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_science_stackexchange_physics_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_science_stackexchange_physics_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_science_stackexchange_physics_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": 5.5014580010614784e+17,
+    "train_loss": 0.43123698242199726,
+    "train_runtime": 12496.7256,
+    "train_samples_per_second": 4.001,
+    "train_steps_per_second": 0.031
+}

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 sha256:8e8a52430f283ddbe4daa15a6fbba1791f2a1aba8d51177080a54734a27bce65
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8dcb398a2f06f144aafb95aa63568373e0ab1820ad30b8abb138b2b303abc5e
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91188b0ce85d31eefc5443828aa83f41d91ca777443e2e9ba1ef499fdcaf3fc1
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d98701e8fac1fa977467b2c8b16f18d5e8ab2318735baac00dff4bd83ca15f7
+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_22-19-53_jrc0924/events.out.tfevents.1747167645.jrc0924.846598.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdac59dc1a18be665b3295fb7daf190bae54fde985083fae51f04ec90862f1f4
+size 5987

runs/May13_22-25-10_jrc0927/events.out.tfevents.1747167960.jrc0927.1537951.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d61fc6583c6d3902e07dba96832f344cdd97c35f67e5fee2093ae13d2252a7d8
+size 87708

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 22:25:09", "end_time": "2025-05-14 01:54:28"}

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": 5.5014580010614784e+17,
+    "train_loss": 0.43123698242199726,
+    "train_runtime": 12496.7256,
+    "train_samples_per_second": 4.001,
+    "train_steps_per_second": 0.031
+}

trainer_log.jsonl

@@ -0,0 +1,393 @@
+{"current_steps": 1, "total_steps": 390, "loss": 1.1001, "lr": 1.0256410256410257e-06, "epoch": 0.0128, "percentage": 0.26, "elapsed_time": "0:01:09", "remaining_time": "7:29:10"}
+{"current_steps": 2, "total_steps": 390, "loss": 1.1295, "lr": 2.0512820512820513e-06, "epoch": 0.0256, "percentage": 0.51, "elapsed_time": "0:02:13", "remaining_time": "7:12:27"}
+{"current_steps": 1, "total_steps": 390, "loss": 1.1003, "lr": 1.0256410256410257e-06, "epoch": 0.0128, "percentage": 0.26, "elapsed_time": "0:00:42", "remaining_time": "4:36:09"}
+{"current_steps": 2, "total_steps": 390, "loss": 1.1344, "lr": 2.0512820512820513e-06, "epoch": 0.0256, "percentage": 0.51, "elapsed_time": "0:01:11", "remaining_time": "3:51:43"}
+{"current_steps": 3, "total_steps": 390, "loss": 1.0523, "lr": 3.0769230769230774e-06, "epoch": 0.0384, "percentage": 0.77, "elapsed_time": "0:01:45", "remaining_time": "3:47:14"}
+{"current_steps": 4, "total_steps": 390, "loss": 1.0893, "lr": 4.102564102564103e-06, "epoch": 0.0512, "percentage": 1.03, "elapsed_time": "0:02:22", "remaining_time": "3:49:04"}
+{"current_steps": 5, "total_steps": 390, "loss": 1.0351, "lr": 5.128205128205128e-06, "epoch": 0.064, "percentage": 1.28, "elapsed_time": "0:02:54", "remaining_time": "3:44:07"}
+{"current_steps": 6, "total_steps": 390, "loss": 1.0068, "lr": 6.153846153846155e-06, "epoch": 0.0768, "percentage": 1.54, "elapsed_time": "0:03:27", "remaining_time": "3:41:10"}
+{"current_steps": 7, "total_steps": 390, "loss": 0.988, "lr": 7.17948717948718e-06, "epoch": 0.0896, "percentage": 1.79, "elapsed_time": "0:03:59", "remaining_time": "3:38:21"}
+{"current_steps": 8, "total_steps": 390, "loss": 1.0507, "lr": 8.205128205128205e-06, "epoch": 0.1024, "percentage": 2.05, "elapsed_time": "0:04:29", "remaining_time": "3:34:13"}
+{"current_steps": 9, "total_steps": 390, "loss": 1.0128, "lr": 9.230769230769232e-06, "epoch": 0.1152, "percentage": 2.31, "elapsed_time": "0:04:56", "remaining_time": "3:29:14"}
+{"current_steps": 10, "total_steps": 390, "loss": 1.0002, "lr": 1.0256410256410256e-05, "epoch": 0.128, "percentage": 2.56, "elapsed_time": "0:05:32", "remaining_time": "3:30:21"}
+{"current_steps": 11, "total_steps": 390, "loss": 0.9686, "lr": 1.1282051282051283e-05, "epoch": 0.1408, "percentage": 2.82, "elapsed_time": "0:06:05", "remaining_time": "3:29:35"}
+{"current_steps": 12, "total_steps": 390, "loss": 0.9299, "lr": 1.230769230769231e-05, "epoch": 0.1536, "percentage": 3.08, "elapsed_time": "0:06:34", "remaining_time": "3:27:10"}
+{"current_steps": 13, "total_steps": 390, "loss": 0.8971, "lr": 1.3333333333333333e-05, "epoch": 0.1664, "percentage": 3.33, "elapsed_time": "0:07:06", "remaining_time": "3:26:20"}
+{"current_steps": 14, "total_steps": 390, "loss": 0.8866, "lr": 1.435897435897436e-05, "epoch": 0.1792, "percentage": 3.59, "elapsed_time": "0:07:36", "remaining_time": "3:24:08"}
+{"current_steps": 15, "total_steps": 390, "loss": 0.9038, "lr": 1.5384615384615387e-05, "epoch": 0.192, "percentage": 3.85, "elapsed_time": "0:08:08", "remaining_time": "3:23:26"}
+{"current_steps": 16, "total_steps": 390, "loss": 0.8545, "lr": 1.641025641025641e-05, "epoch": 0.2048, "percentage": 4.1, "elapsed_time": "0:08:44", "remaining_time": "3:24:18"}
+{"current_steps": 17, "total_steps": 390, "loss": 0.8533, "lr": 1.7435897435897438e-05, "epoch": 0.2176, "percentage": 4.36, "elapsed_time": "0:09:13", "remaining_time": "3:22:25"}
+{"current_steps": 18, "total_steps": 390, "loss": 0.8216, "lr": 1.8461538461538465e-05, "epoch": 0.2304, "percentage": 4.62, "elapsed_time": "0:09:45", "remaining_time": "3:21:42"}
+{"current_steps": 19, "total_steps": 390, "loss": 0.8487, "lr": 1.9487179487179488e-05, "epoch": 0.2432, "percentage": 4.87, "elapsed_time": "0:10:18", "remaining_time": "3:21:22"}
+{"current_steps": 20, "total_steps": 390, "loss": 0.8003, "lr": 2.0512820512820512e-05, "epoch": 0.256, "percentage": 5.13, "elapsed_time": "0:10:58", "remaining_time": "3:23:02"}
+{"current_steps": 21, "total_steps": 390, "loss": 0.8246, "lr": 2.153846153846154e-05, "epoch": 0.2688, "percentage": 5.38, "elapsed_time": "0:11:28", "remaining_time": "3:21:35"}
+{"current_steps": 22, "total_steps": 390, "loss": 0.8272, "lr": 2.2564102564102566e-05, "epoch": 0.2816, "percentage": 5.64, "elapsed_time": "0:11:58", "remaining_time": "3:20:23"}
+{"current_steps": 23, "total_steps": 390, "loss": 0.7995, "lr": 2.3589743589743593e-05, "epoch": 0.2944, "percentage": 5.9, "elapsed_time": "0:12:27", "remaining_time": "3:18:46"}
+{"current_steps": 24, "total_steps": 390, "loss": 0.8149, "lr": 2.461538461538462e-05, "epoch": 0.3072, "percentage": 6.15, "elapsed_time": "0:13:00", "remaining_time": "3:18:15"}
+{"current_steps": 25, "total_steps": 390, "loss": 0.8033, "lr": 2.5641025641025646e-05, "epoch": 0.32, "percentage": 6.41, "elapsed_time": "0:13:30", "remaining_time": "3:17:09"}
+{"current_steps": 26, "total_steps": 390, "loss": 0.8181, "lr": 2.6666666666666667e-05, "epoch": 0.3328, "percentage": 6.67, "elapsed_time": "0:14:02", "remaining_time": "3:16:28"}
+{"current_steps": 27, "total_steps": 390, "loss": 0.7709, "lr": 2.7692307692307694e-05, "epoch": 0.3456, "percentage": 6.92, "elapsed_time": "0:14:34", "remaining_time": "3:15:57"}
+{"current_steps": 28, "total_steps": 390, "loss": 0.7811, "lr": 2.871794871794872e-05, "epoch": 0.3584, "percentage": 7.18, "elapsed_time": "0:15:11", "remaining_time": "3:16:20"}
+{"current_steps": 29, "total_steps": 390, "loss": 0.7508, "lr": 2.9743589743589747e-05, "epoch": 0.3712, "percentage": 7.44, "elapsed_time": "0:15:40", "remaining_time": "3:15:13"}
+{"current_steps": 30, "total_steps": 390, "loss": 0.7576, "lr": 3.0769230769230774e-05, "epoch": 0.384, "percentage": 7.69, "elapsed_time": "0:16:13", "remaining_time": "3:14:47"}
+{"current_steps": 31, "total_steps": 390, "loss": 0.7964, "lr": 3.1794871794871795e-05, "epoch": 0.3968, "percentage": 7.95, "elapsed_time": "0:16:47", "remaining_time": "3:14:27"}
+{"current_steps": 32, "total_steps": 390, "loss": 0.7766, "lr": 3.282051282051282e-05, "epoch": 0.4096, "percentage": 8.21, "elapsed_time": "0:17:27", "remaining_time": "3:15:14"}
+{"current_steps": 33, "total_steps": 390, "loss": 0.7737, "lr": 3.384615384615385e-05, "epoch": 0.4224, "percentage": 8.46, "elapsed_time": "0:17:58", "remaining_time": "3:14:32"}
+{"current_steps": 34, "total_steps": 390, "loss": 0.8219, "lr": 3.4871794871794875e-05, "epoch": 0.4352, "percentage": 8.72, "elapsed_time": "0:18:30", "remaining_time": "3:13:48"}
+{"current_steps": 35, "total_steps": 390, "loss": 0.8048, "lr": 3.58974358974359e-05, "epoch": 0.448, "percentage": 8.97, "elapsed_time": "0:18:59", "remaining_time": "3:12:37"}
+{"current_steps": 36, "total_steps": 390, "loss": 0.7867, "lr": 3.692307692307693e-05, "epoch": 0.4608, "percentage": 9.23, "elapsed_time": "0:19:24", "remaining_time": "3:10:48"}
+{"current_steps": 37, "total_steps": 390, "loss": 0.7734, "lr": 3.794871794871795e-05, "epoch": 0.4736, "percentage": 9.49, "elapsed_time": "0:19:58", "remaining_time": "3:10:32"}
+{"current_steps": 38, "total_steps": 390, "loss": 0.7919, "lr": 3.8974358974358976e-05, "epoch": 0.4864, "percentage": 9.74, "elapsed_time": "0:20:30", "remaining_time": "3:09:59"}
+{"current_steps": 39, "total_steps": 390, "loss": 0.7586, "lr": 4e-05, "epoch": 0.4992, "percentage": 10.0, "elapsed_time": "0:21:02", "remaining_time": "3:09:18"}
+{"current_steps": 40, "total_steps": 390, "loss": 0.7869, "lr": 3.9999198907597046e-05, "epoch": 0.512, "percentage": 10.26, "elapsed_time": "0:21:35", "remaining_time": "3:08:54"}
+{"current_steps": 41, "total_steps": 390, "loss": 0.7541, "lr": 3.9996795694563096e-05, "epoch": 0.5248, "percentage": 10.51, "elapsed_time": "0:22:16", "remaining_time": "3:09:33"}
+{"current_steps": 42, "total_steps": 390, "loss": 0.7729, "lr": 3.999279055341771e-05, "epoch": 0.5376, "percentage": 10.77, "elapsed_time": "0:22:56", "remaining_time": "3:10:05"}
+{"current_steps": 43, "total_steps": 390, "loss": 0.7462, "lr": 3.998718380500971e-05, "epoch": 0.5504, "percentage": 11.03, "elapsed_time": "0:23:36", "remaining_time": "3:10:30"}
+{"current_steps": 44, "total_steps": 390, "loss": 0.7632, "lr": 3.997997589849145e-05, "epoch": 0.5632, "percentage": 11.28, "elapsed_time": "0:24:14", "remaining_time": "3:10:37"}
+{"current_steps": 45, "total_steps": 390, "loss": 0.7893, "lr": 3.9971167411282835e-05, "epoch": 0.576, "percentage": 11.54, "elapsed_time": "0:24:46", "remaining_time": "3:09:54"}
+{"current_steps": 46, "total_steps": 390, "loss": 0.737, "lr": 3.99607590490251e-05, "epoch": 0.5888, "percentage": 11.79, "elapsed_time": "0:25:17", "remaining_time": "3:09:05"}
+{"current_steps": 47, "total_steps": 390, "loss": 0.7661, "lr": 3.9948751645524235e-05, "epoch": 0.6016, "percentage": 12.05, "elapsed_time": "0:25:48", "remaining_time": "3:08:18"}
+{"current_steps": 48, "total_steps": 390, "loss": 0.7558, "lr": 3.9935146162684206e-05, "epoch": 0.6144, "percentage": 12.31, "elapsed_time": "0:26:21", "remaining_time": "3:07:49"}
+{"current_steps": 49, "total_steps": 390, "loss": 0.7851, "lr": 3.9919943690429906e-05, "epoch": 0.6272, "percentage": 12.56, "elapsed_time": "0:26:51", "remaining_time": "3:06:52"}
+{"current_steps": 50, "total_steps": 390, "loss": 0.7523, "lr": 3.9903145446619837e-05, "epoch": 0.64, "percentage": 12.82, "elapsed_time": "0:27:31", "remaining_time": "3:07:11"}
+{"current_steps": 51, "total_steps": 390, "loss": 0.7434, "lr": 3.9884752776948564e-05, "epoch": 0.6528, "percentage": 13.08, "elapsed_time": "0:28:02", "remaining_time": "3:06:20"}
+{"current_steps": 52, "total_steps": 390, "loss": 0.7843, "lr": 3.9864767154838864e-05, "epoch": 0.6656, "percentage": 13.33, "elapsed_time": "0:28:32", "remaining_time": "3:05:32"}
+{"current_steps": 53, "total_steps": 390, "loss": 0.7592, "lr": 3.9843190181323744e-05, "epoch": 0.6784, "percentage": 13.59, "elapsed_time": "0:29:02", "remaining_time": "3:04:42"}
+{"current_steps": 54, "total_steps": 390, "loss": 0.7482, "lr": 3.982002358491817e-05, "epoch": 0.6912, "percentage": 13.85, "elapsed_time": "0:29:38", "remaining_time": "3:04:27"}
+{"current_steps": 55, "total_steps": 390, "loss": 0.7626, "lr": 3.979526922148058e-05, "epoch": 0.704, "percentage": 14.1, "elapsed_time": "0:30:14", "remaining_time": "3:04:12"}
+{"current_steps": 56, "total_steps": 390, "loss": 0.751, "lr": 3.9768929074064206e-05, "epoch": 0.7168, "percentage": 14.36, "elapsed_time": "0:30:51", "remaining_time": "3:03:59"}
+{"current_steps": 57, "total_steps": 390, "loss": 0.7753, "lr": 3.9741005252758255e-05, "epoch": 0.7296, "percentage": 14.62, "elapsed_time": "0:31:28", "remaining_time": "3:03:53"}
+{"current_steps": 58, "total_steps": 390, "loss": 0.7225, "lr": 3.971149999451886e-05, "epoch": 0.7424, "percentage": 14.87, "elapsed_time": "0:32:01", "remaining_time": "3:03:18"}
+{"current_steps": 59, "total_steps": 390, "loss": 0.7754, "lr": 3.9680415662989806e-05, "epoch": 0.7552, "percentage": 15.13, "elapsed_time": "0:32:33", "remaining_time": "3:02:38"}
+{"current_steps": 60, "total_steps": 390, "loss": 0.7447, "lr": 3.9647754748313294e-05, "epoch": 0.768, "percentage": 15.38, "elapsed_time": "0:33:08", "remaining_time": "3:02:18"}
+{"current_steps": 61, "total_steps": 390, "loss": 0.7656, "lr": 3.96135198669304e-05, "epoch": 0.7808, "percentage": 15.64, "elapsed_time": "0:33:45", "remaining_time": "3:02:03"}
+{"current_steps": 62, "total_steps": 390, "loss": 0.7546, "lr": 3.957771376137144e-05, "epoch": 0.7936, "percentage": 15.9, "elapsed_time": "0:34:22", "remaining_time": "3:01:52"}
+{"current_steps": 63, "total_steps": 390, "loss": 0.7738, "lr": 3.954033930003634e-05, "epoch": 0.8064, "percentage": 16.15, "elapsed_time": "0:34:48", "remaining_time": "3:00:40"}
+{"current_steps": 64, "total_steps": 390, "loss": 0.7183, "lr": 3.9501399476964806e-05, "epoch": 0.8192, "percentage": 16.41, "elapsed_time": "0:35:19", "remaining_time": "2:59:54"}
+{"current_steps": 65, "total_steps": 390, "loss": 0.7598, "lr": 3.946089741159648e-05, "epoch": 0.832, "percentage": 16.67, "elapsed_time": "0:35:50", "remaining_time": "2:59:11"}
+{"current_steps": 66, "total_steps": 390, "loss": 0.749, "lr": 3.9418836348521045e-05, "epoch": 0.8448, "percentage": 16.92, "elapsed_time": "0:36:21", "remaining_time": "2:58:30"}
+{"current_steps": 67, "total_steps": 390, "loss": 0.7691, "lr": 3.937521965721831e-05, "epoch": 0.8576, "percentage": 17.18, "elapsed_time": "0:36:47", "remaining_time": "2:57:19"}
+{"current_steps": 68, "total_steps": 390, "loss": 0.7581, "lr": 3.933005083178828e-05, "epoch": 0.8704, "percentage": 17.44, "elapsed_time": "0:37:16", "remaining_time": "2:56:28"}
+{"current_steps": 69, "total_steps": 390, "loss": 0.7587, "lr": 3.928333349067125e-05, "epoch": 0.8832, "percentage": 17.69, "elapsed_time": "0:37:45", "remaining_time": "2:55:40"}
+{"current_steps": 70, "total_steps": 390, "loss": 0.7545, "lr": 3.923507137635792e-05, "epoch": 0.896, "percentage": 17.95, "elapsed_time": "0:38:19", "remaining_time": "2:55:12"}
+{"current_steps": 71, "total_steps": 390, "loss": 0.7284, "lr": 3.9185268355089606e-05, "epoch": 0.9088, "percentage": 18.21, "elapsed_time": "0:38:58", "remaining_time": "2:55:06"}
+{"current_steps": 72, "total_steps": 390, "loss": 0.77, "lr": 3.913392841654851e-05, "epoch": 0.9216, "percentage": 18.46, "elapsed_time": "0:39:29", "remaining_time": "2:54:23"}
+{"current_steps": 73, "total_steps": 390, "loss": 0.7414, "lr": 3.9081055673538093e-05, "epoch": 0.9344, "percentage": 18.72, "elapsed_time": "0:40:00", "remaining_time": "2:53:45"}
+{"current_steps": 74, "total_steps": 390, "loss": 0.7621, "lr": 3.902665436165364e-05, "epoch": 0.9472, "percentage": 18.97, "elapsed_time": "0:40:32", "remaining_time": "2:53:08"}
+{"current_steps": 75, "total_steps": 390, "loss": 0.7438, "lr": 3.897072883894291e-05, "epoch": 0.96, "percentage": 19.23, "elapsed_time": "0:41:10", "remaining_time": "2:52:54"}
+{"current_steps": 76, "total_steps": 390, "loss": 0.7025, "lr": 3.8913283585557054e-05, "epoch": 0.9728, "percentage": 19.49, "elapsed_time": "0:41:45", "remaining_time": "2:52:30"}
+{"current_steps": 77, "total_steps": 390, "loss": 0.7271, "lr": 3.885432320339167e-05, "epoch": 0.9856, "percentage": 19.74, "elapsed_time": "0:42:15", "remaining_time": "2:51:46"}
+{"current_steps": 78, "total_steps": 390, "loss": 0.7361, "lr": 3.879385241571817e-05, "epoch": 0.9984, "percentage": 20.0, "elapsed_time": "0:42:51", "remaining_time": "2:51:25"}
+{"current_steps": 79, "total_steps": 390, "loss": 1.3899, "lr": 3.873187606680543e-05, "epoch": 1.0112, "percentage": 20.26, "elapsed_time": "0:43:45", "remaining_time": "2:52:16"}
+{"current_steps": 80, "total_steps": 390, "loss": 0.5809, "lr": 3.866839912153168e-05, "epoch": 1.024, "percentage": 20.51, "elapsed_time": "0:44:17", "remaining_time": "2:51:36"}
+{"current_steps": 81, "total_steps": 390, "loss": 0.6123, "lr": 3.860342666498677e-05, "epoch": 1.0368, "percentage": 20.77, "elapsed_time": "0:44:56", "remaining_time": "2:51:26"}
+{"current_steps": 82, "total_steps": 390, "loss": 0.6127, "lr": 3.853696390206484e-05, "epoch": 1.0496, "percentage": 21.03, "elapsed_time": "0:45:28", "remaining_time": "2:50:48"}
+{"current_steps": 83, "total_steps": 390, "loss": 0.6589, "lr": 3.846901615704734e-05, "epoch": 1.0624, "percentage": 21.28, "elapsed_time": "0:46:03", "remaining_time": "2:50:22"}
+{"current_steps": 84, "total_steps": 390, "loss": 0.596, "lr": 3.839958887317649e-05, "epoch": 1.0752, "percentage": 21.54, "elapsed_time": "0:46:38", "remaining_time": "2:49:55"}
+{"current_steps": 85, "total_steps": 390, "loss": 0.59, "lr": 3.832868761221926e-05, "epoch": 1.088, "percentage": 21.79, "elapsed_time": "0:47:09", "remaining_time": "2:49:14"}
+{"current_steps": 86, "total_steps": 390, "loss": 0.5864, "lr": 3.825631805402182e-05, "epoch": 1.1008, "percentage": 22.05, "elapsed_time": "0:47:43", "remaining_time": "2:48:41"}
+{"current_steps": 87, "total_steps": 390, "loss": 0.6038, "lr": 3.818248599605448e-05, "epoch": 1.1136, "percentage": 22.31, "elapsed_time": "0:48:17", "remaining_time": "2:48:11"}
+{"current_steps": 88, "total_steps": 390, "loss": 0.5875, "lr": 3.810719735294731e-05, "epoch": 1.1264, "percentage": 22.56, "elapsed_time": "0:48:47", "remaining_time": "2:47:27"}
+{"current_steps": 89, "total_steps": 390, "loss": 0.651, "lr": 3.8030458156016326e-05, "epoch": 1.1392, "percentage": 22.82, "elapsed_time": "0:49:20", "remaining_time": "2:46:52"}
+{"current_steps": 90, "total_steps": 390, "loss": 0.5769, "lr": 3.795227455278029e-05, "epoch": 1.152, "percentage": 23.08, "elapsed_time": "0:49:52", "remaining_time": "2:46:15"}
+{"current_steps": 91, "total_steps": 390, "loss": 0.5703, "lr": 3.787265280646825e-05, "epoch": 1.1648, "percentage": 23.33, "elapsed_time": "0:50:25", "remaining_time": "2:45:39"}
+{"current_steps": 92, "total_steps": 390, "loss": 0.612, "lr": 3.7791599295517825e-05, "epoch": 1.1776, "percentage": 23.59, "elapsed_time": "0:51:02", "remaining_time": "2:45:18"}
+{"current_steps": 93, "total_steps": 390, "loss": 0.5846, "lr": 3.7709120513064196e-05, "epoch": 1.1904, "percentage": 23.85, "elapsed_time": "0:51:30", "remaining_time": "2:44:28"}
+{"current_steps": 94, "total_steps": 390, "loss": 0.5907, "lr": 3.762522306641998e-05, "epoch": 1.2032, "percentage": 24.1, "elapsed_time": "0:52:02", "remaining_time": "2:43:52"}
+{"current_steps": 95, "total_steps": 390, "loss": 0.5919, "lr": 3.7539913676545874e-05, "epoch": 1.216, "percentage": 24.36, "elapsed_time": "0:52:35", "remaining_time": "2:43:18"}
+{"current_steps": 96, "total_steps": 390, "loss": 0.6066, "lr": 3.745319917751229e-05, "epoch": 1.2288000000000001, "percentage": 24.62, "elapsed_time": "0:53:08", "remaining_time": "2:42:45"}
+{"current_steps": 97, "total_steps": 390, "loss": 0.5653, "lr": 3.736508651595188e-05, "epoch": 1.2416, "percentage": 24.87, "elapsed_time": "0:53:41", "remaining_time": "2:42:10"}
+{"current_steps": 98, "total_steps": 390, "loss": 0.5978, "lr": 3.727558275050301e-05, "epoch": 1.2544, "percentage": 25.13, "elapsed_time": "0:54:12", "remaining_time": "2:41:29"}
+{"current_steps": 99, "total_steps": 390, "loss": 0.545, "lr": 3.718469505124434e-05, "epoch": 1.2671999999999999, "percentage": 25.38, "elapsed_time": "0:54:48", "remaining_time": "2:41:05"}
+{"current_steps": 100, "total_steps": 390, "loss": 0.589, "lr": 3.709243069912041e-05, "epoch": 1.28, "percentage": 25.64, "elapsed_time": "0:55:14", "remaining_time": "2:40:13"}
+{"current_steps": 101, "total_steps": 390, "loss": 0.5913, "lr": 3.699879708535838e-05, "epoch": 1.2928, "percentage": 25.9, "elapsed_time": "0:55:49", "remaining_time": "2:39:44"}
+{"current_steps": 102, "total_steps": 390, "loss": 0.5825, "lr": 3.69038017108759e-05, "epoch": 1.3056, "percentage": 26.15, "elapsed_time": "0:56:21", "remaining_time": "2:39:08"}
+{"current_steps": 103, "total_steps": 390, "loss": 0.5649, "lr": 3.680745218568026e-05, "epoch": 1.3184, "percentage": 26.41, "elapsed_time": "0:56:52", "remaining_time": "2:38:28"}
+{"current_steps": 104, "total_steps": 390, "loss": 0.5617, "lr": 3.6709756228258735e-05, "epoch": 1.3312, "percentage": 26.67, "elapsed_time": "0:57:26", "remaining_time": "2:37:57"}
+{"current_steps": 105, "total_steps": 390, "loss": 0.6093, "lr": 3.6610721664960236e-05, "epoch": 1.3439999999999999, "percentage": 26.92, "elapsed_time": "0:57:59", "remaining_time": "2:37:23"}
+{"current_steps": 106, "total_steps": 390, "loss": 0.5517, "lr": 3.65103564293684e-05, "epoch": 1.3568, "percentage": 27.18, "elapsed_time": "0:58:27", "remaining_time": "2:36:36"}
+{"current_steps": 107, "total_steps": 390, "loss": 0.6685, "lr": 3.640866856166601e-05, "epoch": 1.3696, "percentage": 27.44, "elapsed_time": "0:59:01", "remaining_time": "2:36:06"}
+{"current_steps": 108, "total_steps": 390, "loss": 0.558, "lr": 3.6305666207990886e-05, "epoch": 1.3824, "percentage": 27.69, "elapsed_time": "0:59:31", "remaining_time": "2:35:25"}
+{"current_steps": 109, "total_steps": 390, "loss": 0.6059, "lr": 3.6201357619783336e-05, "epoch": 1.3952, "percentage": 27.95, "elapsed_time": "1:00:09", "remaining_time": "2:35:04"}
+{"current_steps": 110, "total_steps": 390, "loss": 0.6013, "lr": 3.609575115312511e-05, "epoch": 1.408, "percentage": 28.21, "elapsed_time": "1:00:41", "remaining_time": "2:34:29"}
+{"current_steps": 111, "total_steps": 390, "loss": 0.5439, "lr": 3.598885526807003e-05, "epoch": 1.4208, "percentage": 28.46, "elapsed_time": "1:01:11", "remaining_time": "2:33:49"}
+{"current_steps": 112, "total_steps": 390, "loss": 0.6107, "lr": 3.5880678527966224e-05, "epoch": 1.4336, "percentage": 28.72, "elapsed_time": "1:01:45", "remaining_time": "2:33:18"}
+{"current_steps": 113, "total_steps": 390, "loss": 0.533, "lr": 3.577122959877017e-05, "epoch": 1.4464000000000001, "percentage": 28.97, "elapsed_time": "1:02:17", "remaining_time": "2:32:41"}
+{"current_steps": 114, "total_steps": 390, "loss": 0.5659, "lr": 3.566051724835245e-05, "epoch": 1.4592, "percentage": 29.23, "elapsed_time": "1:02:48", "remaining_time": "2:32:02"}
+{"current_steps": 115, "total_steps": 390, "loss": 0.5995, "lr": 3.554855034579532e-05, "epoch": 1.472, "percentage": 29.49, "elapsed_time": "1:03:19", "remaining_time": "2:31:26"}
+{"current_steps": 116, "total_steps": 390, "loss": 0.5881, "lr": 3.5435337860682304e-05, "epoch": 1.4848, "percentage": 29.74, "elapsed_time": "1:03:53", "remaining_time": "2:30:54"}
+{"current_steps": 117, "total_steps": 390, "loss": 0.5846, "lr": 3.532088886237956e-05, "epoch": 1.4976, "percentage": 30.0, "elapsed_time": "1:04:19", "remaining_time": "2:30:05"}
+{"current_steps": 118, "total_steps": 390, "loss": 0.6118, "lr": 3.520521251930941e-05, "epoch": 1.5104, "percentage": 30.26, "elapsed_time": "1:04:48", "remaining_time": "2:29:23"}
+{"current_steps": 119, "total_steps": 390, "loss": 0.5948, "lr": 3.5088318098215805e-05, "epoch": 1.5232, "percentage": 30.51, "elapsed_time": "1:05:17", "remaining_time": "2:28:42"}
+{"current_steps": 120, "total_steps": 390, "loss": 0.6025, "lr": 3.497021496342203e-05, "epoch": 1.536, "percentage": 30.77, "elapsed_time": "1:05:48", "remaining_time": "2:28:04"}
+{"current_steps": 121, "total_steps": 390, "loss": 0.5763, "lr": 3.485091257608047e-05, "epoch": 1.5488, "percentage": 31.03, "elapsed_time": "1:06:21", "remaining_time": "2:27:30"}
+{"current_steps": 122, "total_steps": 390, "loss": 0.5707, "lr": 3.473042049341474e-05, "epoch": 1.5615999999999999, "percentage": 31.28, "elapsed_time": "1:06:47", "remaining_time": "2:26:42"}
+{"current_steps": 123, "total_steps": 390, "loss": 0.5983, "lr": 3.4608748367954064e-05, "epoch": 1.5744, "percentage": 31.54, "elapsed_time": "1:07:19", "remaining_time": "2:26:09"}
+{"current_steps": 124, "total_steps": 390, "loss": 0.5747, "lr": 3.4485905946759965e-05, "epoch": 1.5872000000000002, "percentage": 31.79, "elapsed_time": "1:07:53", "remaining_time": "2:25:37"}
+{"current_steps": 125, "total_steps": 390, "loss": 0.5849, "lr": 3.4361903070645484e-05, "epoch": 1.6, "percentage": 32.05, "elapsed_time": "1:08:23", "remaining_time": "2:24:59"}
+{"current_steps": 126, "total_steps": 390, "loss": 0.6005, "lr": 3.423674967338681e-05, "epoch": 1.6128, "percentage": 32.31, "elapsed_time": "1:08:55", "remaining_time": "2:24:24"}
+{"current_steps": 127, "total_steps": 390, "loss": 0.583, "lr": 3.411045578092754e-05, "epoch": 1.6256, "percentage": 32.56, "elapsed_time": "1:09:23", "remaining_time": "2:23:41"}
+{"current_steps": 128, "total_steps": 390, "loss": 0.5885, "lr": 3.398303151057543e-05, "epoch": 1.6383999999999999, "percentage": 32.82, "elapsed_time": "1:09:52", "remaining_time": "2:23:00"}
+{"current_steps": 129, "total_steps": 390, "loss": 0.5849, "lr": 3.385448707019199e-05, "epoch": 1.6512, "percentage": 33.08, "elapsed_time": "1:10:25", "remaining_time": "2:22:28"}
+{"current_steps": 130, "total_steps": 390, "loss": 0.5509, "lr": 3.372483275737468e-05, "epoch": 1.6640000000000001, "percentage": 33.33, "elapsed_time": "1:10:55", "remaining_time": "2:21:51"}
+{"current_steps": 131, "total_steps": 390, "loss": 0.5781, "lr": 3.359407895863199e-05, "epoch": 1.6768, "percentage": 33.59, "elapsed_time": "1:11:29", "remaining_time": "2:21:21"}
+{"current_steps": 132, "total_steps": 390, "loss": 0.5958, "lr": 3.34622361485514e-05, "epoch": 1.6896, "percentage": 33.85, "elapsed_time": "1:11:55", "remaining_time": "2:20:35"}
+{"current_steps": 133, "total_steps": 390, "loss": 0.6211, "lr": 3.332931488896029e-05, "epoch": 1.7024, "percentage": 34.1, "elapsed_time": "1:12:24", "remaining_time": "2:19:54"}
+{"current_steps": 134, "total_steps": 390, "loss": 0.5874, "lr": 3.319532582807977e-05, "epoch": 1.7151999999999998, "percentage": 34.36, "elapsed_time": "1:12:58", "remaining_time": "2:19:24"}
+{"current_steps": 135, "total_steps": 390, "loss": 0.5838, "lr": 3.30602796996717e-05, "epoch": 1.728, "percentage": 34.62, "elapsed_time": "1:13:30", "remaining_time": "2:18:50"}
+{"current_steps": 136, "total_steps": 390, "loss": 0.5913, "lr": 3.2924187322178865e-05, "epoch": 1.7408000000000001, "percentage": 34.87, "elapsed_time": "1:14:00", "remaining_time": "2:18:12"}
+{"current_steps": 137, "total_steps": 390, "loss": 0.5457, "lr": 3.278705959785821e-05, "epoch": 1.7536, "percentage": 35.13, "elapsed_time": "1:14:38", "remaining_time": "2:17:49"}
+{"current_steps": 138, "total_steps": 390, "loss": 0.5893, "lr": 3.2648907511907544e-05, "epoch": 1.7664, "percentage": 35.38, "elapsed_time": "1:15:17", "remaining_time": "2:17:29"}
+{"current_steps": 139, "total_steps": 390, "loss": 0.5857, "lr": 3.250974213158555e-05, "epoch": 1.7792, "percentage": 35.64, "elapsed_time": "1:15:48", "remaining_time": "2:16:53"}
+{"current_steps": 140, "total_steps": 390, "loss": 0.5679, "lr": 3.23695746053251e-05, "epoch": 1.792, "percentage": 35.9, "elapsed_time": "1:16:23", "remaining_time": "2:16:24"}
+{"current_steps": 141, "total_steps": 390, "loss": 0.5549, "lr": 3.222841616184025e-05, "epoch": 1.8048, "percentage": 36.15, "elapsed_time": "1:16:53", "remaining_time": "2:15:47"}
+{"current_steps": 142, "total_steps": 390, "loss": 0.5919, "lr": 3.208627810922665e-05, "epoch": 1.8176, "percentage": 36.41, "elapsed_time": "1:17:27", "remaining_time": "2:15:16"}
+{"current_steps": 143, "total_steps": 390, "loss": 0.5513, "lr": 3.194317183405573e-05, "epoch": 1.8304, "percentage": 36.67, "elapsed_time": "1:18:03", "remaining_time": "2:14:49"}
+{"current_steps": 144, "total_steps": 390, "loss": 0.6127, "lr": 3.1799108800462466e-05, "epoch": 1.8432, "percentage": 36.92, "elapsed_time": "1:18:32", "remaining_time": "2:14:10"}
+{"current_steps": 145, "total_steps": 390, "loss": 0.5912, "lr": 3.1654100549227024e-05, "epoch": 1.8559999999999999, "percentage": 37.18, "elapsed_time": "1:19:06", "remaining_time": "2:13:39"}
+{"current_steps": 146, "total_steps": 390, "loss": 0.6135, "lr": 3.1508158696850275e-05, "epoch": 1.8688, "percentage": 37.44, "elapsed_time": "1:19:35", "remaining_time": "2:13:00"}
+{"current_steps": 147, "total_steps": 390, "loss": 0.6091, "lr": 3.136129493462312e-05, "epoch": 1.8816000000000002, "percentage": 37.69, "elapsed_time": "1:20:04", "remaining_time": "2:12:22"}
+{"current_steps": 148, "total_steps": 390, "loss": 0.5842, "lr": 3.121352102768998e-05, "epoch": 1.8944, "percentage": 37.95, "elapsed_time": "1:20:32", "remaining_time": "2:11:42"}
+{"current_steps": 149, "total_steps": 390, "loss": 0.5895, "lr": 3.106484881410628e-05, "epoch": 1.9072, "percentage": 38.21, "elapsed_time": "1:21:06", "remaining_time": "2:11:11"}
+{"current_steps": 150, "total_steps": 390, "loss": 0.5853, "lr": 3.091529020389009e-05, "epoch": 1.92, "percentage": 38.46, "elapsed_time": "1:21:43", "remaining_time": "2:10:44"}
+{"current_steps": 151, "total_steps": 390, "loss": 0.5877, "lr": 3.076485717806808e-05, "epoch": 1.9327999999999999, "percentage": 38.72, "elapsed_time": "1:22:10", "remaining_time": "2:10:04"}
+{"current_steps": 152, "total_steps": 390, "loss": 0.5679, "lr": 3.061356178771564e-05, "epoch": 1.9456, "percentage": 38.97, "elapsed_time": "1:22:45", "remaining_time": "2:09:34"}
+{"current_steps": 153, "total_steps": 390, "loss": 0.6024, "lr": 3.0461416152991555e-05, "epoch": 1.9584000000000001, "percentage": 39.23, "elapsed_time": "1:23:21", "remaining_time": "2:09:08"}
+{"current_steps": 154, "total_steps": 390, "loss": 0.5926, "lr": 3.0308432462167045e-05, "epoch": 1.9712, "percentage": 39.49, "elapsed_time": "1:23:49", "remaining_time": "2:08:26"}
+{"current_steps": 155, "total_steps": 390, "loss": 0.5512, "lr": 3.015462297064936e-05, "epoch": 1.984, "percentage": 39.74, "elapsed_time": "1:24:19", "remaining_time": "2:07:51"}
+{"current_steps": 156, "total_steps": 390, "loss": 0.5713, "lr": 3.0000000000000004e-05, "epoch": 1.9968, "percentage": 40.0, "elapsed_time": "1:24:55", "remaining_time": "2:07:22"}
+{"current_steps": 157, "total_steps": 390, "loss": 1.0221, "lr": 2.98445759369477e-05, "epoch": 2.0096, "percentage": 40.26, "elapsed_time": "1:25:45", "remaining_time": "2:07:16"}
+{"current_steps": 158, "total_steps": 390, "loss": 0.3996, "lr": 2.9688363232396056e-05, "epoch": 2.0224, "percentage": 40.51, "elapsed_time": "1:26:16", "remaining_time": "2:06:41"}
+{"current_steps": 159, "total_steps": 390, "loss": 0.4314, "lr": 2.9531374400426158e-05, "epoch": 2.0352, "percentage": 40.77, "elapsed_time": "1:26:47", "remaining_time": "2:06:05"}
+{"current_steps": 160, "total_steps": 390, "loss": 0.4054, "lr": 2.9373622017294075e-05, "epoch": 2.048, "percentage": 41.03, "elapsed_time": "1:27:23", "remaining_time": "2:05:37"}
+{"current_steps": 161, "total_steps": 390, "loss": 0.3994, "lr": 2.9215118720423375e-05, "epoch": 2.0608, "percentage": 41.28, "elapsed_time": "1:27:58", "remaining_time": "2:05:07"}
+{"current_steps": 162, "total_steps": 390, "loss": 0.3982, "lr": 2.9055877207392752e-05, "epoch": 2.0736, "percentage": 41.54, "elapsed_time": "1:28:31", "remaining_time": "2:04:35"}
+{"current_steps": 163, "total_steps": 390, "loss": 0.3753, "lr": 2.8895910234918828e-05, "epoch": 2.0864, "percentage": 41.79, "elapsed_time": "1:29:05", "remaining_time": "2:04:04"}
+{"current_steps": 164, "total_steps": 390, "loss": 0.4019, "lr": 2.873523061783426e-05, "epoch": 2.0992, "percentage": 42.05, "elapsed_time": "1:29:38", "remaining_time": "2:03:31"}
+{"current_steps": 165, "total_steps": 390, "loss": 0.401, "lr": 2.8573851228061084e-05, "epoch": 2.112, "percentage": 42.31, "elapsed_time": "1:30:06", "remaining_time": "2:02:52"}
+{"current_steps": 166, "total_steps": 390, "loss": 0.3854, "lr": 2.8411784993579633e-05, "epoch": 2.1248, "percentage": 42.56, "elapsed_time": "1:30:35", "remaining_time": "2:02:15"}
+{"current_steps": 167, "total_steps": 390, "loss": 0.3867, "lr": 2.8249044897392814e-05, "epoch": 2.1376, "percentage": 42.82, "elapsed_time": "1:31:06", "remaining_time": "2:01:39"}
+{"current_steps": 168, "total_steps": 390, "loss": 0.3844, "lr": 2.80856439764861e-05, "epoch": 2.1504, "percentage": 43.08, "elapsed_time": "1:31:38", "remaining_time": "2:01:06"}
+{"current_steps": 169, "total_steps": 390, "loss": 0.3738, "lr": 2.792159532078314e-05, "epoch": 2.1632, "percentage": 43.33, "elapsed_time": "1:32:10", "remaining_time": "2:00:32"}
+{"current_steps": 170, "total_steps": 390, "loss": 0.3821, "lr": 2.77569120720971e-05, "epoch": 2.176, "percentage": 43.59, "elapsed_time": "1:32:38", "remaining_time": "1:59:53"}
+{"current_steps": 171, "total_steps": 390, "loss": 0.377, "lr": 2.7591607423077932e-05, "epoch": 2.1888, "percentage": 43.85, "elapsed_time": "1:33:13", "remaining_time": "1:59:23"}
+{"current_steps": 172, "total_steps": 390, "loss": 0.3559, "lr": 2.7425694616155474e-05, "epoch": 2.2016, "percentage": 44.1, "elapsed_time": "1:33:40", "remaining_time": "1:58:43"}
+{"current_steps": 173, "total_steps": 390, "loss": 0.4095, "lr": 2.7259186942478656e-05, "epoch": 2.2144, "percentage": 44.36, "elapsed_time": "1:34:12", "remaining_time": "1:58:10"}
+{"current_steps": 174, "total_steps": 390, "loss": 0.3796, "lr": 2.7092097740850712e-05, "epoch": 2.2272, "percentage": 44.62, "elapsed_time": "1:34:42", "remaining_time": "1:57:34"}
+{"current_steps": 175, "total_steps": 390, "loss": 0.3837, "lr": 2.692444039666066e-05, "epoch": 2.24, "percentage": 44.87, "elapsed_time": "1:35:14", "remaining_time": "1:57:00"}
+{"current_steps": 176, "total_steps": 390, "loss": 0.3881, "lr": 2.6756228340810946e-05, "epoch": 2.2528, "percentage": 45.13, "elapsed_time": "1:35:45", "remaining_time": "1:56:25"}
+{"current_steps": 177, "total_steps": 390, "loss": 0.373, "lr": 2.6587475048641596e-05, "epoch": 2.2656, "percentage": 45.38, "elapsed_time": "1:36:11", "remaining_time": "1:55:45"}
+{"current_steps": 178, "total_steps": 390, "loss": 0.3681, "lr": 2.6418194038850634e-05, "epoch": 2.2784, "percentage": 45.64, "elapsed_time": "1:36:42", "remaining_time": "1:55:10"}
+{"current_steps": 179, "total_steps": 390, "loss": 0.3934, "lr": 2.624839887241115e-05, "epoch": 2.2912, "percentage": 45.9, "elapsed_time": "1:37:22", "remaining_time": "1:54:47"}
+{"current_steps": 180, "total_steps": 390, "loss": 0.3484, "lr": 2.607810315148494e-05, "epoch": 2.304, "percentage": 46.15, "elapsed_time": "1:37:52", "remaining_time": "1:54:11"}
+{"current_steps": 181, "total_steps": 390, "loss": 0.4197, "lr": 2.5907320518332827e-05, "epoch": 2.3168, "percentage": 46.41, "elapsed_time": "1:38:24", "remaining_time": "1:53:37"}
+{"current_steps": 182, "total_steps": 390, "loss": 0.3731, "lr": 2.5736064654221808e-05, "epoch": 2.3296, "percentage": 46.67, "elapsed_time": "1:38:51", "remaining_time": "1:52:58"}
+{"current_steps": 183, "total_steps": 390, "loss": 0.377, "lr": 2.5564349278329056e-05, "epoch": 2.3424, "percentage": 46.92, "elapsed_time": "1:39:24", "remaining_time": "1:52:27"}
+{"current_steps": 184, "total_steps": 390, "loss": 0.38, "lr": 2.539218814664288e-05, "epoch": 2.3552, "percentage": 47.18, "elapsed_time": "1:39:58", "remaining_time": "1:51:55"}
+{"current_steps": 185, "total_steps": 390, "loss": 0.4006, "lr": 2.521959505086075e-05, "epoch": 2.368, "percentage": 47.44, "elapsed_time": "1:40:33", "remaining_time": "1:51:25"}
+{"current_steps": 186, "total_steps": 390, "loss": 0.3329, "lr": 2.5046583817284437e-05, "epoch": 2.3808, "percentage": 47.69, "elapsed_time": "1:41:10", "remaining_time": "1:50:57"}
+{"current_steps": 187, "total_steps": 390, "loss": 0.387, "lr": 2.487316830571244e-05, "epoch": 2.3936, "percentage": 47.95, "elapsed_time": "1:41:44", "remaining_time": "1:50:27"}
+{"current_steps": 188, "total_steps": 390, "loss": 0.3828, "lr": 2.4699362408329646e-05, "epoch": 2.4064, "percentage": 48.21, "elapsed_time": "1:42:11", "remaining_time": "1:49:47"}
+{"current_steps": 189, "total_steps": 390, "loss": 0.3874, "lr": 2.4525180048594452e-05, "epoch": 2.4192, "percentage": 48.46, "elapsed_time": "1:42:39", "remaining_time": "1:49:10"}
+{"current_steps": 190, "total_steps": 390, "loss": 0.3864, "lr": 2.435063518012335e-05, "epoch": 2.432, "percentage": 48.72, "elapsed_time": "1:43:10", "remaining_time": "1:48:36"}
+{"current_steps": 191, "total_steps": 390, "loss": 0.3824, "lr": 2.4175741785573177e-05, "epoch": 2.4448, "percentage": 48.97, "elapsed_time": "1:43:37", "remaining_time": "1:47:57"}
+{"current_steps": 192, "total_steps": 390, "loss": 0.3884, "lr": 2.4000513875520892e-05, "epoch": 2.4576000000000002, "percentage": 49.23, "elapsed_time": "1:44:10", "remaining_time": "1:47:25"}
+{"current_steps": 193, "total_steps": 390, "loss": 0.3852, "lr": 2.3824965487341247e-05, "epoch": 2.4704, "percentage": 49.49, "elapsed_time": "1:44:37", "remaining_time": "1:46:47"}
+{"current_steps": 194, "total_steps": 390, "loss": 0.3972, "lr": 2.3649110684082258e-05, "epoch": 2.4832, "percentage": 49.74, "elapsed_time": "1:45:14", "remaining_time": "1:46:19"}
+{"current_steps": 195, "total_steps": 390, "loss": 0.3618, "lr": 2.3472963553338614e-05, "epoch": 2.496, "percentage": 50.0, "elapsed_time": "1:45:44", "remaining_time": "1:45:44"}
+{"current_steps": 196, "total_steps": 390, "loss": 0.3939, "lr": 2.3296538206123134e-05, "epoch": 2.5088, "percentage": 50.26, "elapsed_time": "1:46:15", "remaining_time": "1:45:10"}
+{"current_steps": 197, "total_steps": 390, "loss": 0.3479, "lr": 2.311984877573636e-05, "epoch": 2.5216, "percentage": 50.51, "elapsed_time": "1:46:48", "remaining_time": "1:44:38"}
+{"current_steps": 198, "total_steps": 390, "loss": 0.3703, "lr": 2.2942909416634326e-05, "epoch": 2.5343999999999998, "percentage": 50.77, "elapsed_time": "1:47:26", "remaining_time": "1:44:11"}
+{"current_steps": 199, "total_steps": 390, "loss": 0.3996, "lr": 2.2765734303294666e-05, "epoch": 2.5472, "percentage": 51.03, "elapsed_time": "1:47:55", "remaining_time": "1:43:35"}
+{"current_steps": 200, "total_steps": 390, "loss": 0.3772, "lr": 2.2588337629081107e-05, "epoch": 2.56, "percentage": 51.28, "elapsed_time": "1:48:29", "remaining_time": "1:43:04"}
+{"current_steps": 201, "total_steps": 390, "loss": 0.3652, "lr": 2.2410733605106462e-05, "epoch": 2.5728, "percentage": 51.54, "elapsed_time": "1:49:06", "remaining_time": "1:42:36"}
+{"current_steps": 202, "total_steps": 390, "loss": 0.3738, "lr": 2.2232936459094158e-05, "epoch": 2.5856, "percentage": 51.79, "elapsed_time": "1:49:40", "remaining_time": "1:42:04"}
+{"current_steps": 203, "total_steps": 390, "loss": 0.3987, "lr": 2.205496043423849e-05, "epoch": 2.5984, "percentage": 52.05, "elapsed_time": "1:50:14", "remaining_time": "1:41:32"}
+{"current_steps": 204, "total_steps": 390, "loss": 0.362, "lr": 2.1876819788063586e-05, "epoch": 2.6112, "percentage": 52.31, "elapsed_time": "1:50:35", "remaining_time": "1:40:50"}
+{"current_steps": 205, "total_steps": 390, "loss": 0.3858, "lr": 2.16985287912813e-05, "epoch": 2.624, "percentage": 52.56, "elapsed_time": "1:51:08", "remaining_time": "1:40:17"}
+{"current_steps": 206, "total_steps": 390, "loss": 0.3898, "lr": 2.1520101726647922e-05, "epoch": 2.6368, "percentage": 52.82, "elapsed_time": "1:51:39", "remaining_time": "1:39:43"}
+{"current_steps": 207, "total_steps": 390, "loss": 0.3769, "lr": 2.1341552887820048e-05, "epoch": 2.6496, "percentage": 53.08, "elapsed_time": "1:52:10", "remaining_time": "1:39:10"}
+{"current_steps": 208, "total_steps": 390, "loss": 0.3803, "lr": 2.1162896578209517e-05, "epoch": 2.6624, "percentage": 53.33, "elapsed_time": "1:52:40", "remaining_time": "1:38:35"}
+{"current_steps": 209, "total_steps": 390, "loss": 0.3834, "lr": 2.0984147109837564e-05, "epoch": 2.6752000000000002, "percentage": 53.59, "elapsed_time": "1:53:06", "remaining_time": "1:37:57"}
+{"current_steps": 210, "total_steps": 390, "loss": 0.3976, "lr": 2.0805318802188307e-05, "epoch": 2.6879999999999997, "percentage": 53.85, "elapsed_time": "1:53:34", "remaining_time": "1:37:20"}
+{"current_steps": 211, "total_steps": 390, "loss": 0.3553, "lr": 2.0626425981061608e-05, "epoch": 2.7008, "percentage": 54.1, "elapsed_time": "1:54:06", "remaining_time": "1:36:48"}
+{"current_steps": 212, "total_steps": 390, "loss": 0.3911, "lr": 2.0447482977425465e-05, "epoch": 2.7136, "percentage": 54.36, "elapsed_time": "1:54:39", "remaining_time": "1:36:16"}
+{"current_steps": 213, "total_steps": 390, "loss": 0.3771, "lr": 2.0268504126267952e-05, "epoch": 2.7264, "percentage": 54.62, "elapsed_time": "1:55:06", "remaining_time": "1:35:39"}
+{"current_steps": 214, "total_steps": 390, "loss": 0.371, "lr": 2.008950376544887e-05, "epoch": 2.7392, "percentage": 54.87, "elapsed_time": "1:55:44", "remaining_time": "1:35:11"}
+{"current_steps": 215, "total_steps": 390, "loss": 0.3835, "lr": 1.9910496234551132e-05, "epoch": 2.752, "percentage": 55.13, "elapsed_time": "1:56:20", "remaining_time": "1:34:42"}
+{"current_steps": 216, "total_steps": 390, "loss": 0.378, "lr": 1.9731495873732055e-05, "epoch": 2.7648, "percentage": 55.38, "elapsed_time": "1:56:52", "remaining_time": "1:34:09"}
+{"current_steps": 217, "total_steps": 390, "loss": 0.3885, "lr": 1.9552517022574542e-05, "epoch": 2.7776, "percentage": 55.64, "elapsed_time": "1:57:28", "remaining_time": "1:33:38"}
+{"current_steps": 218, "total_steps": 390, "loss": 0.373, "lr": 1.93735740189384e-05, "epoch": 2.7904, "percentage": 55.9, "elapsed_time": "1:58:00", "remaining_time": "1:33:06"}
+{"current_steps": 219, "total_steps": 390, "loss": 0.3717, "lr": 1.9194681197811703e-05, "epoch": 2.8032, "percentage": 56.15, "elapsed_time": "1:58:34", "remaining_time": "1:32:35"}
+{"current_steps": 220, "total_steps": 390, "loss": 0.3744, "lr": 1.901585289016244e-05, "epoch": 2.816, "percentage": 56.41, "elapsed_time": "1:59:08", "remaining_time": "1:32:03"}
+{"current_steps": 221, "total_steps": 390, "loss": 0.4074, "lr": 1.8837103421790486e-05, "epoch": 2.8288, "percentage": 56.67, "elapsed_time": "1:59:35", "remaining_time": "1:31:26"}
+{"current_steps": 222, "total_steps": 390, "loss": 0.4097, "lr": 1.8658447112179952e-05, "epoch": 2.8416, "percentage": 56.92, "elapsed_time": "2:00:04", "remaining_time": "1:30:51"}
+{"current_steps": 223, "total_steps": 390, "loss": 0.385, "lr": 1.8479898273352084e-05, "epoch": 2.8544, "percentage": 57.18, "elapsed_time": "2:00:38", "remaining_time": "1:30:20"}
+{"current_steps": 224, "total_steps": 390, "loss": 0.4015, "lr": 1.83014712087187e-05, "epoch": 2.8672, "percentage": 57.44, "elapsed_time": "2:01:12", "remaining_time": "1:29:49"}
+{"current_steps": 225, "total_steps": 390, "loss": 0.3673, "lr": 1.8123180211936417e-05, "epoch": 2.88, "percentage": 57.69, "elapsed_time": "2:01:50", "remaining_time": "1:29:20"}
+{"current_steps": 226, "total_steps": 390, "loss": 0.3939, "lr": 1.794503956576152e-05, "epoch": 2.8928000000000003, "percentage": 57.95, "elapsed_time": "2:02:15", "remaining_time": "1:28:43"}
+{"current_steps": 227, "total_steps": 390, "loss": 0.4079, "lr": 1.776706354090585e-05, "epoch": 2.9055999999999997, "percentage": 58.21, "elapsed_time": "2:02:43", "remaining_time": "1:28:07"}
+{"current_steps": 228, "total_steps": 390, "loss": 0.3558, "lr": 1.758926639489354e-05, "epoch": 2.9184, "percentage": 58.46, "elapsed_time": "2:03:17", "remaining_time": "1:27:36"}
+{"current_steps": 229, "total_steps": 390, "loss": 0.3774, "lr": 1.7411662370918893e-05, "epoch": 2.9312, "percentage": 58.72, "elapsed_time": "2:03:51", "remaining_time": "1:27:04"}
+{"current_steps": 230, "total_steps": 390, "loss": 0.3751, "lr": 1.7234265696705344e-05, "epoch": 2.944, "percentage": 58.97, "elapsed_time": "2:04:23", "remaining_time": "1:26:32"}
+{"current_steps": 231, "total_steps": 390, "loss": 0.398, "lr": 1.7057090583365678e-05, "epoch": 2.9568, "percentage": 59.23, "elapsed_time": "2:04:56", "remaining_time": "1:25:59"}
+{"current_steps": 232, "total_steps": 390, "loss": 0.3688, "lr": 1.6880151224263646e-05, "epoch": 2.9696, "percentage": 59.49, "elapsed_time": "2:05:25", "remaining_time": "1:25:25"}
+{"current_steps": 233, "total_steps": 390, "loss": 0.392, "lr": 1.6703461793876876e-05, "epoch": 2.9824, "percentage": 59.74, "elapsed_time": "2:05:59", "remaining_time": "1:24:53"}
+{"current_steps": 234, "total_steps": 390, "loss": 0.3634, "lr": 1.6527036446661396e-05, "epoch": 2.9952, "percentage": 60.0, "elapsed_time": "2:06:23", "remaining_time": "1:24:15"}
+{"current_steps": 235, "total_steps": 390, "loss": 0.5682, "lr": 1.635088931591775e-05, "epoch": 3.008, "percentage": 60.26, "elapsed_time": "2:07:06", "remaining_time": "1:23:50"}
+{"current_steps": 236, "total_steps": 390, "loss": 0.214, "lr": 1.6175034512658753e-05, "epoch": 3.0208, "percentage": 60.51, "elapsed_time": "2:07:33", "remaining_time": "1:23:14"}
+{"current_steps": 237, "total_steps": 390, "loss": 0.2233, "lr": 1.5999486124479115e-05, "epoch": 3.0336, "percentage": 60.77, "elapsed_time": "2:08:06", "remaining_time": "1:22:41"}
+{"current_steps": 238, "total_steps": 390, "loss": 0.2289, "lr": 1.5824258214426833e-05, "epoch": 3.0464, "percentage": 61.03, "elapsed_time": "2:08:36", "remaining_time": "1:22:08"}
+{"current_steps": 239, "total_steps": 390, "loss": 0.243, "lr": 1.5649364819876655e-05, "epoch": 3.0592, "percentage": 61.28, "elapsed_time": "2:09:03", "remaining_time": "1:21:32"}
+{"current_steps": 240, "total_steps": 390, "loss": 0.206, "lr": 1.547481995140556e-05, "epoch": 3.072, "percentage": 61.54, "elapsed_time": "2:09:32", "remaining_time": "1:20:57"}
+{"current_steps": 241, "total_steps": 390, "loss": 0.2407, "lr": 1.5300637591670357e-05, "epoch": 3.0848, "percentage": 61.79, "elapsed_time": "2:10:03", "remaining_time": "1:20:24"}
+{"current_steps": 242, "total_steps": 390, "loss": 0.2208, "lr": 1.5126831694287564e-05, "epoch": 3.0976, "percentage": 62.05, "elapsed_time": "2:10:34", "remaining_time": "1:19:51"}
+{"current_steps": 243, "total_steps": 390, "loss": 0.2146, "lr": 1.4953416182715566e-05, "epoch": 3.1104, "percentage": 62.31, "elapsed_time": "2:11:04", "remaining_time": "1:19:17"}
+{"current_steps": 244, "total_steps": 390, "loss": 0.2189, "lr": 1.478040494913926e-05, "epoch": 3.1232, "percentage": 62.56, "elapsed_time": "2:11:38", "remaining_time": "1:18:46"}
+{"current_steps": 245, "total_steps": 390, "loss": 0.2315, "lr": 1.460781185335713e-05, "epoch": 3.136, "percentage": 62.82, "elapsed_time": "2:12:11", "remaining_time": "1:18:14"}
+{"current_steps": 246, "total_steps": 390, "loss": 0.204, "lr": 1.443565072167095e-05, "epoch": 3.1488, "percentage": 63.08, "elapsed_time": "2:12:36", "remaining_time": "1:17:37"}
+{"current_steps": 247, "total_steps": 390, "loss": 0.2311, "lr": 1.4263935345778202e-05, "epoch": 3.1616, "percentage": 63.33, "elapsed_time": "2:13:08", "remaining_time": "1:17:04"}
+{"current_steps": 248, "total_steps": 390, "loss": 0.2111, "lr": 1.409267948166718e-05, "epoch": 3.1744, "percentage": 63.59, "elapsed_time": "2:13:34", "remaining_time": "1:16:29"}
+{"current_steps": 249, "total_steps": 390, "loss": 0.2256, "lr": 1.3921896848515064e-05, "epoch": 3.1872, "percentage": 63.85, "elapsed_time": "2:14:06", "remaining_time": "1:15:56"}
+{"current_steps": 250, "total_steps": 390, "loss": 0.2229, "lr": 1.3751601127588849e-05, "epoch": 3.2, "percentage": 64.1, "elapsed_time": "2:14:35", "remaining_time": "1:15:22"}
+{"current_steps": 251, "total_steps": 390, "loss": 0.2213, "lr": 1.3581805961149371e-05, "epoch": 3.2128, "percentage": 64.36, "elapsed_time": "2:15:07", "remaining_time": "1:14:49"}
+{"current_steps": 252, "total_steps": 390, "loss": 0.2144, "lr": 1.341252495135841e-05, "epoch": 3.2256, "percentage": 64.62, "elapsed_time": "2:15:38", "remaining_time": "1:14:16"}
+{"current_steps": 253, "total_steps": 390, "loss": 0.2259, "lr": 1.324377165918906e-05, "epoch": 3.2384, "percentage": 64.87, "elapsed_time": "2:16:03", "remaining_time": "1:13:40"}
+{"current_steps": 254, "total_steps": 390, "loss": 0.213, "lr": 1.3075559603339354e-05, "epoch": 3.2512, "percentage": 65.13, "elapsed_time": "2:16:33", "remaining_time": "1:13:06"}
+{"current_steps": 255, "total_steps": 390, "loss": 0.2083, "lr": 1.2907902259149287e-05, "epoch": 3.2640000000000002, "percentage": 65.38, "elapsed_time": "2:17:02", "remaining_time": "1:12:32"}
+{"current_steps": 256, "total_steps": 390, "loss": 0.2054, "lr": 1.274081305752135e-05, "epoch": 3.2768, "percentage": 65.64, "elapsed_time": "2:17:31", "remaining_time": "1:11:59"}
+{"current_steps": 257, "total_steps": 390, "loss": 0.2389, "lr": 1.2574305383844528e-05, "epoch": 3.2896, "percentage": 65.9, "elapsed_time": "2:18:07", "remaining_time": "1:11:28"}
+{"current_steps": 258, "total_steps": 390, "loss": 0.2106, "lr": 1.2408392576922075e-05, "epoch": 3.3024, "percentage": 66.15, "elapsed_time": "2:18:34", "remaining_time": "1:10:53"}
+{"current_steps": 259, "total_steps": 390, "loss": 0.21, "lr": 1.2243087927902905e-05, "epoch": 3.3152, "percentage": 66.41, "elapsed_time": "2:19:07", "remaining_time": "1:10:21"}
+{"current_steps": 260, "total_steps": 390, "loss": 0.2577, "lr": 1.2078404679216864e-05, "epoch": 3.328, "percentage": 66.67, "elapsed_time": "2:19:42", "remaining_time": "1:09:51"}
+{"current_steps": 261, "total_steps": 390, "loss": 0.1915, "lr": 1.1914356023513904e-05, "epoch": 3.3407999999999998, "percentage": 66.92, "elapsed_time": "2:20:10", "remaining_time": "1:09:16"}
+{"current_steps": 262, "total_steps": 390, "loss": 0.2122, "lr": 1.1750955102607193e-05, "epoch": 3.3536, "percentage": 67.18, "elapsed_time": "2:20:42", "remaining_time": "1:08:44"}
+{"current_steps": 263, "total_steps": 390, "loss": 0.2054, "lr": 1.1588215006420374e-05, "epoch": 3.3664, "percentage": 67.44, "elapsed_time": "2:21:14", "remaining_time": "1:08:12"}
+{"current_steps": 264, "total_steps": 390, "loss": 0.2133, "lr": 1.1426148771938915e-05, "epoch": 3.3792, "percentage": 67.69, "elapsed_time": "2:21:42", "remaining_time": "1:07:37"}
+{"current_steps": 265, "total_steps": 390, "loss": 0.2167, "lr": 1.1264769382165748e-05, "epoch": 3.392, "percentage": 67.95, "elapsed_time": "2:22:18", "remaining_time": "1:07:07"}
+{"current_steps": 266, "total_steps": 390, "loss": 0.2234, "lr": 1.110408976508118e-05, "epoch": 3.4048, "percentage": 68.21, "elapsed_time": "2:22:56", "remaining_time": "1:06:38"}
+{"current_steps": 267, "total_steps": 390, "loss": 0.2165, "lr": 1.094412279260726e-05, "epoch": 3.4176, "percentage": 68.46, "elapsed_time": "2:23:34", "remaining_time": "1:06:08"}
+{"current_steps": 268, "total_steps": 390, "loss": 0.1905, "lr": 1.0784881279576635e-05, "epoch": 3.4304, "percentage": 68.72, "elapsed_time": "2:24:03", "remaining_time": "1:05:34"}
+{"current_steps": 269, "total_steps": 390, "loss": 0.2352, "lr": 1.0626377982705929e-05, "epoch": 3.4432, "percentage": 68.97, "elapsed_time": "2:24:36", "remaining_time": "1:05:02"}
+{"current_steps": 270, "total_steps": 390, "loss": 0.206, "lr": 1.0468625599573842e-05, "epoch": 3.456, "percentage": 69.23, "elapsed_time": "2:25:05", "remaining_time": "1:04:28"}
+{"current_steps": 271, "total_steps": 390, "loss": 0.2147, "lr": 1.0311636767603952e-05, "epoch": 3.4688, "percentage": 69.49, "elapsed_time": "2:25:33", "remaining_time": "1:03:55"}
+{"current_steps": 272, "total_steps": 390, "loss": 0.2272, "lr": 1.0155424063052306e-05, "epoch": 3.4816, "percentage": 69.74, "elapsed_time": "2:26:03", "remaining_time": "1:03:21"}
+{"current_steps": 273, "total_steps": 390, "loss": 0.202, "lr": 1.0000000000000006e-05, "epoch": 3.4944, "percentage": 70.0, "elapsed_time": "2:26:32", "remaining_time": "1:02:48"}
+{"current_steps": 274, "total_steps": 390, "loss": 0.1878, "lr": 9.84537702935065e-06, "epoch": 3.5072, "percentage": 70.26, "elapsed_time": "2:27:03", "remaining_time": "1:02:15"}
+{"current_steps": 275, "total_steps": 390, "loss": 0.2058, "lr": 9.691567537832964e-06, "epoch": 3.52, "percentage": 70.51, "elapsed_time": "2:27:33", "remaining_time": "1:01:42"}
+{"current_steps": 276, "total_steps": 390, "loss": 0.2397, "lr": 9.538583847008452e-06, "epoch": 3.5328, "percentage": 70.77, "elapsed_time": "2:28:02", "remaining_time": "1:01:09"}
+{"current_steps": 277, "total_steps": 390, "loss": 0.2038, "lr": 9.386438212284372e-06, "epoch": 3.5456, "percentage": 71.03, "elapsed_time": "2:28:35", "remaining_time": "1:00:37"}
+{"current_steps": 278, "total_steps": 390, "loss": 0.2401, "lr": 9.235142821931928e-06, "epoch": 3.5584, "percentage": 71.28, "elapsed_time": "2:29:02", "remaining_time": "1:00:02"}
+{"current_steps": 279, "total_steps": 390, "loss": 0.2069, "lr": 9.084709796109907e-06, "epoch": 3.5712, "percentage": 71.54, "elapsed_time": "2:29:37", "remaining_time": "0:59:31"}
+{"current_steps": 280, "total_steps": 390, "loss": 0.2319, "lr": 8.93515118589373e-06, "epoch": 3.584, "percentage": 71.79, "elapsed_time": "2:30:11", "remaining_time": "0:59:00"}
+{"current_steps": 281, "total_steps": 390, "loss": 0.1986, "lr": 8.786478972310023e-06, "epoch": 3.5968, "percentage": 72.05, "elapsed_time": "2:30:41", "remaining_time": "0:58:27"}
+{"current_steps": 282, "total_steps": 390, "loss": 0.1838, "lr": 8.638705065376887e-06, "epoch": 3.6096, "percentage": 72.31, "elapsed_time": "2:31:11", "remaining_time": "0:57:54"}
+{"current_steps": 283, "total_steps": 390, "loss": 0.1885, "lr": 8.491841303149728e-06, "epoch": 3.6224, "percentage": 72.56, "elapsed_time": "2:31:37", "remaining_time": "0:57:19"}
+{"current_steps": 284, "total_steps": 390, "loss": 0.2378, "lr": 8.345899450772975e-06, "epoch": 3.6352, "percentage": 72.82, "elapsed_time": "2:32:10", "remaining_time": "0:56:47"}
+{"current_steps": 285, "total_steps": 390, "loss": 0.2177, "lr": 8.200891199537549e-06, "epoch": 3.648, "percentage": 73.08, "elapsed_time": "2:32:44", "remaining_time": "0:56:16"}
+{"current_steps": 286, "total_steps": 390, "loss": 0.2078, "lr": 8.056828165944282e-06, "epoch": 3.6608, "percentage": 73.33, "elapsed_time": "2:33:16", "remaining_time": "0:55:44"}
+{"current_steps": 287, "total_steps": 390, "loss": 0.2146, "lr": 7.913721890773354e-06, "epoch": 3.6736, "percentage": 73.59, "elapsed_time": "2:33:49", "remaining_time": "0:55:12"}
+{"current_steps": 288, "total_steps": 390, "loss": 0.2129, "lr": 7.771583838159756e-06, "epoch": 3.6864, "percentage": 73.85, "elapsed_time": "2:34:16", "remaining_time": "0:54:38"}
+{"current_steps": 289, "total_steps": 390, "loss": 0.2172, "lr": 7.630425394674903e-06, "epoch": 3.6992000000000003, "percentage": 74.1, "elapsed_time": "2:34:46", "remaining_time": "0:54:05"}
+{"current_steps": 290, "total_steps": 390, "loss": 0.2266, "lr": 7.49025786841445e-06, "epoch": 3.7119999999999997, "percentage": 74.36, "elapsed_time": "2:35:22", "remaining_time": "0:53:34"}
+{"current_steps": 291, "total_steps": 390, "loss": 0.2098, "lr": 7.3510924880924575e-06, "epoch": 3.7248, "percentage": 74.62, "elapsed_time": "2:35:51", "remaining_time": "0:53:01"}
+{"current_steps": 292, "total_steps": 390, "loss": 0.2154, "lr": 7.212940402141808e-06, "epoch": 3.7376, "percentage": 74.87, "elapsed_time": "2:36:24", "remaining_time": "0:52:29"}
+{"current_steps": 293, "total_steps": 390, "loss": 0.1979, "lr": 7.075812677821145e-06, "epoch": 3.7504, "percentage": 75.13, "elapsed_time": "2:36:50", "remaining_time": "0:51:55"}
+{"current_steps": 294, "total_steps": 390, "loss": 0.2229, "lr": 6.939720300328303e-06, "epoch": 3.7632, "percentage": 75.38, "elapsed_time": "2:37:22", "remaining_time": "0:51:23"}
+{"current_steps": 295, "total_steps": 390, "loss": 0.2182, "lr": 6.8046741719202385e-06, "epoch": 3.776, "percentage": 75.64, "elapsed_time": "2:37:56", "remaining_time": "0:50:51"}
+{"current_steps": 296, "total_steps": 390, "loss": 0.2112, "lr": 6.67068511103971e-06, "epoch": 3.7888, "percentage": 75.9, "elapsed_time": "2:38:30", "remaining_time": "0:50:20"}
+{"current_steps": 297, "total_steps": 390, "loss": 0.1985, "lr": 6.537763851448593e-06, "epoch": 3.8016, "percentage": 76.15, "elapsed_time": "2:38:57", "remaining_time": "0:49:46"}
+{"current_steps": 298, "total_steps": 390, "loss": 0.2085, "lr": 6.4059210413680175e-06, "epoch": 3.8144, "percentage": 76.41, "elapsed_time": "2:39:33", "remaining_time": "0:49:15"}
+{"current_steps": 299, "total_steps": 390, "loss": 0.2109, "lr": 6.275167242625331e-06, "epoch": 3.8272, "percentage": 76.67, "elapsed_time": "2:40:05", "remaining_time": "0:48:43"}
+{"current_steps": 300, "total_steps": 390, "loss": 0.219, "lr": 6.145512929808013e-06, "epoch": 3.84, "percentage": 76.92, "elapsed_time": "2:40:40", "remaining_time": "0:48:12"}
+{"current_steps": 301, "total_steps": 390, "loss": 0.2092, "lr": 6.016968489424572e-06, "epoch": 3.8528000000000002, "percentage": 77.18, "elapsed_time": "2:41:13", "remaining_time": "0:47:40"}
+{"current_steps": 302, "total_steps": 390, "loss": 0.1991, "lr": 5.889544219072465e-06, "epoch": 3.8656, "percentage": 77.44, "elapsed_time": "2:41:51", "remaining_time": "0:47:09"}
+{"current_steps": 303, "total_steps": 390, "loss": 0.2197, "lr": 5.7632503266131925e-06, "epoch": 3.8784, "percentage": 77.69, "elapsed_time": "2:42:25", "remaining_time": "0:46:38"}
+{"current_steps": 304, "total_steps": 390, "loss": 0.2147, "lr": 5.638096929354522e-06, "epoch": 3.8912, "percentage": 77.95, "elapsed_time": "2:42:50", "remaining_time": "0:46:03"}
+{"current_steps": 305, "total_steps": 390, "loss": 0.2193, "lr": 5.514094053240035e-06, "epoch": 3.904, "percentage": 78.21, "elapsed_time": "2:43:19", "remaining_time": "0:45:30"}
+{"current_steps": 306, "total_steps": 390, "loss": 0.2213, "lr": 5.39125163204594e-06, "epoch": 3.9168, "percentage": 78.46, "elapsed_time": "2:43:55", "remaining_time": "0:44:59"}
+{"current_steps": 307, "total_steps": 390, "loss": 0.1955, "lr": 5.269579506585259e-06, "epoch": 3.9295999999999998, "percentage": 78.72, "elapsed_time": "2:44:27", "remaining_time": "0:44:27"}
+{"current_steps": 308, "total_steps": 390, "loss": 0.2248, "lr": 5.149087423919541e-06, "epoch": 3.9424, "percentage": 78.97, "elapsed_time": "2:45:02", "remaining_time": "0:43:56"}
+{"current_steps": 309, "total_steps": 390, "loss": 0.2519, "lr": 5.029785036577976e-06, "epoch": 3.9552, "percentage": 79.23, "elapsed_time": "2:45:42", "remaining_time": "0:43:26"}
+{"current_steps": 310, "total_steps": 390, "loss": 0.2115, "lr": 4.911681901784198e-06, "epoch": 3.968, "percentage": 79.49, "elapsed_time": "2:46:15", "remaining_time": "0:42:54"}
+{"current_steps": 311, "total_steps": 390, "loss": 0.1975, "lr": 4.794787480690597e-06, "epoch": 3.9808, "percentage": 79.74, "elapsed_time": "2:46:43", "remaining_time": "0:42:21"}
+{"current_steps": 312, "total_steps": 390, "loss": 0.213, "lr": 4.679111137620442e-06, "epoch": 3.9936, "percentage": 80.0, "elapsed_time": "2:47:13", "remaining_time": "0:41:48"}
+{"current_steps": 313, "total_steps": 390, "loss": 0.3211, "lr": 4.5646621393177e-06, "epoch": 4.0064, "percentage": 80.26, "elapsed_time": "2:48:03", "remaining_time": "0:41:20"}
+{"current_steps": 314, "total_steps": 390, "loss": 0.1589, "lr": 4.451449654204685e-06, "epoch": 4.0192, "percentage": 80.51, "elapsed_time": "2:48:32", "remaining_time": "0:40:47"}
+{"current_steps": 315, "total_steps": 390, "loss": 0.1342, "lr": 4.339482751647557e-06, "epoch": 4.032, "percentage": 80.77, "elapsed_time": "2:49:04", "remaining_time": "0:40:15"}
+{"current_steps": 316, "total_steps": 390, "loss": 0.1318, "lr": 4.228770401229824e-06, "epoch": 4.0448, "percentage": 81.03, "elapsed_time": "2:49:36", "remaining_time": "0:39:42"}
+{"current_steps": 317, "total_steps": 390, "loss": 0.1161, "lr": 4.119321472033779e-06, "epoch": 4.0576, "percentage": 81.28, "elapsed_time": "2:50:00", "remaining_time": "0:39:09"}
+{"current_steps": 318, "total_steps": 390, "loss": 0.1389, "lr": 4.011144731929981e-06, "epoch": 4.0704, "percentage": 81.54, "elapsed_time": "2:50:37", "remaining_time": "0:38:37"}
+{"current_steps": 319, "total_steps": 390, "loss": 0.1316, "lr": 3.904248846874894e-06, "epoch": 4.0832, "percentage": 81.79, "elapsed_time": "2:51:03", "remaining_time": "0:38:04"}
+{"current_steps": 320, "total_steps": 390, "loss": 0.1364, "lr": 3.7986423802166705e-06, "epoch": 4.096, "percentage": 82.05, "elapsed_time": "2:51:39", "remaining_time": "0:37:32"}
+{"current_steps": 321, "total_steps": 390, "loss": 0.1232, "lr": 3.694333792009115e-06, "epoch": 4.1088, "percentage": 82.31, "elapsed_time": "2:52:09", "remaining_time": "0:37:00"}
+{"current_steps": 322, "total_steps": 390, "loss": 0.1202, "lr": 3.5913314383339937e-06, "epoch": 4.1216, "percentage": 82.56, "elapsed_time": "2:52:38", "remaining_time": "0:36:27"}
+{"current_steps": 323, "total_steps": 390, "loss": 0.138, "lr": 3.4896435706316e-06, "epoch": 4.1344, "percentage": 82.82, "elapsed_time": "2:53:11", "remaining_time": "0:35:55"}
+{"current_steps": 324, "total_steps": 390, "loss": 0.1416, "lr": 3.3892783350397675e-06, "epoch": 4.1472, "percentage": 83.08, "elapsed_time": "2:53:49", "remaining_time": "0:35:24"}
+{"current_steps": 325, "total_steps": 390, "loss": 0.1315, "lr": 3.290243771741275e-06, "epoch": 4.16, "percentage": 83.33, "elapsed_time": "2:54:22", "remaining_time": "0:34:52"}
+{"current_steps": 326, "total_steps": 390, "loss": 0.119, "lr": 3.1925478143197418e-06, "epoch": 4.1728, "percentage": 83.59, "elapsed_time": "2:54:51", "remaining_time": "0:34:19"}
+{"current_steps": 327, "total_steps": 390, "loss": 0.1151, "lr": 3.0961982891241083e-06, "epoch": 4.1856, "percentage": 83.85, "elapsed_time": "2:55:22", "remaining_time": "0:33:47"}
+{"current_steps": 328, "total_steps": 390, "loss": 0.1359, "lr": 3.001202914641628e-06, "epoch": 4.1984, "percentage": 84.1, "elapsed_time": "2:55:58", "remaining_time": "0:33:15"}
+{"current_steps": 329, "total_steps": 390, "loss": 0.1232, "lr": 2.907569300879596e-06, "epoch": 4.2112, "percentage": 84.36, "elapsed_time": "2:56:26", "remaining_time": "0:32:42"}
+{"current_steps": 330, "total_steps": 390, "loss": 0.1193, "lr": 2.815304948755664e-06, "epoch": 4.224, "percentage": 84.62, "elapsed_time": "2:57:00", "remaining_time": "0:32:11"}
+{"current_steps": 331, "total_steps": 390, "loss": 0.1152, "lr": 2.7244172494969978e-06, "epoch": 4.2368, "percentage": 84.87, "elapsed_time": "2:57:27", "remaining_time": "0:31:37"}
+{"current_steps": 332, "total_steps": 390, "loss": 0.1426, "lr": 2.6349134840481294e-06, "epoch": 4.2496, "percentage": 85.13, "elapsed_time": "2:57:59", "remaining_time": "0:31:05"}
+{"current_steps": 333, "total_steps": 390, "loss": 0.1387, "lr": 2.546800822487714e-06, "epoch": 4.2624, "percentage": 85.38, "elapsed_time": "2:58:33", "remaining_time": "0:30:33"}
+{"current_steps": 334, "total_steps": 390, "loss": 0.1144, "lr": 2.4600863234541338e-06, "epoch": 4.2752, "percentage": 85.64, "elapsed_time": "2:59:02", "remaining_time": "0:30:01"}
+{"current_steps": 335, "total_steps": 390, "loss": 0.1485, "lr": 2.374776933580025e-06, "epoch": 4.288, "percentage": 85.9, "elapsed_time": "2:59:35", "remaining_time": "0:29:29"}
+{"current_steps": 336, "total_steps": 390, "loss": 0.1147, "lr": 2.2908794869358044e-06, "epoch": 4.3008, "percentage": 86.15, "elapsed_time": "3:00:05", "remaining_time": "0:28:56"}
+{"current_steps": 337, "total_steps": 390, "loss": 0.1508, "lr": 2.2084007044821764e-06, "epoch": 4.3136, "percentage": 86.41, "elapsed_time": "3:00:37", "remaining_time": "0:28:24"}
+{"current_steps": 338, "total_steps": 390, "loss": 0.1169, "lr": 2.127347193531757e-06, "epoch": 4.3264, "percentage": 86.67, "elapsed_time": "3:01:10", "remaining_time": "0:27:52"}
+{"current_steps": 339, "total_steps": 390, "loss": 0.1327, "lr": 2.0477254472197237e-06, "epoch": 4.3392, "percentage": 86.92, "elapsed_time": "3:01:44", "remaining_time": "0:27:20"}
+{"current_steps": 340, "total_steps": 390, "loss": 0.1373, "lr": 1.96954184398368e-06, "epoch": 4.352, "percentage": 87.18, "elapsed_time": "3:02:14", "remaining_time": "0:26:47"}
+{"current_steps": 341, "total_steps": 390, "loss": 0.1291, "lr": 1.8928026470526917e-06, "epoch": 4.3648, "percentage": 87.44, "elapsed_time": "3:02:46", "remaining_time": "0:26:15"}
+{"current_steps": 342, "total_steps": 390, "loss": 0.1255, "lr": 1.817514003945524e-06, "epoch": 4.3776, "percentage": 87.69, "elapsed_time": "3:03:19", "remaining_time": "0:25:43"}
+{"current_steps": 343, "total_steps": 390, "loss": 0.1417, "lr": 1.743681945978184e-06, "epoch": 4.3904, "percentage": 87.95, "elapsed_time": "3:03:46", "remaining_time": "0:25:10"}
+{"current_steps": 344, "total_steps": 390, "loss": 0.1057, "lr": 1.6713123877807413e-06, "epoch": 4.4032, "percentage": 88.21, "elapsed_time": "3:04:16", "remaining_time": "0:24:38"}
+{"current_steps": 345, "total_steps": 390, "loss": 0.1333, "lr": 1.6004111268235156e-06, "epoch": 4.416, "percentage": 88.46, "elapsed_time": "3:04:42", "remaining_time": "0:24:05"}
+{"current_steps": 346, "total_steps": 390, "loss": 0.1175, "lr": 1.5309838429526714e-06, "epoch": 4.4288, "percentage": 88.72, "elapsed_time": "3:05:08", "remaining_time": "0:23:32"}
+{"current_steps": 347, "total_steps": 390, "loss": 0.1388, "lr": 1.4630360979351644e-06, "epoch": 4.4416, "percentage": 88.97, "elapsed_time": "3:05:40", "remaining_time": "0:23:00"}
+{"current_steps": 348, "total_steps": 390, "loss": 0.1314, "lr": 1.396573335013236e-06, "epoch": 4.4544, "percentage": 89.23, "elapsed_time": "3:06:12", "remaining_time": "0:22:28"}
+{"current_steps": 349, "total_steps": 390, "loss": 0.1383, "lr": 1.3316008784683265e-06, "epoch": 4.4672, "percentage": 89.49, "elapsed_time": "3:06:56", "remaining_time": "0:21:57"}
+{"current_steps": 350, "total_steps": 390, "loss": 0.1043, "lr": 1.2681239331945695e-06, "epoch": 4.48, "percentage": 89.74, "elapsed_time": "3:07:26", "remaining_time": "0:21:25"}
+{"current_steps": 351, "total_steps": 390, "loss": 0.1352, "lr": 1.2061475842818337e-06, "epoch": 4.4928, "percentage": 90.0, "elapsed_time": "3:07:57", "remaining_time": "0:20:53"}
+{"current_steps": 352, "total_steps": 390, "loss": 0.1309, "lr": 1.1456767966083393e-06, "epoch": 4.5056, "percentage": 90.26, "elapsed_time": "3:08:27", "remaining_time": "0:20:20"}
+{"current_steps": 353, "total_steps": 390, "loss": 0.1175, "lr": 1.086716414442952e-06, "epoch": 4.5184, "percentage": 90.51, "elapsed_time": "3:08:59", "remaining_time": "0:19:48"}
+{"current_steps": 354, "total_steps": 390, "loss": 0.1371, "lr": 1.0292711610570904e-06, "epoch": 4.5312, "percentage": 90.77, "elapsed_time": "3:09:26", "remaining_time": "0:19:15"}
+{"current_steps": 355, "total_steps": 390, "loss": 0.1351, "lr": 9.733456383463658e-07, "epoch": 4.5440000000000005, "percentage": 91.03, "elapsed_time": "3:10:00", "remaining_time": "0:18:43"}
+{"current_steps": 356, "total_steps": 390, "loss": 0.1229, "lr": 9.189443264619102e-07, "epoch": 4.5568, "percentage": 91.28, "elapsed_time": "3:10:34", "remaining_time": "0:18:12"}
+{"current_steps": 357, "total_steps": 390, "loss": 0.1188, "lr": 8.660715834514977e-07, "epoch": 4.5696, "percentage": 91.54, "elapsed_time": "3:11:00", "remaining_time": "0:17:39"}
+{"current_steps": 358, "total_steps": 390, "loss": 0.1283, "lr": 8.147316449103959e-07, "epoch": 4.5824, "percentage": 91.79, "elapsed_time": "3:11:30", "remaining_time": "0:17:07"}
+{"current_steps": 359, "total_steps": 390, "loss": 0.1147, "lr": 7.649286236420806e-07, "epoch": 4.5952, "percentage": 92.05, "elapsed_time": "3:11:58", "remaining_time": "0:16:34"}
+{"current_steps": 360, "total_steps": 390, "loss": 0.1139, "lr": 7.166665093287539e-07, "epoch": 4.608, "percentage": 92.31, "elapsed_time": "3:12:30", "remaining_time": "0:16:02"}
+{"current_steps": 361, "total_steps": 390, "loss": 0.1382, "lr": 6.69949168211721e-07, "epoch": 4.6208, "percentage": 92.56, "elapsed_time": "3:13:10", "remaining_time": "0:15:31"}
+{"current_steps": 362, "total_steps": 390, "loss": 0.1379, "lr": 6.247803427816945e-07, "epoch": 4.6336, "percentage": 92.82, "elapsed_time": "3:13:47", "remaining_time": "0:14:59"}
+{"current_steps": 363, "total_steps": 390, "loss": 0.0996, "lr": 5.811636514789598e-07, "epoch": 4.6464, "percentage": 93.08, "elapsed_time": "3:14:14", "remaining_time": "0:14:26"}
+{"current_steps": 364, "total_steps": 390, "loss": 0.1302, "lr": 5.391025884035239e-07, "epoch": 4.6592, "percentage": 93.33, "elapsed_time": "3:14:45", "remaining_time": "0:13:54"}
+{"current_steps": 365, "total_steps": 390, "loss": 0.1294, "lr": 4.986005230351954e-07, "epoch": 4.672, "percentage": 93.59, "elapsed_time": "3:15:14", "remaining_time": "0:13:22"}
+{"current_steps": 366, "total_steps": 390, "loss": 0.1322, "lr": 4.5966069996365993e-07, "epoch": 4.6848, "percentage": 93.85, "elapsed_time": "3:15:45", "remaining_time": "0:12:50"}
+{"current_steps": 367, "total_steps": 390, "loss": 0.105, "lr": 4.22286238628562e-07, "epoch": 4.6975999999999996, "percentage": 94.1, "elapsed_time": "3:16:10", "remaining_time": "0:12:17"}
+{"current_steps": 368, "total_steps": 390, "loss": 0.1076, "lr": 3.8648013306960664e-07, "epoch": 4.7104, "percentage": 94.36, "elapsed_time": "3:16:39", "remaining_time": "0:11:45"}
+{"current_steps": 369, "total_steps": 390, "loss": 0.1415, "lr": 3.522452516867048e-07, "epoch": 4.7232, "percentage": 94.62, "elapsed_time": "3:17:15", "remaining_time": "0:11:13"}
+{"current_steps": 370, "total_steps": 390, "loss": 0.1097, "lr": 3.1958433701019697e-07, "epoch": 4.736, "percentage": 94.87, "elapsed_time": "3:17:45", "remaining_time": "0:10:41"}
+{"current_steps": 371, "total_steps": 390, "loss": 0.1388, "lr": 2.8850000548115155e-07, "epoch": 4.7488, "percentage": 95.13, "elapsed_time": "3:18:16", "remaining_time": "0:10:09"}
+{"current_steps": 372, "total_steps": 390, "loss": 0.1244, "lr": 2.5899474724174313e-07, "epoch": 4.7616, "percentage": 95.38, "elapsed_time": "3:18:50", "remaining_time": "0:09:37"}
+{"current_steps": 373, "total_steps": 390, "loss": 0.1321, "lr": 2.3107092593579905e-07, "epoch": 4.7744, "percentage": 95.64, "elapsed_time": "3:19:16", "remaining_time": "0:09:04"}
+{"current_steps": 374, "total_steps": 390, "loss": 0.1317, "lr": 2.0473077851942858e-07, "epoch": 4.7872, "percentage": 95.9, "elapsed_time": "3:19:44", "remaining_time": "0:08:32"}
+{"current_steps": 375, "total_steps": 390, "loss": 0.1096, "lr": 1.799764150818306e-07, "epoch": 4.8, "percentage": 96.15, "elapsed_time": "3:20:06", "remaining_time": "0:08:00"}
+{"current_steps": 376, "total_steps": 390, "loss": 0.1252, "lr": 1.5680981867625566e-07, "epoch": 4.8128, "percentage": 96.41, "elapsed_time": "3:20:37", "remaining_time": "0:07:28"}
+{"current_steps": 377, "total_steps": 390, "loss": 0.1276, "lr": 1.3523284516113955e-07, "epoch": 4.8256, "percentage": 96.67, "elapsed_time": "3:21:05", "remaining_time": "0:06:56"}
+{"current_steps": 378, "total_steps": 390, "loss": 0.1165, "lr": 1.1524722305144231e-07, "epoch": 4.8384, "percentage": 96.92, "elapsed_time": "3:21:33", "remaining_time": "0:06:23"}
+{"current_steps": 379, "total_steps": 390, "loss": 0.1744, "lr": 9.685455338016347e-08, "epoch": 4.8512, "percentage": 97.18, "elapsed_time": "3:22:05", "remaining_time": "0:05:51"}
+{"current_steps": 380, "total_steps": 390, "loss": 0.1389, "lr": 8.005630957010014e-08, "epoch": 4.864, "percentage": 97.44, "elapsed_time": "3:22:35", "remaining_time": "0:05:19"}
+{"current_steps": 381, "total_steps": 390, "loss": 0.1246, "lr": 6.485383731580142e-08, "epoch": 4.8768, "percentage": 97.69, "elapsed_time": "3:23:10", "remaining_time": "0:04:47"}
+{"current_steps": 382, "total_steps": 390, "loss": 0.0861, "lr": 5.1248354475768034e-08, "epoch": 4.8896, "percentage": 97.95, "elapsed_time": "3:23:34", "remaining_time": "0:04:15"}
+{"current_steps": 383, "total_steps": 390, "loss": 0.1286, "lr": 3.924095097489922e-08, "epoch": 4.9024, "percentage": 98.21, "elapsed_time": "3:24:09", "remaining_time": "0:03:43"}
+{"current_steps": 384, "total_steps": 390, "loss": 0.131, "lr": 2.8832588717164766e-08, "epoch": 4.9152000000000005, "percentage": 98.46, "elapsed_time": "3:24:40", "remaining_time": "0:03:11"}
+{"current_steps": 385, "total_steps": 390, "loss": 0.1211, "lr": 2.0024101508555604e-08, "epoch": 4.928, "percentage": 98.72, "elapsed_time": "3:25:08", "remaining_time": "0:02:39"}
+{"current_steps": 386, "total_steps": 390, "loss": 0.1235, "lr": 1.281619499029274e-08, "epoch": 4.9408, "percentage": 98.97, "elapsed_time": "3:25:39", "remaining_time": "0:02:07"}
+{"current_steps": 387, "total_steps": 390, "loss": 0.1375, "lr": 7.209446582292501e-09, "epoch": 4.9536, "percentage": 99.23, "elapsed_time": "3:26:17", "remaining_time": "0:01:35"}
+{"current_steps": 388, "total_steps": 390, "loss": 0.1114, "lr": 3.2043054369057523e-09, "epoch": 4.9664, "percentage": 99.49, "elapsed_time": "3:26:45", "remaining_time": "0:01:03"}
+{"current_steps": 389, "total_steps": 390, "loss": 0.118, "lr": 8.010924029533406e-10, "epoch": 4.9792, "percentage": 99.74, "elapsed_time": "3:27:12", "remaining_time": "0:00:31"}
+{"current_steps": 390, "total_steps": 390, "loss": 0.1122, "lr": 0.0, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "3:27:43", "remaining_time": "0:00:00"}
+{"current_steps": 390, "total_steps": 390, "epoch": 4.992, "percentage": 100.0, "elapsed_time": "3:28:14", "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": 6.743006231739799,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.1003,
+      "step": 1
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 7.066286014281838,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.1344,
+      "step": 2
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.543734967748884,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0523,
+      "step": 3
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 6.185415104515718,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.0893,
+      "step": 4
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.582062974941517,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.0351,
+      "step": 5
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.3568103459171583,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.0068,
+      "step": 6
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.5486373319600255,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.988,
+      "step": 7
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 5.769645380344031,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.0507,
+      "step": 8
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 5.6323306570246325,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.0128,
+      "step": 9
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 5.034537420737593,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 1.0002,
+      "step": 10
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 4.556789451306426,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.9686,
+      "step": 11
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 3.229368642334696,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.9299,
+      "step": 12
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 3.1268213512826746,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8971,
+      "step": 13
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.951138666945752,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.8866,
+      "step": 14
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.41084492250629,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.9038,
+      "step": 15
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.060994232629589,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.8545,
+      "step": 16
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.6912470046476213,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.8533,
+      "step": 17
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.5308450851167237,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.8216,
+      "step": 18
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.385196954892433,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.8487,
+      "step": 19
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.2589226982770487,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.8003,
+      "step": 20
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.2997833638343748,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.8246,
+      "step": 21
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.0461089510811301,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.8272,
+      "step": 22
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 1.0623497495645282,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.7995,
+      "step": 23
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 1.0704270544562335,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.8149,
+      "step": 24
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8403873554483027,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.8033,
+      "step": 25
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 1.0269434671234854,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.8181,
+      "step": 26
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.8842728867166108,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.7709,
+      "step": 27
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.9042028317551858,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.7811,
+      "step": 28
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.8407412456402528,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.7508,
+      "step": 29
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.8157749681184576,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.7576,
+      "step": 30
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.9468804541268353,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.7964,
+      "step": 31
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 1.0448319108433777,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.7766,
+      "step": 32
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.7882626770790413,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.7737,
+      "step": 33
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.8958010350698487,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.8219,
+      "step": 34
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.4119578800425963,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.8048,
+      "step": 35
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.8316426603132697,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.7867,
+      "step": 36
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 1.336306206226515,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.7734,
+      "step": 37
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 1.0920004406942043,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.7919,
+      "step": 38
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.8324477174992415,
+      "learning_rate": 4e-05,
+      "loss": 0.7586,
+      "step": 39
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.926352188723838,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.7869,
+      "step": 40
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.8488036255028663,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.7541,
+      "step": 41
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.7884338459657835,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.7729,
+      "step": 42
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.7660390848779673,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.7462,
+      "step": 43
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.8459027435862373,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.7632,
+      "step": 44
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.8940377382439754,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.7893,
+      "step": 45
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.7691283272808419,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.737,
+      "step": 46
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.7921502079565289,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.7661,
+      "step": 47
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.9143606168163722,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.7558,
+      "step": 48
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.9877743040894182,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.7851,
+      "step": 49
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9395935087181746,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.7523,
+      "step": 50
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 1.1034303852147378,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.7434,
+      "step": 51
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.7098514072807728,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.7843,
+      "step": 52
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 1.0931874133665274,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.7592,
+      "step": 53
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.9828066396874403,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.7482,
+      "step": 54
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.6682720185189923,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.7626,
+      "step": 55
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.7623480077792901,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.751,
+      "step": 56
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.7968316892104372,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.7753,
+      "step": 57
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.5762932932825403,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.7225,
+      "step": 58
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.7952251676163752,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.7754,
+      "step": 59
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.6249399230071104,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.7447,
+      "step": 60
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.8465014898118267,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.7656,
+      "step": 61
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.6536819865801515,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.7546,
+      "step": 62
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.7728254294403446,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.7738,
+      "step": 63
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.6941269136417265,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.7183,
+      "step": 64
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.8916566066459013,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7598,
+      "step": 65
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.6562173051774739,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.749,
+      "step": 66
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.8602294775597902,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.7691,
+      "step": 67
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.7992151009323316,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.7581,
+      "step": 68
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 1.111188791424757,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.7587,
+      "step": 69
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.6333849705792798,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.7545,
+      "step": 70
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 1.0937257594146246,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.7284,
+      "step": 71
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.657569069134253,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.77,
+      "step": 72
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.8702144228865949,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.7414,
+      "step": 73
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.7687916347429037,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.7621,
+      "step": 74
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.671104158214696,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.7438,
+      "step": 75
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.9382269183222769,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.7025,
+      "step": 76
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.6033043618678664,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.7271,
+      "step": 77
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.897003755370354,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7361,
+      "step": 78
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 1.8771881274834503,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 1.3899,
+      "step": 79
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.907363205256343,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.5809,
+      "step": 80
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.940232637744297,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.6123,
+      "step": 81
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.8930763483311664,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.6127,
+      "step": 82
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 1.507694121126766,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.6589,
+      "step": 83
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.7755213064005223,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.596,
+      "step": 84
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 1.2871383748816294,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.59,
+      "step": 85
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.8175075701162106,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.5864,
+      "step": 86
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.8218461363219056,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.6038,
+      "step": 87
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.8418104777895835,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.5875,
+      "step": 88
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.8962853320304269,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.651,
+      "step": 89
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.7989348132161029,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.5769,
+      "step": 90
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.7088816222868426,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5703,
+      "step": 91
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.7933044148537098,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.612,
+      "step": 92
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.7204524029756719,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.5846,
+      "step": 93
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.6533305458934046,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.5907,
+      "step": 94
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.7539358884159101,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.5919,
+      "step": 95
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.6508422039152885,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.6066,
+      "step": 96
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.7974059960507492,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.5653,
+      "step": 97
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.7914453037246021,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.5978,
+      "step": 98
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.5902327458916147,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.545,
+      "step": 99
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8491865230527035,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.589,
+      "step": 100
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.6508850179168086,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.5913,
+      "step": 101
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.6195750115582437,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.5825,
+      "step": 102
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.6210267450632905,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.5649,
+      "step": 103
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.6195321419367635,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5617,
+      "step": 104
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.6664655520277717,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.6093,
+      "step": 105
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.6724564262285763,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.5517,
+      "step": 106
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.7760308995461608,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.6685,
+      "step": 107
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.8218867430785521,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.558,
+      "step": 108
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.7574410585643446,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.6059,
+      "step": 109
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.7539813169545178,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.6013,
+      "step": 110
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.6089110994269326,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.5439,
+      "step": 111
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.6928109289974452,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.6107,
+      "step": 112
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.6164707103669036,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.533,
+      "step": 113
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.619207292989488,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.5659,
+      "step": 114
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.6351887384287722,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.5995,
+      "step": 115
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.6246030634486776,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.5881,
+      "step": 116
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.5957184370507052,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5846,
+      "step": 117
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.6369490928018712,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.6118,
+      "step": 118
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.6494942161963383,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.5948,
+      "step": 119
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.6765364072530382,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.6025,
+      "step": 120
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.7107319821307428,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.5763,
+      "step": 121
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.5918446868458465,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.5707,
+      "step": 122
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.6455256575805056,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.5983,
+      "step": 123
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.5605150476130768,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.5747,
+      "step": 124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.594883992048509,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.5849,
+      "step": 125
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.662261932659435,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.6005,
+      "step": 126
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.5722295160795337,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.583,
+      "step": 127
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.6157569357554913,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.5885,
+      "step": 128
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.5904010232809053,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.5849,
+      "step": 129
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.6475524668269238,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5509,
+      "step": 130
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.6006178341941506,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.5781,
+      "step": 131
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.671130758174799,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.5958,
+      "step": 132
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.7819794686641027,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.6211,
+      "step": 133
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.5956306386437098,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.5874,
+      "step": 134
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.7957925424808077,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.5838,
+      "step": 135
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.6085084247177972,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.5913,
+      "step": 136
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.6252461504282083,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.5457,
+      "step": 137
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.6360224703384407,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.5893,
+      "step": 138
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.5609394341072801,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.5857,
+      "step": 139
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.5597885839381709,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.5679,
+      "step": 140
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.5654056984793283,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.5549,
+      "step": 141
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.5353166027031147,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.5919,
+      "step": 142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.5673649604383674,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5513,
+      "step": 143
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.5568470979222744,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.6127,
+      "step": 144
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.6216818580700804,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.5912,
+      "step": 145
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.5721417318565675,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.6135,
+      "step": 146
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.6615855063383611,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.6091,
+      "step": 147
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.5395153603818907,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.5842,
+      "step": 148
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.6408686398302472,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.5895,
+      "step": 149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5889563874637227,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.5853,
+      "step": 150
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.5156351709913245,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.5877,
+      "step": 151
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.5583616485524463,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.5679,
+      "step": 152
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.5539646170731631,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.6024,
+      "step": 153
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.6478719841524553,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.5926,
+      "step": 154
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.5564712641538095,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.5512,
+      "step": 155
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.6098587876755319,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5713,
+      "step": 156
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 1.6360603508559461,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 1.0221,
+      "step": 157
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.844315850579978,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.3996,
+      "step": 158
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 2.177605398314913,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.4314,
+      "step": 159
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.8303057826140725,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.4054,
+      "step": 160
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 1.1569565232191912,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.3994,
+      "step": 161
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.8242245471099799,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.3982,
+      "step": 162
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.7034044545384213,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.3753,
+      "step": 163
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.7468137668014582,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.4019,
+      "step": 164
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.7517716901503451,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.401,
+      "step": 165
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.6982235176553026,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.3854,
+      "step": 166
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.6400957668045134,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.3867,
+      "step": 167
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.705563669363062,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.3844,
+      "step": 168
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.6284567085755858,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3738,
+      "step": 169
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.6744209614158668,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.3821,
+      "step": 170
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.6979614990180333,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.377,
+      "step": 171
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.670547945119299,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.3559,
+      "step": 172
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.6281958157845674,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.4095,
+      "step": 173
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.6582267965689705,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.3796,
+      "step": 174
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6637685940523228,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.3837,
+      "step": 175
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.6527949439216556,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.3881,
+      "step": 176
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.6788814167344014,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.373,
+      "step": 177
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.6818863973058077,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.3681,
+      "step": 178
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.681200434346096,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.3934,
+      "step": 179
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.6453945980118074,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.3484,
+      "step": 180
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.7100914038130671,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.4197,
+      "step": 181
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.6234442007153514,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3731,
+      "step": 182
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.5590667449119111,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.377,
+      "step": 183
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.6311500054971748,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.38,
+      "step": 184
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.5197607018582181,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.4006,
+      "step": 185
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.5106539014684699,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.3329,
+      "step": 186
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.5924312364064122,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.387,
+      "step": 187
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.571324244177717,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.3828,
+      "step": 188
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.589281934993984,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.3874,
+      "step": 189
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.5271593205315481,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.3864,
+      "step": 190
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.6322216610338603,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.3824,
+      "step": 191
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.6066288351705504,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.3884,
+      "step": 192
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.5718959522094771,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.3852,
+      "step": 193
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.6608915863373546,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.3972,
+      "step": 194
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.4933212354635612,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3618,
+      "step": 195
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.5346243499647223,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.3939,
+      "step": 196
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.4804380842427941,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.3479,
+      "step": 197
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.4786440319184376,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.3703,
+      "step": 198
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.5081730950084173,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.3996,
+      "step": 199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.4863387669882436,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.3772,
+      "step": 200
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.4585598944887773,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.3652,
+      "step": 201
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.46908195967328453,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.3738,
+      "step": 202
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.47868045783941926,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.3987,
+      "step": 203
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.4890016383050143,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.362,
+      "step": 204
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.4981270526528196,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.3858,
+      "step": 205
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.5483243106048328,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.3898,
+      "step": 206
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.47404834053941164,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.3769,
+      "step": 207
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.5475057821595155,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3803,
+      "step": 208
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.49706261114284817,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.3834,
+      "step": 209
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.5796670143141532,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.3976,
+      "step": 210
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.4530782176566248,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.3553,
+      "step": 211
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.5325770424086235,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.3911,
+      "step": 212
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.4679734004445458,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.3771,
+      "step": 213
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.4685529885675748,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.371,
+      "step": 214
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.47232100341684985,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.3835,
+      "step": 215
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.46822632920098184,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.378,
+      "step": 216
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.47839667947352366,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.3885,
+      "step": 217
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.4645311550420786,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.373,
+      "step": 218
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.43135939445857385,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.3717,
+      "step": 219
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.4674798986428613,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.3744,
+      "step": 220
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.5040146879951358,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4074,
+      "step": 221
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.4595163087112345,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.4097,
+      "step": 222
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.4662977257148966,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.385,
+      "step": 223
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.47411238425927493,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.4015,
+      "step": 224
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4876533414692807,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.3673,
+      "step": 225
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.4672217677139193,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.3939,
+      "step": 226
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.5825431531781076,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.4079,
+      "step": 227
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.4247186703468168,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.3558,
+      "step": 228
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.46615689942131117,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.3774,
+      "step": 229
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.42946239153200483,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.3751,
+      "step": 230
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.4752383842164273,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.398,
+      "step": 231
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.4411936381497578,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.3688,
+      "step": 232
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.42760972370688155,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.392,
+      "step": 233
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.4525407628538123,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3634,
+      "step": 234
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 1.4177139618970116,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.5682,
+      "step": 235
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.6925085800018793,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.214,
+      "step": 236
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.47177896225343874,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.2233,
+      "step": 237
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.9195986135605155,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.2289,
+      "step": 238
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.9763397031801505,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.243,
+      "step": 239
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.4901086617633099,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.206,
+      "step": 240
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.6475683122770262,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.2407,
+      "step": 241
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.6873161453472533,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.2208,
+      "step": 242
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.5706446658942862,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.2146,
+      "step": 243
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.4854726342930153,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.2189,
+      "step": 244
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.5542526554905188,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.2315,
+      "step": 245
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.5803953763775979,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.204,
+      "step": 246
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.6195239135124986,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2311,
+      "step": 247
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.496000653739078,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.2111,
+      "step": 248
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.48230141394722,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.2256,
+      "step": 249
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.49374512839849205,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.2229,
+      "step": 250
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.4413939159306726,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.2213,
+      "step": 251
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.4566325569264846,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.2144,
+      "step": 252
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.454463055182607,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.2259,
+      "step": 253
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.45013586999773686,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.213,
+      "step": 254
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.4648885493853386,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.2083,
+      "step": 255
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.4714390232083337,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.2054,
+      "step": 256
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.4647603965979526,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.2389,
+      "step": 257
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.4344248129572138,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.2106,
+      "step": 258
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.45809728424572677,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.21,
+      "step": 259
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.4752245096183222,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2577,
+      "step": 260
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.4400282968628508,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.1915,
+      "step": 261
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.4611980618891619,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.2122,
+      "step": 262
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.41614703493642785,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.2054,
+      "step": 263
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.46161255039575166,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.2133,
+      "step": 264
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.4714992002753008,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.2167,
+      "step": 265
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.44481976643296856,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.2234,
+      "step": 266
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.4198371195512831,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.2165,
+      "step": 267
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.4437172019256421,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.1905,
+      "step": 268
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.4709696188583375,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.2352,
+      "step": 269
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.38894349869277145,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.206,
+      "step": 270
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.43718838213894007,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.2147,
+      "step": 271
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.4461692155707171,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.2272,
+      "step": 272
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.402592295437286,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.202,
+      "step": 273
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.39384261611649546,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.1878,
+      "step": 274
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.41422600777590796,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.2058,
+      "step": 275
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.4439390165906728,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.2397,
+      "step": 276
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.4031486863365962,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.2038,
+      "step": 277
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.7147750966976936,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.2401,
+      "step": 278
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.4003166986954898,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.2069,
+      "step": 279
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.42348274815871273,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.2319,
+      "step": 280
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.40999530293219816,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.1986,
+      "step": 281
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.38609378609685124,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.1838,
+      "step": 282
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.37108947957292604,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.1885,
+      "step": 283
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.4150399899467208,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.2378,
+      "step": 284
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.3986819830927738,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.2177,
+      "step": 285
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.4254688138180027,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2078,
+      "step": 286
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.3954538549677782,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.2146,
+      "step": 287
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.4038148287692982,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.2129,
+      "step": 288
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.4048053050314327,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.2172,
+      "step": 289
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.41653866927378147,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.2266,
+      "step": 290
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.4212285780098467,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.2098,
+      "step": 291
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.40360029339874726,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.2154,
+      "step": 292
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.39176031323013305,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.1979,
+      "step": 293
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.42267810769634195,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.2229,
+      "step": 294
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.3917595112976641,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.2182,
+      "step": 295
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.3821745046473022,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.2112,
+      "step": 296
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.39665391314289067,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.1985,
+      "step": 297
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.400874736229629,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.2085,
+      "step": 298
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.3985352271461248,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2109,
+      "step": 299
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3920132547502682,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.219,
+      "step": 300
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.4180980895330953,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.2092,
+      "step": 301
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.3789531909907328,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.1991,
+      "step": 302
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.40812199028199514,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.2197,
+      "step": 303
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.40060267236946917,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.2147,
+      "step": 304
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.41032206046684666,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.2193,
+      "step": 305
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.398157330900452,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.2213,
+      "step": 306
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.35423101855129724,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.1955,
+      "step": 307
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.396728316281466,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.2248,
+      "step": 308
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.39836211069827154,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.2519,
+      "step": 309
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.36733985532484675,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.2115,
+      "step": 310
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.3603200874699631,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.1975,
+      "step": 311
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.38770712770963794,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.213,
+      "step": 312
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 1.072034754235289,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.3211,
+      "step": 313
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.6394912724198267,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.1589,
+      "step": 314
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.49997281078511197,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.1342,
+      "step": 315
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.40893610951063547,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.1318,
+      "step": 316
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.34138336309330364,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.1161,
+      "step": 317
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.41908803395125693,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.1389,
+      "step": 318
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.44663662995763065,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.1316,
+      "step": 319
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.5359505822842496,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.1364,
+      "step": 320
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.47817526273731126,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.1232,
+      "step": 321
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.4210004785129665,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.1202,
+      "step": 322
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.44191997573689457,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.138,
+      "step": 323
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.38935195899174607,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.1416,
+      "step": 324
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.36520114071871523,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1315,
+      "step": 325
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.3354853117344792,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.119,
+      "step": 326
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.3561785138404257,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.1151,
+      "step": 327
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.3668451616935275,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.1359,
+      "step": 328
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.36324608187244656,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.1232,
+      "step": 329
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.3422841292949907,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.1193,
+      "step": 330
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.34359701372833973,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.1152,
+      "step": 331
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.35372436751185904,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.1426,
+      "step": 332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.34512898590726343,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.1387,
+      "step": 333
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.31051516944180213,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.1144,
+      "step": 334
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.37044519269634696,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.1485,
+      "step": 335
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.32295564558688744,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.1147,
+      "step": 336
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.355009655758483,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.1508,
+      "step": 337
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.3221882214241443,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1169,
+      "step": 338
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.3664741225985812,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.1327,
+      "step": 339
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.35678234543817344,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.1373,
+      "step": 340
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.32451399470512954,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.1291,
+      "step": 341
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.3307494465400305,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.1255,
+      "step": 342
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.3475294950385137,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.1417,
+      "step": 343
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.3101664364832486,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.1057,
+      "step": 344
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.3353207317042542,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.1333,
+      "step": 345
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.33385957933824684,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.1175,
+      "step": 346
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.3444353012345218,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.1388,
+      "step": 347
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.3287399358054545,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.1314,
+      "step": 348
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.3294039561981482,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.1383,
+      "step": 349
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.3104503581687655,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.1043,
+      "step": 350
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.3462247785261336,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1352,
+      "step": 351
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.3327974507931004,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.1309,
+      "step": 352
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.30930647306060244,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.1175,
+      "step": 353
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.3335253186073523,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.1371,
+      "step": 354
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.3313582392385476,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.1351,
+      "step": 355
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.30606212930146565,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.1229,
+      "step": 356
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.3290780507607063,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.1188,
+      "step": 357
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.33344821500866495,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.1283,
+      "step": 358
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.3168113373308338,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.1147,
+      "step": 359
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.3255011918336976,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.1139,
+      "step": 360
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.34397046460345937,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.1382,
+      "step": 361
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.33503574592858115,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.1379,
+      "step": 362
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.289760702079406,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.0996,
+      "step": 363
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.3383072364750368,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1302,
+      "step": 364
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.3451561371037132,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.1294,
+      "step": 365
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.3497736067051506,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.1322,
+      "step": 366
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.3127099192907463,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.105,
+      "step": 367
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.31779656453817845,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.1076,
+      "step": 368
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.3450591969260004,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.1415,
+      "step": 369
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.29266572844817823,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.1097,
+      "step": 370
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.33225273193579985,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.1388,
+      "step": 371
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.3299428434253753,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.1244,
+      "step": 372
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.3380340225556441,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.1321,
+      "step": 373
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.3279798716332824,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.1317,
+      "step": 374
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3209590648883009,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.1096,
+      "step": 375
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.3221077195107284,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.1252,
+      "step": 376
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.33153730514168395,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1276,
+      "step": 377
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.31665557445843256,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.1165,
+      "step": 378
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.3851601916353107,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.1744,
+      "step": 379
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.32566731341703226,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.1389,
+      "step": 380
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.3334744365055278,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.1246,
+      "step": 381
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.2865317871862437,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.0861,
+      "step": 382
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.32529973656028704,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.1286,
+      "step": 383
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.3307741400659206,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.131,
+      "step": 384
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.3146121058990516,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.1211,
+      "step": 385
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.32441693670590377,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.1235,
+      "step": 386
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.3429018065496982,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.1375,
+      "step": 387
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.31757890825952534,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.1114,
+      "step": 388
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.3351016325671888,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.118,
+      "step": 389
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.3213164821733735,
+      "learning_rate": 0.0,
+      "loss": 0.1122,
+      "step": 390
+    },
+    {
+      "epoch": 4.992,
+      "step": 390,
+      "total_flos": 5.5014580010614784e+17,
+      "train_loss": 0.43123698242199726,
+      "train_runtime": 12496.7256,
+      "train_samples_per_second": 4.001,
+      "train_steps_per_second": 0.031
+    }
+  ],
+  "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": 5.5014580010614784e+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 sha256:40f87001334c54a85298ca510aaebaa07eef8f91a471f6754900cb7f62f3f696
+size 7352