining_time": "3:54:32"}
+{"current_steps": 5, "total_steps": 390, "loss": 0.9483, "lr": 5.128205128205128e-06, "epoch": 0.064, "percentage": 1.28, "elapsed_time": "0:03:00", "remaining_time": "3:51:36"}
+{"current_steps": 6, "total_steps": 390, "loss": 0.9022, "lr": 6.153846153846155e-06, "epoch": 0.0768, "percentage": 1.54, "elapsed_time": "0:03:33", "remaining_time": "3:47:52"}
+{"current_steps": 7, "total_steps": 390, "loss": 0.8501, "lr": 7.17948717948718e-06, "epoch": 0.0896, "percentage": 1.79, "elapsed_time": "0:04:15", "remaining_time": "3:53:08"}
+{"current_steps": 8, "total_steps": 390, "loss": 0.8768, "lr": 8.205128205128205e-06, "epoch": 0.1024, "percentage": 2.05, "elapsed_time": "0:04:54", "remaining_time": "3:54:42"}
+{"current_steps": 9, "total_steps": 390, "loss": 0.8887, "lr": 9.230769230769232e-06, "epoch": 0.1152, "percentage": 2.31, "elapsed_time": "0:05:32", "remaining_time": "3:54:15"}
+{"current_steps": 10, "total_steps": 390, "loss": 0.8776, "lr": 1.0256410256410256e-05, "epoch": 0.128, "percentage": 2.56, "elapsed_time": "0:06:04", "remaining_time": "3:51:05"}
+{"current_steps": 11, "total_steps": 390, "loss": 0.8272, "lr": 1.1282051282051283e-05, "epoch": 0.1408, "percentage": 2.82, "elapsed_time": "0:06:40", "remaining_time": "3:50:15"}
+{"current_steps": 12, "total_steps": 390, "loss": 0.8057, "lr": 1.230769230769231e-05, "epoch": 0.1536, "percentage": 3.08, "elapsed_time": "0:07:17", "remaining_time": "3:49:53"}
+{"current_steps": 13, "total_steps": 390, "loss": 0.8367, "lr": 1.3333333333333333e-05, "epoch": 0.1664, "percentage": 3.33, "elapsed_time": "0:07:47", "remaining_time": "3:45:53"}
+{"current_steps": 14, "total_steps": 390, "loss": 0.7608, "lr": 1.435897435897436e-05, "epoch": 0.1792, "percentage": 3.59, "elapsed_time": "0:08:31", "remaining_time": "3:49:05"}
+{"current_steps": 15, "total_steps": 390, "loss": 0.7656, "lr": 1.5384615384615387e-05, "epoch": 0.192, "percentage": 3.85, "elapsed_time": "0:09:13", "remaining_time": "3:50:31"}
+{"current_steps": 16, "total_steps": 390, "loss": 0.7609, "lr": 1.641025641025641e-05, "epoch": 0.2048, "percentage": 4.1, "elapsed_time": "0:09:51", "remaining_time": "3:50:27"}
+{"current_steps": 17, "total_steps": 390, "loss": 0.732, "lr": 1.7435897435897438e-05, "epoch": 0.2176, "percentage": 4.36, "elapsed_time": "0:10:25", "remaining_time": "3:48:54"}
+{"current_steps": 18, "total_steps": 390, "loss": 0.7513, "lr": 1.8461538461538465e-05, "epoch": 0.2304, "percentage": 4.62, "elapsed_time": "0:10:55", "remaining_time": "3:45:55"}
+{"current_steps": 19, "total_steps": 390, "loss": 0.7624, "lr": 1.9487179487179488e-05, "epoch": 0.2432, "percentage": 4.87, "elapsed_time": "0:11:35", "remaining_time": "3:46:12"}
+{"current_steps": 20, "total_steps": 390, "loss": 0.7191, "lr": 2.0512820512820512e-05, "epoch": 0.256, "percentage": 5.13, "elapsed_time": "0:12:14", "remaining_time": "3:46:19"}
+{"current_steps": 21, "total_steps": 390, "loss": 0.7181, "lr": 2.153846153846154e-05, "epoch": 0.2688, "percentage": 5.38, "elapsed_time": "0:12:53", "remaining_time": "3:46:22"}
+{"current_steps": 22, "total_steps": 390, "loss": 0.6926, "lr": 2.2564102564102566e-05, "epoch": 0.2816, "percentage": 5.64, "elapsed_time": "0:13:31", "remaining_time": "3:46:17"}
+{"current_steps": 23, "total_steps": 390, "loss": 0.7319, "lr": 2.3589743589743593e-05, "epoch": 0.2944, "percentage": 5.9, "elapsed_time": "0:14:03", "remaining_time": "3:44:19"}
+{"current_steps": 24, "total_steps": 390, "loss": 0.6766, "lr": 2.461538461538462e-05, "epoch": 0.3072, "percentage": 6.15, "elapsed_time": "0:14:45", "remaining_time": "3:45:00"}
+{"current_steps": 25, "total_steps": 390, "loss": 0.7083, "lr": 2.5641025641025646e-05, "epoch": 0.32, "percentage": 6.41, "elapsed_time": "0:15:21", "remaining_time": "3:44:14"}
+{"current_steps": 26, "total_steps": 390, "loss": 0.686, "lr": 2.6666666666666667e-05, "epoch": 0.3328, "percentage": 6.67, "elapsed_time": "0:15:57", "remaining_time": "3:43:29"}
+{"current_steps": 27, "total_steps": 390, "loss": 0.6769, "lr": 2.7692307692307694e-05, "epoch": 0.3456, "percentage": 6.92, "elapsed_time": "0:16:31", "remaining_time": "3:42:09"}
+{"current_steps": 28, "total_steps": 390, "loss": 0.6911, "lr": 2.871794871794872e-05, "epoch": 0.3584, "percentage": 7.18, "elapsed_time": "0:17:05", "remaining_time": "3:40:54"}
+{"current_steps": 29, "total_steps": 390, "loss": 0.6811, "lr": 2.9743589743589747e-05, "epoch": 0.3712, "percentage": 7.44, "elapsed_time": "0:17:44", "remaining_time": "3:40:46"}
+{"current_steps": 30, "total_steps": 390, "loss": 0.6782, "lr": 3.0769230769230774e-05, "epoch": 0.384, "percentage": 7.69, "elapsed_time": "0:18:28", "remaining_time": "3:41:38"}
+{"current_steps": 31, "total_steps": 390, "loss": 0.686, "lr": 3.1794871794871795e-05, "epoch": 0.3968, "percentage": 7.95, "elapsed_time": "0:18:59", "remaining_time": "3:40:00"}
+{"current_steps": 32, "total_steps": 390, "loss": 0.6669, "lr": 3.282051282051282e-05, "epoch": 0.4096, "percentage": 8.21, "elapsed_time": "0:19:32", "remaining_time": "3:38:40"}
+{"current_steps": 33, "total_steps": 390, "loss": 0.6685, "lr": 3.384615384615385e-05, "epoch": 0.4224, "percentage": 8.46, "elapsed_time": "0:20:13", "remaining_time": "3:38:52"}
+{"current_steps": 34, "total_steps": 390, "loss": 0.6574, "lr": 3.4871794871794875e-05, "epoch": 0.4352, "percentage": 8.72, "elapsed_time": "0:20:52", "remaining_time": "3:38:35"}
+{"current_steps": 35, "total_steps": 390, "loss": 0.6749, "lr": 3.58974358974359e-05, "epoch": 0.448, "percentage": 8.97, "elapsed_time": "0:21:30", "remaining_time": "3:38:10"}
+{"current_steps": 36, "total_steps": 390, "loss": 0.6632, "lr": 3.692307692307693e-05, "epoch": 0.4608, "percentage": 9.23, "elapsed_time": "0:22:08", "remaining_time": "3:37:41"}
+{"current_steps": 37, "total_steps": 390, "loss": 0.6512, "lr": 3.794871794871795e-05, "epoch": 0.4736, "percentage": 9.49, "elapsed_time": "0:22:47", "remaining_time": "3:37:31"}
+{"current_steps": 38, "total_steps": 390, "loss": 0.6663, "lr": 3.8974358974358976e-05, "epoch": 0.4864, "percentage": 9.74, "elapsed_time": "0:23:22", "remaining_time": "3:36:30"}
+{"current_steps": 39, "total_steps": 390, "loss": 0.6699, "lr": 4e-05, "epoch": 0.4992, "percentage": 10.0, "elapsed_time": "0:23:58", "remaining_time": "3:35:47"}
+{"current_steps": 40, "total_steps": 390, "loss": 0.6103, "lr": 3.9999198907597046e-05, "epoch": 0.512, "percentage": 10.26, "elapsed_time": "0:24:38", "remaining_time": "3:35:40"}
+{"current_steps": 41, "total_steps": 390, "loss": 0.6294, "lr": 3.9996795694563096e-05, "epoch": 0.5248, "percentage": 10.51, "elapsed_time": "0:25:10", "remaining_time": "3:34:21"}
+{"current_steps": 42, "total_steps": 390, "loss": 0.63, "lr": 3.999279055341771e-05, "epoch": 0.5376, "percentage": 10.77, "elapsed_time": "0:25:50", "remaining_time": "3:34:04"}
+{"current_steps": 43, "total_steps": 390, "loss": 0.6555, "lr": 3.998718380500971e-05, "epoch": 0.5504, "percentage": 11.03, "elapsed_time": "0:26:28", "remaining_time": "3:33:37"}
+{"current_steps": 44, "total_steps": 390, "loss": 0.6747, "lr": 3.997997589849145e-05, "epoch": 0.5632, "percentage": 11.28, "elapsed_time": "0:27:04", "remaining_time": "3:32:56"}
+{"current_steps": 45, "total_steps": 390, "loss": 0.6803, "lr": 3.9971167411282835e-05, "epoch": 0.576, "percentage": 11.54, "elapsed_time": "0:27:38", "remaining_time": "3:31:51"}
+{"current_steps": 46, "total_steps": 390, "loss": 0.6544, "lr": 3.99607590490251e-05, "epoch": 0.5888, "percentage": 11.79, "elapsed_time": "0:28:16", "remaining_time": "3:31:28"}
+{"current_steps": 47, "total_steps": 390, "loss": 0.6655, "lr": 3.9948751645524235e-05, "epoch": 0.6016, "percentage": 12.05, "elapsed_time": "0:28:50", "remaining_time": "3:30:28"}
+{"current_steps": 48, "total_steps": 390, "loss": 0.6521, "lr": 3.9935146162684206e-05, "epoch": 0.6144, "percentage": 12.31, "elapsed_time": "0:29:28", "remaining_time": "3:29:59"}
+{"current_steps": 49, "total_steps": 390, "loss": 0.6357, "lr": 3.9919943690429906e-05, "epoch": 0.6272, "percentage": 12.56, "elapsed_time": "0:30:09", "remaining_time": "3:29:52"}
+{"current_steps": 50, "total_steps": 390, "loss": 0.6443, "lr": 3.9903145446619837e-05, "epoch": 0.64, "percentage": 12.82, "elapsed_time": "0:30:44", "remaining_time": "3:29:01"}
+{"current_steps": 51, "total_steps": 390, "loss": 0.6465, "lr": 3.9884752776948564e-05, "epoch": 0.6528, "percentage": 13.08, "elapsed_time": "0:31:24", "remaining_time": "3:28:43"}
+{"current_steps": 52, "total_steps": 390, "loss": 0.5951, "lr": 3.9864767154838864e-05, "epoch": 0.6656, "percentage": 13.33, "elapsed_time": "0:32:00", "remaining_time": "3:28:04"}
+{"current_steps": 53, "total_steps": 390, "loss": 0.6357, "lr": 3.9843190181323744e-05, "epoch": 0.6784, "percentage": 13.59, "elapsed_time": "0:32:34", "remaining_time": "3:27:09"}
+{"current_steps": 54, "total_steps": 390, "loss": 0.6657, "lr": 3.982002358491817e-05, "epoch": 0.6912, "percentage": 13.85, "elapsed_time": "0:33:08", "remaining_time": "3:26:10"}
+{"current_steps": 55, "total_steps": 390, "loss": 0.62, "lr": 3.979526922148058e-05, "epoch": 0.704, "percentage": 14.1, "elapsed_time": "0:33:44", "remaining_time": "3:25:30"}
+{"current_steps": 56, "total_steps": 390, "loss": 0.6552, "lr": 3.9768929074064206e-05, "epoch": 0.7168, "percentage": 14.36, "elapsed_time": "0:34:22", "remaining_time": "3:25:01"}
+{"current_steps": 57, "total_steps": 390, "loss": 0.6265, "lr": 3.9741005252758255e-05, "epoch": 0.7296, "percentage": 14.62, "elapsed_time": "0:34:58", "remaining_time": "3:24:18"}
+{"current_steps": 58, "total_steps": 390, "loss": 0.6316, "lr": 3.971149999451886e-05, "epoch": 0.7424, "percentage": 14.87, "elapsed_time": "0:35:32", "remaining_time": "3:23:25"}
+{"current_steps": 59, "total_steps": 390, "loss": 0.634, "lr": 3.9680415662989806e-05, "epoch": 0.7552, "percentage": 15.13, "elapsed_time": "0:36:10", "remaining_time": "3:22:56"}
+{"current_steps": 60, "total_steps": 390, "loss": 0.6566, "lr": 3.9647754748313294e-05, "epoch": 0.768, "percentage": 15.38, "elapsed_time": "0:36:48", "remaining_time": "3:22:25"}
+{"current_steps": 61, "total_steps": 390, "loss": 0.6418, "lr": 3.96135198669304e-05, "epoch": 0.7808, "percentage": 15.64, "elapsed_time": "0:37:21", "remaining_time": "3:21:28"}
+{"current_steps": 62, "total_steps": 390, "loss": 0.6489, "lr": 3.957771376137144e-05, "epoch": 0.7936, "percentage": 15.9, "elapsed_time": "0:38:00", "remaining_time": "3:21:07"}
+{"current_steps": 63, "total_steps": 390, "loss": 0.6625, "lr": 3.954033930003634e-05, "epoch": 0.8064, "percentage": 16.15, "elapsed_time": "0:38:32", "remaining_time": "3:20:03"}
+{"current_steps": 64, "total_steps": 390, "loss": 0.6287, "lr": 3.9501399476964806e-05, "epoch": 0.8192, "percentage": 16.41, "elapsed_time": "0:39:06", "remaining_time": "3:19:13"}
+{"current_steps": 65, "total_steps": 390, "loss": 0.615, "lr": 3.946089741159648e-05, "epoch": 0.832, "percentage": 16.67, "elapsed_time": "0:39:42", "remaining_time": "3:18:30"}
+{"current_steps": 66, "total_steps": 390, "loss": 0.6237, "lr": 3.9418836348521045e-05, "epoch": 0.8448, "percentage": 16.92, "elapsed_time": "0:40:20", "remaining_time": "3:18:02"}
+{"current_steps": 67, "total_steps": 390, "loss": 0.6343, "lr": 3.937521965721831e-05, "epoch": 0.8576, "percentage": 17.18, "elapsed_time": "0:41:00", "remaining_time": "3:17:42"}
+{"current_steps": 68, "total_steps": 390, "loss": 0.5971, "lr": 3.933005083178828e-05, "epoch": 0.8704, "percentage": 17.44, "elapsed_time": "0:41:37", "remaining_time": "3:17:04"}
+{"current_steps": 69, "total_steps": 390, "loss": 0.6243, "lr": 3.928333349067125e-05, "epoch": 0.8832, "percentage": 17.69, "elapsed_time": "0:42:13", "remaining_time": "3:16:24"}
+{"current_steps": 70, "total_steps": 390, "loss": 0.6372, "lr": 3.923507137635792e-05, "epoch": 0.896, "percentage": 17.95, "elapsed_time": "0:42:54", "remaining_time": "3:16:09"}
+{"current_steps": 71, "total_steps": 390, "loss": 0.6413, "lr": 3.9185268355089606e-05, "epoch": 0.9088, "percentage": 18.21, "elapsed_time": "0:43:26", "remaining_time": "3:15:09"}
+{"current_steps": 72, "total_steps": 390, "loss": 0.6426, "lr": 3.913392841654851e-05, "epoch": 0.9216, "percentage": 18.46, "elapsed_time": "0:43:59", "remaining_time": "3:14:16"}
+{"current_steps": 73, "total_steps": 390, "loss": 0.6203, "lr": 3.9081055673538093e-05, "epoch": 0.9344, "percentage": 18.72, "elapsed_time": "0:44:36", "remaining_time": "3:13:41"}
+{"current_steps": 74, "total_steps": 390, "loss": 0.6412, "lr": 3.902665436165364e-05, "epoch": 0.9472, "percentage": 18.97, "elapsed_time": "0:45:12", "remaining_time": "3:13:03"}
+{"current_steps": 75, "total_steps": 390, "loss": 0.6329, "lr": 3.897072883894291e-05, "epoch": 0.96, "percentage": 19.23, "elapsed_time": "0:45:54", "remaining_time": "3:12:49"}
+{"current_steps": 76, "total_steps": 390, "loss": 0.6135, "lr": 3.8913283585557054e-05, "epoch": 0.9728, "percentage": 19.49, "elapsed_time": "0:46:29", "remaining_time": "3:12:06"}
+{"current_steps": 77, "total_steps": 390, "loss": 0.5911, "lr": 3.885432320339167e-05, "epoch": 0.9856, "percentage": 19.74, "elapsed_time": "0:47:09", "remaining_time": "3:11:42"}
+{"current_steps": 78, "total_steps": 390, "loss": 0.6083, "lr": 3.879385241571817e-05, "epoch": 0.9984, "percentage": 20.0, "elapsed_time": "0:47:45", "remaining_time": "3:11:02"}
+{"current_steps": 79, "total_steps": 390, "loss": 1.2361, "lr": 3.873187606680543e-05, "epoch": 1.0112, "percentage": 20.26, "elapsed_time": "0:48:45", "remaining_time": "3:11:56"}
+{"current_steps": 80, "total_steps": 390, "loss": 0.5482, "lr": 3.866839912153168e-05, "epoch": 1.024, "percentage": 20.51, "elapsed_time": "0:49:21", "remaining_time": "3:11:15"}
+{"current_steps": 81, "total_steps": 390, "loss": 0.5149, "lr": 3.860342666498677e-05, "epoch": 1.0368, "percentage": 20.77, "elapsed_time": "0:49:59", "remaining_time": "3:10:41"}
+{"current_steps": 82, "total_steps": 390, "loss": 0.4693, "lr": 3.853696390206484e-05, "epoch": 1.0496, "percentage": 21.03, "elapsed_time": "0:50:35", "remaining_time": "3:10:00"}
+{"current_steps": 83, "total_steps": 390, "loss": 0.5323, "lr": 3.846901615704734e-05, "epoch": 1.0624, "percentage": 21.28, "elapsed_time": "0:51:17", "remaining_time": "3:09:43"}
+{"current_steps": 84, "total_steps": 390, "loss": 0.5492, "lr": 3.839958887317649e-05, "epoch": 1.0752, "percentage": 21.54, "elapsed_time": "0:51:54", "remaining_time": "3:09:05"}
+{"current_steps": 85, "total_steps": 390, "loss": 0.5168, "lr": 3.832868761221926e-05, "epoch": 1.088, "percentage": 21.79, "elapsed_time": "0:52:31", "remaining_time": "3:08:27"}
+{"current_steps": 86, "total_steps": 390, "loss": 0.5329, "lr": 3.825631805402182e-05, "epoch": 1.1008, "percentage": 22.05, "elapsed_time": "0:53:08", "remaining_time": "3:07:51"}
+{"current_steps": 87, "total_steps": 390, "loss": 0.5412, "lr": 3.818248599605448e-05, "epoch": 1.1136, "percentage": 22.31, "elapsed_time": "0:53:41", "remaining_time": "3:06:59"}
+{"current_steps": 88, "total_steps": 390, "loss": 0.5296, "lr": 3.810719735294731e-05, "epoch": 1.1264, "percentage": 22.56, "elapsed_time": "0:54:19", "remaining_time": "3:06:27"}
+{"current_steps": 89, "total_steps": 390, "loss": 0.534, "lr": 3.8030458156016326e-05, "epoch": 1.1392, "percentage": 22.82, "elapsed_time": "0:54:51", "remaining_time": "3:05:32"}
+{"current_steps": 90, "total_steps": 390, "loss": 0.5218, "lr": 3.795227455278029e-05, "epoch": 1.152, "percentage": 23.08, "elapsed_time": "0:55:28", "remaining_time": "3:04:55"}
+{"current_steps": 91, "total_steps": 390, "loss": 0.5101, "lr": 3.787265280646825e-05, "epoch": 1.1648, "percentage": 23.33, "elapsed_time": "0:56:04", "remaining_time": "3:04:15"}
+{"current_steps": 92, "total_steps": 390, "loss": 0.5607, "lr": 3.7791599295517825e-05, "epoch": 1.1776, "percentage": 23.59, "elapsed_time": "0:56:44", "remaining_time": "3:03:46"}
+{"current_steps": 93, "total_steps": 390, "loss": 0.4887, "lr": 3.7709120513064196e-05, "epoch": 1.1904, "percentage": 23.85, "elapsed_time": "0:57:12", "remaining_time": "3:02:42"}
+{"current_steps": 94, "total_steps": 390, "loss": 0.5348, "lr": 3.762522306641998e-05, "epoch": 1.2032, "percentage": 24.1, "elapsed_time": "0:57:53", "remaining_time": "3:02:18"}
+{"current_steps": 95, "total_steps": 390, "loss": 0.5112, "lr": 3.7539913676545874e-05, "epoch": 1.216, "percentage": 24.36, "elapsed_time": "0:58:29", "remaining_time": "3:01:39"}
+{"current_steps": 96, "total_steps": 390, "loss": 0.5239, "lr": 3.745319917751229e-05, "epoch": 1.2288000000000001, "percentage": 24.62, "elapsed_time": "0:59:03", "remaining_time": "3:00:51"}
+{"current_steps": 97, "total_steps": 390, "loss": 0.5178, "lr": 3.736508651595188e-05, "epoch": 1.2416, "percentage": 24.87, "elapsed_time": "0:59:39", "remaining_time": "3:00:13"}
+{"current_steps": 98, "total_steps": 390, "loss": 0.4831, "lr": 3.727558275050301e-05, "epoch": 1.2544, "percentage": 25.13, "elapsed_time": "1:00:17", "remaining_time": "2:59:38"}
+{"current_steps": 99, "total_steps": 390, "loss": 0.5062, "lr": 3.718469505124434e-05, "epoch": 1.2671999999999999, "percentage": 25.38, "elapsed_time": "1:00:55", "remaining_time": "2:59:04"}
+{"current_steps": 100, "total_steps": 390, "loss": 0.5333, "lr": 3.709243069912041e-05, "epoch": 1.28, "percentage": 25.64, "elapsed_time": "1:01:29", "remaining_time": "2:58:19"}
+{"current_steps": 101, "total_steps": 390, "loss": 0.5002, "lr": 3.699879708535838e-05, "epoch": 1.2928, "percentage": 25.9, "elapsed_time": "1:02:07", "remaining_time": "2:57:46"}
+{"current_steps": 102, "total_steps": 390, "loss": 0.5048, "lr": 3.69038017108759e-05, "epoch": 1.3056, "percentage": 26.15, "elapsed_time": "1:02:41", "remaining_time": "2:57:01"}
+{"current_steps": 103, "total_steps": 390, "loss": 0.5119, "lr": 3.680745218568026e-05, "epoch": 1.3184, "percentage": 26.41, "elapsed_time": "1:03:17", "remaining_time": "2:56:21"}
+{"current_steps": 104, "total_steps": 390, "loss": 0.5088, "lr": 3.6709756228258735e-05, "epoch": 1.3312, "percentage": 26.67, "elapsed_time": "1:03:54", "remaining_time": "2:55:45"}
+{"current_steps": 105, "total_steps": 390, "loss": 0.5037, "lr": 3.6610721664960236e-05, "epoch": 1.3439999999999999, "percentage": 26.92, "elapsed_time": "1:04:34", "remaining_time": "2:55:16"}
+{"current_steps": 106, "total_steps": 390, "loss": 0.5186, "lr": 3.65103564293684e-05, "epoch": 1.3568, "percentage": 27.18, "elapsed_time": "1:05:05", "remaining_time": "2:54:24"}
+{"current_steps": 107, "total_steps": 390, "loss": 0.5176, "lr": 3.640866856166601e-05, "epoch": 1.3696, "percentage": 27.44, "elapsed_time": "1:05:43", "remaining_time": "2:53:49"}
+{"current_steps": 108, "total_steps": 390, "loss": 0.4977, "lr": 3.6305666207990886e-05, "epoch": 1.3824, "percentage": 27.69, "elapsed_time": "1:06:20", "remaining_time": "2:53:12"}
+{"current_steps": 109, "total_steps": 390, "loss": 0.4976, "lr": 3.6201357619783336e-05, "epoch": 1.3952, "percentage": 27.95, "elapsed_time": "1:06:58", "remaining_time": "2:52:39"}
+{"current_steps": 110, "total_steps": 390, "loss": 0.5085, "lr": 3.609575115312511e-05, "epoch": 1.408, "percentage": 28.21, "elapsed_time": "1:07:36", "remaining_time": "2:52:04"}
+{"current_steps": 111, "total_steps": 390, "loss": 0.5002, "lr": 3.598885526807003e-05, "epoch": 1.4208, "percentage": 28.46, "elapsed_time": "1:08:12", "remaining_time": "2:51:25"}
+{"current_steps": 112, "total_steps": 390, "loss": 0.4976, "lr": 3.5880678527966224e-05, "epoch": 1.4336, "percentage": 28.72, "elapsed_time": "1:08:48", "remaining_time": "2:50:47"}
+{"current_steps": 113, "total_steps": 390, "loss": 0.4968, "lr": 3.577122959877017e-05, "epoch": 1.4464000000000001, "percentage": 28.97, "elapsed_time": "1:09:27", "remaining_time": "2:50:15"}
+{"current_steps": 114, "total_steps": 390, "loss": 0.5287, "lr": 3.566051724835245e-05, "epoch": 1.4592, "percentage": 29.23, "elapsed_time": "1:10:03", "remaining_time": "2:49:37"}
+{"current_steps": 115, "total_steps": 390, "loss": 0.4974, "lr": 3.554855034579532e-05, "epoch": 1.472, "percentage": 29.49, "elapsed_time": "1:10:45", "remaining_time": "2:49:11"}
+{"current_steps": 116, "total_steps": 390, "loss": 0.4622, "lr": 3.5435337860682304e-05, "epoch": 1.4848, "percentage": 29.74, "elapsed_time": "1:11:16", "remaining_time": "2:48:21"}
+{"current_steps": 117, "total_steps": 390, "loss": 0.5288, "lr": 3.532088886237956e-05, "epoch": 1.4976, "percentage": 30.0, "elapsed_time": "1:11:58", "remaining_time": "2:47:55"}
+{"current_steps": 118, "total_steps": 390, "loss": 0.5132, "lr": 3.520521251930941e-05, "epoch": 1.5104, "percentage": 30.26, "elapsed_time": "1:12:37", "remaining_time": "2:47:24"}
+{"current_steps": 119, "total_steps": 390, "loss": 0.4653, "lr": 3.5088318098215805e-05, "epoch": 1.5232, "percentage": 30.51, "elapsed_time": "1:13:12", "remaining_time": "2:46:42"}
+{"current_steps": 120, "total_steps": 390, "loss": 0.527, "lr": 3.497021496342203e-05, "epoch": 1.536, "percentage": 30.77, "elapsed_time": "1:13:47", "remaining_time": "2:46:02"}
+{"current_steps": 121, "total_steps": 390, "loss": 0.522, "lr": 3.485091257608047e-05, "epoch": 1.5488, "percentage": 31.03, "elapsed_time": "1:14:25", "remaining_time": "2:45:27"}
+{"current_steps": 122, "total_steps": 390, "loss": 0.516, "lr": 3.473042049341474e-05, "epoch": 1.5615999999999999, "percentage": 31.28, "elapsed_time": "1:15:04", "remaining_time": "2:44:54"}
+{"current_steps": 123, "total_steps": 390, "loss": 0.4763, "lr": 3.4608748367954064e-05, "epoch": 1.5744, "percentage": 31.54, "elapsed_time": "1:15:41", "remaining_time": "2:44:17"}
+{"current_steps": 124, "total_steps": 390, "loss": 0.4972, "lr": 3.4485905946759965e-05, "epoch": 1.5872000000000002, "percentage": 31.79, "elapsed_time": "1:16:20", "remaining_time": "2:43:46"}
+{"current_steps": 125, "total_steps": 390, "loss": 0.5135, "lr": 3.4361903070645484e-05, "epoch": 1.6, "percentage": 32.05, "elapsed_time": "1:16:58", "remaining_time": "2:43:10"}
+{"current_steps": 126, "total_steps": 390, "loss": 0.5075, "lr": 3.423674967338681e-05, "epoch": 1.6128, "percentage": 32.31, "elapsed_time": "1:17:28", "remaining_time": "2:42:18"}
+{"current_steps": 127, "total_steps": 390, "loss": 0.4865, "lr": 3.411045578092754e-05, "epoch": 1.6256, "percentage": 32.56, "elapsed_time": "1:18:03", "remaining_time": "2:41:39"}
+{"current_steps": 128, "total_steps": 390, "loss": 0.5053, "lr": 3.398303151057543e-05, "epoch": 1.6383999999999999, "percentage": 32.82, "elapsed_time": "1:18:34", "remaining_time": "2:40:50"}
+{"current_steps": 129, "total_steps": 390, "loss": 0.4876, "lr": 3.385448707019199e-05, "epoch": 1.6512, "percentage": 33.08, "elapsed_time": "1:19:09", "remaining_time": "2:40:09"}
+{"current_steps": 130, "total_steps": 390, "loss": 0.5109, "lr": 3.372483275737468e-05, "epoch": 1.6640000000000001, "percentage": 33.33, "elapsed_time": "1:19:50", "remaining_time": "2:39:41"}
+{"current_steps": 131, "total_steps": 390, "loss": 0.5027, "lr": 3.359407895863199e-05, "epoch": 1.6768, "percentage": 33.59, "elapsed_time": "1:20:29", "remaining_time": "2:39:08"}
+{"current_steps": 132, "total_steps": 390, "loss": 0.4732, "lr": 3.34622361485514e-05, "epoch": 1.6896, "percentage": 33.85, "elapsed_time": "1:21:04", "remaining_time": "2:38:27"}
+{"current_steps": 133, "total_steps": 390, "loss": 0.4814, "lr": 3.332931488896029e-05, "epoch": 1.7024, "percentage": 34.1, "elapsed_time": "1:21:38", "remaining_time": "2:37:44"}
+{"current_steps": 134, "total_steps": 390, "loss": 0.5222, "lr": 3.319532582807977e-05, "epoch": 1.7151999999999998, "percentage": 34.36, "elapsed_time": "1:22:12", "remaining_time": "2:37:03"}
+{"current_steps": 135, "total_steps": 390, "loss": 0.5069, "lr": 3.30602796996717e-05, "epoch": 1.728, "percentage": 34.62, "elapsed_time": "1:22:45", "remaining_time": "2:36:18"}
+{"current_steps": 136, "total_steps": 390, "loss": 0.5171, "lr": 3.2924187322178865e-05, "epoch": 1.7408000000000001, "percentage": 34.87, "elapsed_time": "1:23:27", "remaining_time": "2:35:51"}
+{"current_steps": 137, "total_steps": 390, "loss": 0.5164, "lr": 3.278705959785821e-05, "epoch": 1.7536, "percentage": 35.13, "elapsed_time": "1:24:07", "remaining_time": "2:35:21"}
+{"current_steps": 138, "total_steps": 390, "loss": 0.4934, "lr": 3.2648907511907544e-05, "epoch": 1.7664, "percentage": 35.38, "elapsed_time": "1:24:48", "remaining_time": "2:34:51"}
+{"current_steps": 139, "total_steps": 390, "loss": 0.5172, "lr": 3.250974213158555e-05, "epoch": 1.7792, "percentage": 35.64, "elapsed_time": "1:25:28", "remaining_time": "2:34:20"}
+{"current_steps": 140, "total_steps": 390, "loss": 0.5196, "lr": 3.23695746053251e-05, "epoch": 1.792, "percentage": 35.9, "elapsed_time": "1:26:04", "remaining_time": "2:33:42"}
+{"current_steps": 141, "total_steps": 390, "loss": 0.4828, "lr": 3.222841616184025e-05, "epoch": 1.8048, "percentage": 36.15, "elapsed_time": "1:26:36", "remaining_time": "2:32:57"}
+{"current_steps": 142, "total_steps": 390, "loss": 0.5066, "lr": 3.208627810922665e-05, "epoch": 1.8176, "percentage": 36.41, "elapsed_time": "1:27:16", "remaining_time": "2:32:26"}
+{"current_steps": 143, "total_steps": 390, "loss": 0.4885, "lr": 3.194317183405573e-05, "epoch": 1.8304, "percentage": 36.67, "elapsed_time": "1:27:54", "remaining_time": "2:31:50"}
+{"current_steps": 144, "total_steps": 390, "loss": 0.4752, "lr": 3.1799108800462466e-05, "epoch": 1.8432, "percentage": 36.92, "elapsed_time": "1:28:24", "remaining_time": "2:31:02"}
+{"current_steps": 145, "total_steps": 390, "loss": 0.5025, "lr": 3.1654100549227024e-05, "epoch": 1.8559999999999999, "percentage": 37.18, "elapsed_time": "1:29:00", "remaining_time": "2:30:22"}
+{"current_steps": 146, "total_steps": 390, "loss": 0.4949, "lr": 3.1508158696850275e-05, "epoch": 1.8688, "percentage": 37.44, "elapsed_time": "1:29:39", "remaining_time": "2:29:50"}
+{"current_steps": 147, "total_steps": 390, "loss": 0.5056, "lr": 3.136129493462312e-05, "epoch": 1.8816000000000002, "percentage": 37.69, "elapsed_time": "1:30:21", "remaining_time": "2:29:21"}
+{"current_steps": 148, "total_steps": 390, "loss": 0.4953, "lr": 3.121352102768998e-05, "epoch": 1.8944, "percentage": 37.95, "elapsed_time": "1:30:53", "remaining_time": "2:28:37"}
+{"current_steps": 149, "total_steps": 390, "loss": 0.4972, "lr": 3.106484881410628e-05, "epoch": 1.9072, "percentage": 38.21, "elapsed_time": "1:31:29", "remaining_time": "2:27:59"}
+{"current_steps": 150, "total_steps": 390, "loss": 0.5273, "lr": 3.091529020389009e-05, "epoch": 1.92, "percentage": 38.46, "elapsed_time": "1:32:03", "remaining_time": "2:27:17"}
+{"current_steps": 151, "total_steps": 390, "loss": 0.4691, "lr": 3.076485717806808e-05, "epoch": 1.9327999999999999, "percentage": 38.72, "elapsed_time": "1:32:38", "remaining_time": "2:26:38"}
+{"current_steps": 152, "total_steps": 390, "loss": 0.5203, "lr": 3.061356178771564e-05, "epoch": 1.9456, "percentage": 38.97, "elapsed_time": "1:33:19", "remaining_time": "2:26:08"}
+{"current_steps": 153, "total_steps": 390, "loss": 0.5122, "lr": 3.0461416152991555e-05, "epoch": 1.9584000000000001, "percentage": 39.23, "elapsed_time": "1:34:01", "remaining_time": "2:25:39"}
+{"current_steps": 154, "total_steps": 390, "loss": 0.5395, "lr": 3.0308432462167045e-05, "epoch": 1.9712, "percentage": 39.49, "elapsed_time": "1:34:31", "remaining_time": "2:24:51"}
+{"current_steps": 155, "total_steps": 390, "loss": 0.5042, "lr": 3.015462297064936e-05, "epoch": 1.984, "percentage": 39.74, "elapsed_time": "1:35:05", "remaining_time": "2:24:10"}
+{"current_steps": 156, "total_steps": 390, "loss": 0.479, "lr": 3.0000000000000004e-05, "epoch": 1.9968, "percentage": 40.0, "elapsed_time": "1:35:44", "remaining_time": "2:23:36"}
+{"current_steps": 157, "total_steps": 390, "loss": 0.9111, "lr": 2.98445759369477e-05, "epoch": 2.0096, "percentage": 40.26, "elapsed_time": "1:36:46", "remaining_time": "2:23:37"}
+{"current_steps": 158, "total_steps": 390, "loss": 0.3983, "lr": 2.9688363232396056e-05, "epoch": 2.0224, "percentage": 40.51, "elapsed_time": "1:37:25", "remaining_time": "2:23:03"}
+{"current_steps": 159, "total_steps": 390, "loss": 0.3809, "lr": 2.9531374400426158e-05, "epoch": 2.0352, "percentage": 40.77, "elapsed_time": "1:38:03", "remaining_time": "2:22:27"}
+{"current_steps": 160, "total_steps": 390, "loss": 0.3784, "lr": 2.9373622017294075e-05, "epoch": 2.048, "percentage": 41.03, "elapsed_time": "1:38:41", "remaining_time": "2:21:52"}
+{"current_steps": 161, "total_steps": 390, "loss": 0.3789, "lr": 2.9215118720423375e-05, "epoch": 2.0608, "percentage": 41.28, "elapsed_time": "1:39:22", "remaining_time": "2:21:20"}
+{"current_steps": 162, "total_steps": 390, "loss": 0.3872, "lr": 2.9055877207392752e-05, "epoch": 2.0736, "percentage": 41.54, "elapsed_time": "1:39:51", "remaining_time": "2:20:32"}
+{"current_steps": 163, "total_steps": 390, "loss": 0.3535, "lr": 2.8895910234918828e-05, "epoch": 2.0864, "percentage": 41.79, "elapsed_time": "1:40:24", "remaining_time": "2:19:50"}
+{"current_steps": 164, "total_steps": 390, "loss": 0.3413, "lr": 2.873523061783426e-05, "epoch": 2.0992, "percentage": 42.05, "elapsed_time": "1:41:03", "remaining_time": "2:19:15"}
+{"current_steps": 165, "total_steps": 390, "loss": 0.3907, "lr": 2.8573851228061084e-05, "epoch": 2.112, "percentage": 42.31, "elapsed_time": "1:41:42", "remaining_time": "2:18:41"}
+{"current_steps": 166, "total_steps": 390, "loss": 0.3687, "lr": 2.8411784993579633e-05, "epoch": 2.1248, "percentage": 42.56, "elapsed_time": "1:42:21", "remaining_time": "2:18:07"}
+{"current_steps": 167, "total_steps": 390, "loss": 0.3874, "lr": 2.8249044897392814e-05, "epoch": 2.1376, "percentage": 42.82, "elapsed_time": "1:42:55", "remaining_time": "2:17:26"}
+{"current_steps": 168, "total_steps": 390, "loss": 0.3429, "lr": 2.80856439764861e-05, "epoch": 2.1504, "percentage": 43.08, "elapsed_time": "1:43:35", "remaining_time": "2:16:53"}
+{"current_steps": 169, "total_steps": 390, "loss": 0.4248, "lr": 2.792159532078314e-05, "epoch": 2.1632, "percentage": 43.33, "elapsed_time": "1:44:14", "remaining_time": "2:16:18"}
+{"current_steps": 170, "total_steps": 390, "loss": 0.4148, "lr": 2.77569120720971e-05, "epoch": 2.176, "percentage": 43.59, "elapsed_time": "1:44:49", "remaining_time": "2:15:39"}
+{"current_steps": 171, "total_steps": 390, "loss": 0.3839, "lr": 2.7591607423077932e-05, "epoch": 2.1888, "percentage": 43.85, "elapsed_time": "1:45:31", "remaining_time": "2:15:08"}
+{"current_steps": 172, "total_steps": 390, "loss": 0.422, "lr": 2.7425694616155474e-05, "epoch": 2.2016, "percentage": 44.1, "elapsed_time": "1:46:00", "remaining_time": "2:14:20"}
+{"current_steps": 173, "total_steps": 390, "loss": 0.4676, "lr": 2.7259186942478656e-05, "epoch": 2.2144, "percentage": 44.36, "elapsed_time": "1:46:39", "remaining_time": "2:13:47"}
+{"current_steps": 174, "total_steps": 390, "loss": 0.4401, "lr": 2.7092097740850712e-05, "epoch": 2.2272, "percentage": 44.62, "elapsed_time": "1:47:13", "remaining_time": "2:13:06"}
+{"current_steps": 175, "total_steps": 390, "loss": 0.4232, "lr": 2.692444039666066e-05, "epoch": 2.24, "percentage": 44.87, "elapsed_time": "1:47:49", "remaining_time": "2:12:28"}
+{"current_steps": 176, "total_steps": 390, "loss": 0.3999, "lr": 2.6756228340810946e-05, "epoch": 2.2528, "percentage": 45.13, "elapsed_time": "1:48:27", "remaining_time": "2:11:52"}
+{"current_steps": 177, "total_steps": 390, "loss": 0.3703, "lr": 2.6587475048641596e-05, "epoch": 2.2656, "percentage": 45.38, "elapsed_time": "1:49:02", "remaining_time": "2:11:13"}
+{"current_steps": 178, "total_steps": 390, "loss": 0.4428, "lr": 2.6418194038850634e-05, "epoch": 2.2784, "percentage": 45.64, "elapsed_time": "1:49:38", "remaining_time": "2:10:35"}
+{"current_steps": 179, "total_steps": 390, "loss": 0.3908, "lr": 2.624839887241115e-05, "epoch": 2.2912, "percentage": 45.9, "elapsed_time": "1:50:23", "remaining_time": "2:10:07"}
+{"current_steps": 180, "total_steps": 390, "loss": 0.3526, "lr": 2.607810315148494e-05, "epoch": 2.304, "percentage": 46.15, "elapsed_time": "1:50:53", "remaining_time": "2:09:22"}
+{"current_steps": 181, "total_steps": 390, "loss": 0.3676, "lr": 2.5907320518332827e-05, "epoch": 2.3168, "percentage": 46.41, "elapsed_time": "1:51:32", "remaining_time": "2:08:48"}
+{"current_steps": 182, "total_steps": 390, "loss": 0.391, "lr": 2.5736064654221808e-05, "epoch": 2.3296, "percentage": 46.67, "elapsed_time": "1:52:10", "remaining_time": "2:08:11"}
+{"current_steps": 183, "total_steps": 390, "loss": 0.3627, "lr": 2.5564349278329056e-05, "epoch": 2.3424, "percentage": 46.92, "elapsed_time": "1:52:46", "remaining_time": "2:07:33"}
+{"current_steps": 184, "total_steps": 390, "loss": 0.3823, "lr": 2.539218814664288e-05, "epoch": 2.3552, "percentage": 47.18, "elapsed_time": "1:53:26", "remaining_time": "2:06:59"}
+{"current_steps": 185, "total_steps": 390, "loss": 0.379, "lr": 2.521959505086075e-05, "epoch": 2.368, "percentage": 47.44, "elapsed_time": "1:54:03", "remaining_time": "2:06:23"}
+{"current_steps": 186, "total_steps": 390, "loss": 0.3519, "lr": 2.5046583817284437e-05, "epoch": 2.3808, "percentage": 47.69, "elapsed_time": "1:54:34", "remaining_time": "2:05:39"}
+{"current_steps": 187, "total_steps": 390, "loss": 0.3657, "lr": 2.487316830571244e-05, "epoch": 2.3936, "percentage": 47.95, "elapsed_time": "1:55:13", "remaining_time": "2:05:04"}
+{"current_steps": 188, "total_steps": 390, "loss": 0.345, "lr": 2.4699362408329646e-05, "epoch": 2.4064, "percentage": 48.21, "elapsed_time": "1:55:51", "remaining_time": "2:04:29"}
+{"current_steps": 189, "total_steps": 390, "loss": 0.3873, "lr": 2.4525180048594452e-05, "epoch": 2.4192, "percentage": 48.46, "elapsed_time": "1:56:24", "remaining_time": "2:03:48"}
+{"current_steps": 190, "total_steps": 390, "loss": 0.3686, "lr": 2.435063518012335e-05, "epoch": 2.432, "percentage": 48.72, "elapsed_time": "1:57:05", "remaining_time": "2:03:15"}
+{"current_steps": 191, "total_steps": 390, "loss": 0.3363, "lr": 2.4175741785573177e-05, "epoch": 2.4448, "percentage": 48.97, "elapsed_time": "1:57:47", "remaining_time": "2:02:43"}
+{"current_steps": 192, "total_steps": 390, "loss": 0.3713, "lr": 2.4000513875520892e-05, "epoch": 2.4576000000000002, "percentage": 49.23, "elapsed_time": "1:58:30", "remaining_time": "2:02:12"}
+{"current_steps": 193, "total_steps": 390, "loss": 0.3343, "lr": 2.3824965487341247e-05, "epoch": 2.4704, "percentage": 49.49, "elapsed_time": "1:59:02", "remaining_time": "2:01:30"}
+{"current_steps": 194, "total_steps": 390, "loss": 0.3895, "lr": 2.3649110684082258e-05, "epoch": 2.4832, "percentage": 49.74, "elapsed_time": "1:59:33", "remaining_time": "2:00:47"}
+{"current_steps": 195, "total_steps": 390, "loss": 0.3458, "lr": 2.3472963553338614e-05, "epoch": 2.496, "percentage": 50.0, "elapsed_time": "2:00:07", "remaining_time": "2:00:07"}
+{"current_steps": 196, "total_steps": 390, "loss": 0.3981, "lr": 2.3296538206123134e-05, "epoch": 2.5088, "percentage": 50.26, "elapsed_time": "2:00:47", "remaining_time": "1:59:33"}
+{"current_steps": 197, "total_steps": 390, "loss": 0.3588, "lr": 2.311984877573636e-05, "epoch": 2.5216, "percentage": 50.51, "elapsed_time": "2:01:24", "remaining_time": "1:58:56"}
+{"current_steps": 198, "total_steps": 390, "loss": 0.389, "lr": 2.2942909416634326e-05, "epoch": 2.5343999999999998, "percentage": 50.77, "elapsed_time": "2:02:03", "remaining_time": "1:58:21"}
+{"current_steps": 199, "total_steps": 390, "loss": 0.3461, "lr": 2.2765734303294666e-05, "epoch": 2.5472, "percentage": 51.03, "elapsed_time": "2:02:40", "remaining_time": "1:57:44"}
+{"current_steps": 200, "total_steps": 390, "loss": 0.3636, "lr": 2.2588337629081107e-05, "epoch": 2.56, "percentage": 51.28, "elapsed_time": "2:03:22", "remaining_time": "1:57:12"}
+{"current_steps": 201, "total_steps": 390, "loss": 0.3426, "lr": 2.2410733605106462e-05, "epoch": 2.5728, "percentage": 51.54, "elapsed_time": "2:03:52", "remaining_time": "1:56:28"}
+{"current_steps": 202, "total_steps": 390, "loss": 0.3837, "lr": 2.2232936459094158e-05, "epoch": 2.5856, "percentage": 51.79, "elapsed_time": "2:04:25", "remaining_time": "1:55:47"}
+{"current_steps": 203, "total_steps": 390, "loss": 0.3753, "lr": 2.205496043423849e-05, "epoch": 2.5984, "percentage": 52.05, "elapsed_time": "2:04:59", "remaining_time": "1:55:08"}
+{"current_steps": 204, "total_steps": 390, "loss": 0.3587, "lr": 2.1876819788063586e-05, "epoch": 2.6112, "percentage": 52.31, "elapsed_time": "2:05:28", "remaining_time": "1:54:24"}
+{"current_steps": 205, "total_steps": 390, "loss": 0.3793, "lr": 2.16985287912813e-05, "epoch": 2.624, "percentage": 52.56, "elapsed_time": "2:06:05", "remaining_time": "1:53:47"}
+{"current_steps": 206, "total_steps": 390, "loss": 0.358, "lr": 2.1520101726647922e-05, "epoch": 2.6368, "percentage": 52.82, "elapsed_time": "2:06:40", "remaining_time": "1:53:08"}
+{"current_steps": 207, "total_steps": 390, "loss": 0.3506, "lr": 2.1341552887820048e-05, "epoch": 2.6496, "percentage": 53.08, "elapsed_time": "2:07:21", "remaining_time": "1:52:35"}
+{"current_steps": 208, "total_steps": 390, "loss": 0.3725, "lr": 2.1162896578209517e-05, "epoch": 2.6624, "percentage": 53.33, "elapsed_time": "2:08:00", "remaining_time": "1:52:00"}
+{"current_steps": 209, "total_steps": 390, "loss": 0.3448, "lr": 2.0984147109837564e-05, "epoch": 2.6752000000000002, "percentage": 53.59, "elapsed_time": "2:08:36", "remaining_time": "1:51:22"}
+{"current_steps": 210, "total_steps": 390, "loss": 0.354, "lr": 2.0805318802188307e-05, "epoch": 2.6879999999999997, "percentage": 53.85, "elapsed_time": "2:09:14", "remaining_time": "1:50:46"}
+{"current_steps": 211, "total_steps": 390, "loss": 0.3588, "lr": 2.0626425981061608e-05, "epoch": 2.7008, "percentage": 54.1, "elapsed_time": "2:09:50", "remaining_time": "1:50:08"}
+{"current_steps": 212, "total_steps": 390, "loss": 0.385, "lr": 2.0447482977425465e-05, "epoch": 2.7136, "percentage": 54.36, "elapsed_time": "2:10:30", "remaining_time": "1:49:34"}
+{"current_steps": 213, "total_steps": 390, "loss": 0.3644, "lr": 2.0268504126267952e-05, "epoch": 2.7264, "percentage": 54.62, "elapsed_time": "2:11:08", "remaining_time": "1:48:58"}
+{"current_steps": 214, "total_steps": 390, "loss": 0.3639, "lr": 2.008950376544887e-05, "epoch": 2.7392, "percentage": 54.87, "elapsed_time": "2:11:44", "remaining_time": "1:48:20"}
+{"current_steps": 215, "total_steps": 390, "loss": 0.3564, "lr": 1.9910496234551132e-05, "epoch": 2.752, "percentage": 55.13, "elapsed_time": "2:12:20", "remaining_time": "1:47:42"}
+{"current_steps": 216, "total_steps": 390, "loss": 0.3585, "lr": 1.9731495873732055e-05, "epoch": 2.7648, "percentage": 55.38, "elapsed_time": "2:12:57", "remaining_time": "1:47:06"}
+{"current_steps": 217, "total_steps": 390, "loss": 0.386, "lr": 1.9552517022574542e-05, "epoch": 2.7776, "percentage": 55.64, "elapsed_time": "2:13:38", "remaining_time": "1:46:32"}
+{"current_steps": 218, "total_steps": 390, "loss": 0.3527, "lr": 1.93735740189384e-05, "epoch": 2.7904, "percentage": 55.9, "elapsed_time": "2:14:11", "remaining_time": "1:45:52"}
+{"current_steps": 219, "total_steps": 390, "loss": 0.3483, "lr": 1.9194681197811703e-05, "epoch": 2.8032, "percentage": 56.15, "elapsed_time": "2:14:49", "remaining_time": "1:45:16"}
+{"current_steps": 220, "total_steps": 390, "loss": 0.3554, "lr": 1.901585289016244e-05, "epoch": 2.816, "percentage": 56.41, "elapsed_time": "2:15:25", "remaining_time": "1:44:38"}
+{"current_steps": 221, "total_steps": 390, "loss": 0.3702, "lr": 1.8837103421790486e-05, "epoch": 2.8288, "percentage": 56.67, "elapsed_time": "2:15:53", "remaining_time": "1:43:54"}
+{"current_steps": 222, "total_steps": 390, "loss": 0.351, "lr": 1.8658447112179952e-05, "epoch": 2.8416, "percentage": 56.92, "elapsed_time": "2:16:28", "remaining_time": "1:43:16"}
+{"current_steps": 223, "total_steps": 390, "loss": 0.3752, "lr": 1.8479898273352084e-05, "epoch": 2.8544, "percentage": 57.18, "elapsed_time": "2:17:03", "remaining_time": "1:42:38"}
+{"current_steps": 224, "total_steps": 390, "loss": 0.3791, "lr": 1.83014712087187e-05, "epoch": 2.8672, "percentage": 57.44, "elapsed_time": "2:17:37", "remaining_time": "1:41:59"}
+{"current_steps": 225, "total_steps": 390, "loss": 0.3442, "lr": 1.8123180211936417e-05, "epoch": 2.88, "percentage": 57.69, "elapsed_time": "2:18:14", "remaining_time": "1:41:22"}
+{"current_steps": 226, "total_steps": 390, "loss": 0.3729, "lr": 1.794503956576152e-05, "epoch": 2.8928000000000003, "percentage": 57.95, "elapsed_time": "2:18:44", "remaining_time": "1:40:40"}
+{"current_steps": 227, "total_steps": 390, "loss": 0.3545, "lr": 1.776706354090585e-05, "epoch": 2.9055999999999997, "percentage": 58.21, "elapsed_time": "2:19:17", "remaining_time": "1:40:01"}
+{"current_steps": 228, "total_steps": 390, "loss": 0.3596, "lr": 1.758926639489354e-05, "epoch": 2.9184, "percentage": 58.46, "elapsed_time": "2:19:55", "remaining_time": "1:39:25"}
+{"current_steps": 229, "total_steps": 390, "loss": 0.3365, "lr": 1.7411662370918893e-05, "epoch": 2.9312, "percentage": 58.72, "elapsed_time": "2:20:29", "remaining_time": "1:38:46"}
+{"current_steps": 230, "total_steps": 390, "loss": 0.3664, "lr": 1.7234265696705344e-05, "epoch": 2.944, "percentage": 58.97, "elapsed_time": "2:21:02", "remaining_time": "1:38:06"}
+{"current_steps": 231, "total_steps": 390, "loss": 0.3612, "lr": 1.7057090583365678e-05, "epoch": 2.9568, "percentage": 59.23, "elapsed_time": "2:21:35", "remaining_time": "1:37:27"}
+{"current_steps": 232, "total_steps": 390, "loss": 0.3524, "lr": 1.6880151224263646e-05, "epoch": 2.9696, "percentage": 59.49, "elapsed_time": "2:22:21", "remaining_time": "1:36:57"}
+{"current_steps": 233, "total_steps": 390, "loss": 0.401, "lr": 1.6703461793876876e-05, "epoch": 2.9824, "percentage": 59.74, "elapsed_time": "2:22:56", "remaining_time": "1:36:19"}
+{"current_steps": 234, "total_steps": 390, "loss": 0.3454, "lr": 1.6527036446661396e-05, "epoch": 2.9952, "percentage": 60.0, "elapsed_time": "2:23:31", "remaining_time": "1:35:40"}
+{"current_steps": 235, "total_steps": 390, "loss": 0.583, "lr": 1.635088931591775e-05, "epoch": 3.008, "percentage": 60.26, "elapsed_time": "2:24:50", "remaining_time": "1:35:31"}
+{"current_steps": 236, "total_steps": 390, "loss": 0.2506, "lr": 1.6175034512658753e-05, "epoch": 3.0208, "percentage": 60.51, "elapsed_time": "2:25:22", "remaining_time": "1:34:51"}
+{"current_steps": 237, "total_steps": 390, "loss": 0.247, "lr": 1.5999486124479115e-05, "epoch": 3.0336, "percentage": 60.77, "elapsed_time": "2:25:52", "remaining_time": "1:34:10"}
+{"current_steps": 238, "total_steps": 390, "loss": 0.2848, "lr": 1.5824258214426833e-05, "epoch": 3.0464, "percentage": 61.03, "elapsed_time": "2:26:28", "remaining_time": "1:33:32"}
+{"current_steps": 239, "total_steps": 390, "loss": 0.2767, "lr": 1.5649364819876655e-05, "epoch": 3.0592, "percentage": 61.28, "elapsed_time": "2:27:08", "remaining_time": "1:32:58"}
+{"current_steps": 240, "total_steps": 390, "loss": 0.2324, "lr": 1.547481995140556e-05, "epoch": 3.072, "percentage": 61.54, "elapsed_time": "2:27:47", "remaining_time": "1:32:21"}
+{"current_steps": 241, "total_steps": 390, "loss": 0.2555, "lr": 1.5300637591670357e-05, "epoch": 3.0848, "percentage": 61.79, "elapsed_time": "2:28:23", "remaining_time": "1:31:44"}
+{"current_steps": 242, "total_steps": 390, "loss": 0.2895, "lr": 1.5126831694287564e-05, "epoch": 3.0976, "percentage": 62.05, "elapsed_time": "2:29:04", "remaining_time": "1:31:10"}
+{"current_steps": 243, "total_steps": 390, "loss": 0.2508, "lr": 1.4953416182715566e-05, "epoch": 3.1104, "percentage": 62.31, "elapsed_time": "2:29:38", "remaining_time": "1:30:31"}
+{"current_steps": 244, "total_steps": 390, "loss": 0.2685, "lr": 1.478040494913926e-05, "epoch": 3.1232, "percentage": 62.56, "elapsed_time": "2:30:14", "remaining_time": "1:29:54"}
+{"current_steps": 245, "total_steps": 390, "loss": 0.2709, "lr": 1.460781185335713e-05, "epoch": 3.136, "percentage": 62.82, "elapsed_time": "2:30:49", "remaining_time": "1:29:15"}
+{"current_steps": 246, "total_steps": 390, "loss": 0.2517, "lr": 1.443565072167095e-05, "epoch": 3.1488, "percentage": 63.08, "elapsed_time": "2:31:23", "remaining_time": "1:28:37"}
+{"current_steps": 247, "total_steps": 390, "loss": 0.2569, "lr": 1.4263935345778202e-05, "epoch": 3.1616, "percentage": 63.33, "elapsed_time": "2:32:03", "remaining_time": "1:28:02"}
+{"current_steps": 248, "total_steps": 390, "loss": 0.2543, "lr": 1.409267948166718e-05, "epoch": 3.1744, "percentage": 63.59, "elapsed_time": "2:32:36", "remaining_time": "1:27:23"}
+{"current_steps": 249, "total_steps": 390, "loss": 0.2762, "lr": 1.3921896848515064e-05, "epoch": 3.1872, "percentage": 63.85, "elapsed_time": "2:33:18", "remaining_time": "1:26:48"}
+{"current_steps": 250, "total_steps": 390, "loss": 0.235, "lr": 1.3751601127588849e-05, "epoch": 3.2, "percentage": 64.1, "elapsed_time": "2:33:56", "remaining_time": "1:26:12"}
+{"current_steps": 251, "total_steps": 390, "loss": 0.2531, "lr": 1.3581805961149371e-05, "epoch": 3.2128, "percentage": 64.36, "elapsed_time": "2:34:31", "remaining_time": "1:25:34"}
+{"current_steps": 252, "total_steps": 390, "loss": 0.2678, "lr": 1.341252495135841e-05, "epoch": 3.2256, "percentage": 64.62, "elapsed_time": "2:35:13", "remaining_time": "1:25:00"}
+{"current_steps": 253, "total_steps": 390, "loss": 0.2341, "lr": 1.324377165918906e-05, "epoch": 3.2384, "percentage": 64.87, "elapsed_time": "2:35:48", "remaining_time": "1:24:22"}
+{"current_steps": 254, "total_steps": 390, "loss": 0.2428, "lr": 1.3075559603339354e-05, "epoch": 3.2512, "percentage": 65.13, "elapsed_time": "2:36:20", "remaining_time": "1:23:42"}
+{"current_steps": 255, "total_steps": 390, "loss": 0.2615, "lr": 1.2907902259149287e-05, "epoch": 3.2640000000000002, "percentage": 65.38, "elapsed_time": "2:36:59", "remaining_time": "1:23:06"}
+{"current_steps": 256, "total_steps": 390, "loss": 0.255, "lr": 1.274081305752135e-05, "epoch": 3.2768, "percentage": 65.64, "elapsed_time": "2:37:32", "remaining_time": "1:22:28"}
+{"current_steps": 257, "total_steps": 390, "loss": 0.2396, "lr": 1.2574305383844528e-05, "epoch": 3.2896, "percentage": 65.9, "elapsed_time": "2:38:09", "remaining_time": "1:21:51"}
+{"current_steps": 258, "total_steps": 390, "loss": 0.2774, "lr": 1.2408392576922075e-05, "epoch": 3.3024, "percentage": 66.15, "elapsed_time": "2:38:46", "remaining_time": "1:21:14"}
+{"current_steps": 259, "total_steps": 390, "loss": 0.2476, "lr": 1.2243087927902905e-05, "epoch": 3.3152, "percentage": 66.41, "elapsed_time": "2:39:29", "remaining_time": "1:20:40"}
+{"current_steps": 260, "total_steps": 390, "loss": 0.266, "lr": 1.2078404679216864e-05, "epoch": 3.328, "percentage": 66.67, "elapsed_time": "2:40:08", "remaining_time": "1:20:04"}
+{"current_steps": 261, "total_steps": 390, "loss": 0.2668, "lr": 1.1914356023513904e-05, "epoch": 3.3407999999999998, "percentage": 66.92, "elapsed_time": "2:40:44", "remaining_time": "1:19:26"}
+{"current_steps": 262, "total_steps": 390, "loss": 0.2305, "lr": 1.1750955102607193e-05, "epoch": 3.3536, "percentage": 67.18, "elapsed_time": "2:41:18", "remaining_time": "1:18:48"}
+{"current_steps": 263, "total_steps": 390, "loss": 0.2739, "lr": 1.1588215006420374e-05, "epoch": 3.3664, "percentage": 67.44, "elapsed_time": "2:41:59", "remaining_time": "1:18:13"}
+{"current_steps": 264, "total_steps": 390, "loss": 0.2472, "lr": 1.1426148771938915e-05, "epoch": 3.3792, "percentage": 67.69, "elapsed_time": "2:42:37", "remaining_time": "1:17:36"}
+{"current_steps": 265, "total_steps": 390, "loss": 0.2442, "lr": 1.1264769382165748e-05, "epoch": 3.392, "percentage": 67.95, "elapsed_time": "2:43:12", "remaining_time": "1:16:58"}
+{"current_steps": 266, "total_steps": 390, "loss": 0.2434, "lr": 1.110408976508118e-05, "epoch": 3.4048, "percentage": 68.21, "elapsed_time": "2:43:53", "remaining_time": "1:16:24"}
+{"current_steps": 267, "total_steps": 390, "loss": 0.2335, "lr": 1.094412279260726e-05, "epoch": 3.4176, "percentage": 68.46, "elapsed_time": "2:44:31", "remaining_time": "1:15:47"}
+{"current_steps": 268, "total_steps": 390, "loss": 0.2563, "lr": 1.0784881279576635e-05, "epoch": 3.4304, "percentage": 68.72, "elapsed_time": "2:45:07", "remaining_time": "1:15:09"}
+{"current_steps": 269, "total_steps": 390, "loss": 0.237, "lr": 1.0626377982705929e-05, "epoch": 3.4432, "percentage": 68.97, "elapsed_time": "2:45:45", "remaining_time": "1:14:33"}
+{"current_steps": 270, "total_steps": 390, "loss": 0.2521, "lr": 1.0468625599573842e-05, "epoch": 3.456, "percentage": 69.23, "elapsed_time": "2:46:26", "remaining_time": "1:13:58"}
+{"current_steps": 271, "total_steps": 390, "loss": 0.2521, "lr": 1.0311636767603952e-05, "epoch": 3.4688, "percentage": 69.49, "elapsed_time": "2:46:59", "remaining_time": "1:13:19"}
+{"current_steps": 272, "total_steps": 390, "loss": 0.2486, "lr": 1.0155424063052306e-05, "epoch": 3.4816, "percentage": 69.74, "elapsed_time": "2:47:35", "remaining_time": "1:12:42"}
+{"current_steps": 273, "total_steps": 390, "loss": 0.2634, "lr": 1.0000000000000006e-05, "epoch": 3.4944, "percentage": 70.0, "elapsed_time": "2:48:11", "remaining_time": "1:12:05"}
+{"current_steps": 274, "total_steps": 390, "loss": 0.2499, "lr": 9.84537702935065e-06, "epoch": 3.5072, "percentage": 70.26, "elapsed_time": "2:48:49", "remaining_time": "1:11:28"}
+{"current_steps": 275, "total_steps": 390, "loss": 0.2596, "lr": 9.691567537832964e-06, "epoch": 3.52, "percentage": 70.51, "elapsed_time": "2:49:29", "remaining_time": "1:10:52"}
+{"current_steps": 276, "total_steps": 390, "loss": 0.2273, "lr": 9.538583847008452e-06, "epoch": 3.5328, "percentage": 70.77, "elapsed_time": "2:50:07", "remaining_time": "1:10:16"}
+{"current_steps": 277, "total_steps": 390, "loss": 0.2573, "lr": 9.386438212284372e-06, "epoch": 3.5456, "percentage": 71.03, "elapsed_time": "2:50:46", "remaining_time": "1:09:39"}
+{"current_steps": 278, "total_steps": 390, "loss": 0.2408, "lr": 9.235142821931928e-06, "epoch": 3.5584, "percentage": 71.28, "elapsed_time": "2:51:19", "remaining_time": "1:09:01"}
+{"current_steps": 279, "total_steps": 390, "loss": 0.2707, "lr": 9.084709796109907e-06, "epoch": 3.5712, "percentage": 71.54, "elapsed_time": "2:51:58", "remaining_time": "1:08:25"}
+{"current_steps": 280, "total_steps": 390, "loss": 0.2669, "lr": 8.93515118589373e-06, "epoch": 3.584, "percentage": 71.79, "elapsed_time": "2:52:28", "remaining_time": "1:07:45"}
+{"current_steps": 281, "total_steps": 390, "loss": 0.2354, "lr": 8.786478972310023e-06, "epoch": 3.5968, "percentage": 72.05, "elapsed_time": "2:53:04", "remaining_time": "1:07:08"}
+{"current_steps": 282, "total_steps": 390, "loss": 0.2355, "lr": 8.638705065376887e-06, "epoch": 3.6096, "percentage": 72.31, "elapsed_time": "2:53:41", "remaining_time": "1:06:31"}
+{"current_steps": 283, "total_steps": 390, "loss": 0.2664, "lr": 8.491841303149728e-06, "epoch": 3.6224, "percentage": 72.56, "elapsed_time": "2:54:23", "remaining_time": "1:05:56"}
+{"current_steps": 284, "total_steps": 390, "loss": 0.2501, "lr": 8.345899450772975e-06, "epoch": 3.6352, "percentage": 72.82, "elapsed_time": "2:54:59", "remaining_time": "1:05:18"}
+{"current_steps": 285, "total_steps": 390, "loss": 0.2543, "lr": 8.200891199537549e-06, "epoch": 3.648, "percentage": 73.08, "elapsed_time": "2:55:31", "remaining_time": "1:04:39"}
+{"current_steps": 286, "total_steps": 390, "loss": 0.261, "lr": 8.056828165944282e-06, "epoch": 3.6608, "percentage": 73.33, "elapsed_time": "2:56:10", "remaining_time": "1:04:03"}
+{"current_steps": 287, "total_steps": 390, "loss": 0.2405, "lr": 7.913721890773354e-06, "epoch": 3.6736, "percentage": 73.59, "elapsed_time": "2:56:40", "remaining_time": "1:03:24"}
+{"current_steps": 288, "total_steps": 390, "loss": 0.2732, "lr": 7.771583838159756e-06, "epoch": 3.6864, "percentage": 73.85, "elapsed_time": "2:57:13", "remaining_time": "1:02:46"}
+{"current_steps": 289, "total_steps": 390, "loss": 0.2399, "lr": 7.630425394674903e-06, "epoch": 3.6992000000000003, "percentage": 74.1, "elapsed_time": "2:57:49", "remaining_time": "1:02:08"}
+{"current_steps": 290, "total_steps": 390, "loss": 0.2517, "lr": 7.49025786841445e-06, "epoch": 3.7119999999999997, "percentage": 74.36, "elapsed_time": "2:58:29", "remaining_time": "1:01:33"}
+{"current_steps": 291, "total_steps": 390, "loss": 0.2321, "lr": 7.3510924880924575e-06, "epoch": 3.7248, "percentage": 74.62, "elapsed_time": "2:59:05", "remaining_time": "1:00:55"}
+{"current_steps": 292, "total_steps": 390, "loss": 0.2338, "lr": 7.212940402141808e-06, "epoch": 3.7376, "percentage": 74.87, "elapsed_time": "2:59:38", "remaining_time": "1:00:17"}
+{"current_steps": 293, "total_steps": 390, "loss": 0.2871, "lr": 7.075812677821145e-06, "epoch": 3.7504, "percentage": 75.13, "elapsed_time": "3:00:22", "remaining_time": "0:59:42"}
+{"current_steps": 294, "total_steps": 390, "loss": 0.263, "lr": 6.939720300328303e-06, "epoch": 3.7632, "percentage": 75.38, "elapsed_time": "3:00:59", "remaining_time": "0:59:05"}
+{"current_steps": 295, "total_steps": 390, "loss": 0.2309, "lr": 6.8046741719202385e-06, "epoch": 3.776, "percentage": 75.64, "elapsed_time": "3:01:24", "remaining_time": "0:58:25"}
+{"current_steps": 296, "total_steps": 390, "loss": 0.264, "lr": 6.67068511103971e-06, "epoch": 3.7888, "percentage": 75.9, "elapsed_time": "3:01:59", "remaining_time": "0:57:47"}
+{"current_steps": 297, "total_steps": 390, "loss": 0.2379, "lr": 6.537763851448593e-06, "epoch": 3.8016, "percentage": 76.15, "elapsed_time": "3:02:30", "remaining_time": "0:57:09"}
+{"current_steps": 298, "total_steps": 390, "loss": 0.2465, "lr": 6.4059210413680175e-06, "epoch": 3.8144, "percentage": 76.41, "elapsed_time": "3:03:05", "remaining_time": "0:56:31"}
+{"current_steps": 299, "total_steps": 390, "loss": 0.2542, "lr": 6.275167242625331e-06, "epoch": 3.8272, "percentage": 76.67, "elapsed_time": "3:03:42", "remaining_time": "0:55:54"}
+{"current_steps": 300, "total_steps": 390, "loss": 0.2461, "lr": 6.145512929808013e-06, "epoch": 3.84, "percentage": 76.92, "elapsed_time": "3:04:19", "remaining_time": "0:55:17"}
+{"current_steps": 301, "total_steps": 390, "loss": 0.2409, "lr": 6.016968489424572e-06, "epoch": 3.8528000000000002, "percentage": 77.18, "elapsed_time": "3:04:57", "remaining_time": "0:54:41"}
+{"current_steps": 302, "total_steps": 390, "loss": 0.2515, "lr": 5.889544219072465e-06, "epoch": 3.8656, "percentage": 77.44, "elapsed_time": "3:05:37", "remaining_time": "0:54:05"}
+{"current_steps": 303, "total_steps": 390, "loss": 0.2466, "lr": 5.7632503266131925e-06, "epoch": 3.8784, "percentage": 77.69, "elapsed_time": "3:06:13", "remaining_time": "0:53:28"}
+{"current_steps": 304, "total_steps": 390, "loss": 0.2592, "lr": 5.638096929354522e-06, "epoch": 3.8912, "percentage": 77.95, "elapsed_time": "3:06:55", "remaining_time": "0:52:52"}
+{"current_steps": 305, "total_steps": 390, "loss": 0.2538, "lr": 5.514094053240035e-06, "epoch": 3.904, "percentage": 78.21, "elapsed_time": "3:07:28", "remaining_time": "0:52:14"}
+{"current_steps": 306, "total_steps": 390, "loss": 0.2324, "lr": 5.39125163204594e-06, "epoch": 3.9168, "percentage": 78.46, "elapsed_time": "3:07:55", "remaining_time": "0:51:35"}
+{"current_steps": 307, "total_steps": 390, "loss": 0.2573, "lr": 5.269579506585259e-06, "epoch": 3.9295999999999998, "percentage": 78.72, "elapsed_time": "3:08:29", "remaining_time": "0:50:57"}
+{"current_steps": 308, "total_steps": 390, "loss": 0.2761, "lr": 5.149087423919541e-06, "epoch": 3.9424, "percentage": 78.97, "elapsed_time": "3:09:05", "remaining_time": "0:50:20"}
+{"current_steps": 309, "total_steps": 390, "loss": 0.2614, "lr": 5.029785036577976e-06, "epoch": 3.9552, "percentage": 79.23, "elapsed_time": "3:09:41", "remaining_time": "0:49:43"}
+{"current_steps": 310, "total_steps": 390, "loss": 0.2495, "lr": 4.911681901784198e-06, "epoch": 3.968, "percentage": 79.49, "elapsed_time": "3:10:17", "remaining_time": "0:49:06"}
+{"current_steps": 311, "total_steps": 390, "loss": 0.2475, "lr": 4.794787480690597e-06, "epoch": 3.9808, "percentage": 79.74, "elapsed_time": "3:10:57", "remaining_time": "0:48:30"}
+{"current_steps": 312, "total_steps": 390, "loss": 0.2643, "lr": 4.679111137620442e-06, "epoch": 3.9936, "percentage": 80.0, "elapsed_time": "3:11:35", "remaining_time": "0:47:53"}
+{"current_steps": 313, "total_steps": 390, "loss": 0.4669, "lr": 4.5646621393177e-06, "epoch": 4.0064, "percentage": 80.26, "elapsed_time": "3:12:59", "remaining_time": "0:47:28"}
+{"current_steps": 314, "total_steps": 390, "loss": 0.1854, "lr": 4.451449654204685e-06, "epoch": 4.0192, "percentage": 80.51, "elapsed_time": "3:13:41", "remaining_time": "0:46:52"}
+{"current_steps": 315, "total_steps": 390, "loss": 0.1981, "lr": 4.339482751647557e-06, "epoch": 4.032, "percentage": 80.77, "elapsed_time": "3:14:20", "remaining_time": "0:46:16"}
+{"current_steps": 316, "total_steps": 390, "loss": 0.1925, "lr": 4.228770401229824e-06, "epoch": 4.0448, "percentage": 81.03, "elapsed_time": "3:14:54", "remaining_time": "0:45:38"}
+{"current_steps": 317, "total_steps": 390, "loss": 0.1865, "lr": 4.119321472033779e-06, "epoch": 4.0576, "percentage": 81.28, "elapsed_time": "3:15:32", "remaining_time": "0:45:01"}
+{"current_steps": 318, "total_steps": 390, "loss": 0.1717, "lr": 4.011144731929981e-06, "epoch": 4.0704, "percentage": 81.54, "elapsed_time": "3:16:12", "remaining_time": "0:44:25"}
+{"current_steps": 319, "total_steps": 390, "loss": 0.198, "lr": 3.904248846874894e-06, "epoch": 4.0832, "percentage": 81.79, "elapsed_time": "3:16:50", "remaining_time": "0:43:48"}
+{"current_steps": 320, "total_steps": 390, "loss": 0.1765, "lr": 3.7986423802166705e-06, "epoch": 4.096, "percentage": 82.05, "elapsed_time": "3:17:26", "remaining_time": "0:43:11"}
+{"current_steps": 321, "total_steps": 390, "loss": 0.1988, "lr": 3.694333792009115e-06, "epoch": 4.1088, "percentage": 82.31, "elapsed_time": "3:17:57", "remaining_time": "0:42:33"}
+{"current_steps": 322, "total_steps": 390, "loss": 0.1705, "lr": 3.5913314383339937e-06, "epoch": 4.1216, "percentage": 82.56, "elapsed_time": "3:18:28", "remaining_time": "0:41:54"}
+{"current_steps": 323, "total_steps": 390, "loss": 0.1817, "lr": 3.4896435706316e-06, "epoch": 4.1344, "percentage": 82.82, "elapsed_time": "3:19:07", "remaining_time": "0:41:18"}
+{"current_steps": 324, "total_steps": 390, "loss": 0.2028, "lr": 3.3892783350397675e-06, "epoch": 4.1472, "percentage": 83.08, "elapsed_time": "3:19:46", "remaining_time": "0:40:41"}
+{"current_steps": 325, "total_steps": 390, "loss": 0.187, "lr": 3.290243771741275e-06, "epoch": 4.16, "percentage": 83.33, "elapsed_time": "3:20:20", "remaining_time": "0:40:04"}
+{"current_steps": 326, "total_steps": 390, "loss": 0.1656, "lr": 3.1925478143197418e-06, "epoch": 4.1728, "percentage": 83.59, "elapsed_time": "3:20:54", "remaining_time": "0:39:26"}
+{"current_steps": 327, "total_steps": 390, "loss": 0.1909, "lr": 3.0961982891241083e-06, "epoch": 4.1856, "percentage": 83.85, "elapsed_time": "3:21:34", "remaining_time": "0:38:50"}
+{"current_steps": 328, "total_steps": 390, "loss": 0.1664, "lr": 3.001202914641628e-06, "epoch": 4.1984, "percentage": 84.1, "elapsed_time": "3:22:13", "remaining_time": "0:38:13"}
+{"current_steps": 329, "total_steps": 390, "loss": 0.1779, "lr": 2.907569300879596e-06, "epoch": 4.2112, "percentage": 84.36, "elapsed_time": "3:22:48", "remaining_time": "0:37:36"}
+{"current_steps": 330, "total_steps": 390, "loss": 0.1848, "lr": 2.815304948755664e-06, "epoch": 4.224, "percentage": 84.62, "elapsed_time": "3:23:22", "remaining_time": "0:36:58"}
+{"current_steps": 331, "total_steps": 390, "loss": 0.1492, "lr": 2.7244172494969978e-06, "epoch": 4.2368, "percentage": 84.87, "elapsed_time": "3:23:55", "remaining_time": "0:36:20"}
+{"current_steps": 332, "total_steps": 390, "loss": 0.1934, "lr": 2.6349134840481294e-06, "epoch": 4.2496, "percentage": 85.13, "elapsed_time": "3:24:33", "remaining_time": "0:35:44"}
+{"current_steps": 333, "total_steps": 390, "loss": 0.1672, "lr": 2.546800822487714e-06, "epoch": 4.2624, "percentage": 85.38, "elapsed_time": "3:25:10", "remaining_time": "0:35:07"}
+{"current_steps": 334, "total_steps": 390, "loss": 0.1709, "lr": 2.4600863234541338e-06, "epoch": 4.2752, "percentage": 85.64, "elapsed_time": "3:25:46", "remaining_time": "0:34:30"}
+{"current_steps": 335, "total_steps": 390, "loss": 0.1784, "lr": 2.374776933580025e-06, "epoch": 4.288, "percentage": 85.9, "elapsed_time": "3:26:22", "remaining_time": "0:33:52"}
+{"current_steps": 336, "total_steps": 390, "loss": 0.1883, "lr": 2.2908794869358044e-06, "epoch": 4.3008, "percentage": 86.15, "elapsed_time": "3:27:05", "remaining_time": "0:33:16"}
+{"current_steps": 337, "total_steps": 390, "loss": 0.19, "lr": 2.2084007044821764e-06, "epoch": 4.3136, "percentage": 86.41, "elapsed_time": "3:27:38", "remaining_time": "0:32:39"}
+{"current_steps": 338, "total_steps": 390, "loss": 0.1805, "lr": 2.127347193531757e-06, "epoch": 4.3264, "percentage": 86.67, "elapsed_time": "3:28:18", "remaining_time": "0:32:02"}
+{"current_steps": 339, "total_steps": 390, "loss": 0.1561, "lr": 2.0477254472197237e-06, "epoch": 4.3392, "percentage": 86.92, "elapsed_time": "3:28:58", "remaining_time": "0:31:26"}
+{"current_steps": 340, "total_steps": 390, "loss": 0.187, "lr": 1.96954184398368e-06, "epoch": 4.352, "percentage": 87.18, "elapsed_time": "3:29:33", "remaining_time": "0:30:49"}
+{"current_steps": 341, "total_steps": 390, "loss": 0.18, "lr": 1.8928026470526917e-06, "epoch": 4.3648, "percentage": 87.44, "elapsed_time": "3:30:10", "remaining_time": "0:30:12"}
+{"current_steps": 342, "total_steps": 390, "loss": 0.1962, "lr": 1.817514003945524e-06, "epoch": 4.3776, "percentage": 87.69, "elapsed_time": "3:30:46", "remaining_time": "0:29:34"}
+{"current_steps": 343, "total_steps": 390, "loss": 0.1859, "lr": 1.743681945978184e-06, "epoch": 4.3904, "percentage": 87.95, "elapsed_time": "3:31:25", "remaining_time": "0:28:58"}
+{"current_steps": 344, "total_steps": 390, "loss": 0.1708, "lr": 1.6713123877807413e-06, "epoch": 4.4032, "percentage": 88.21, "elapsed_time": "3:32:07", "remaining_time": "0:28:21"}
+{"current_steps": 345, "total_steps": 390, "loss": 0.1631, "lr": 1.6004111268235156e-06, "epoch": 4.416, "percentage": 88.46, "elapsed_time": "3:32:45", "remaining_time": "0:27:45"}
+{"current_steps": 346, "total_steps": 390, "loss": 0.175, "lr": 1.5309838429526714e-06, "epoch": 4.4288, "percentage": 88.72, "elapsed_time": "3:33:23", "remaining_time": "0:27:08"}
+{"current_steps": 347, "total_steps": 390, "loss": 0.1765, "lr": 1.4630360979351644e-06, "epoch": 4.4416, "percentage": 88.97, "elapsed_time": "3:34:00", "remaining_time": "0:26:31"}
+{"current_steps": 348, "total_steps": 390, "loss": 0.1886, "lr": 1.396573335013236e-06, "epoch": 4.4544, "percentage": 89.23, "elapsed_time": "3:34:37", "remaining_time": "0:25:54"}
+{"current_steps": 349, "total_steps": 390, "loss": 0.1824, "lr": 1.3316008784683265e-06, "epoch": 4.4672, "percentage": 89.49, "elapsed_time": "3:35:10", "remaining_time": "0:25:16"}
+{"current_steps": 350, "total_steps": 390, "loss": 0.1828, "lr": 1.2681239331945695e-06, "epoch": 4.48, "percentage": 89.74, "elapsed_time": "3:35:47", "remaining_time": "0:24:39"}
+{"current_steps": 351, "total_steps": 390, "loss": 0.1722, "lr": 1.2061475842818337e-06, "epoch": 4.4928, "percentage": 90.0, "elapsed_time": "3:36:26", "remaining_time": "0:24:02"}
+{"current_steps": 352, "total_steps": 390, "loss": 0.1666, "lr": 1.1456767966083393e-06, "epoch": 4.5056, "percentage": 90.26, "elapsed_time": "3:36:58", "remaining_time": "0:23:25"}
+{"current_steps": 353, "total_steps": 390, "loss": 0.179, "lr": 1.086716414442952e-06, "epoch": 4.5184, "percentage": 90.51, "elapsed_time": "3:37:35", "remaining_time": "0:22:48"}
+{"current_steps": 354, "total_steps": 390, "loss": 0.2035, "lr": 1.0292711610570904e-06, "epoch": 4.5312, "percentage": 90.77, "elapsed_time": "3:38:15", "remaining_time": "0:22:11"}
+{"current_steps": 355, "total_steps": 390, "loss": 0.1784, "lr": 9.733456383463658e-07, "epoch": 4.5440000000000005, "percentage": 91.03, "elapsed_time": "3:38:47", "remaining_time": "0:21:34"}
+{"current_steps": 356, "total_steps": 390, "loss": 0.1758, "lr": 9.189443264619102e-07, "epoch": 4.5568, "percentage": 91.28, "elapsed_time": "3:39:25", "remaining_time": "0:20:57"}
+{"current_steps": 357, "total_steps": 390, "loss": 0.174, "lr": 8.660715834514977e-07, "epoch": 4.5696, "percentage": 91.54, "elapsed_time": "3:40:05", "remaining_time": "0:20:20"}
+{"current_steps": 358, "total_steps": 390, "loss": 0.1868, "lr": 8.147316449103959e-07, "epoch": 4.5824, "percentage": 91.79, "elapsed_time": "3:40:44", "remaining_time": "0:19:43"}
+{"current_steps": 359, "total_steps": 390, "loss": 0.1781, "lr": 7.649286236420806e-07, "epoch": 4.5952, "percentage": 92.05, "elapsed_time": "3:41:16", "remaining_time": "0:19:06"}
+{"current_steps": 360, "total_steps": 390, "loss": 0.2014, "lr": 7.166665093287539e-07, "epoch": 4.608, "percentage": 92.31, "elapsed_time": "3:41:57", "remaining_time": "0:18:29"}
+{"current_steps": 361, "total_steps": 390, "loss": 0.1841, "lr": 6.69949168211721e-07, "epoch": 4.6208, "percentage": 92.56, "elapsed_time": "3:42:34", "remaining_time": "0:17:52"}
+{"current_steps": 362, "total_steps": 390, "loss": 0.1771, "lr": 6.247803427816945e-07, "epoch": 4.6336, "percentage": 92.82, "elapsed_time": "3:43:12", "remaining_time": "0:17:15"}
+{"current_steps": 363, "total_steps": 390, "loss": 0.1739, "lr": 5.811636514789598e-07, "epoch": 4.6464, "percentage": 93.08, "elapsed_time": "3:43:51", "remaining_time": "0:16:39"}
+{"current_steps": 364, "total_steps": 390, "loss": 0.1891, "lr": 5.391025884035239e-07, "epoch": 4.6592, "percentage": 93.33, "elapsed_time": "3:44:26", "remaining_time": "0:16:01"}
+{"current_steps": 365, "total_steps": 390, "loss": 0.1692, "lr": 4.986005230351954e-07, "epoch": 4.672, "percentage": 93.59, "elapsed_time": "3:44:59", "remaining_time": "0:15:24"}
+{"current_steps": 366, "total_steps": 390, "loss": 0.172, "lr": 4.5966069996365993e-07, "epoch": 4.6848, "percentage": 93.85, "elapsed_time": "3:45:34", "remaining_time": "0:14:47"}
+{"current_steps": 367, "total_steps": 390, "loss": 0.1522, "lr": 4.22286238628562e-07, "epoch": 4.6975999999999996, "percentage": 94.1, "elapsed_time": "3:46:04", "remaining_time": "0:14:10"}
+{"current_steps": 368, "total_steps": 390, "loss": 0.1645, "lr": 3.8648013306960664e-07, "epoch": 4.7104, "percentage": 94.36, "elapsed_time": "3:46:36", "remaining_time": "0:13:32"}
+{"current_steps": 369, "total_steps": 390, "loss": 0.1962, "lr": 3.522452516867048e-07, "epoch": 4.7232, "percentage": 94.62, "elapsed_time": "3:47:16", "remaining_time": "0:12:56"}
+{"current_steps": 370, "total_steps": 390, "loss": 0.1991, "lr": 3.1958433701019697e-07, "epoch": 4.736, "percentage": 94.87, "elapsed_time": "3:47:47", "remaining_time": "0:12:18"}
+{"current_steps": 371, "total_steps": 390, "loss": 0.1592, "lr": 2.8850000548115155e-07, "epoch": 4.7488, "percentage": 95.13, "elapsed_time": "3:48:18", "remaining_time": "0:11:41"}
+{"current_steps": 372, "total_steps": 390, "loss": 0.1608, "lr": 2.5899474724174313e-07, "epoch": 4.7616, "percentage": 95.38, "elapsed_time": "3:48:54", "remaining_time": "0:11:04"}
+{"current_steps": 373, "total_steps": 390, "loss": 0.1766, "lr": 2.3107092593579905e-07, "epoch": 4.7744, "percentage": 95.64, "elapsed_time": "3:49:28", "remaining_time": "0:10:27"}
+{"current_steps": 374, "total_steps": 390, "loss": 0.1676, "lr": 2.0473077851942858e-07, "epoch": 4.7872, "percentage": 95.9, "elapsed_time": "3:50:06", "remaining_time": "0:09:50"}
+{"current_steps": 375, "total_steps": 390, "loss": 0.1708, "lr": 1.799764150818306e-07, "epoch": 4.8, "percentage": 96.15, "elapsed_time": "3:50:41", "remaining_time": "0:09:13"}
+{"current_steps": 376, "total_steps": 390, "loss": 0.1707, "lr": 1.5680981867625566e-07, "epoch": 4.8128, "percentage": 96.41, "elapsed_time": "3:51:19", "remaining_time": "0:08:36"}
+{"current_steps": 377, "total_steps": 390, "loss": 0.1752, "lr": 1.3523284516113955e-07, "epoch": 4.8256, "percentage": 96.67, "elapsed_time": "3:51:56", "remaining_time": "0:07:59"}
+{"current_steps": 378, "total_steps": 390, "loss": 0.178, "lr": 1.1524722305144231e-07, "epoch": 4.8384, "percentage": 96.92, "elapsed_time": "3:52:32", "remaining_time": "0:07:22"}
+{"current_steps": 379, "total_steps": 390, "loss": 0.1859, "lr": 9.685455338016347e-08, "epoch": 4.8512, "percentage": 97.18, "elapsed_time": "3:53:08", "remaining_time": "0:06:46"}
+{"current_steps": 380, "total_steps": 390, "loss": 0.1667, "lr": 8.005630957010014e-08, "epoch": 4.864, "percentage": 97.44, "elapsed_time": "3:53:47", "remaining_time": "0:06:09"}
+{"current_steps": 381, "total_steps": 390, "loss": 0.1828, "lr": 6.485383731580142e-08, "epoch": 4.8768, "percentage": 97.69, "elapsed_time": "3:54:27", "remaining_time": "0:05:32"}
+{"current_steps": 382, "total_steps": 390, "loss": 0.1761, "lr": 5.1248354475768034e-08, "epoch": 4.8896, "percentage": 97.95, "elapsed_time": "3:55:04", "remaining_time": "0:04:55"}
+{"current_steps": 383, "total_steps": 390, "loss": 0.1705, "lr": 3.924095097489922e-08, "epoch": 4.9024, "percentage": 98.21, "elapsed_time": "3:55:42", "remaining_time": "0:04:18"}
+{"current_steps": 384, "total_steps": 390, "loss": 0.181, "lr": 2.8832588717164766e-08, "epoch": 4.9152000000000005, "percentage": 98.46, "elapsed_time": "3:56:21", "remaining_time": "0:03:41"}
+{"current_steps": 385, "total_steps": 390, "loss": 0.1646, "lr": 2.0024101508555604e-08, "epoch": 4.928, "percentage": 98.72, "elapsed_time": "3:56:58", "remaining_time": "0:03:04"}
+{"current_steps": 386, "total_steps": 390, "loss": 0.1792, "lr": 1.281619499029274e-08, "epoch": 4.9408, "percentage": 98.97, "elapsed_time": "3:57:35", "remaining_time": "0:02:27"}
+{"current_steps": 387, "total_steps": 390, "loss": 0.1696, "lr": 7.209446582292501e-09, "epoch": 4.9536, "percentage": 99.23, "elapsed_time": "3:58:09", "remaining_time": "0:01:50"}
+{"current_steps": 388, "total_steps": 390, "loss": 0.1696, "lr": 3.2043054369057523e-09, "epoch": 4.9664, "percentage": 99.49, "elapsed_time": "3:58:45", "remaining_time": "0:01:13"}
+{"current_steps": 389, "total_steps": 390, "loss": 0.1822, "lr": 8.010924029533406e-10, "epoch": 4.9792, "percentage": 99.74, "elapsed_time": "3:59:23", "remaining_time": "0:00:36"}
+{"current_steps": 390, "total_steps": 390, "loss": 0.172, "lr": 0.0, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "4:00:04", "remaining_time": "0:00:00"}
+{"current_steps": 390, "total_steps": 390, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "4:01:17", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2772 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.992,
+  "eval_steps": 500,
+  "global_step": 390,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0128,
+      "grad_norm": 5.863900157318113,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.9596,
+      "step": 1
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 6.109932746524333,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.0162,
+      "step": 2
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.20387498758172,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0178,
+      "step": 3
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 5.790412950499749,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.9835,
+      "step": 4
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.158416716109019,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.9483,
+      "step": 5
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.0993792148946016,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9022,
+      "step": 6
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.058882280839032,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.8501,
+      "step": 7
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 4.522457391421698,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.8768,
+      "step": 8
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 4.921125491878129,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.8887,
+      "step": 9
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 4.719086125625584,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.8776,
+      "step": 10
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 4.00819662188803,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.8272,
+      "step": 11
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 3.050397009508778,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.8057,
+      "step": 12
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 2.547425655198485,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8367,
+      "step": 13
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.239649674283963,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.7608,
+      "step": 14
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.4398965199344405,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.7656,
+      "step": 15
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.9379897743248704,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.7609,
+      "step": 16
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.4123344749511757,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.732,
+      "step": 17
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.560699485232345,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.7513,
+      "step": 18
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.2910905898117273,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.7624,
+      "step": 19
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.9767602523000253,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.7191,
+      "step": 20
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.113128043736566,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.7181,
+      "step": 21
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.762279274050193,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.6926,
+      "step": 22
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.8970542754675157,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.7319,
+      "step": 23
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.8154373738338958,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.6766,
+      "step": 24
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9043268429451219,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.7083,
+      "step": 25
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.7790288918469499,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.686,
+      "step": 26
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.7493033550636032,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.6769,
+      "step": 27
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.7814414824186039,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.6911,
+      "step": 28
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.8858153230819887,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.6811,
+      "step": 29
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.7747815330420422,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.6782,
+      "step": 30
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.6964900372216322,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.686,
+      "step": 31
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.7667266533907808,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.6669,
+      "step": 32
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.8501524649816993,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6685,
+      "step": 33
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.7828592371443549,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.6574,
+      "step": 34
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.8724394670805821,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.6749,
+      "step": 35
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.7512321558447025,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.6632,
+      "step": 36
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.7750132744507578,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.6512,
+      "step": 37
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.7967123908817402,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.6663,
+      "step": 38
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.6947515112348691,
+      "learning_rate": 4e-05,
+      "loss": 0.6699,
+      "step": 39
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.8838121858158148,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.6103,
+      "step": 40
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.937953074232824,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.6294,
+      "step": 41
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.6663100984929431,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.63,
+      "step": 42
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.9686394127674997,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.6555,
+      "step": 43
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 1.0948792609440945,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.6747,
+      "step": 44
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.7646908276106144,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.6803,
+      "step": 45
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 1.0848540792128565,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.6544,
+      "step": 46
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 1.0115954713396924,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.6655,
+      "step": 47
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.9120073804372092,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.6521,
+      "step": 48
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 1.050855454401179,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.6357,
+      "step": 49
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5475520022510691,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.6443,
+      "step": 50
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.8273962306006929,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.6465,
+      "step": 51
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.7881464919038017,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.5951,
+      "step": 52
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.9537798288964097,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.6357,
+      "step": 53
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.579153228532081,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.6657,
+      "step": 54
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.074195784702514,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.62,
+      "step": 55
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.7439084879563846,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.6552,
+      "step": 56
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.9296398158867684,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.6265,
+      "step": 57
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 1.0194408640683725,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.6316,
+      "step": 58
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.7355167158335016,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.634,
+      "step": 59
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.906333179886472,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.6566,
+      "step": 60
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.9984252494246183,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.6418,
+      "step": 61
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.6288544699214411,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.6489,
+      "step": 62
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.9808612656456591,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.6625,
+      "step": 63
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.7864506596926686,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.6287,
+      "step": 64
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.7673819391613449,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.615,
+      "step": 65
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 1.0093619808484628,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.6237,
+      "step": 66
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.9977549136570084,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.6343,
+      "step": 67
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.6027710823264074,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.5971,
+      "step": 68
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.6159304324960394,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.6243,
+      "step": 69
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.6213381460421435,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.6372,
+      "step": 70
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.8756713112352614,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.6413,
+      "step": 71
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.7675343635238856,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.6426,
+      "step": 72
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.8581644262943975,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.6203,
+      "step": 73
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.9530066853833877,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.6412,
+      "step": 74
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6732284794508566,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.6329,
+      "step": 75
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.6765379432047343,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.6135,
+      "step": 76
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.6269936249438118,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.5911,
+      "step": 77
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.5241257450024129,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6083,
+      "step": 78
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 1.6719128755921155,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 1.2361,
+      "step": 79
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.8044098947452024,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.5482,
+      "step": 80
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 1.0946885767510257,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.5149,
+      "step": 81
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.8598307990653626,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.4693,
+      "step": 82
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.8552065496627894,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.5323,
+      "step": 83
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.8239284111177897,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.5492,
+      "step": 84
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.6347944568201517,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.5168,
+      "step": 85
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.653364996874192,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.5329,
+      "step": 86
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.6989776738975938,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.5412,
+      "step": 87
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.7181508475505547,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.5296,
+      "step": 88
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.6911928102780305,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.534,
+      "step": 89
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.6436708383669849,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.5218,
+      "step": 90
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.6396133336267467,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5101,
+      "step": 91
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.6423502214891712,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.5607,
+      "step": 92
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.6214000215124265,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.4887,
+      "step": 93
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.5503754472052366,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.5348,
+      "step": 94
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.5973312350989958,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.5112,
+      "step": 95
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.485838378982901,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.5239,
+      "step": 96
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.5077143548051274,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.5178,
+      "step": 97
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.5689308324907205,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.4831,
+      "step": 98
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.6449782413165691,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.5062,
+      "step": 99
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6021971176344099,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.5333,
+      "step": 100
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.7193952024706654,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.5002,
+      "step": 101
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.49547113809381366,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.5048,
+      "step": 102
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.6618235319736041,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.5119,
+      "step": 103
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.5644535979761948,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5088,
+      "step": 104
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.5525405461144406,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.5037,
+      "step": 105
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.8144469278978772,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.5186,
+      "step": 106
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.44891431256783426,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.5176,
+      "step": 107
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.6221611200366272,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.4977,
+      "step": 108
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.43947676722734164,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.4976,
+      "step": 109
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.5218160924853039,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.5085,
+      "step": 110
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.5034372949967536,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.5002,
+      "step": 111
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.45895268200491246,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.4976,
+      "step": 112
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.46321651746461323,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.4968,
+      "step": 113
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.47870495552411885,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.5287,
+      "step": 114
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.474762382449506,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.4974,
+      "step": 115
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.47651416342315034,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.4622,
+      "step": 116
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.5410630445106255,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5288,
+      "step": 117
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.5637184608262586,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.5132,
+      "step": 118
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.5727458424587676,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.4653,
+      "step": 119
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.5802567977235668,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.527,
+      "step": 120
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.5707103092411799,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.522,
+      "step": 121
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.5896233440772041,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.516,
+      "step": 122
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.47476490550748923,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.4763,
+      "step": 123
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.4995541864941269,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.4972,
+      "step": 124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5157397952449607,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.5135,
+      "step": 125
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.5265684040798319,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.5075,
+      "step": 126
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.6509664756017955,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.4865,
+      "step": 127
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.5230304058607859,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.5053,
+      "step": 128
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.564502117499931,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.4876,
+      "step": 129
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.5313724748821392,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5109,
+      "step": 130
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.49048494571733037,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.5027,
+      "step": 131
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.483002582063503,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.4732,
+      "step": 132
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.5536908988980134,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.4814,
+      "step": 133
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.5135758035148594,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.5222,
+      "step": 134
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.6693583165294315,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.5069,
+      "step": 135
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.6703936126866795,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.5171,
+      "step": 136
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.4402756674176734,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.5164,
+      "step": 137
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.4834750840488784,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.4934,
+      "step": 138
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.5001628764630827,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.5172,
+      "step": 139
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.5156475824631777,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.5196,
+      "step": 140
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.46421714247369034,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.4828,
+      "step": 141
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.4435566342919553,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.5066,
+      "step": 142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.4472454124285204,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4885,
+      "step": 143
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.45149483986705774,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.4752,
+      "step": 144
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.43737137429775913,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.5025,
+      "step": 145
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.4144061152549108,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.4949,
+      "step": 146
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.45685221044206653,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.5056,
+      "step": 147
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.4384227669456062,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.4953,
+      "step": 148
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.4583195214456875,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.4972,
+      "step": 149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5036176164018994,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.5273,
+      "step": 150
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.3900927170850886,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.4691,
+      "step": 151
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.5187118022065168,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.5203,
+      "step": 152
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.4231705325668886,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.5122,
+      "step": 153
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.44469869668183354,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.5395,
+      "step": 154
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.41913855903509245,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.5042,
+      "step": 155
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.44923975878690897,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.479,
+      "step": 156
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 1.265886045380679,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.9111,
+      "step": 157
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.7788939887932237,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.3983,
+      "step": 158
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 1.5814266423710819,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.3809,
+      "step": 159
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.6541946579068361,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.3784,
+      "step": 160
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 1.2254090579151116,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.3789,
+      "step": 161
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.6517013681171183,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.3872,
+      "step": 162
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.7885832265283199,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.3535,
+      "step": 163
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.5877804328511399,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.3413,
+      "step": 164
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.6404825701419146,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.3907,
+      "step": 165
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.6911953292508461,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.3687,
+      "step": 166
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.5199471791956721,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.3874,
+      "step": 167
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.6434304308376638,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.3429,
+      "step": 168
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 5.039222272531735,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4248,
+      "step": 169
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 1.5682575966736636,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.4148,
+      "step": 170
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 2.2116325260594603,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.3839,
+      "step": 171
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 4.965840848460932,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.422,
+      "step": 172
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 8.972880867629566,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.4676,
+      "step": 173
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 6.120391510692015,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.4401,
+      "step": 174
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.7351836374094791,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.4232,
+      "step": 175
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 3.5854369520818015,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.3999,
+      "step": 176
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 1.9878205895223442,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.3703,
+      "step": 177
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 33.41847469258458,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.4428,
+      "step": 178
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.7737787985631442,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.3908,
+      "step": 179
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.5908933797743031,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.3526,
+      "step": 180
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.8625462111400396,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.3676,
+      "step": 181
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.7499809401337406,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.391,
+      "step": 182
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.49289242477309925,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.3627,
+      "step": 183
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.6436201609705232,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.3823,
+      "step": 184
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.7416369441658253,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.379,
+      "step": 185
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.49608853671818276,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.3519,
+      "step": 186
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.5592841113161919,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.3657,
+      "step": 187
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.5517950962622802,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.345,
+      "step": 188
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.5506982072115814,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.3873,
+      "step": 189
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.5523277228718634,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.3686,
+      "step": 190
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.4836562777887041,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.3363,
+      "step": 191
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.4856236540285901,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.3713,
+      "step": 192
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.43993841071604617,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.3343,
+      "step": 193
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.5734289420004812,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.3895,
+      "step": 194
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 1.0507912669903445,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3458,
+      "step": 195
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.6631940633125842,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.3981,
+      "step": 196
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.46515671795501223,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.3588,
+      "step": 197
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.4918452599317938,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.389,
+      "step": 198
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.4554316499011092,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.3461,
+      "step": 199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.455662913331897,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.3636,
+      "step": 200
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.3935422562046005,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.3426,
+      "step": 201
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.5002040423927975,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.3837,
+      "step": 202
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.43073063190022365,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.3753,
+      "step": 203
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.41187216224269874,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.3587,
+      "step": 204
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.43523550336085925,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.3793,
+      "step": 205
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.40840561472711745,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.358,
+      "step": 206
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.4271901296090871,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.3506,
+      "step": 207
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.42938531914891453,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3725,
+      "step": 208
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.385551093697211,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.3448,
+      "step": 209
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.42732610929649756,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.354,
+      "step": 210
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.4363697439406535,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.3588,
+      "step": 211
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.4455666934289362,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.385,
+      "step": 212
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.48232062459395736,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.3644,
+      "step": 213
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.43697749519928014,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.3639,
+      "step": 214
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.42215823476760755,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.3564,
+      "step": 215
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.4533057037652281,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.3585,
+      "step": 216
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.412936501647393,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.386,
+      "step": 217
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.44281905988539394,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.3527,
+      "step": 218
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.3804050542534375,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.3483,
+      "step": 219
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.41179401319943626,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.3554,
+      "step": 220
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.43315272736795446,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3702,
+      "step": 221
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.398555958378086,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.351,
+      "step": 222
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.42222659633240006,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.3752,
+      "step": 223
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.4358304314931661,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.3791,
+      "step": 224
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.39597183289865395,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.3442,
+      "step": 225
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.4041807692290575,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.3729,
+      "step": 226
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.4300731136350461,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.3545,
+      "step": 227
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.3879260964768763,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.3596,
+      "step": 228
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.4055432164000725,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.3365,
+      "step": 229
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.3958328462192638,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.3664,
+      "step": 230
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.38826305855875354,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.3612,
+      "step": 231
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.445023011920815,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.3524,
+      "step": 232
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.40354636978208697,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.401,
+      "step": 233
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.3723224522633442,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3454,
+      "step": 234
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 1.0760263945169621,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.583,
+      "step": 235
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.57256564444408,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.2506,
+      "step": 236
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.5147206452202644,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.247,
+      "step": 237
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.7716271271290052,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.2848,
+      "step": 238
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.6412161087083598,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.2767,
+      "step": 239
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.46626647972419805,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.2324,
+      "step": 240
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.5448589396475085,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.2555,
+      "step": 241
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.585804456518818,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.2895,
+      "step": 242
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.4300412895380309,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.2508,
+      "step": 243
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.5168234192574198,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.2685,
+      "step": 244
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.4893794924489026,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.2709,
+      "step": 245
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.4968317843194202,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.2517,
+      "step": 246
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.3928163060384461,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2569,
+      "step": 247
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.4237663552092556,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.2543,
+      "step": 248
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.5067424096750924,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.2762,
+      "step": 249
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.36213911630724266,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.235,
+      "step": 250
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.45366427130401843,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.2531,
+      "step": 251
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.43311027908123373,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.2678,
+      "step": 252
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.37955430937803575,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.2341,
+      "step": 253
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.4332081369900881,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.2428,
+      "step": 254
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.40689868552476915,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.2615,
+      "step": 255
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.37956577708548206,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.255,
+      "step": 256
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.38931315034089564,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.2396,
+      "step": 257
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.3867181588805245,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.2774,
+      "step": 258
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.36661214027127537,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.2476,
+      "step": 259
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.3917198545486744,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.266,
+      "step": 260
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.38107425210510537,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.2668,
+      "step": 261
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.3639150748646448,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.2305,
+      "step": 262
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.3895921125067074,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.2739,
+      "step": 263
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.3604470198674436,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.2472,
+      "step": 264
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.3573866946305205,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.2442,
+      "step": 265
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.3751341303884494,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.2434,
+      "step": 266
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.3533824095734159,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.2335,
+      "step": 267
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.38925546709181347,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.2563,
+      "step": 268
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.3667682256849455,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.237,
+      "step": 269
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.3501415228169262,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.2521,
+      "step": 270
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.34288244851982835,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.2521,
+      "step": 271
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.3343160035531691,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.2486,
+      "step": 272
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.35266964091059416,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2634,
+      "step": 273
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.3310664566174733,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.2499,
+      "step": 274
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3347015557547574,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.2596,
+      "step": 275
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.32304663423209445,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.2273,
+      "step": 276
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.34703664582821353,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.2573,
+      "step": 277
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.34287290395368464,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.2408,
+      "step": 278
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.34811520345371805,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.2707,
+      "step": 279
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.36299928886945515,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.2669,
+      "step": 280
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.3168750185679657,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.2354,
+      "step": 281
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.3350612454059626,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.2355,
+      "step": 282
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.3251488389428146,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.2664,
+      "step": 283
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.3181776200149048,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.2501,
+      "step": 284
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.3283512698070235,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.2543,
+      "step": 285
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.3330946669696271,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.261,
+      "step": 286
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.3190143994069841,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.2405,
+      "step": 287
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.3762307158093425,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.2732,
+      "step": 288
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.3241138302819481,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.2399,
+      "step": 289
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.3168032723774227,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.2517,
+      "step": 290
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.3266848288775348,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.2321,
+      "step": 291
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.3385676680973875,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.2338,
+      "step": 292
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.3460231212436604,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.2871,
+      "step": 293
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.32517381734622913,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.263,
+      "step": 294
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.30847719176822114,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.2309,
+      "step": 295
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.3369417730320049,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.264,
+      "step": 296
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.31869551674524277,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.2379,
+      "step": 297
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.31609783042835005,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.2465,
+      "step": 298
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.32882321257536323,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2542,
+      "step": 299
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3068370521320545,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.2461,
+      "step": 300
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.32758411147114447,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.2409,
+      "step": 301
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.3221173256989185,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.2515,
+      "step": 302
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.32264423655414426,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.2466,
+      "step": 303
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.32539030618048703,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.2592,
+      "step": 304
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.32643194959120964,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.2538,
+      "step": 305
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.32112558543154684,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.2324,
+      "step": 306
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.32459137125078197,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.2573,
+      "step": 307
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.32316349415665657,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.2761,
+      "step": 308
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.3350750218103054,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.2614,
+      "step": 309
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.32730727205744853,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.2495,
+      "step": 310
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.333725463894041,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.2475,
+      "step": 311
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.3294470948013299,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2643,
+      "step": 312
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 1.2839108579019607,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.4669,
+      "step": 313
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.5107764857543692,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.1854,
+      "step": 314
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.5032474964702015,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.1981,
+      "step": 315
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.3739097179041935,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.1925,
+      "step": 316
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.30220129362139286,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.1865,
+      "step": 317
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.3292300542252776,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.1717,
+      "step": 318
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.46968083528040183,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.198,
+      "step": 319
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.498790951481086,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.1765,
+      "step": 320
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.521705304577139,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.1988,
+      "step": 321
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.3679435482726114,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.1705,
+      "step": 322
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.32621795497574985,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.1817,
+      "step": 323
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.35216362102409743,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.2028,
+      "step": 324
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.35599057654806565,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.187,
+      "step": 325
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.37254584373028504,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.1656,
+      "step": 326
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.37540403302100134,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.1909,
+      "step": 327
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.31221887227619116,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.1664,
+      "step": 328
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.31542463544830635,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.1779,
+      "step": 329
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.3211841194629076,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.1848,
+      "step": 330
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.27762081759466123,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.1492,
+      "step": 331
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.3385807046048009,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.1934,
+      "step": 332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.30736132267170674,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.1672,
+      "step": 333
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.3419430497892458,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.1709,
+      "step": 334
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.32901820205339777,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.1784,
+      "step": 335
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.32284606187029924,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.1883,
+      "step": 336
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.2979899558075267,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.19,
+      "step": 337
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.29358109019278905,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1805,
+      "step": 338
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.27168467699556326,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.1561,
+      "step": 339
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.29855700253581063,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.187,
+      "step": 340
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.3102102832565782,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.18,
+      "step": 341
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.32589748777310773,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.1962,
+      "step": 342
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.29069882328179,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.1859,
+      "step": 343
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.27290994923559936,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.1708,
+      "step": 344
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.2803971413210815,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.1631,
+      "step": 345
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.2877557641748574,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.175,
+      "step": 346
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.28713797935091373,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.1765,
+      "step": 347
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.2961951071739156,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.1886,
+      "step": 348
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.31406853686902136,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.1824,
+      "step": 349
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.30517288114113955,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.1828,
+      "step": 350
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.2832178998141666,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1722,
+      "step": 351
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.2815109264602268,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.1666,
+      "step": 352
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.29166166476356714,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.179,
+      "step": 353
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.2997842315599802,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.2035,
+      "step": 354
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.30570020154695926,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.1784,
+      "step": 355
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.2887441465885464,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.1758,
+      "step": 356
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.28667025350190384,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.174,
+      "step": 357
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.2759770219797655,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.1868,
+      "step": 358
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.2759288320823624,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.1781,
+      "step": 359
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.2861440779101426,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.2014,
+      "step": 360
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.27680717610008887,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.1841,
+      "step": 361
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.27828400360484457,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.1771,
+      "step": 362
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.2922440545224223,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.1739,
+      "step": 363
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.2888552049188564,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1891,
+      "step": 364
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.28350616794758077,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.1692,
+      "step": 365
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.288102760229299,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.172,
+      "step": 366
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.27176895224481923,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.1522,
+      "step": 367
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.28088791328196594,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.1645,
+      "step": 368
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.30491036385171727,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.1962,
+      "step": 369
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.29595033711883767,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.1991,
+      "step": 370
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.2844453957047236,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.1592,
+      "step": 371
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.2730226751995552,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.1608,
+      "step": 372
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.30054267330755824,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.1766,
+      "step": 373
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.27825341435155493,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.1676,
+      "step": 374
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.27728341090713626,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.1708,
+      "step": 375
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.27689669432403957,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.1707,
+      "step": 376
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.27564867108595503,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1752,
+      "step": 377
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.28092938065824846,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.178,
+      "step": 378
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.2985374876648191,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.1859,
+      "step": 379
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.270848132846425,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.1667,
+      "step": 380
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.29173327551585365,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.1828,
+      "step": 381
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.2751833092982596,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.1761,
+      "step": 382
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.2731068396465843,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.1705,
+      "step": 383
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.2832971823575287,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.181,
+      "step": 384
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.27417127684146547,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.1646,
+      "step": 385
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.2839052962915241,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.1792,
+      "step": 386
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.2736035642134928,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.1696,
+      "step": 387
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.2849611699605065,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.1696,
+      "step": 388
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.285288808839962,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.1822,
+      "step": 389
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.2833048397064369,
+      "learning_rate": 0.0,
+      "loss": 0.172,
+      "step": 390
+    },
+    {
+      "epoch": 4.992,
+      "step": 390,
+      "total_flos": 8.005281734805422e+17,
+      "train_loss": 0.40785385312942357,
+      "train_runtime": 14479.4896,
+      "train_samples_per_second": 3.453,
+      "train_steps_per_second": 0.027
+    }
+  ],
+  "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": 8.005281734805422e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a84d594b3f52acefb2e105cc9016236691367d0552009855f4363caf3d3de72
+size 7352