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: a1_math_open2math_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. -->
+
+# a1_math_open2math_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/a1_math_open2math_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 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 2.5.0a0+b465a5843b.nv24.09
+- 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": 6.874049729748009e+17,
+    "train_loss": 0.30696559743239327,
+    "train_runtime": 15599.7333,
+    "train_samples_per_second": 3.205,
+    "train_steps_per_second": 0.025
+}

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 sha256:28ba6311a4eeed86cd10322f3a51edba4565f19688ed084df63398bfd5d0ea09
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8bba3a43fd4b395d0fa342b3422f7e631490be73c10d064540116e49ec719b0
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a65871add0b0e7c8132b9c49531012097000494a20a9c3d89957149ceca4955
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fb8f7397847996125c652abe4f5acaf84f456fb15c60c6c760c5ec81cd27902
+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/May13_22-27-40_jrc0926/events.out.tfevents.1747168096.jrc0926.4089719.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8dafd8a896dd42d9537ce2d183ce3ad72770358a763db75e0c9c1a50d726808
+size 87648

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": "2025-05-13 22:27:40", "end_time": "2025-05-14 02:48:27"}

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": 6.874049729748009e+17,
+    "train_loss": 0.30696559743239327,
+    "train_runtime": 15599.7333,
+    "train_samples_per_second": 3.205,
+    "train_steps_per_second": 0.025
+}

trainer_log.jsonl

@@ -0,0 +1,391 @@
+{"current_steps": 1, "total_steps": 390, "loss": 0.7972, "lr": 1.0256410256410257e-06, "epoch": 0.0128, "percentage": 0.26, "elapsed_time": "0:00:53", "remaining_time": "5:46:31"}
+{"current_steps": 2, "total_steps": 390, "loss": 0.8223, "lr": 2.0512820512820513e-06, "epoch": 0.0256, "percentage": 0.51, "elapsed_time": "0:01:30", "remaining_time": "4:53:42"}
+{"current_steps": 3, "total_steps": 390, "loss": 0.8336, "lr": 3.0769230769230774e-06, "epoch": 0.0384, "percentage": 0.77, "elapsed_time": "0:02:12", "remaining_time": "4:45:25"}
+{"current_steps": 4, "total_steps": 390, "loss": 0.794, "lr": 4.102564102564103e-06, "epoch": 0.0512, "percentage": 1.03, "elapsed_time": "0:02:52", "remaining_time": "4:36:59"}
+{"current_steps": 5, "total_steps": 390, "loss": 0.7651, "lr": 5.128205128205128e-06, "epoch": 0.064, "percentage": 1.28, "elapsed_time": "0:03:26", "remaining_time": "4:25:30"}
+{"current_steps": 6, "total_steps": 390, "loss": 0.6905, "lr": 6.153846153846155e-06, "epoch": 0.0768, "percentage": 1.54, "elapsed_time": "0:04:17", "remaining_time": "4:34:31"}
+{"current_steps": 7, "total_steps": 390, "loss": 0.6952, "lr": 7.17948717948718e-06, "epoch": 0.0896, "percentage": 1.79, "elapsed_time": "0:04:53", "remaining_time": "4:27:54"}
+{"current_steps": 8, "total_steps": 390, "loss": 0.7169, "lr": 8.205128205128205e-06, "epoch": 0.1024, "percentage": 2.05, "elapsed_time": "0:05:25", "remaining_time": "4:18:51"}
+{"current_steps": 9, "total_steps": 390, "loss": 0.6915, "lr": 9.230769230769232e-06, "epoch": 0.1152, "percentage": 2.31, "elapsed_time": "0:06:03", "remaining_time": "4:16:42"}
+{"current_steps": 10, "total_steps": 390, "loss": 0.6619, "lr": 1.0256410256410256e-05, "epoch": 0.128, "percentage": 2.56, "elapsed_time": "0:06:46", "remaining_time": "4:17:34"}
+{"current_steps": 11, "total_steps": 390, "loss": 0.6244, "lr": 1.1282051282051283e-05, "epoch": 0.1408, "percentage": 2.82, "elapsed_time": "0:07:23", "remaining_time": "4:14:40"}
+{"current_steps": 12, "total_steps": 390, "loss": 0.6451, "lr": 1.230769230769231e-05, "epoch": 0.1536, "percentage": 3.08, "elapsed_time": "0:08:03", "remaining_time": "4:13:38"}
+{"current_steps": 13, "total_steps": 390, "loss": 0.5837, "lr": 1.3333333333333333e-05, "epoch": 0.1664, "percentage": 3.33, "elapsed_time": "0:08:40", "remaining_time": "4:11:27"}
+{"current_steps": 14, "total_steps": 390, "loss": 0.6217, "lr": 1.435897435897436e-05, "epoch": 0.1792, "percentage": 3.59, "elapsed_time": "0:09:20", "remaining_time": "4:10:40"}
+{"current_steps": 15, "total_steps": 390, "loss": 0.5798, "lr": 1.5384615384615387e-05, "epoch": 0.192, "percentage": 3.85, "elapsed_time": "0:09:52", "remaining_time": "4:06:58"}
+{"current_steps": 16, "total_steps": 390, "loss": 0.5836, "lr": 1.641025641025641e-05, "epoch": 0.2048, "percentage": 4.1, "elapsed_time": "0:10:38", "remaining_time": "4:08:46"}
+{"current_steps": 17, "total_steps": 390, "loss": 0.5692, "lr": 1.7435897435897438e-05, "epoch": 0.2176, "percentage": 4.36, "elapsed_time": "0:11:21", "remaining_time": "4:09:20"}
+{"current_steps": 18, "total_steps": 390, "loss": 0.5896, "lr": 1.8461538461538465e-05, "epoch": 0.2304, "percentage": 4.62, "elapsed_time": "0:11:59", "remaining_time": "4:07:56"}
+{"current_steps": 19, "total_steps": 390, "loss": 0.5607, "lr": 1.9487179487179488e-05, "epoch": 0.2432, "percentage": 4.87, "elapsed_time": "0:12:42", "remaining_time": "4:08:00"}
+{"current_steps": 20, "total_steps": 390, "loss": 0.5338, "lr": 2.0512820512820512e-05, "epoch": 0.256, "percentage": 5.13, "elapsed_time": "0:13:19", "remaining_time": "4:06:36"}
+{"current_steps": 21, "total_steps": 390, "loss": 0.5467, "lr": 2.153846153846154e-05, "epoch": 0.2688, "percentage": 5.38, "elapsed_time": "0:13:56", "remaining_time": "4:04:56"}
+{"current_steps": 22, "total_steps": 390, "loss": 0.5649, "lr": 2.2564102564102566e-05, "epoch": 0.2816, "percentage": 5.64, "elapsed_time": "0:14:41", "remaining_time": "4:05:51"}
+{"current_steps": 23, "total_steps": 390, "loss": 0.5375, "lr": 2.3589743589743593e-05, "epoch": 0.2944, "percentage": 5.9, "elapsed_time": "0:15:24", "remaining_time": "4:05:48"}
+{"current_steps": 24, "total_steps": 390, "loss": 0.4995, "lr": 2.461538461538462e-05, "epoch": 0.3072, "percentage": 6.15, "elapsed_time": "0:16:06", "remaining_time": "4:05:34"}
+{"current_steps": 25, "total_steps": 390, "loss": 0.5509, "lr": 2.5641025641025646e-05, "epoch": 0.32, "percentage": 6.41, "elapsed_time": "0:16:42", "remaining_time": "4:03:56"}
+{"current_steps": 26, "total_steps": 390, "loss": 0.5251, "lr": 2.6666666666666667e-05, "epoch": 0.3328, "percentage": 6.67, "elapsed_time": "0:17:18", "remaining_time": "4:02:22"}
+{"current_steps": 27, "total_steps": 390, "loss": 0.493, "lr": 2.7692307692307694e-05, "epoch": 0.3456, "percentage": 6.92, "elapsed_time": "0:17:54", "remaining_time": "4:00:42"}
+{"current_steps": 28, "total_steps": 390, "loss": 0.5124, "lr": 2.871794871794872e-05, "epoch": 0.3584, "percentage": 7.18, "elapsed_time": "0:18:35", "remaining_time": "4:00:27"}
+{"current_steps": 29, "total_steps": 390, "loss": 0.4934, "lr": 2.9743589743589747e-05, "epoch": 0.3712, "percentage": 7.44, "elapsed_time": "0:19:17", "remaining_time": "4:00:03"}
+{"current_steps": 30, "total_steps": 390, "loss": 0.4959, "lr": 3.0769230769230774e-05, "epoch": 0.384, "percentage": 7.69, "elapsed_time": "0:19:59", "remaining_time": "3:59:49"}
+{"current_steps": 31, "total_steps": 390, "loss": 0.5276, "lr": 3.1794871794871795e-05, "epoch": 0.3968, "percentage": 7.95, "elapsed_time": "0:20:40", "remaining_time": "3:59:23"}
+{"current_steps": 32, "total_steps": 390, "loss": 0.5056, "lr": 3.282051282051282e-05, "epoch": 0.4096, "percentage": 8.21, "elapsed_time": "0:21:17", "remaining_time": "3:58:17"}
+{"current_steps": 33, "total_steps": 390, "loss": 0.4775, "lr": 3.384615384615385e-05, "epoch": 0.4224, "percentage": 8.46, "elapsed_time": "0:21:59", "remaining_time": "3:57:54"}
+{"current_steps": 34, "total_steps": 390, "loss": 0.5025, "lr": 3.4871794871794875e-05, "epoch": 0.4352, "percentage": 8.72, "elapsed_time": "0:22:40", "remaining_time": "3:57:21"}
+{"current_steps": 35, "total_steps": 390, "loss": 0.4956, "lr": 3.58974358974359e-05, "epoch": 0.448, "percentage": 8.97, "elapsed_time": "0:23:21", "remaining_time": "3:56:55"}
+{"current_steps": 36, "total_steps": 390, "loss": 0.4795, "lr": 3.692307692307693e-05, "epoch": 0.4608, "percentage": 9.23, "elapsed_time": "0:24:02", "remaining_time": "3:56:29"}
+{"current_steps": 37, "total_steps": 390, "loss": 0.4886, "lr": 3.794871794871795e-05, "epoch": 0.4736, "percentage": 9.49, "elapsed_time": "0:24:45", "remaining_time": "3:56:11"}
+{"current_steps": 38, "total_steps": 390, "loss": 0.4662, "lr": 3.8974358974358976e-05, "epoch": 0.4864, "percentage": 9.74, "elapsed_time": "0:25:22", "remaining_time": "3:55:06"}
+{"current_steps": 39, "total_steps": 390, "loss": 0.4845, "lr": 4e-05, "epoch": 0.4992, "percentage": 10.0, "elapsed_time": "0:25:57", "remaining_time": "3:53:33"}
+{"current_steps": 40, "total_steps": 390, "loss": 0.4946, "lr": 3.9999198907597046e-05, "epoch": 0.512, "percentage": 10.26, "elapsed_time": "0:26:37", "remaining_time": "3:53:00"}
+{"current_steps": 41, "total_steps": 390, "loss": 0.4751, "lr": 3.9996795694563096e-05, "epoch": 0.5248, "percentage": 10.51, "elapsed_time": "0:27:21", "remaining_time": "3:52:49"}
+{"current_steps": 42, "total_steps": 390, "loss": 0.4879, "lr": 3.999279055341771e-05, "epoch": 0.5376, "percentage": 10.77, "elapsed_time": "0:28:01", "remaining_time": "3:52:10"}
+{"current_steps": 43, "total_steps": 390, "loss": 0.4667, "lr": 3.998718380500971e-05, "epoch": 0.5504, "percentage": 11.03, "elapsed_time": "0:28:43", "remaining_time": "3:51:46"}
+{"current_steps": 44, "total_steps": 390, "loss": 0.4761, "lr": 3.997997589849145e-05, "epoch": 0.5632, "percentage": 11.28, "elapsed_time": "0:29:26", "remaining_time": "3:51:31"}
+{"current_steps": 45, "total_steps": 390, "loss": 0.4654, "lr": 3.9971167411282835e-05, "epoch": 0.576, "percentage": 11.54, "elapsed_time": "0:30:04", "remaining_time": "3:50:37"}
+{"current_steps": 46, "total_steps": 390, "loss": 0.4543, "lr": 3.99607590490251e-05, "epoch": 0.5888, "percentage": 11.79, "elapsed_time": "0:30:48", "remaining_time": "3:50:25"}
+{"current_steps": 47, "total_steps": 390, "loss": 0.4758, "lr": 3.9948751645524235e-05, "epoch": 0.6016, "percentage": 12.05, "elapsed_time": "0:31:32", "remaining_time": "3:50:08"}
+{"current_steps": 48, "total_steps": 390, "loss": 0.4906, "lr": 3.9935146162684206e-05, "epoch": 0.6144, "percentage": 12.31, "elapsed_time": "0:32:15", "remaining_time": "3:49:51"}
+{"current_steps": 49, "total_steps": 390, "loss": 0.4632, "lr": 3.9919943690429906e-05, "epoch": 0.6272, "percentage": 12.56, "elapsed_time": "0:32:47", "remaining_time": "3:48:13"}
+{"current_steps": 50, "total_steps": 390, "loss": 0.4629, "lr": 3.9903145446619837e-05, "epoch": 0.64, "percentage": 12.82, "elapsed_time": "0:33:25", "remaining_time": "3:47:20"}
+{"current_steps": 51, "total_steps": 390, "loss": 0.4996, "lr": 3.9884752776948564e-05, "epoch": 0.6528, "percentage": 13.08, "elapsed_time": "0:34:07", "remaining_time": "3:46:47"}
+{"current_steps": 52, "total_steps": 390, "loss": 0.4734, "lr": 3.9864767154838864e-05, "epoch": 0.6656, "percentage": 13.33, "elapsed_time": "0:34:45", "remaining_time": "3:45:58"}
+{"current_steps": 53, "total_steps": 390, "loss": 0.4543, "lr": 3.9843190181323744e-05, "epoch": 0.6784, "percentage": 13.59, "elapsed_time": "0:35:26", "remaining_time": "3:45:19"}
+{"current_steps": 54, "total_steps": 390, "loss": 0.4908, "lr": 3.982002358491817e-05, "epoch": 0.6912, "percentage": 13.85, "elapsed_time": "0:36:07", "remaining_time": "3:44:44"}
+{"current_steps": 55, "total_steps": 390, "loss": 0.4663, "lr": 3.979526922148058e-05, "epoch": 0.704, "percentage": 14.1, "elapsed_time": "0:36:41", "remaining_time": "3:43:26"}
+{"current_steps": 56, "total_steps": 390, "loss": 0.4809, "lr": 3.9768929074064206e-05, "epoch": 0.7168, "percentage": 14.36, "elapsed_time": "0:37:16", "remaining_time": "3:42:18"}
+{"current_steps": 57, "total_steps": 390, "loss": 0.4784, "lr": 3.9741005252758255e-05, "epoch": 0.7296, "percentage": 14.62, "elapsed_time": "0:38:02", "remaining_time": "3:42:16"}
+{"current_steps": 58, "total_steps": 390, "loss": 0.4806, "lr": 3.971149999451886e-05, "epoch": 0.7424, "percentage": 14.87, "elapsed_time": "0:38:48", "remaining_time": "3:42:08"}
+{"current_steps": 59, "total_steps": 390, "loss": 0.4854, "lr": 3.9680415662989806e-05, "epoch": 0.7552, "percentage": 15.13, "elapsed_time": "0:39:33", "remaining_time": "3:41:53"}
+{"current_steps": 60, "total_steps": 390, "loss": 0.4903, "lr": 3.9647754748313294e-05, "epoch": 0.768, "percentage": 15.38, "elapsed_time": "0:40:19", "remaining_time": "3:41:47"}
+{"current_steps": 61, "total_steps": 390, "loss": 0.4707, "lr": 3.96135198669304e-05, "epoch": 0.7808, "percentage": 15.64, "elapsed_time": "0:41:02", "remaining_time": "3:41:23"}
+{"current_steps": 62, "total_steps": 390, "loss": 0.4474, "lr": 3.957771376137144e-05, "epoch": 0.7936, "percentage": 15.9, "elapsed_time": "0:41:40", "remaining_time": "3:40:28"}
+{"current_steps": 63, "total_steps": 390, "loss": 0.4529, "lr": 3.954033930003634e-05, "epoch": 0.8064, "percentage": 16.15, "elapsed_time": "0:42:13", "remaining_time": "3:39:10"}
+{"current_steps": 64, "total_steps": 390, "loss": 0.445, "lr": 3.9501399476964806e-05, "epoch": 0.8192, "percentage": 16.41, "elapsed_time": "0:42:56", "remaining_time": "3:38:46"}
+{"current_steps": 65, "total_steps": 390, "loss": 0.4589, "lr": 3.946089741159648e-05, "epoch": 0.832, "percentage": 16.67, "elapsed_time": "0:43:32", "remaining_time": "3:37:44"}
+{"current_steps": 66, "total_steps": 390, "loss": 0.4857, "lr": 3.9418836348521045e-05, "epoch": 0.8448, "percentage": 16.92, "elapsed_time": "0:44:19", "remaining_time": "3:37:33"}
+{"current_steps": 67, "total_steps": 390, "loss": 0.4618, "lr": 3.937521965721831e-05, "epoch": 0.8576, "percentage": 17.18, "elapsed_time": "0:45:00", "remaining_time": "3:36:56"}
+{"current_steps": 68, "total_steps": 390, "loss": 0.4492, "lr": 3.933005083178828e-05, "epoch": 0.8704, "percentage": 17.44, "elapsed_time": "0:45:36", "remaining_time": "3:35:56"}
+{"current_steps": 69, "total_steps": 390, "loss": 0.4429, "lr": 3.928333349067125e-05, "epoch": 0.8832, "percentage": 17.69, "elapsed_time": "0:46:15", "remaining_time": "3:35:10"}
+{"current_steps": 70, "total_steps": 390, "loss": 0.4579, "lr": 3.923507137635792e-05, "epoch": 0.896, "percentage": 17.95, "elapsed_time": "0:46:54", "remaining_time": "3:34:24"}
+{"current_steps": 71, "total_steps": 390, "loss": 0.4445, "lr": 3.9185268355089606e-05, "epoch": 0.9088, "percentage": 18.21, "elapsed_time": "0:47:26", "remaining_time": "3:33:11"}
+{"current_steps": 72, "total_steps": 390, "loss": 0.4653, "lr": 3.913392841654851e-05, "epoch": 0.9216, "percentage": 18.46, "elapsed_time": "0:48:10", "remaining_time": "3:32:47"}
+{"current_steps": 73, "total_steps": 390, "loss": 0.4588, "lr": 3.9081055673538093e-05, "epoch": 0.9344, "percentage": 18.72, "elapsed_time": "0:48:51", "remaining_time": "3:32:11"}
+{"current_steps": 74, "total_steps": 390, "loss": 0.4429, "lr": 3.902665436165364e-05, "epoch": 0.9472, "percentage": 18.97, "elapsed_time": "0:49:34", "remaining_time": "3:31:41"}
+{"current_steps": 75, "total_steps": 390, "loss": 0.4826, "lr": 3.897072883894291e-05, "epoch": 0.96, "percentage": 19.23, "elapsed_time": "0:50:21", "remaining_time": "3:31:30"}
+{"current_steps": 76, "total_steps": 390, "loss": 0.4544, "lr": 3.8913283585557054e-05, "epoch": 0.9728, "percentage": 19.49, "elapsed_time": "0:50:59", "remaining_time": "3:30:39"}
+{"current_steps": 77, "total_steps": 390, "loss": 0.4613, "lr": 3.885432320339167e-05, "epoch": 0.9856, "percentage": 19.74, "elapsed_time": "0:51:47", "remaining_time": "3:30:33"}
+{"current_steps": 78, "total_steps": 390, "loss": 0.4449, "lr": 3.879385241571817e-05, "epoch": 0.9984, "percentage": 20.0, "elapsed_time": "0:52:27", "remaining_time": "3:29:51"}
+{"current_steps": 79, "total_steps": 390, "loss": 0.7069, "lr": 3.873187606680543e-05, "epoch": 1.0112, "percentage": 20.26, "elapsed_time": "0:53:27", "remaining_time": "3:30:27"}
+{"current_steps": 80, "total_steps": 390, "loss": 0.3875, "lr": 3.866839912153168e-05, "epoch": 1.024, "percentage": 20.51, "elapsed_time": "0:54:07", "remaining_time": "3:29:44"}
+{"current_steps": 81, "total_steps": 390, "loss": 0.3507, "lr": 3.860342666498677e-05, "epoch": 1.0368, "percentage": 20.77, "elapsed_time": "0:54:50", "remaining_time": "3:29:14"}
+{"current_steps": 82, "total_steps": 390, "loss": 0.4196, "lr": 3.853696390206484e-05, "epoch": 1.0496, "percentage": 21.03, "elapsed_time": "0:55:35", "remaining_time": "3:28:49"}
+{"current_steps": 83, "total_steps": 390, "loss": 0.3929, "lr": 3.846901615704734e-05, "epoch": 1.0624, "percentage": 21.28, "elapsed_time": "0:56:15", "remaining_time": "3:28:06"}
+{"current_steps": 84, "total_steps": 390, "loss": 0.4045, "lr": 3.839958887317649e-05, "epoch": 1.0752, "percentage": 21.54, "elapsed_time": "0:56:59", "remaining_time": "3:27:37"}
+{"current_steps": 85, "total_steps": 390, "loss": 0.3792, "lr": 3.832868761221926e-05, "epoch": 1.088, "percentage": 21.79, "elapsed_time": "0:57:40", "remaining_time": "3:26:57"}
+{"current_steps": 86, "total_steps": 390, "loss": 0.3781, "lr": 3.825631805402182e-05, "epoch": 1.1008, "percentage": 22.05, "elapsed_time": "0:58:13", "remaining_time": "3:25:49"}
+{"current_steps": 87, "total_steps": 390, "loss": 0.3997, "lr": 3.818248599605448e-05, "epoch": 1.1136, "percentage": 22.31, "elapsed_time": "0:58:53", "remaining_time": "3:25:05"}
+{"current_steps": 88, "total_steps": 390, "loss": 0.416, "lr": 3.810719735294731e-05, "epoch": 1.1264, "percentage": 22.56, "elapsed_time": "0:59:32", "remaining_time": "3:24:18"}
+{"current_steps": 89, "total_steps": 390, "loss": 0.3655, "lr": 3.8030458156016326e-05, "epoch": 1.1392, "percentage": 22.82, "elapsed_time": "1:00:15", "remaining_time": "3:23:48"}
+{"current_steps": 90, "total_steps": 390, "loss": 0.3895, "lr": 3.795227455278029e-05, "epoch": 1.152, "percentage": 23.08, "elapsed_time": "1:00:57", "remaining_time": "3:23:12"}
+{"current_steps": 91, "total_steps": 390, "loss": 0.403, "lr": 3.787265280646825e-05, "epoch": 1.1648, "percentage": 23.33, "elapsed_time": "1:01:33", "remaining_time": "3:22:17"}
+{"current_steps": 92, "total_steps": 390, "loss": 0.3995, "lr": 3.7791599295517825e-05, "epoch": 1.1776, "percentage": 23.59, "elapsed_time": "1:02:19", "remaining_time": "3:21:52"}
+{"current_steps": 93, "total_steps": 390, "loss": 0.3953, "lr": 3.7709120513064196e-05, "epoch": 1.1904, "percentage": 23.85, "elapsed_time": "1:03:01", "remaining_time": "3:21:15"}
+{"current_steps": 94, "total_steps": 390, "loss": 0.4154, "lr": 3.762522306641998e-05, "epoch": 1.2032, "percentage": 24.1, "elapsed_time": "1:03:45", "remaining_time": "3:20:46"}
+{"current_steps": 95, "total_steps": 390, "loss": 0.3607, "lr": 3.7539913676545874e-05, "epoch": 1.216, "percentage": 24.36, "elapsed_time": "1:04:29", "remaining_time": "3:20:17"}
+{"current_steps": 96, "total_steps": 390, "loss": 0.416, "lr": 3.745319917751229e-05, "epoch": 1.2288000000000001, "percentage": 24.62, "elapsed_time": "1:05:04", "remaining_time": "3:19:17"}
+{"current_steps": 97, "total_steps": 390, "loss": 0.3556, "lr": 3.736508651595188e-05, "epoch": 1.2416, "percentage": 24.87, "elapsed_time": "1:05:44", "remaining_time": "3:18:34"}
+{"current_steps": 98, "total_steps": 390, "loss": 0.4022, "lr": 3.727558275050301e-05, "epoch": 1.2544, "percentage": 25.13, "elapsed_time": "1:06:25", "remaining_time": "3:17:55"}
+{"current_steps": 99, "total_steps": 390, "loss": 0.3802, "lr": 3.718469505124434e-05, "epoch": 1.2671999999999999, "percentage": 25.38, "elapsed_time": "1:06:59", "remaining_time": "3:16:55"}
+{"current_steps": 100, "total_steps": 390, "loss": 0.4032, "lr": 3.709243069912041e-05, "epoch": 1.28, "percentage": 25.64, "elapsed_time": "1:07:41", "remaining_time": "3:16:17"}
+{"current_steps": 101, "total_steps": 390, "loss": 0.3724, "lr": 3.699879708535838e-05, "epoch": 1.2928, "percentage": 25.9, "elapsed_time": "1:08:26", "remaining_time": "3:15:49"}
+{"current_steps": 102, "total_steps": 390, "loss": 0.3848, "lr": 3.69038017108759e-05, "epoch": 1.3056, "percentage": 26.15, "elapsed_time": "1:09:08", "remaining_time": "3:15:13"}
+{"current_steps": 103, "total_steps": 390, "loss": 0.4048, "lr": 3.680745218568026e-05, "epoch": 1.3184, "percentage": 26.41, "elapsed_time": "1:09:50", "remaining_time": "3:14:36"}
+{"current_steps": 104, "total_steps": 390, "loss": 0.3684, "lr": 3.6709756228258735e-05, "epoch": 1.3312, "percentage": 26.67, "elapsed_time": "1:10:29", "remaining_time": "3:13:50"}
+{"current_steps": 105, "total_steps": 390, "loss": 0.3689, "lr": 3.6610721664960236e-05, "epoch": 1.3439999999999999, "percentage": 26.92, "elapsed_time": "1:11:08", "remaining_time": "3:13:06"}
+{"current_steps": 106, "total_steps": 390, "loss": 0.3929, "lr": 3.65103564293684e-05, "epoch": 1.3568, "percentage": 27.18, "elapsed_time": "1:11:49", "remaining_time": "3:12:25"}
+{"current_steps": 107, "total_steps": 390, "loss": 0.3837, "lr": 3.640866856166601e-05, "epoch": 1.3696, "percentage": 27.44, "elapsed_time": "1:12:29", "remaining_time": "3:11:44"}
+{"current_steps": 108, "total_steps": 390, "loss": 0.3822, "lr": 3.6305666207990886e-05, "epoch": 1.3824, "percentage": 27.69, "elapsed_time": "1:13:09", "remaining_time": "3:11:00"}
+{"current_steps": 109, "total_steps": 390, "loss": 0.391, "lr": 3.6201357619783336e-05, "epoch": 1.3952, "percentage": 27.95, "elapsed_time": "1:13:51", "remaining_time": "3:10:23"}
+{"current_steps": 110, "total_steps": 390, "loss": 0.3987, "lr": 3.609575115312511e-05, "epoch": 1.408, "percentage": 28.21, "elapsed_time": "1:14:33", "remaining_time": "3:09:48"}
+{"current_steps": 111, "total_steps": 390, "loss": 0.3669, "lr": 3.598885526807003e-05, "epoch": 1.4208, "percentage": 28.46, "elapsed_time": "1:15:11", "remaining_time": "3:08:58"}
+{"current_steps": 112, "total_steps": 390, "loss": 0.403, "lr": 3.5880678527966224e-05, "epoch": 1.4336, "percentage": 28.72, "elapsed_time": "1:15:50", "remaining_time": "3:08:15"}
+{"current_steps": 113, "total_steps": 390, "loss": 0.4006, "lr": 3.577122959877017e-05, "epoch": 1.4464000000000001, "percentage": 28.97, "elapsed_time": "1:16:28", "remaining_time": "3:07:29"}
+{"current_steps": 114, "total_steps": 390, "loss": 0.3685, "lr": 3.566051724835245e-05, "epoch": 1.4592, "percentage": 29.23, "elapsed_time": "1:17:12", "remaining_time": "3:06:55"}
+{"current_steps": 115, "total_steps": 390, "loss": 0.3875, "lr": 3.554855034579532e-05, "epoch": 1.472, "percentage": 29.49, "elapsed_time": "1:17:52", "remaining_time": "3:06:13"}
+{"current_steps": 116, "total_steps": 390, "loss": 0.3653, "lr": 3.5435337860682304e-05, "epoch": 1.4848, "percentage": 29.74, "elapsed_time": "1:18:28", "remaining_time": "3:05:22"}
+{"current_steps": 117, "total_steps": 390, "loss": 0.4005, "lr": 3.532088886237956e-05, "epoch": 1.4976, "percentage": 30.0, "elapsed_time": "1:19:08", "remaining_time": "3:04:39"}
+{"current_steps": 118, "total_steps": 390, "loss": 0.3654, "lr": 3.520521251930941e-05, "epoch": 1.5104, "percentage": 30.26, "elapsed_time": "1:19:47", "remaining_time": "3:03:56"}
+{"current_steps": 119, "total_steps": 390, "loss": 0.4001, "lr": 3.5088318098215805e-05, "epoch": 1.5232, "percentage": 30.51, "elapsed_time": "1:20:37", "remaining_time": "3:03:36"}
+{"current_steps": 120, "total_steps": 390, "loss": 0.3572, "lr": 3.497021496342203e-05, "epoch": 1.536, "percentage": 30.77, "elapsed_time": "1:21:16", "remaining_time": "3:02:52"}
+{"current_steps": 121, "total_steps": 390, "loss": 0.3467, "lr": 3.485091257608047e-05, "epoch": 1.5488, "percentage": 31.03, "elapsed_time": "1:21:56", "remaining_time": "3:02:09"}
+{"current_steps": 122, "total_steps": 390, "loss": 0.4082, "lr": 3.473042049341474e-05, "epoch": 1.5615999999999999, "percentage": 31.28, "elapsed_time": "1:22:43", "remaining_time": "3:01:42"}
+{"current_steps": 123, "total_steps": 390, "loss": 0.3347, "lr": 3.4608748367954064e-05, "epoch": 1.5744, "percentage": 31.54, "elapsed_time": "1:23:21", "remaining_time": "3:00:57"}
+{"current_steps": 124, "total_steps": 390, "loss": 0.3879, "lr": 3.4485905946759965e-05, "epoch": 1.5872000000000002, "percentage": 31.79, "elapsed_time": "1:24:06", "remaining_time": "3:00:24"}
+{"current_steps": 125, "total_steps": 390, "loss": 0.3642, "lr": 3.4361903070645484e-05, "epoch": 1.6, "percentage": 32.05, "elapsed_time": "1:24:49", "remaining_time": "2:59:50"}
+{"current_steps": 126, "total_steps": 390, "loss": 0.4074, "lr": 3.423674967338681e-05, "epoch": 1.6128, "percentage": 32.31, "elapsed_time": "1:25:31", "remaining_time": "2:59:11"}
+{"current_steps": 127, "total_steps": 390, "loss": 0.3801, "lr": 3.411045578092754e-05, "epoch": 1.6256, "percentage": 32.56, "elapsed_time": "1:26:09", "remaining_time": "2:58:25"}
+{"current_steps": 128, "total_steps": 390, "loss": 0.3757, "lr": 3.398303151057543e-05, "epoch": 1.6383999999999999, "percentage": 32.82, "elapsed_time": "1:26:49", "remaining_time": "2:57:43"}
+{"current_steps": 129, "total_steps": 390, "loss": 0.4022, "lr": 3.385448707019199e-05, "epoch": 1.6512, "percentage": 33.08, "elapsed_time": "1:27:28", "remaining_time": "2:56:59"}
+{"current_steps": 130, "total_steps": 390, "loss": 0.3948, "lr": 3.372483275737468e-05, "epoch": 1.6640000000000001, "percentage": 33.33, "elapsed_time": "1:28:09", "remaining_time": "2:56:19"}
+{"current_steps": 131, "total_steps": 390, "loss": 0.3889, "lr": 3.359407895863199e-05, "epoch": 1.6768, "percentage": 33.59, "elapsed_time": "1:28:51", "remaining_time": "2:55:41"}
+{"current_steps": 132, "total_steps": 390, "loss": 0.3741, "lr": 3.34622361485514e-05, "epoch": 1.6896, "percentage": 33.85, "elapsed_time": "1:29:33", "remaining_time": "2:55:03"}
+{"current_steps": 133, "total_steps": 390, "loss": 0.3891, "lr": 3.332931488896029e-05, "epoch": 1.7024, "percentage": 34.1, "elapsed_time": "1:30:06", "remaining_time": "2:54:06"}
+{"current_steps": 134, "total_steps": 390, "loss": 0.3739, "lr": 3.319532582807977e-05, "epoch": 1.7151999999999998, "percentage": 34.36, "elapsed_time": "1:30:43", "remaining_time": "2:53:19"}
+{"current_steps": 135, "total_steps": 390, "loss": 0.3685, "lr": 3.30602796996717e-05, "epoch": 1.728, "percentage": 34.62, "elapsed_time": "1:31:27", "remaining_time": "2:52:46"}
+{"current_steps": 136, "total_steps": 390, "loss": 0.3637, "lr": 3.2924187322178865e-05, "epoch": 1.7408000000000001, "percentage": 34.87, "elapsed_time": "1:32:06", "remaining_time": "2:52:01"}
+{"current_steps": 137, "total_steps": 390, "loss": 0.3995, "lr": 3.278705959785821e-05, "epoch": 1.7536, "percentage": 35.13, "elapsed_time": "1:32:43", "remaining_time": "2:51:13"}
+{"current_steps": 138, "total_steps": 390, "loss": 0.3717, "lr": 3.2648907511907544e-05, "epoch": 1.7664, "percentage": 35.38, "elapsed_time": "1:33:20", "remaining_time": "2:50:26"}
+{"current_steps": 139, "total_steps": 390, "loss": 0.3746, "lr": 3.250974213158555e-05, "epoch": 1.7792, "percentage": 35.64, "elapsed_time": "1:34:01", "remaining_time": "2:49:47"}
+{"current_steps": 140, "total_steps": 390, "loss": 0.3919, "lr": 3.23695746053251e-05, "epoch": 1.792, "percentage": 35.9, "elapsed_time": "1:34:44", "remaining_time": "2:49:11"}
+{"current_steps": 141, "total_steps": 390, "loss": 0.3711, "lr": 3.222841616184025e-05, "epoch": 1.8048, "percentage": 36.15, "elapsed_time": "1:35:21", "remaining_time": "2:48:24"}
+{"current_steps": 142, "total_steps": 390, "loss": 0.3822, "lr": 3.208627810922665e-05, "epoch": 1.8176, "percentage": 36.41, "elapsed_time": "1:35:55", "remaining_time": "2:47:31"}
+{"current_steps": 143, "total_steps": 390, "loss": 0.3996, "lr": 3.194317183405573e-05, "epoch": 1.8304, "percentage": 36.67, "elapsed_time": "1:36:33", "remaining_time": "2:46:46"}
+{"current_steps": 144, "total_steps": 390, "loss": 0.3613, "lr": 3.1799108800462466e-05, "epoch": 1.8432, "percentage": 36.92, "elapsed_time": "1:37:17", "remaining_time": "2:46:11"}
+{"current_steps": 145, "total_steps": 390, "loss": 0.3801, "lr": 3.1654100549227024e-05, "epoch": 1.8559999999999999, "percentage": 37.18, "elapsed_time": "1:37:54", "remaining_time": "2:45:25"}
+{"current_steps": 146, "total_steps": 390, "loss": 0.3793, "lr": 3.1508158696850275e-05, "epoch": 1.8688, "percentage": 37.44, "elapsed_time": "1:38:33", "remaining_time": "2:44:43"}
+{"current_steps": 147, "total_steps": 390, "loss": 0.3649, "lr": 3.136129493462312e-05, "epoch": 1.8816000000000002, "percentage": 37.69, "elapsed_time": "1:39:17", "remaining_time": "2:44:08"}
+{"current_steps": 148, "total_steps": 390, "loss": 0.3448, "lr": 3.121352102768998e-05, "epoch": 1.8944, "percentage": 37.95, "elapsed_time": "1:39:56", "remaining_time": "2:43:24"}
+{"current_steps": 149, "total_steps": 390, "loss": 0.3763, "lr": 3.106484881410628e-05, "epoch": 1.9072, "percentage": 38.21, "elapsed_time": "1:40:34", "remaining_time": "2:42:40"}
+{"current_steps": 150, "total_steps": 390, "loss": 0.3694, "lr": 3.091529020389009e-05, "epoch": 1.92, "percentage": 38.46, "elapsed_time": "1:41:10", "remaining_time": "2:41:53"}
+{"current_steps": 151, "total_steps": 390, "loss": 0.3931, "lr": 3.076485717806808e-05, "epoch": 1.9327999999999999, "percentage": 38.72, "elapsed_time": "1:41:47", "remaining_time": "2:41:06"}
+{"current_steps": 152, "total_steps": 390, "loss": 0.3628, "lr": 3.061356178771564e-05, "epoch": 1.9456, "percentage": 38.97, "elapsed_time": "1:42:27", "remaining_time": "2:40:25"}
+{"current_steps": 153, "total_steps": 390, "loss": 0.4089, "lr": 3.0461416152991555e-05, "epoch": 1.9584000000000001, "percentage": 39.23, "elapsed_time": "1:43:13", "remaining_time": "2:39:53"}
+{"current_steps": 154, "total_steps": 390, "loss": 0.3163, "lr": 3.0308432462167045e-05, "epoch": 1.9712, "percentage": 39.49, "elapsed_time": "1:43:53", "remaining_time": "2:39:12"}
+{"current_steps": 155, "total_steps": 390, "loss": 0.413, "lr": 3.015462297064936e-05, "epoch": 1.984, "percentage": 39.74, "elapsed_time": "1:44:32", "remaining_time": "2:38:30"}
+{"current_steps": 156, "total_steps": 390, "loss": 0.3589, "lr": 3.0000000000000004e-05, "epoch": 1.9968, "percentage": 40.0, "elapsed_time": "1:45:06", "remaining_time": "2:37:39"}
+{"current_steps": 157, "total_steps": 390, "loss": 0.616, "lr": 2.98445759369477e-05, "epoch": 2.0096, "percentage": 40.26, "elapsed_time": "1:46:02", "remaining_time": "2:37:21"}
+{"current_steps": 158, "total_steps": 390, "loss": 0.2665, "lr": 2.9688363232396056e-05, "epoch": 2.0224, "percentage": 40.51, "elapsed_time": "1:46:38", "remaining_time": "2:36:34"}
+{"current_steps": 159, "total_steps": 390, "loss": 0.3094, "lr": 2.9531374400426158e-05, "epoch": 2.0352, "percentage": 40.77, "elapsed_time": "1:47:19", "remaining_time": "2:35:55"}
+{"current_steps": 160, "total_steps": 390, "loss": 0.2876, "lr": 2.9373622017294075e-05, "epoch": 2.048, "percentage": 41.03, "elapsed_time": "1:48:07", "remaining_time": "2:35:25"}
+{"current_steps": 161, "total_steps": 390, "loss": 0.3013, "lr": 2.9215118720423375e-05, "epoch": 2.0608, "percentage": 41.28, "elapsed_time": "1:48:43", "remaining_time": "2:34:38"}
+{"current_steps": 162, "total_steps": 390, "loss": 0.2793, "lr": 2.9055877207392752e-05, "epoch": 2.0736, "percentage": 41.54, "elapsed_time": "1:49:24", "remaining_time": "2:33:58"}
+{"current_steps": 163, "total_steps": 390, "loss": 0.2753, "lr": 2.8895910234918828e-05, "epoch": 2.0864, "percentage": 41.79, "elapsed_time": "1:50:07", "remaining_time": "2:33:21"}
+{"current_steps": 164, "total_steps": 390, "loss": 0.2938, "lr": 2.873523061783426e-05, "epoch": 2.0992, "percentage": 42.05, "elapsed_time": "1:50:47", "remaining_time": "2:32:40"}
+{"current_steps": 165, "total_steps": 390, "loss": 0.3133, "lr": 2.8573851228061084e-05, "epoch": 2.112, "percentage": 42.31, "elapsed_time": "1:51:34", "remaining_time": "2:32:08"}
+{"current_steps": 166, "total_steps": 390, "loss": 0.2637, "lr": 2.8411784993579633e-05, "epoch": 2.1248, "percentage": 42.56, "elapsed_time": "1:52:10", "remaining_time": "2:31:22"}
+{"current_steps": 167, "total_steps": 390, "loss": 0.3112, "lr": 2.8249044897392814e-05, "epoch": 2.1376, "percentage": 42.82, "elapsed_time": "1:52:42", "remaining_time": "2:30:30"}
+{"current_steps": 168, "total_steps": 390, "loss": 0.3028, "lr": 2.80856439764861e-05, "epoch": 2.1504, "percentage": 43.08, "elapsed_time": "1:53:27", "remaining_time": "2:29:56"}
+{"current_steps": 169, "total_steps": 390, "loss": 0.2822, "lr": 2.792159532078314e-05, "epoch": 2.1632, "percentage": 43.33, "elapsed_time": "1:54:07", "remaining_time": "2:29:14"}
+{"current_steps": 170, "total_steps": 390, "loss": 0.3134, "lr": 2.77569120720971e-05, "epoch": 2.176, "percentage": 43.59, "elapsed_time": "1:54:50", "remaining_time": "2:28:37"}
+{"current_steps": 171, "total_steps": 390, "loss": 0.2833, "lr": 2.7591607423077932e-05, "epoch": 2.1888, "percentage": 43.85, "elapsed_time": "1:55:35", "remaining_time": "2:28:01"}
+{"current_steps": 172, "total_steps": 390, "loss": 0.2672, "lr": 2.7425694616155474e-05, "epoch": 2.2016, "percentage": 44.1, "elapsed_time": "1:56:12", "remaining_time": "2:27:16"}
+{"current_steps": 173, "total_steps": 390, "loss": 0.3206, "lr": 2.7259186942478656e-05, "epoch": 2.2144, "percentage": 44.36, "elapsed_time": "1:56:54", "remaining_time": "2:26:37"}
+{"current_steps": 174, "total_steps": 390, "loss": 0.2456, "lr": 2.7092097740850712e-05, "epoch": 2.2272, "percentage": 44.62, "elapsed_time": "1:57:31", "remaining_time": "2:25:53"}
+{"current_steps": 175, "total_steps": 390, "loss": 0.2885, "lr": 2.692444039666066e-05, "epoch": 2.24, "percentage": 44.87, "elapsed_time": "1:58:15", "remaining_time": "2:25:16"}
+{"current_steps": 176, "total_steps": 390, "loss": 0.3178, "lr": 2.6756228340810946e-05, "epoch": 2.2528, "percentage": 45.13, "elapsed_time": "1:58:50", "remaining_time": "2:24:29"}
+{"current_steps": 177, "total_steps": 390, "loss": 0.2685, "lr": 2.6587475048641596e-05, "epoch": 2.2656, "percentage": 45.38, "elapsed_time": "1:59:29", "remaining_time": "2:23:47"}
+{"current_steps": 178, "total_steps": 390, "loss": 0.3006, "lr": 2.6418194038850634e-05, "epoch": 2.2784, "percentage": 45.64, "elapsed_time": "2:00:08", "remaining_time": "2:23:05"}
+{"current_steps": 179, "total_steps": 390, "loss": 0.2559, "lr": 2.624839887241115e-05, "epoch": 2.2912, "percentage": 45.9, "elapsed_time": "2:00:47", "remaining_time": "2:22:22"}
+{"current_steps": 180, "total_steps": 390, "loss": 0.3054, "lr": 2.607810315148494e-05, "epoch": 2.304, "percentage": 46.15, "elapsed_time": "2:01:24", "remaining_time": "2:21:38"}
+{"current_steps": 181, "total_steps": 390, "loss": 0.2812, "lr": 2.5907320518332827e-05, "epoch": 2.3168, "percentage": 46.41, "elapsed_time": "2:02:08", "remaining_time": "2:21:02"}
+{"current_steps": 182, "total_steps": 390, "loss": 0.3065, "lr": 2.5736064654221808e-05, "epoch": 2.3296, "percentage": 46.67, "elapsed_time": "2:02:50", "remaining_time": "2:20:23"}
+{"current_steps": 183, "total_steps": 390, "loss": 0.3054, "lr": 2.5564349278329056e-05, "epoch": 2.3424, "percentage": 46.92, "elapsed_time": "2:03:32", "remaining_time": "2:19:44"}
+{"current_steps": 184, "total_steps": 390, "loss": 0.2787, "lr": 2.539218814664288e-05, "epoch": 2.3552, "percentage": 47.18, "elapsed_time": "2:04:08", "remaining_time": "2:18:58"}
+{"current_steps": 185, "total_steps": 390, "loss": 0.2745, "lr": 2.521959505086075e-05, "epoch": 2.368, "percentage": 47.44, "elapsed_time": "2:04:48", "remaining_time": "2:18:17"}
+{"current_steps": 186, "total_steps": 390, "loss": 0.2837, "lr": 2.5046583817284437e-05, "epoch": 2.3808, "percentage": 47.69, "elapsed_time": "2:05:27", "remaining_time": "2:17:36"}
+{"current_steps": 187, "total_steps": 390, "loss": 0.2776, "lr": 2.487316830571244e-05, "epoch": 2.3936, "percentage": 47.95, "elapsed_time": "2:06:06", "remaining_time": "2:16:54"}
+{"current_steps": 188, "total_steps": 390, "loss": 0.289, "lr": 2.4699362408329646e-05, "epoch": 2.4064, "percentage": 48.21, "elapsed_time": "2:06:49", "remaining_time": "2:16:16"}
+{"current_steps": 189, "total_steps": 390, "loss": 0.3012, "lr": 2.4525180048594452e-05, "epoch": 2.4192, "percentage": 48.46, "elapsed_time": "2:07:23", "remaining_time": "2:15:28"}
+{"current_steps": 190, "total_steps": 390, "loss": 0.2589, "lr": 2.435063518012335e-05, "epoch": 2.432, "percentage": 48.72, "elapsed_time": "2:07:57", "remaining_time": "2:14:41"}
+{"current_steps": 191, "total_steps": 390, "loss": 0.2996, "lr": 2.4175741785573177e-05, "epoch": 2.4448, "percentage": 48.97, "elapsed_time": "2:08:37", "remaining_time": "2:14:00"}
+{"current_steps": 192, "total_steps": 390, "loss": 0.2895, "lr": 2.4000513875520892e-05, "epoch": 2.4576000000000002, "percentage": 49.23, "elapsed_time": "2:09:23", "remaining_time": "2:13:26"}
+{"current_steps": 193, "total_steps": 390, "loss": 0.3278, "lr": 2.3824965487341247e-05, "epoch": 2.4704, "percentage": 49.49, "elapsed_time": "2:10:10", "remaining_time": "2:12:52"}
+{"current_steps": 194, "total_steps": 390, "loss": 0.2592, "lr": 2.3649110684082258e-05, "epoch": 2.4832, "percentage": 49.74, "elapsed_time": "2:10:47", "remaining_time": "2:12:08"}
+{"current_steps": 195, "total_steps": 390, "loss": 0.2738, "lr": 2.3472963553338614e-05, "epoch": 2.496, "percentage": 50.0, "elapsed_time": "2:11:28", "remaining_time": "2:11:28"}
+{"current_steps": 196, "total_steps": 390, "loss": 0.2889, "lr": 2.3296538206123134e-05, "epoch": 2.5088, "percentage": 50.26, "elapsed_time": "2:12:08", "remaining_time": "2:10:47"}
+{"current_steps": 197, "total_steps": 390, "loss": 0.3022, "lr": 2.311984877573636e-05, "epoch": 2.5216, "percentage": 50.51, "elapsed_time": "2:12:50", "remaining_time": "2:10:08"}
+{"current_steps": 198, "total_steps": 390, "loss": 0.2595, "lr": 2.2942909416634326e-05, "epoch": 2.5343999999999998, "percentage": 50.77, "elapsed_time": "2:13:32", "remaining_time": "2:09:29"}
+{"current_steps": 199, "total_steps": 390, "loss": 0.2641, "lr": 2.2765734303294666e-05, "epoch": 2.5472, "percentage": 51.03, "elapsed_time": "2:14:12", "remaining_time": "2:08:48"}
+{"current_steps": 200, "total_steps": 390, "loss": 0.2933, "lr": 2.2588337629081107e-05, "epoch": 2.56, "percentage": 51.28, "elapsed_time": "2:14:51", "remaining_time": "2:08:07"}
+{"current_steps": 201, "total_steps": 390, "loss": 0.304, "lr": 2.2410733605106462e-05, "epoch": 2.5728, "percentage": 51.54, "elapsed_time": "2:15:26", "remaining_time": "2:07:21"}
+{"current_steps": 202, "total_steps": 390, "loss": 0.2741, "lr": 2.2232936459094158e-05, "epoch": 2.5856, "percentage": 51.79, "elapsed_time": "2:16:04", "remaining_time": "2:06:38"}
+{"current_steps": 203, "total_steps": 390, "loss": 0.2851, "lr": 2.205496043423849e-05, "epoch": 2.5984, "percentage": 52.05, "elapsed_time": "2:16:47", "remaining_time": "2:06:00"}
+{"current_steps": 204, "total_steps": 390, "loss": 0.2889, "lr": 2.1876819788063586e-05, "epoch": 2.6112, "percentage": 52.31, "elapsed_time": "2:17:28", "remaining_time": "2:05:20"}
+{"current_steps": 205, "total_steps": 390, "loss": 0.2663, "lr": 2.16985287912813e-05, "epoch": 2.624, "percentage": 52.56, "elapsed_time": "2:18:01", "remaining_time": "2:04:33"}
+{"current_steps": 206, "total_steps": 390, "loss": 0.2848, "lr": 2.1520101726647922e-05, "epoch": 2.6368, "percentage": 52.82, "elapsed_time": "2:18:44", "remaining_time": "2:03:55"}
+{"current_steps": 207, "total_steps": 390, "loss": 0.2879, "lr": 2.1341552887820048e-05, "epoch": 2.6496, "percentage": 53.08, "elapsed_time": "2:19:23", "remaining_time": "2:03:14"}
+{"current_steps": 208, "total_steps": 390, "loss": 0.27, "lr": 2.1162896578209517e-05, "epoch": 2.6624, "percentage": 53.33, "elapsed_time": "2:20:06", "remaining_time": "2:02:35"}
+{"current_steps": 209, "total_steps": 390, "loss": 0.2778, "lr": 2.0984147109837564e-05, "epoch": 2.6752000000000002, "percentage": 53.59, "elapsed_time": "2:20:51", "remaining_time": "2:01:58"}
+{"current_steps": 210, "total_steps": 390, "loss": 0.2938, "lr": 2.0805318802188307e-05, "epoch": 2.6879999999999997, "percentage": 53.85, "elapsed_time": "2:21:24", "remaining_time": "2:01:12"}
+{"current_steps": 211, "total_steps": 390, "loss": 0.2721, "lr": 2.0626425981061608e-05, "epoch": 2.7008, "percentage": 54.1, "elapsed_time": "2:22:04", "remaining_time": "2:00:32"}
+{"current_steps": 212, "total_steps": 390, "loss": 0.276, "lr": 2.0447482977425465e-05, "epoch": 2.7136, "percentage": 54.36, "elapsed_time": "2:22:45", "remaining_time": "1:59:51"}
+{"current_steps": 213, "total_steps": 390, "loss": 0.2421, "lr": 2.0268504126267952e-05, "epoch": 2.7264, "percentage": 54.62, "elapsed_time": "2:23:22", "remaining_time": "1:59:08"}
+{"current_steps": 214, "total_steps": 390, "loss": 0.2822, "lr": 2.008950376544887e-05, "epoch": 2.7392, "percentage": 54.87, "elapsed_time": "2:23:58", "remaining_time": "1:58:24"}
+{"current_steps": 215, "total_steps": 390, "loss": 0.2569, "lr": 1.9910496234551132e-05, "epoch": 2.752, "percentage": 55.13, "elapsed_time": "2:24:32", "remaining_time": "1:57:39"}
+{"current_steps": 216, "total_steps": 390, "loss": 0.2765, "lr": 1.9731495873732055e-05, "epoch": 2.7648, "percentage": 55.38, "elapsed_time": "2:25:12", "remaining_time": "1:56:58"}
+{"current_steps": 217, "total_steps": 390, "loss": 0.2883, "lr": 1.9552517022574542e-05, "epoch": 2.7776, "percentage": 55.64, "elapsed_time": "2:25:50", "remaining_time": "1:56:16"}
+{"current_steps": 218, "total_steps": 390, "loss": 0.2828, "lr": 1.93735740189384e-05, "epoch": 2.7904, "percentage": 55.9, "elapsed_time": "2:26:31", "remaining_time": "1:55:36"}
+{"current_steps": 219, "total_steps": 390, "loss": 0.2852, "lr": 1.9194681197811703e-05, "epoch": 2.8032, "percentage": 56.15, "elapsed_time": "2:27:05", "remaining_time": "1:54:51"}
+{"current_steps": 220, "total_steps": 390, "loss": 0.2689, "lr": 1.901585289016244e-05, "epoch": 2.816, "percentage": 56.41, "elapsed_time": "2:27:49", "remaining_time": "1:54:13"}
+{"current_steps": 221, "total_steps": 390, "loss": 0.2986, "lr": 1.8837103421790486e-05, "epoch": 2.8288, "percentage": 56.67, "elapsed_time": "2:28:37", "remaining_time": "1:53:39"}
+{"current_steps": 222, "total_steps": 390, "loss": 0.314, "lr": 1.8658447112179952e-05, "epoch": 2.8416, "percentage": 56.92, "elapsed_time": "2:29:17", "remaining_time": "1:52:58"}
+{"current_steps": 223, "total_steps": 390, "loss": 0.2704, "lr": 1.8479898273352084e-05, "epoch": 2.8544, "percentage": 57.18, "elapsed_time": "2:29:56", "remaining_time": "1:52:17"}
+{"current_steps": 224, "total_steps": 390, "loss": 0.3129, "lr": 1.83014712087187e-05, "epoch": 2.8672, "percentage": 57.44, "elapsed_time": "2:30:42", "remaining_time": "1:51:40"}
+{"current_steps": 225, "total_steps": 390, "loss": 0.2541, "lr": 1.8123180211936417e-05, "epoch": 2.88, "percentage": 57.69, "elapsed_time": "2:31:19", "remaining_time": "1:50:58"}
+{"current_steps": 226, "total_steps": 390, "loss": 0.2781, "lr": 1.794503956576152e-05, "epoch": 2.8928000000000003, "percentage": 57.95, "elapsed_time": "2:31:55", "remaining_time": "1:50:15"}
+{"current_steps": 227, "total_steps": 390, "loss": 0.2693, "lr": 1.776706354090585e-05, "epoch": 2.9055999999999997, "percentage": 58.21, "elapsed_time": "2:32:27", "remaining_time": "1:49:28"}
+{"current_steps": 228, "total_steps": 390, "loss": 0.2794, "lr": 1.758926639489354e-05, "epoch": 2.9184, "percentage": 58.46, "elapsed_time": "2:33:02", "remaining_time": "1:48:44"}
+{"current_steps": 229, "total_steps": 390, "loss": 0.2702, "lr": 1.7411662370918893e-05, "epoch": 2.9312, "percentage": 58.72, "elapsed_time": "2:33:42", "remaining_time": "1:48:04"}
+{"current_steps": 230, "total_steps": 390, "loss": 0.2684, "lr": 1.7234265696705344e-05, "epoch": 2.944, "percentage": 58.97, "elapsed_time": "2:34:22", "remaining_time": "1:47:23"}
+{"current_steps": 231, "total_steps": 390, "loss": 0.2886, "lr": 1.7057090583365678e-05, "epoch": 2.9568, "percentage": 59.23, "elapsed_time": "2:35:01", "remaining_time": "1:46:42"}
+{"current_steps": 232, "total_steps": 390, "loss": 0.2757, "lr": 1.6880151224263646e-05, "epoch": 2.9696, "percentage": 59.49, "elapsed_time": "2:35:46", "remaining_time": "1:46:05"}
+{"current_steps": 233, "total_steps": 390, "loss": 0.2928, "lr": 1.6703461793876876e-05, "epoch": 2.9824, "percentage": 59.74, "elapsed_time": "2:36:23", "remaining_time": "1:45:22"}
+{"current_steps": 234, "total_steps": 390, "loss": 0.2677, "lr": 1.6527036446661396e-05, "epoch": 2.9952, "percentage": 60.0, "elapsed_time": "2:37:07", "remaining_time": "1:44:45"}
+{"current_steps": 235, "total_steps": 390, "loss": 0.4456, "lr": 1.635088931591775e-05, "epoch": 3.008, "percentage": 60.26, "elapsed_time": "2:37:59", "remaining_time": "1:44:12"}
+{"current_steps": 236, "total_steps": 390, "loss": 0.1978, "lr": 1.6175034512658753e-05, "epoch": 3.0208, "percentage": 60.51, "elapsed_time": "2:38:37", "remaining_time": "1:43:30"}
+{"current_steps": 237, "total_steps": 390, "loss": 0.1948, "lr": 1.5999486124479115e-05, "epoch": 3.0336, "percentage": 60.77, "elapsed_time": "2:39:08", "remaining_time": "1:42:44"}
+{"current_steps": 238, "total_steps": 390, "loss": 0.2023, "lr": 1.5824258214426833e-05, "epoch": 3.0464, "percentage": 61.03, "elapsed_time": "2:39:48", "remaining_time": "1:42:03"}
+{"current_steps": 239, "total_steps": 390, "loss": 0.1948, "lr": 1.5649364819876655e-05, "epoch": 3.0592, "percentage": 61.28, "elapsed_time": "2:40:29", "remaining_time": "1:41:24"}
+{"current_steps": 240, "total_steps": 390, "loss": 0.2137, "lr": 1.547481995140556e-05, "epoch": 3.072, "percentage": 61.54, "elapsed_time": "2:41:12", "remaining_time": "1:40:45"}
+{"current_steps": 241, "total_steps": 390, "loss": 0.1894, "lr": 1.5300637591670357e-05, "epoch": 3.0848, "percentage": 61.79, "elapsed_time": "2:41:54", "remaining_time": "1:40:06"}
+{"current_steps": 242, "total_steps": 390, "loss": 0.1736, "lr": 1.5126831694287564e-05, "epoch": 3.0976, "percentage": 62.05, "elapsed_time": "2:42:31", "remaining_time": "1:39:23"}
+{"current_steps": 243, "total_steps": 390, "loss": 0.2129, "lr": 1.4953416182715566e-05, "epoch": 3.1104, "percentage": 62.31, "elapsed_time": "2:43:14", "remaining_time": "1:38:45"}
+{"current_steps": 244, "total_steps": 390, "loss": 0.2238, "lr": 1.478040494913926e-05, "epoch": 3.1232, "percentage": 62.56, "elapsed_time": "2:43:53", "remaining_time": "1:38:04"}
+{"current_steps": 245, "total_steps": 390, "loss": 0.2079, "lr": 1.460781185335713e-05, "epoch": 3.136, "percentage": 62.82, "elapsed_time": "2:44:31", "remaining_time": "1:37:22"}
+{"current_steps": 246, "total_steps": 390, "loss": 0.2064, "lr": 1.443565072167095e-05, "epoch": 3.1488, "percentage": 63.08, "elapsed_time": "2:45:10", "remaining_time": "1:36:41"}
+{"current_steps": 247, "total_steps": 390, "loss": 0.1754, "lr": 1.4263935345778202e-05, "epoch": 3.1616, "percentage": 63.33, "elapsed_time": "2:45:48", "remaining_time": "1:35:59"}
+{"current_steps": 248, "total_steps": 390, "loss": 0.1975, "lr": 1.409267948166718e-05, "epoch": 3.1744, "percentage": 63.59, "elapsed_time": "2:46:26", "remaining_time": "1:35:18"}
+{"current_steps": 249, "total_steps": 390, "loss": 0.2116, "lr": 1.3921896848515064e-05, "epoch": 3.1872, "percentage": 63.85, "elapsed_time": "2:47:09", "remaining_time": "1:34:39"}
+{"current_steps": 250, "total_steps": 390, "loss": 0.1939, "lr": 1.3751601127588849e-05, "epoch": 3.2, "percentage": 64.1, "elapsed_time": "2:47:40", "remaining_time": "1:33:54"}
+{"current_steps": 251, "total_steps": 390, "loss": 0.2053, "lr": 1.3581805961149371e-05, "epoch": 3.2128, "percentage": 64.36, "elapsed_time": "2:48:19", "remaining_time": "1:33:13"}
+{"current_steps": 252, "total_steps": 390, "loss": 0.189, "lr": 1.341252495135841e-05, "epoch": 3.2256, "percentage": 64.62, "elapsed_time": "2:49:00", "remaining_time": "1:32:33"}
+{"current_steps": 253, "total_steps": 390, "loss": 0.2025, "lr": 1.324377165918906e-05, "epoch": 3.2384, "percentage": 64.87, "elapsed_time": "2:49:38", "remaining_time": "1:31:51"}
+{"current_steps": 254, "total_steps": 390, "loss": 0.2034, "lr": 1.3075559603339354e-05, "epoch": 3.2512, "percentage": 65.13, "elapsed_time": "2:50:17", "remaining_time": "1:31:10"}
+{"current_steps": 255, "total_steps": 390, "loss": 0.1916, "lr": 1.2907902259149287e-05, "epoch": 3.2640000000000002, "percentage": 65.38, "elapsed_time": "2:51:02", "remaining_time": "1:30:33"}
+{"current_steps": 256, "total_steps": 390, "loss": 0.2182, "lr": 1.274081305752135e-05, "epoch": 3.2768, "percentage": 65.64, "elapsed_time": "2:51:40", "remaining_time": "1:29:51"}
+{"current_steps": 257, "total_steps": 390, "loss": 0.1814, "lr": 1.2574305383844528e-05, "epoch": 3.2896, "percentage": 65.9, "elapsed_time": "2:52:22", "remaining_time": "1:29:12"}
+{"current_steps": 258, "total_steps": 390, "loss": 0.2023, "lr": 1.2408392576922075e-05, "epoch": 3.3024, "percentage": 66.15, "elapsed_time": "2:53:05", "remaining_time": "1:28:33"}
+{"current_steps": 259, "total_steps": 390, "loss": 0.187, "lr": 1.2243087927902905e-05, "epoch": 3.3152, "percentage": 66.41, "elapsed_time": "2:53:40", "remaining_time": "1:27:50"}
+{"current_steps": 260, "total_steps": 390, "loss": 0.1638, "lr": 1.2078404679216864e-05, "epoch": 3.328, "percentage": 66.67, "elapsed_time": "2:54:17", "remaining_time": "1:27:08"}
+{"current_steps": 261, "total_steps": 390, "loss": 0.1862, "lr": 1.1914356023513904e-05, "epoch": 3.3407999999999998, "percentage": 66.92, "elapsed_time": "2:54:59", "remaining_time": "1:26:29"}
+{"current_steps": 262, "total_steps": 390, "loss": 0.206, "lr": 1.1750955102607193e-05, "epoch": 3.3536, "percentage": 67.18, "elapsed_time": "2:55:41", "remaining_time": "1:25:49"}
+{"current_steps": 263, "total_steps": 390, "loss": 0.1778, "lr": 1.1588215006420374e-05, "epoch": 3.3664, "percentage": 67.44, "elapsed_time": "2:56:19", "remaining_time": "1:25:08"}
+{"current_steps": 264, "total_steps": 390, "loss": 0.2185, "lr": 1.1426148771938915e-05, "epoch": 3.3792, "percentage": 67.69, "elapsed_time": "2:57:00", "remaining_time": "1:24:28"}
+{"current_steps": 265, "total_steps": 390, "loss": 0.1874, "lr": 1.1264769382165748e-05, "epoch": 3.392, "percentage": 67.95, "elapsed_time": "2:57:38", "remaining_time": "1:23:47"}
+{"current_steps": 266, "total_steps": 390, "loss": 0.2084, "lr": 1.110408976508118e-05, "epoch": 3.4048, "percentage": 68.21, "elapsed_time": "2:58:20", "remaining_time": "1:23:08"}
+{"current_steps": 267, "total_steps": 390, "loss": 0.1635, "lr": 1.094412279260726e-05, "epoch": 3.4176, "percentage": 68.46, "elapsed_time": "2:58:57", "remaining_time": "1:22:26"}
+{"current_steps": 268, "total_steps": 390, "loss": 0.1842, "lr": 1.0784881279576635e-05, "epoch": 3.4304, "percentage": 68.72, "elapsed_time": "2:59:31", "remaining_time": "1:21:43"}
+{"current_steps": 269, "total_steps": 390, "loss": 0.2027, "lr": 1.0626377982705929e-05, "epoch": 3.4432, "percentage": 68.97, "elapsed_time": "3:00:18", "remaining_time": "1:21:06"}
+{"current_steps": 270, "total_steps": 390, "loss": 0.1928, "lr": 1.0468625599573842e-05, "epoch": 3.456, "percentage": 69.23, "elapsed_time": "3:01:00", "remaining_time": "1:20:26"}
+{"current_steps": 271, "total_steps": 390, "loss": 0.1914, "lr": 1.0311636767603952e-05, "epoch": 3.4688, "percentage": 69.49, "elapsed_time": "3:01:41", "remaining_time": "1:19:46"}
+{"current_steps": 272, "total_steps": 390, "loss": 0.1857, "lr": 1.0155424063052306e-05, "epoch": 3.4816, "percentage": 69.74, "elapsed_time": "3:02:26", "remaining_time": "1:19:08"}
+{"current_steps": 273, "total_steps": 390, "loss": 0.1992, "lr": 1.0000000000000006e-05, "epoch": 3.4944, "percentage": 70.0, "elapsed_time": "3:03:07", "remaining_time": "1:18:29"}
+{"current_steps": 274, "total_steps": 390, "loss": 0.1926, "lr": 9.84537702935065e-06, "epoch": 3.5072, "percentage": 70.26, "elapsed_time": "3:03:47", "remaining_time": "1:17:48"}
+{"current_steps": 275, "total_steps": 390, "loss": 0.1809, "lr": 9.691567537832964e-06, "epoch": 3.52, "percentage": 70.51, "elapsed_time": "3:04:29", "remaining_time": "1:17:09"}
+{"current_steps": 276, "total_steps": 390, "loss": 0.2143, "lr": 9.538583847008452e-06, "epoch": 3.5328, "percentage": 70.77, "elapsed_time": "3:05:10", "remaining_time": "1:16:29"}
+{"current_steps": 277, "total_steps": 390, "loss": 0.1767, "lr": 9.386438212284372e-06, "epoch": 3.5456, "percentage": 71.03, "elapsed_time": "3:05:55", "remaining_time": "1:15:50"}
+{"current_steps": 278, "total_steps": 390, "loss": 0.1928, "lr": 9.235142821931928e-06, "epoch": 3.5584, "percentage": 71.28, "elapsed_time": "3:06:32", "remaining_time": "1:15:09"}
+{"current_steps": 279, "total_steps": 390, "loss": 0.1775, "lr": 9.084709796109907e-06, "epoch": 3.5712, "percentage": 71.54, "elapsed_time": "3:07:07", "remaining_time": "1:14:26"}
+{"current_steps": 280, "total_steps": 390, "loss": 0.2056, "lr": 8.93515118589373e-06, "epoch": 3.584, "percentage": 71.79, "elapsed_time": "3:07:48", "remaining_time": "1:13:46"}
+{"current_steps": 281, "total_steps": 390, "loss": 0.1774, "lr": 8.786478972310023e-06, "epoch": 3.5968, "percentage": 72.05, "elapsed_time": "3:08:24", "remaining_time": "1:13:05"}
+{"current_steps": 282, "total_steps": 390, "loss": 0.1832, "lr": 8.638705065376887e-06, "epoch": 3.6096, "percentage": 72.31, "elapsed_time": "3:09:07", "remaining_time": "1:12:25"}
+{"current_steps": 283, "total_steps": 390, "loss": 0.2018, "lr": 8.491841303149728e-06, "epoch": 3.6224, "percentage": 72.56, "elapsed_time": "3:09:51", "remaining_time": "1:11:46"}
+{"current_steps": 284, "total_steps": 390, "loss": 0.1753, "lr": 8.345899450772975e-06, "epoch": 3.6352, "percentage": 72.82, "elapsed_time": "3:10:32", "remaining_time": "1:11:06"}
+{"current_steps": 285, "total_steps": 390, "loss": 0.1857, "lr": 8.200891199537549e-06, "epoch": 3.648, "percentage": 73.08, "elapsed_time": "3:11:12", "remaining_time": "1:10:26"}
+{"current_steps": 286, "total_steps": 390, "loss": 0.1887, "lr": 8.056828165944282e-06, "epoch": 3.6608, "percentage": 73.33, "elapsed_time": "3:11:53", "remaining_time": "1:09:46"}
+{"current_steps": 287, "total_steps": 390, "loss": 0.1732, "lr": 7.913721890773354e-06, "epoch": 3.6736, "percentage": 73.59, "elapsed_time": "3:12:33", "remaining_time": "1:09:06"}
+{"current_steps": 288, "total_steps": 390, "loss": 0.1835, "lr": 7.771583838159756e-06, "epoch": 3.6864, "percentage": 73.85, "elapsed_time": "3:13:12", "remaining_time": "1:08:25"}
+{"current_steps": 289, "total_steps": 390, "loss": 0.1975, "lr": 7.630425394674903e-06, "epoch": 3.6992000000000003, "percentage": 74.1, "elapsed_time": "3:13:48", "remaining_time": "1:07:44"}
+{"current_steps": 290, "total_steps": 390, "loss": 0.1742, "lr": 7.49025786841445e-06, "epoch": 3.7119999999999997, "percentage": 74.36, "elapsed_time": "3:14:27", "remaining_time": "1:07:03"}
+{"current_steps": 291, "total_steps": 390, "loss": 0.1773, "lr": 7.3510924880924575e-06, "epoch": 3.7248, "percentage": 74.62, "elapsed_time": "3:15:00", "remaining_time": "1:06:20"}
+{"current_steps": 292, "total_steps": 390, "loss": 0.1854, "lr": 7.212940402141808e-06, "epoch": 3.7376, "percentage": 74.87, "elapsed_time": "3:15:40", "remaining_time": "1:05:40"}
+{"current_steps": 293, "total_steps": 390, "loss": 0.1913, "lr": 7.075812677821145e-06, "epoch": 3.7504, "percentage": 75.13, "elapsed_time": "3:16:20", "remaining_time": "1:05:00"}
+{"current_steps": 294, "total_steps": 390, "loss": 0.1897, "lr": 6.939720300328303e-06, "epoch": 3.7632, "percentage": 75.38, "elapsed_time": "3:16:59", "remaining_time": "1:04:19"}
+{"current_steps": 295, "total_steps": 390, "loss": 0.1834, "lr": 6.8046741719202385e-06, "epoch": 3.776, "percentage": 75.64, "elapsed_time": "3:17:38", "remaining_time": "1:03:38"}
+{"current_steps": 296, "total_steps": 390, "loss": 0.1979, "lr": 6.67068511103971e-06, "epoch": 3.7888, "percentage": 75.9, "elapsed_time": "3:18:19", "remaining_time": "1:02:58"}
+{"current_steps": 297, "total_steps": 390, "loss": 0.1795, "lr": 6.537763851448593e-06, "epoch": 3.8016, "percentage": 76.15, "elapsed_time": "3:18:55", "remaining_time": "1:02:17"}
+{"current_steps": 298, "total_steps": 390, "loss": 0.2122, "lr": 6.4059210413680175e-06, "epoch": 3.8144, "percentage": 76.41, "elapsed_time": "3:19:34", "remaining_time": "1:01:36"}
+{"current_steps": 299, "total_steps": 390, "loss": 0.1689, "lr": 6.275167242625331e-06, "epoch": 3.8272, "percentage": 76.67, "elapsed_time": "3:20:09", "remaining_time": "1:00:55"}
+{"current_steps": 300, "total_steps": 390, "loss": 0.1721, "lr": 6.145512929808013e-06, "epoch": 3.84, "percentage": 76.92, "elapsed_time": "3:20:45", "remaining_time": "1:00:13"}
+{"current_steps": 301, "total_steps": 390, "loss": 0.2155, "lr": 6.016968489424572e-06, "epoch": 3.8528000000000002, "percentage": 77.18, "elapsed_time": "3:21:24", "remaining_time": "0:59:33"}
+{"current_steps": 302, "total_steps": 390, "loss": 0.1855, "lr": 5.889544219072465e-06, "epoch": 3.8656, "percentage": 77.44, "elapsed_time": "3:22:04", "remaining_time": "0:58:53"}
+{"current_steps": 303, "total_steps": 390, "loss": 0.1899, "lr": 5.7632503266131925e-06, "epoch": 3.8784, "percentage": 77.69, "elapsed_time": "3:22:42", "remaining_time": "0:58:12"}
+{"current_steps": 304, "total_steps": 390, "loss": 0.1703, "lr": 5.638096929354522e-06, "epoch": 3.8912, "percentage": 77.95, "elapsed_time": "3:23:27", "remaining_time": "0:57:33"}
+{"current_steps": 305, "total_steps": 390, "loss": 0.201, "lr": 5.514094053240035e-06, "epoch": 3.904, "percentage": 78.21, "elapsed_time": "3:24:05", "remaining_time": "0:56:52"}
+{"current_steps": 306, "total_steps": 390, "loss": 0.1994, "lr": 5.39125163204594e-06, "epoch": 3.9168, "percentage": 78.46, "elapsed_time": "3:24:47", "remaining_time": "0:56:13"}
+{"current_steps": 307, "total_steps": 390, "loss": 0.183, "lr": 5.269579506585259e-06, "epoch": 3.9295999999999998, "percentage": 78.72, "elapsed_time": "3:25:31", "remaining_time": "0:55:33"}
+{"current_steps": 308, "total_steps": 390, "loss": 0.1912, "lr": 5.149087423919541e-06, "epoch": 3.9424, "percentage": 78.97, "elapsed_time": "3:26:08", "remaining_time": "0:54:52"}
+{"current_steps": 309, "total_steps": 390, "loss": 0.1889, "lr": 5.029785036577976e-06, "epoch": 3.9552, "percentage": 79.23, "elapsed_time": "3:26:45", "remaining_time": "0:54:11"}
+{"current_steps": 310, "total_steps": 390, "loss": 0.1804, "lr": 4.911681901784198e-06, "epoch": 3.968, "percentage": 79.49, "elapsed_time": "3:27:21", "remaining_time": "0:53:30"}
+{"current_steps": 311, "total_steps": 390, "loss": 0.1988, "lr": 4.794787480690597e-06, "epoch": 3.9808, "percentage": 79.74, "elapsed_time": "3:28:06", "remaining_time": "0:52:51"}
+{"current_steps": 312, "total_steps": 390, "loss": 0.1933, "lr": 4.679111137620442e-06, "epoch": 3.9936, "percentage": 80.0, "elapsed_time": "3:28:44", "remaining_time": "0:52:11"}
+{"current_steps": 313, "total_steps": 390, "loss": 0.3106, "lr": 4.5646621393177e-06, "epoch": 4.0064, "percentage": 80.26, "elapsed_time": "3:29:35", "remaining_time": "0:51:33"}
+{"current_steps": 314, "total_steps": 390, "loss": 0.1257, "lr": 4.451449654204685e-06, "epoch": 4.0192, "percentage": 80.51, "elapsed_time": "3:30:08", "remaining_time": "0:50:51"}
+{"current_steps": 315, "total_steps": 390, "loss": 0.1444, "lr": 4.339482751647557e-06, "epoch": 4.032, "percentage": 80.77, "elapsed_time": "3:30:48", "remaining_time": "0:50:11"}
+{"current_steps": 316, "total_steps": 390, "loss": 0.1559, "lr": 4.228770401229824e-06, "epoch": 4.0448, "percentage": 81.03, "elapsed_time": "3:31:28", "remaining_time": "0:49:31"}
+{"current_steps": 317, "total_steps": 390, "loss": 0.1183, "lr": 4.119321472033779e-06, "epoch": 4.0576, "percentage": 81.28, "elapsed_time": "3:31:58", "remaining_time": "0:48:48"}
+{"current_steps": 318, "total_steps": 390, "loss": 0.1478, "lr": 4.011144731929981e-06, "epoch": 4.0704, "percentage": 81.54, "elapsed_time": "3:32:37", "remaining_time": "0:48:08"}
+{"current_steps": 319, "total_steps": 390, "loss": 0.1135, "lr": 3.904248846874894e-06, "epoch": 4.0832, "percentage": 81.79, "elapsed_time": "3:33:08", "remaining_time": "0:47:26"}
+{"current_steps": 320, "total_steps": 390, "loss": 0.1565, "lr": 3.7986423802166705e-06, "epoch": 4.096, "percentage": 82.05, "elapsed_time": "3:33:51", "remaining_time": "0:46:46"}
+{"current_steps": 321, "total_steps": 390, "loss": 0.1394, "lr": 3.694333792009115e-06, "epoch": 4.1088, "percentage": 82.31, "elapsed_time": "3:34:31", "remaining_time": "0:46:06"}
+{"current_steps": 322, "total_steps": 390, "loss": 0.1577, "lr": 3.5913314383339937e-06, "epoch": 4.1216, "percentage": 82.56, "elapsed_time": "3:35:17", "remaining_time": "0:45:27"}
+{"current_steps": 323, "total_steps": 390, "loss": 0.1368, "lr": 3.4896435706316e-06, "epoch": 4.1344, "percentage": 82.82, "elapsed_time": "3:35:58", "remaining_time": "0:44:48"}
+{"current_steps": 324, "total_steps": 390, "loss": 0.1494, "lr": 3.3892783350397675e-06, "epoch": 4.1472, "percentage": 83.08, "elapsed_time": "3:36:41", "remaining_time": "0:44:08"}
+{"current_steps": 325, "total_steps": 390, "loss": 0.1328, "lr": 3.290243771741275e-06, "epoch": 4.16, "percentage": 83.33, "elapsed_time": "3:37:21", "remaining_time": "0:43:28"}
+{"current_steps": 326, "total_steps": 390, "loss": 0.1379, "lr": 3.1925478143197418e-06, "epoch": 4.1728, "percentage": 83.59, "elapsed_time": "3:38:03", "remaining_time": "0:42:48"}
+{"current_steps": 327, "total_steps": 390, "loss": 0.1294, "lr": 3.0961982891241083e-06, "epoch": 4.1856, "percentage": 83.85, "elapsed_time": "3:38:42", "remaining_time": "0:42:08"}
+{"current_steps": 328, "total_steps": 390, "loss": 0.1412, "lr": 3.001202914641628e-06, "epoch": 4.1984, "percentage": 84.1, "elapsed_time": "3:39:17", "remaining_time": "0:41:27"}
+{"current_steps": 329, "total_steps": 390, "loss": 0.1187, "lr": 2.907569300879596e-06, "epoch": 4.2112, "percentage": 84.36, "elapsed_time": "3:40:00", "remaining_time": "0:40:47"}
+{"current_steps": 330, "total_steps": 390, "loss": 0.1256, "lr": 2.815304948755664e-06, "epoch": 4.224, "percentage": 84.62, "elapsed_time": "3:40:43", "remaining_time": "0:40:07"}
+{"current_steps": 331, "total_steps": 390, "loss": 0.1462, "lr": 2.7244172494969978e-06, "epoch": 4.2368, "percentage": 84.87, "elapsed_time": "3:41:18", "remaining_time": "0:39:26"}
+{"current_steps": 332, "total_steps": 390, "loss": 0.1346, "lr": 2.6349134840481294e-06, "epoch": 4.2496, "percentage": 85.13, "elapsed_time": "3:41:58", "remaining_time": "0:38:46"}
+{"current_steps": 333, "total_steps": 390, "loss": 0.1181, "lr": 2.546800822487714e-06, "epoch": 4.2624, "percentage": 85.38, "elapsed_time": "3:42:42", "remaining_time": "0:38:07"}
+{"current_steps": 334, "total_steps": 390, "loss": 0.1264, "lr": 2.4600863234541338e-06, "epoch": 4.2752, "percentage": 85.64, "elapsed_time": "3:43:18", "remaining_time": "0:37:26"}
+{"current_steps": 335, "total_steps": 390, "loss": 0.1352, "lr": 2.374776933580025e-06, "epoch": 4.288, "percentage": 85.9, "elapsed_time": "3:43:53", "remaining_time": "0:36:45"}
+{"current_steps": 336, "total_steps": 390, "loss": 0.111, "lr": 2.2908794869358044e-06, "epoch": 4.3008, "percentage": 86.15, "elapsed_time": "3:44:27", "remaining_time": "0:36:04"}
+{"current_steps": 337, "total_steps": 390, "loss": 0.1304, "lr": 2.2084007044821764e-06, "epoch": 4.3136, "percentage": 86.41, "elapsed_time": "3:45:03", "remaining_time": "0:35:23"}
+{"current_steps": 338, "total_steps": 390, "loss": 0.1442, "lr": 2.127347193531757e-06, "epoch": 4.3264, "percentage": 86.67, "elapsed_time": "3:45:44", "remaining_time": "0:34:43"}
+{"current_steps": 339, "total_steps": 390, "loss": 0.1325, "lr": 2.0477254472197237e-06, "epoch": 4.3392, "percentage": 86.92, "elapsed_time": "3:46:23", "remaining_time": "0:34:03"}
+{"current_steps": 340, "total_steps": 390, "loss": 0.1259, "lr": 1.96954184398368e-06, "epoch": 4.352, "percentage": 87.18, "elapsed_time": "3:47:02", "remaining_time": "0:33:23"}
+{"current_steps": 341, "total_steps": 390, "loss": 0.1401, "lr": 1.8928026470526917e-06, "epoch": 4.3648, "percentage": 87.44, "elapsed_time": "3:47:45", "remaining_time": "0:32:43"}
+{"current_steps": 342, "total_steps": 390, "loss": 0.1285, "lr": 1.817514003945524e-06, "epoch": 4.3776, "percentage": 87.69, "elapsed_time": "3:48:19", "remaining_time": "0:32:02"}
+{"current_steps": 343, "total_steps": 390, "loss": 0.1465, "lr": 1.743681945978184e-06, "epoch": 4.3904, "percentage": 87.95, "elapsed_time": "3:48:58", "remaining_time": "0:31:22"}
+{"current_steps": 344, "total_steps": 390, "loss": 0.1279, "lr": 1.6713123877807413e-06, "epoch": 4.4032, "percentage": 88.21, "elapsed_time": "3:49:42", "remaining_time": "0:30:42"}
+{"current_steps": 345, "total_steps": 390, "loss": 0.1243, "lr": 1.6004111268235156e-06, "epoch": 4.416, "percentage": 88.46, "elapsed_time": "3:50:16", "remaining_time": "0:30:02"}
+{"current_steps": 346, "total_steps": 390, "loss": 0.1417, "lr": 1.5309838429526714e-06, "epoch": 4.4288, "percentage": 88.72, "elapsed_time": "3:50:57", "remaining_time": "0:29:22"}
+{"current_steps": 347, "total_steps": 390, "loss": 0.1368, "lr": 1.4630360979351644e-06, "epoch": 4.4416, "percentage": 88.97, "elapsed_time": "3:51:42", "remaining_time": "0:28:42"}
+{"current_steps": 348, "total_steps": 390, "loss": 0.1451, "lr": 1.396573335013236e-06, "epoch": 4.4544, "percentage": 89.23, "elapsed_time": "3:52:21", "remaining_time": "0:28:02"}
+{"current_steps": 349, "total_steps": 390, "loss": 0.122, "lr": 1.3316008784683265e-06, "epoch": 4.4672, "percentage": 89.49, "elapsed_time": "3:53:00", "remaining_time": "0:27:22"}
+{"current_steps": 350, "total_steps": 390, "loss": 0.1203, "lr": 1.2681239331945695e-06, "epoch": 4.48, "percentage": 89.74, "elapsed_time": "3:53:37", "remaining_time": "0:26:42"}
+{"current_steps": 351, "total_steps": 390, "loss": 0.1366, "lr": 1.2061475842818337e-06, "epoch": 4.4928, "percentage": 90.0, "elapsed_time": "3:54:13", "remaining_time": "0:26:01"}
+{"current_steps": 352, "total_steps": 390, "loss": 0.1446, "lr": 1.1456767966083393e-06, "epoch": 4.5056, "percentage": 90.26, "elapsed_time": "3:54:48", "remaining_time": "0:25:20"}
+{"current_steps": 353, "total_steps": 390, "loss": 0.1444, "lr": 1.086716414442952e-06, "epoch": 4.5184, "percentage": 90.51, "elapsed_time": "3:55:29", "remaining_time": "0:24:40"}
+{"current_steps": 354, "total_steps": 390, "loss": 0.1483, "lr": 1.0292711610570904e-06, "epoch": 4.5312, "percentage": 90.77, "elapsed_time": "3:56:11", "remaining_time": "0:24:01"}
+{"current_steps": 355, "total_steps": 390, "loss": 0.1464, "lr": 9.733456383463658e-07, "epoch": 4.5440000000000005, "percentage": 91.03, "elapsed_time": "3:56:49", "remaining_time": "0:23:20"}
+{"current_steps": 356, "total_steps": 390, "loss": 0.1296, "lr": 9.189443264619102e-07, "epoch": 4.5568, "percentage": 91.28, "elapsed_time": "3:57:22", "remaining_time": "0:22:40"}
+{"current_steps": 357, "total_steps": 390, "loss": 0.1367, "lr": 8.660715834514977e-07, "epoch": 4.5696, "percentage": 91.54, "elapsed_time": "3:58:06", "remaining_time": "0:22:00"}
+{"current_steps": 358, "total_steps": 390, "loss": 0.1376, "lr": 8.147316449103959e-07, "epoch": 4.5824, "percentage": 91.79, "elapsed_time": "3:58:46", "remaining_time": "0:21:20"}
+{"current_steps": 359, "total_steps": 390, "loss": 0.1234, "lr": 7.649286236420806e-07, "epoch": 4.5952, "percentage": 92.05, "elapsed_time": "3:59:23", "remaining_time": "0:20:40"}
+{"current_steps": 360, "total_steps": 390, "loss": 0.1272, "lr": 7.166665093287539e-07, "epoch": 4.608, "percentage": 92.31, "elapsed_time": "4:00:04", "remaining_time": "0:20:00"}
+{"current_steps": 361, "total_steps": 390, "loss": 0.1409, "lr": 6.69949168211721e-07, "epoch": 4.6208, "percentage": 92.56, "elapsed_time": "4:00:43", "remaining_time": "0:19:20"}
+{"current_steps": 362, "total_steps": 390, "loss": 0.1184, "lr": 6.247803427816945e-07, "epoch": 4.6336, "percentage": 92.82, "elapsed_time": "4:01:24", "remaining_time": "0:18:40"}
+{"current_steps": 363, "total_steps": 390, "loss": 0.137, "lr": 5.811636514789598e-07, "epoch": 4.6464, "percentage": 93.08, "elapsed_time": "4:02:01", "remaining_time": "0:18:00"}
+{"current_steps": 364, "total_steps": 390, "loss": 0.1343, "lr": 5.391025884035239e-07, "epoch": 4.6592, "percentage": 93.33, "elapsed_time": "4:02:41", "remaining_time": "0:17:20"}
+{"current_steps": 365, "total_steps": 390, "loss": 0.1384, "lr": 4.986005230351954e-07, "epoch": 4.672, "percentage": 93.59, "elapsed_time": "4:03:22", "remaining_time": "0:16:40"}
+{"current_steps": 366, "total_steps": 390, "loss": 0.1183, "lr": 4.5966069996365993e-07, "epoch": 4.6848, "percentage": 93.85, "elapsed_time": "4:03:58", "remaining_time": "0:15:59"}
+{"current_steps": 367, "total_steps": 390, "loss": 0.1222, "lr": 4.22286238628562e-07, "epoch": 4.6975999999999996, "percentage": 94.1, "elapsed_time": "4:04:28", "remaining_time": "0:15:19"}
+{"current_steps": 368, "total_steps": 390, "loss": 0.1415, "lr": 3.8648013306960664e-07, "epoch": 4.7104, "percentage": 94.36, "elapsed_time": "4:05:04", "remaining_time": "0:14:39"}
+{"current_steps": 369, "total_steps": 390, "loss": 0.1234, "lr": 3.522452516867048e-07, "epoch": 4.7232, "percentage": 94.62, "elapsed_time": "4:05:45", "remaining_time": "0:13:59"}
+{"current_steps": 370, "total_steps": 390, "loss": 0.1197, "lr": 3.1958433701019697e-07, "epoch": 4.736, "percentage": 94.87, "elapsed_time": "4:06:21", "remaining_time": "0:13:18"}
+{"current_steps": 371, "total_steps": 390, "loss": 0.1261, "lr": 2.8850000548115155e-07, "epoch": 4.7488, "percentage": 95.13, "elapsed_time": "4:06:59", "remaining_time": "0:12:38"}
+{"current_steps": 372, "total_steps": 390, "loss": 0.1454, "lr": 2.5899474724174313e-07, "epoch": 4.7616, "percentage": 95.38, "elapsed_time": "4:07:37", "remaining_time": "0:11:58"}
+{"current_steps": 373, "total_steps": 390, "loss": 0.1444, "lr": 2.3107092593579905e-07, "epoch": 4.7744, "percentage": 95.64, "elapsed_time": "4:08:20", "remaining_time": "0:11:19"}
+{"current_steps": 374, "total_steps": 390, "loss": 0.129, "lr": 2.0473077851942858e-07, "epoch": 4.7872, "percentage": 95.9, "elapsed_time": "4:08:56", "remaining_time": "0:10:39"}
+{"current_steps": 375, "total_steps": 390, "loss": 0.1268, "lr": 1.799764150818306e-07, "epoch": 4.8, "percentage": 96.15, "elapsed_time": "4:09:37", "remaining_time": "0:09:59"}
+{"current_steps": 376, "total_steps": 390, "loss": 0.1374, "lr": 1.5680981867625566e-07, "epoch": 4.8128, "percentage": 96.41, "elapsed_time": "4:10:16", "remaining_time": "0:09:19"}
+{"current_steps": 377, "total_steps": 390, "loss": 0.1344, "lr": 1.3523284516113955e-07, "epoch": 4.8256, "percentage": 96.67, "elapsed_time": "4:11:00", "remaining_time": "0:08:39"}
+{"current_steps": 378, "total_steps": 390, "loss": 0.1445, "lr": 1.1524722305144231e-07, "epoch": 4.8384, "percentage": 96.92, "elapsed_time": "4:11:44", "remaining_time": "0:07:59"}
+{"current_steps": 379, "total_steps": 390, "loss": 0.126, "lr": 9.685455338016347e-08, "epoch": 4.8512, "percentage": 97.18, "elapsed_time": "4:12:17", "remaining_time": "0:07:19"}
+{"current_steps": 380, "total_steps": 390, "loss": 0.1449, "lr": 8.005630957010014e-08, "epoch": 4.864, "percentage": 97.44, "elapsed_time": "4:13:03", "remaining_time": "0:06:39"}
+{"current_steps": 381, "total_steps": 390, "loss": 0.1231, "lr": 6.485383731580142e-08, "epoch": 4.8768, "percentage": 97.69, "elapsed_time": "4:13:39", "remaining_time": "0:05:59"}
+{"current_steps": 382, "total_steps": 390, "loss": 0.1322, "lr": 5.1248354475768034e-08, "epoch": 4.8896, "percentage": 97.95, "elapsed_time": "4:14:22", "remaining_time": "0:05:19"}
+{"current_steps": 383, "total_steps": 390, "loss": 0.1363, "lr": 3.924095097489922e-08, "epoch": 4.9024, "percentage": 98.21, "elapsed_time": "4:15:02", "remaining_time": "0:04:39"}
+{"current_steps": 384, "total_steps": 390, "loss": 0.1184, "lr": 2.8832588717164766e-08, "epoch": 4.9152000000000005, "percentage": 98.46, "elapsed_time": "4:15:46", "remaining_time": "0:03:59"}
+{"current_steps": 385, "total_steps": 390, "loss": 0.1392, "lr": 2.0024101508555604e-08, "epoch": 4.928, "percentage": 98.72, "elapsed_time": "4:16:26", "remaining_time": "0:03:19"}
+{"current_steps": 386, "total_steps": 390, "loss": 0.1204, "lr": 1.281619499029274e-08, "epoch": 4.9408, "percentage": 98.97, "elapsed_time": "4:17:04", "remaining_time": "0:02:39"}
+{"current_steps": 387, "total_steps": 390, "loss": 0.1354, "lr": 7.209446582292501e-09, "epoch": 4.9536, "percentage": 99.23, "elapsed_time": "4:17:44", "remaining_time": "0:01:59"}
+{"current_steps": 388, "total_steps": 390, "loss": 0.1161, "lr": 3.2043054369057523e-09, "epoch": 4.9664, "percentage": 99.49, "elapsed_time": "4:18:15", "remaining_time": "0:01:19"}
+{"current_steps": 389, "total_steps": 390, "loss": 0.1446, "lr": 8.010924029533406e-10, "epoch": 4.9792, "percentage": 99.74, "elapsed_time": "4:18:53", "remaining_time": "0:00:39"}
+{"current_steps": 390, "total_steps": 390, "loss": 0.1216, "lr": 0.0, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "4:19:28", "remaining_time": "0:00:00"}
+{"current_steps": 390, "total_steps": 390, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "4:19:58", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2772 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.992,
+  "eval_steps": 500,
+  "global_step": 390,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0128,
+      "grad_norm": 6.115225457318272,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.7972,
+      "step": 1
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 6.4778649832938,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.8223,
+      "step": 2
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.439432589198134,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.8336,
+      "step": 3
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 5.916085347437607,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.794,
+      "step": 4
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.649690599890324,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.7651,
+      "step": 5
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.3516014711724074,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.6905,
+      "step": 6
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.0210867581306613,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.6952,
+      "step": 7
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 4.1647802159924066,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.7169,
+      "step": 8
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 4.285557439271826,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.6915,
+      "step": 9
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 4.742700936931223,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.6619,
+      "step": 10
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 4.345916968810201,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.6244,
+      "step": 11
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 3.1198253262604476,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.6451,
+      "step": 12
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.854722364751461,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5837,
+      "step": 13
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.6202210991889734,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.6217,
+      "step": 14
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.467827204969234,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.5798,
+      "step": 15
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.2901839247708227,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.5836,
+      "step": 16
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.0524795187246585,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.5692,
+      "step": 17
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.6043224605320856,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.5896,
+      "step": 18
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.083519275986709,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.5607,
+      "step": 19
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.1676451919151445,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.5338,
+      "step": 20
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.1649317866906146,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.5467,
+      "step": 21
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.0070139617846363,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.5649,
+      "step": 22
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.8808640916185065,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.5375,
+      "step": 23
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.6891911050508679,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.4995,
+      "step": 24
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8034299306922472,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.5509,
+      "step": 25
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.6810478528470438,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5251,
+      "step": 26
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.6628937158980546,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.493,
+      "step": 27
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.7071258758781976,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.5124,
+      "step": 28
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.5730112406456523,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.4934,
+      "step": 29
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.7591360874451095,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.4959,
+      "step": 30
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.7433414478275835,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.5276,
+      "step": 31
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.5230218083058822,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.5056,
+      "step": 32
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.5788315940951135,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.4775,
+      "step": 33
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.628996593385306,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.5025,
+      "step": 34
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.6364819152597294,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.4956,
+      "step": 35
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.6952439014406236,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.4795,
+      "step": 36
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.5879249245725728,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.4886,
+      "step": 37
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.7143567047490685,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.4662,
+      "step": 38
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.754270169595988,
+      "learning_rate": 4e-05,
+      "loss": 0.4845,
+      "step": 39
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.6916230753184576,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.4946,
+      "step": 40
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.6655929376045815,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.4751,
+      "step": 41
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.7019272485859203,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.4879,
+      "step": 42
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.8696723571749855,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.4667,
+      "step": 43
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.7429977461233525,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.4761,
+      "step": 44
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.6918166294558654,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.4654,
+      "step": 45
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.7171915391313837,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.4543,
+      "step": 46
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.9295428779797352,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.4758,
+      "step": 47
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.7984379574635976,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.4906,
+      "step": 48
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.9454295106930493,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.4632,
+      "step": 49
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8457972079509006,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.4629,
+      "step": 50
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.8777267910421587,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.4996,
+      "step": 51
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.7765226591426262,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.4734,
+      "step": 52
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.9927734864349484,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.4543,
+      "step": 53
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.9443366075284568,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.4908,
+      "step": 54
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.0531525900028036,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.4663,
+      "step": 55
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.799320060729199,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.4809,
+      "step": 56
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 1.1131030340708192,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.4784,
+      "step": 57
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.7567089549779984,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.4806,
+      "step": 58
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 1.0788676254710285,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.4854,
+      "step": 59
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.7991461438625849,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.4903,
+      "step": 60
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.6199089194762689,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.4707,
+      "step": 61
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.7896919100970462,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.4474,
+      "step": 62
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.45780085481027216,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.4529,
+      "step": 63
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.7221706609456762,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.445,
+      "step": 64
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.5165115410834582,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.4589,
+      "step": 65
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.7202953596591821,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.4857,
+      "step": 66
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.5705334637961889,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.4618,
+      "step": 67
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.5500129664597081,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.4492,
+      "step": 68
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.5744757957783347,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.4429,
+      "step": 69
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.5802945866116739,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.4579,
+      "step": 70
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.7235041394840783,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.4445,
+      "step": 71
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.5041980422363816,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.4653,
+      "step": 72
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.660792488065925,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.4588,
+      "step": 73
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.5486994848030444,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.4429,
+      "step": 74
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7034073232264785,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.4826,
+      "step": 75
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.5059700872962927,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.4544,
+      "step": 76
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.6369868161296368,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.4613,
+      "step": 77
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.48870471079255756,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4449,
+      "step": 78
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 1.2635183664979024,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 0.7069,
+      "step": 79
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.8268755545602047,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.3875,
+      "step": 80
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.6721621975488452,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.3507,
+      "step": 81
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.6347440615314595,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.4196,
+      "step": 82
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.7367054221518949,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.3929,
+      "step": 83
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.6821773954639229,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.4045,
+      "step": 84
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.7612172229842501,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.3792,
+      "step": 85
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.5906116108616847,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.3781,
+      "step": 86
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.6312131712709363,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.3997,
+      "step": 87
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.6329032589193756,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.416,
+      "step": 88
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.5997371849044506,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.3655,
+      "step": 89
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.6121363195901529,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.3895,
+      "step": 90
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.7461859611022873,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.403,
+      "step": 91
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.6457462210535605,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.3995,
+      "step": 92
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.6276460566924693,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.3953,
+      "step": 93
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.6651405841975484,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.4154,
+      "step": 94
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.5545620200098397,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.3607,
+      "step": 95
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.6397790668991523,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.416,
+      "step": 96
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.4720374118182995,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.3556,
+      "step": 97
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.6346584954709299,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.4022,
+      "step": 98
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.46025361233210943,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.3802,
+      "step": 99
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6179922371263346,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.4032,
+      "step": 100
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.5123787588585933,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.3724,
+      "step": 101
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.5016395997791134,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.3848,
+      "step": 102
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.5275969489140777,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.4048,
+      "step": 103
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.4274247976232669,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.3684,
+      "step": 104
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.43690661889705235,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.3689,
+      "step": 105
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.5027782707045861,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.3929,
+      "step": 106
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.4043604339052812,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.3837,
+      "step": 107
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.45253437084829845,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.3822,
+      "step": 108
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.5111440287154948,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.391,
+      "step": 109
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.43656648959033406,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.3987,
+      "step": 110
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.447739635005765,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.3669,
+      "step": 111
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.4891443852288491,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.403,
+      "step": 112
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.5354777825209106,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.4006,
+      "step": 113
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.44665401713182695,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.3685,
+      "step": 114
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.5653760562391243,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.3875,
+      "step": 115
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.4540472373860856,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.3653,
+      "step": 116
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.6113566659285303,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4005,
+      "step": 117
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.4168488729472874,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.3654,
+      "step": 118
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.4620977772662231,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.4001,
+      "step": 119
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.476420893899808,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.3572,
+      "step": 120
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.381303454265223,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.3467,
+      "step": 121
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.4863320007755061,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.4082,
+      "step": 122
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.381600758560905,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.3347,
+      "step": 123
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.394620847076037,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.3879,
+      "step": 124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44406295699549964,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.3642,
+      "step": 125
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.39408607155522185,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.4074,
+      "step": 126
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.49314747982309703,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.3801,
+      "step": 127
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.36062471985657507,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.3757,
+      "step": 128
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.48026342944805833,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.4022,
+      "step": 129
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.4489396246814716,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.3948,
+      "step": 130
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.47748548923970513,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.3889,
+      "step": 131
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.45412392757138254,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.3741,
+      "step": 132
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.49636144268343996,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.3891,
+      "step": 133
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.4390905719231864,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.3739,
+      "step": 134
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.5158833144634472,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.3685,
+      "step": 135
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.4094994605665896,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.3637,
+      "step": 136
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.46870316387196487,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.3995,
+      "step": 137
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.39827114114257894,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.3717,
+      "step": 138
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.452460089054947,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.3746,
+      "step": 139
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.4742389634137148,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.3919,
+      "step": 140
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.3725927079643266,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.3711,
+      "step": 141
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.4331691811604596,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.3822,
+      "step": 142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.4419833649320582,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.3996,
+      "step": 143
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.42256653425091645,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.3613,
+      "step": 144
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.48066707088078137,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.3801,
+      "step": 145
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.43537246948490294,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.3793,
+      "step": 146
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.4202591485432943,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.3649,
+      "step": 147
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.394523012460404,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.3448,
+      "step": 148
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.39136239240309023,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.3763,
+      "step": 149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.41271571999739015,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.3694,
+      "step": 150
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.40875329398478466,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.3931,
+      "step": 151
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.3883380354441773,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.3628,
+      "step": 152
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.4091328267428603,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.4089,
+      "step": 153
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.3570274352018894,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.3163,
+      "step": 154
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.443729606957663,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.413,
+      "step": 155
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.3818874557795846,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3589,
+      "step": 156
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 1.029327384874714,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.616,
+      "step": 157
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.7112912072965769,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.2665,
+      "step": 158
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 1.3570251966441835,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.3094,
+      "step": 159
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.6606729021015588,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.2876,
+      "step": 160
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 0.6071440995571158,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.3013,
+      "step": 161
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.5168033718618819,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.2793,
+      "step": 162
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.5597858928265351,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.2753,
+      "step": 163
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.4676542419821463,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.2938,
+      "step": 164
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.624694783602782,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.3133,
+      "step": 165
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.4508439076138672,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.2637,
+      "step": 166
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.5508607119104917,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.3112,
+      "step": 167
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.5334197735316446,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.3028,
+      "step": 168
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.5087020493527552,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.2822,
+      "step": 169
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.5113084597267213,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.3134,
+      "step": 170
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.5879580442573731,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.2833,
+      "step": 171
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.4073874911657843,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.2672,
+      "step": 172
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.6308940531620888,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.3206,
+      "step": 173
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.48145709078390986,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.2456,
+      "step": 174
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5086807605931843,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.2885,
+      "step": 175
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.5065318303439502,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.3178,
+      "step": 176
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.5214188023044475,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.2685,
+      "step": 177
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.48564632364190025,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.3006,
+      "step": 178
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.5081858291521164,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.2559,
+      "step": 179
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.4873213781242658,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.3054,
+      "step": 180
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.4970841886138032,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.2812,
+      "step": 181
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.48557798780934575,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3065,
+      "step": 182
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.4882395532803412,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.3054,
+      "step": 183
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.42507926185662265,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.2787,
+      "step": 184
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.4522538873448131,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.2745,
+      "step": 185
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.39747296793816345,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.2837,
+      "step": 186
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.43180929878370683,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.2776,
+      "step": 187
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.422046755560608,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.289,
+      "step": 188
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.3939134040727102,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.3012,
+      "step": 189
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.4360103933699082,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.2589,
+      "step": 190
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.43222381961527234,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.2996,
+      "step": 191
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.39947532282855386,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.2895,
+      "step": 192
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.468193590172406,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.3278,
+      "step": 193
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.37028122335191704,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.2592,
+      "step": 194
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.40521601819241465,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2738,
+      "step": 195
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.39880849277987285,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.2889,
+      "step": 196
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.4360800776688632,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.3022,
+      "step": 197
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.332952630190702,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.2595,
+      "step": 198
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.35302482121437867,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.2641,
+      "step": 199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3671929433847647,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.2933,
+      "step": 200
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.3810125942961015,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.304,
+      "step": 201
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.3571039273418628,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.2741,
+      "step": 202
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.3797320221202612,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.2851,
+      "step": 203
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.36043504682960464,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.2889,
+      "step": 204
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.40755688277419305,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.2663,
+      "step": 205
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.34514855584879406,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.2848,
+      "step": 206
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.3838068439029518,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.2879,
+      "step": 207
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.3847498131506985,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.27,
+      "step": 208
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.34940071010776225,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.2778,
+      "step": 209
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.4223538244246661,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.2938,
+      "step": 210
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.35980587807766456,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.2721,
+      "step": 211
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.36119284083682335,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.276,
+      "step": 212
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.3682108196916936,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.2421,
+      "step": 213
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.3538115363318626,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.2822,
+      "step": 214
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.36504269855271493,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.2569,
+      "step": 215
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.36439562322493063,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.2765,
+      "step": 216
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.3366573070490289,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.2883,
+      "step": 217
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.3366306462927351,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.2828,
+      "step": 218
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.36513014328007964,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.2852,
+      "step": 219
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.3117418449051438,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.2689,
+      "step": 220
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.339060400239575,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2986,
+      "step": 221
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.3666591657683088,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.314,
+      "step": 222
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.32066906189526106,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.2704,
+      "step": 223
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.33772346517378604,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.3129,
+      "step": 224
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.35495336862264104,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.2541,
+      "step": 225
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.3010062534177195,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.2781,
+      "step": 226
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.37288601253513487,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.2693,
+      "step": 227
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.3280334445303061,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.2794,
+      "step": 228
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.34011244409998703,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.2702,
+      "step": 229
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.3294569180096905,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.2684,
+      "step": 230
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.3418098557951546,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.2886,
+      "step": 231
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.322402298730085,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.2757,
+      "step": 232
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.34587493019466187,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.2928,
+      "step": 233
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.3177547577906029,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2677,
+      "step": 234
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 1.0570934459380867,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.4456,
+      "step": 235
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.4983219007754451,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.1978,
+      "step": 236
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.40594125810558024,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.1948,
+      "step": 237
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.6097176434241903,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.2023,
+      "step": 238
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.540156981197435,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.1948,
+      "step": 239
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.449124003378837,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.2137,
+      "step": 240
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.44005144456712886,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.1894,
+      "step": 241
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.38990337829232735,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.1736,
+      "step": 242
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.5144918977495327,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.2129,
+      "step": 243
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.3957954029581393,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.2238,
+      "step": 244
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.3991613810037469,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.2079,
+      "step": 245
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.4364261994509852,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.2064,
+      "step": 246
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.36711860268112606,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1754,
+      "step": 247
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.41000829446821285,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.1975,
+      "step": 248
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.37471080060425027,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.2116,
+      "step": 249
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3752427978399487,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.1939,
+      "step": 250
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.38504540888291255,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.2053,
+      "step": 251
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.344247206773842,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.189,
+      "step": 252
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.3655250175724291,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.2025,
+      "step": 253
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.3845876904971133,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.2034,
+      "step": 254
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.3554121162054489,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.1916,
+      "step": 255
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.37347476325986834,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.2182,
+      "step": 256
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.3341073684904624,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.1814,
+      "step": 257
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.3504981331843485,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.2023,
+      "step": 258
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.34781101624044813,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.187,
+      "step": 259
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.3127735331627072,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1638,
+      "step": 260
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.33265237990198615,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.1862,
+      "step": 261
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.35015832867925345,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.206,
+      "step": 262
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.32908095503184986,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.1778,
+      "step": 263
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.37537224375824596,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.2185,
+      "step": 264
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.30543957154824575,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.1874,
+      "step": 265
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.33672682853021696,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.2084,
+      "step": 266
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.2955685048428476,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.1635,
+      "step": 267
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.3171529594970444,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.1842,
+      "step": 268
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.34736581552759127,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.2027,
+      "step": 269
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.3150102183187486,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.1928,
+      "step": 270
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.3329004389577658,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.1914,
+      "step": 271
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.3423245818888095,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.1857,
+      "step": 272
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.3075317039384064,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1992,
+      "step": 273
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.31724733790920756,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.1926,
+      "step": 274
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.30076557190569597,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.1809,
+      "step": 275
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.30519863258037544,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.2143,
+      "step": 276
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.29506570414695543,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.1767,
+      "step": 277
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.37820461333233696,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.1928,
+      "step": 278
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.30779340816976763,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.1775,
+      "step": 279
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.34605248083621654,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.2056,
+      "step": 280
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.3114467514691629,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.1774,
+      "step": 281
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.3115269201330818,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.1832,
+      "step": 282
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.3450165912378589,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.2018,
+      "step": 283
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.3227156104513421,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.1753,
+      "step": 284
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.3029791254137128,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.1857,
+      "step": 285
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.3088465643618022,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1887,
+      "step": 286
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.305263460649981,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.1732,
+      "step": 287
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.30080533998033043,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.1835,
+      "step": 288
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.3101978906920816,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.1975,
+      "step": 289
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.31151270349369586,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.1742,
+      "step": 290
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.31759513152883234,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.1773,
+      "step": 291
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.32097522054765193,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.1854,
+      "step": 292
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.30042618451980785,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.1913,
+      "step": 293
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.30588959543731825,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.1897,
+      "step": 294
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.2877406552784323,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.1834,
+      "step": 295
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.30319113463160563,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.1979,
+      "step": 296
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.3026029014241974,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.1795,
+      "step": 297
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.3220822432074593,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.2122,
+      "step": 298
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.286174071468055,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1689,
+      "step": 299
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.2976224108721899,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.1721,
+      "step": 300
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.3026819830570282,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.2155,
+      "step": 301
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.2771619310052799,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.1855,
+      "step": 302
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.3110088221545863,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.1899,
+      "step": 303
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.27848627362089545,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.1703,
+      "step": 304
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.32841441923674536,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.201,
+      "step": 305
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.29364084745073127,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.1994,
+      "step": 306
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.28912769750239053,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.183,
+      "step": 307
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.29953889869329897,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.1912,
+      "step": 308
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.30085837546384075,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.1889,
+      "step": 309
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.28765182024887026,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.1804,
+      "step": 310
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.28983659076395446,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.1988,
+      "step": 311
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.2797250120247476,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1933,
+      "step": 312
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 0.8579904807362196,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.3106,
+      "step": 313
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.3686807073622712,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.1257,
+      "step": 314
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.37815873740386835,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.1444,
+      "step": 315
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.3134039760178023,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.1559,
+      "step": 316
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.259911145416525,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.1183,
+      "step": 317
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.3560684167412651,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.1478,
+      "step": 318
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.3388086260272147,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.1135,
+      "step": 319
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.4595029727093649,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.1565,
+      "step": 320
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.3829975297149748,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.1394,
+      "step": 321
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.3557753474699705,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.1577,
+      "step": 322
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.277898611629972,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.1368,
+      "step": 323
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.3009807450851778,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.1494,
+      "step": 324
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.27962571461409447,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1328,
+      "step": 325
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.29841854123122513,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.1379,
+      "step": 326
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.31715843446203096,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.1294,
+      "step": 327
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.2866957811931399,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.1412,
+      "step": 328
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.27339643221481014,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.1187,
+      "step": 329
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.2767366552244504,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.1256,
+      "step": 330
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.3235338230264642,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.1462,
+      "step": 331
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.29220332383844744,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.1346,
+      "step": 332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.2620875042907759,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.1181,
+      "step": 333
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.2753430473262122,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.1264,
+      "step": 334
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.3055838045897359,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.1352,
+      "step": 335
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.26108815755898396,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.111,
+      "step": 336
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.27344902870130877,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.1304,
+      "step": 337
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.28104922812306843,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1442,
+      "step": 338
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.2589803368614472,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.1325,
+      "step": 339
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.26619056445422595,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.1259,
+      "step": 340
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.2717324953630556,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.1401,
+      "step": 341
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.2608867229914016,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.1285,
+      "step": 342
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.26744213800042205,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.1465,
+      "step": 343
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.2654610799356427,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.1279,
+      "step": 344
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.2537264472875052,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.1243,
+      "step": 345
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.27063300969206494,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.1417,
+      "step": 346
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.2715344987158459,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.1368,
+      "step": 347
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.2774586022594309,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.1451,
+      "step": 348
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.25823280715305535,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.122,
+      "step": 349
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.24994646655316943,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.1203,
+      "step": 350
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.26102830614544353,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1366,
+      "step": 351
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.2621874678042739,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.1446,
+      "step": 352
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.2523200217151637,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.1444,
+      "step": 353
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.26357906948775117,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.1483,
+      "step": 354
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.254595565148833,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.1464,
+      "step": 355
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.25541870607323525,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.1296,
+      "step": 356
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.2678438520080487,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.1367,
+      "step": 357
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.27246834098654793,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.1376,
+      "step": 358
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.2545316898387819,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.1234,
+      "step": 359
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.23980914455097604,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.1272,
+      "step": 360
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.27533603789694006,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.1409,
+      "step": 361
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.24256452436055875,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.1184,
+      "step": 362
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.26512113226814327,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.137,
+      "step": 363
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.25057089470542104,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1343,
+      "step": 364
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.26905994457509647,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.1384,
+      "step": 365
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.2515567887129915,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.1183,
+      "step": 366
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.25860107318444187,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.1222,
+      "step": 367
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.2647431542641901,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.1415,
+      "step": 368
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.24625643686363832,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.1234,
+      "step": 369
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.24904863719886972,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.1197,
+      "step": 370
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.23781029435536066,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.1261,
+      "step": 371
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.2634085996145061,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.1454,
+      "step": 372
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.2533454000515048,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.1444,
+      "step": 373
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.24985204283286847,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.129,
+      "step": 374
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.2559467668931813,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.1268,
+      "step": 375
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.2626131517894016,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.1374,
+      "step": 376
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.25223516188684386,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1344,
+      "step": 377
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.24653382543023958,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.1445,
+      "step": 378
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.2487358839721311,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.126,
+      "step": 379
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.27092325982020843,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.1449,
+      "step": 380
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.2419114914044283,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.1231,
+      "step": 381
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.28368695654925413,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.1322,
+      "step": 382
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.2615847280303799,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.1363,
+      "step": 383
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.23348986784414286,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.1184,
+      "step": 384
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.2612408310339395,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.1392,
+      "step": 385
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.2624637156145282,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.1204,
+      "step": 386
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.2800766114957828,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.1354,
+      "step": 387
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.25414283692757583,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.1161,
+      "step": 388
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.2637008974252742,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.1446,
+      "step": 389
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.2402179661973949,
+      "learning_rate": 0.0,
+      "loss": 0.1216,
+      "step": 390
+    },
+    {
+      "epoch": 4.992,
+      "step": 390,
+      "total_flos": 6.874049729748009e+17,
+      "train_loss": 0.30696559743239327,
+      "train_runtime": 15599.7333,
+      "train_samples_per_second": 3.205,
+      "train_steps_per_second": 0.025
+    }
+  ],
+  "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": 6.874049729748009e+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:d576672b285375adf19f6a76c4a28df87ca579bac8aaf70380a0b48841ef1edc
+size 7352