Upload folder using huggingface_hub

.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 +1,60 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen2.5-72B-Instruct
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: DCFT-Stratos-Verified-114k-qwen-2.5-72B
+  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. -->
+
+# DCFT-Stratos-Verified-114k-qwen-2.5-72B
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-72B-Instruct](https://huggingface.co/Qwen/Qwen2.5-72B-Instruct) on the mlfoundations-dev/stratos_verified_mix 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: 1e-05
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 1024
+- total_train_batch_size: 1024
+- total_eval_batch_size: 8192
+- 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: 3.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.46.1
+- Pytorch 2.5.0a0+b465a5843b.nv24.09
+- Datasets 3.0.2
+- 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": 3.0,
+    "total_flos": 2.982866014384947e+16,
+    "train_loss": 0.3923433954339652,
+    "train_runtime": 48299.7361,
+    "train_samples_per_second": 7.078,
+    "train_steps_per_second": 0.007
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "/leonardo_work/EUHPC_E03_068/marianna/models/Qwen/Qwen2.5-72B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 8192,
+  "initializer_range": 0.02,
+  "intermediate_size": 29568,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 64,
+  "num_hidden_layers": 80,
+  "num_key_value_heads": 8,
+  "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-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9bdb27cc612a16cc63a39b862c63f6b5fb45343d8b342b4366f91b4d4b3bdf2
+size 4548798728

model-00002-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68ff5818be595ff1eff848bc670348f0511c09e7d352aa6301a4a6ac88230200
+size 4964101384

model-00003-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db5d96f741d70485a683d6f666ced58c70e70f662148f65e1f7def7147c17ea1
+size 4781637328

model-00004-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ccd56bf1f72713ca8802bf8254d98b0f3a541b1de3422ae60d86a5539ba145e
+size 4781670320

model-00005-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dda77ac89e4e46ee7b7f10cb78bb18d99b4cc610c7d9545887ce383e888a629e
+size 4781670360

model-00006-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99ebc8cecde4d02dbb745f3d18f2ac8d1b83d4ddeb1205c74c8b405618f6fca7
+size 4964101416

model-00007-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:487b76a2263823462f7bd6d5b0c1838d1c893d2998fb0f943ce41d0c2c3859a8
+size 4781637360

model-00008-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f8bc0a2e675299ceaee7b696020d1363ea9ca4074c408dab6e3c0f9f12f321e
+size 4781670360

model-00009-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:327f43ac6461cf87a950b5473e06ed6a4a3691f4a2c9e49b37cb2145dab94689
+size 4781670360

model-00010-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd5fbe672e2301863490bf5a6fdbc3899f8c009c7d93c7601306c7248b080853
+size 4964101416

model-00011-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce5e57aa90e5f5c11fcbf911fedb3d835ec56e92f20a582637d48819143712ab
+size 4781637360

model-00012-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a2b052e6e4b2c61e2bab59da439a8840fdb7089c4c066c3cc2b6b61f28c67fd
+size 4781670360

model-00013-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27ef49de075d6346f9de00e9ba1973a7eb82a26465ea0c1eaa2e467666f718c8
+size 4781670360

model-00014-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dc6c94886cf1660538dc1672a760cb2d5176f9d9f4107acf04a7cf36c2ffb63
+size 4964101416

model-00015-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd8812142bd2190095d4042fd377b8fef36eeab88d0e6f09ef37fe8c97fd940
+size 4781637360

model-00016-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ababc50b0e59f270fd79492ceabaf1fd3a6ce031f13183d2d241795b42f7576
+size 4781670360

model-00017-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e26dec7e8d6e385c4f02ee7d6325393959b88537f5d44879e81f55a05f75a60e
+size 4781670360

model-00018-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc9a1de0afd16bf75dd4064e21e2961d4c07a846572b4d4a00b7380862e19da5
+size 4964101416

model-00019-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:163e2633f049d7d033536eb652593d8e2f6a7c161a3dcd9cc5f0ca2cd532d0f5
+size 4781637360

model-00020-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4238250cbc4e86b30d05acc439febec1017daa148740c31161ac6c7ee812fc06
+size 4781670360

model-00021-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbbd57d3d15f079461caa303f816ccc476d9a4557ffa07e9344bc3fbd9e1e37d
+size 4781670360

model-00022-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b4417b6e72ed238b92c743bec942acdb25dc44cfd20d2d0629fca5e062dca66
+size 4964101416

model-00023-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c812f168e7a2e993ecc4c451a8c8fe8f8d04423b33f751a536d77065c3284633
+size 4781637360

model-00024-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53636d881515eadf29ae38e6d17cb4b640589058297b45ee85e102fce522f790
+size 4781670360

model-00025-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e9a2117494921569a5996e6e8dc8ce144f5ec7016b0d3dacbc6e5cd925cdfc5
+size 4781670360

model-00026-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1560bfc35768fcf3772d098e420a92eaa964a7cacbef2bbce5cad37bc38ceb9
+size 4964101416

model-00027-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b11de21ef0d97a96a0b45c11fca8216546724ef20bff4515b590aa7ae1a3d7d5
+size 4781637360

model-00028-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc494cdd4685054cb2474502d02257ca6abc94ec059da9ee2144a34e46c48e13
+size 4781670360

model-00029-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67ef57cf06c8d0f05ce1ebdfc5b092c2c29715a74073d3c56ea356ee0b69fe76
+size 4781670360

model-00030-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54a744276a59b6467b2c5e3ae20f3034b36365c828f8d159d6584316b0a1c335
+size 3208747032

model-00031-of-00031.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:095b512ad81a5199c2fe1ce203fb840a34d9051d9d983ba0b80ea9980ca022a1
+size 2491416704

model.safetensors.index.json

@@ -0,0 +1,970 @@
+{
+  "metadata": {
+    "total_size": 145412407296
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00031-of-00031.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00031.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00002-of-00031.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00031.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00031.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00031.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00031.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00031.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00006-of-00031.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00006-of-00031.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00005-of-00031.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00006-of-00031.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00006-of-00031.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00006-of-00031.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00008-of-00031.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00008-of-00031.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00007-of-00031.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00007-of-00031.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00002-of-00031.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00009-of-00031.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00009-of-00031.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00008-of-00031.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00008-of-00031.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00010-of-00031.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00010-of-00031.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00009-of-00031.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00009-of-00031.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00010-of-00031.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00010-of-00031.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00010-of-00031.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00012-of-00031.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00012-of-00031.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00011-of-00031.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00011-of-00031.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00002-of-00031.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00031.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00013-of-00031.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00013-of-00031.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00012-of-00031.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00012-of-00031.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00014-of-00031.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00014-of-00031.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00013-of-00031.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00013-of-00031.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00014-of-00031.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.input_layernorm.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.mlp.down_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.mlp.gate_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.mlp.up_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.post_attention_layernorm.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.k_proj.bias": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.k_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.o_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.q_proj.bias": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.q_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.v_proj.bias": "model-00014-of-00031.safetensors",
+    "model.layers.36.self_attn.v_proj.weight": "model-00014-of-00031.safetensors",
+    "model.layers.37.input_layernorm.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.mlp.down_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.mlp.gate_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.mlp.up_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.post_attention_layernorm.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.k_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.k_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.o_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.q_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.q_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.v_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.37.self_attn.v_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.input_layernorm.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.mlp.down_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.mlp.gate_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.mlp.up_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.post_attention_layernorm.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.k_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.k_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.o_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.q_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.q_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.v_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.38.self_attn.v_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.39.input_layernorm.weight": "model-00016-of-00031.safetensors",
+    "model.layers.39.mlp.down_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.39.mlp.gate_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.39.mlp.up_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.39.post_attention_layernorm.weight": "model-00016-of-00031.safetensors",
+    "model.layers.39.self_attn.k_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.39.self_attn.k_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.39.self_attn.o_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.39.self_attn.q_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.39.self_attn.q_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.39.self_attn.v_proj.bias": "model-00015-of-00031.safetensors",
+    "model.layers.39.self_attn.v_proj.weight": "model-00015-of-00031.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.40.input_layernorm.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.mlp.down_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.mlp.gate_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.mlp.up_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.post_attention_layernorm.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.k_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.k_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.o_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.q_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.q_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.v_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.40.self_attn.v_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.input_layernorm.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.mlp.down_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.mlp.gate_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.mlp.up_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.post_attention_layernorm.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.k_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.k_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.o_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.q_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.q_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.v_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.41.self_attn.v_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.42.input_layernorm.weight": "model-00017-of-00031.safetensors",
+    "model.layers.42.mlp.down_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.42.mlp.gate_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.42.mlp.up_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.42.post_attention_layernorm.weight": "model-00017-of-00031.safetensors",
+    "model.layers.42.self_attn.k_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.42.self_attn.k_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.42.self_attn.o_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.42.self_attn.q_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.42.self_attn.q_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.42.self_attn.v_proj.bias": "model-00016-of-00031.safetensors",
+    "model.layers.42.self_attn.v_proj.weight": "model-00016-of-00031.safetensors",
+    "model.layers.43.input_layernorm.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.mlp.down_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.mlp.gate_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.mlp.up_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.post_attention_layernorm.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.k_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.k_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.o_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.q_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.q_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.v_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.43.self_attn.v_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.input_layernorm.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.mlp.down_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.mlp.gate_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.mlp.up_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.post_attention_layernorm.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.k_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.k_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.o_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.q_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.q_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.v_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.44.self_attn.v_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.45.input_layernorm.weight": "model-00018-of-00031.safetensors",
+    "model.layers.45.mlp.down_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.45.mlp.gate_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.45.mlp.up_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.45.post_attention_layernorm.weight": "model-00018-of-00031.safetensors",
+    "model.layers.45.self_attn.k_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.45.self_attn.k_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.45.self_attn.o_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.45.self_attn.q_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.45.self_attn.q_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.45.self_attn.v_proj.bias": "model-00017-of-00031.safetensors",
+    "model.layers.45.self_attn.v_proj.weight": "model-00017-of-00031.safetensors",
+    "model.layers.46.input_layernorm.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.mlp.down_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.mlp.gate_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.mlp.up_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.post_attention_layernorm.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.k_proj.bias": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.k_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.o_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.q_proj.bias": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.q_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.v_proj.bias": "model-00018-of-00031.safetensors",
+    "model.layers.46.self_attn.v_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.input_layernorm.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.mlp.down_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.mlp.gate_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.mlp.up_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.post_attention_layernorm.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.k_proj.bias": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.k_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.o_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.q_proj.bias": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.q_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.v_proj.bias": "model-00018-of-00031.safetensors",
+    "model.layers.47.self_attn.v_proj.weight": "model-00018-of-00031.safetensors",
+    "model.layers.48.input_layernorm.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.mlp.down_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.mlp.gate_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.mlp.up_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.post_attention_layernorm.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.k_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.k_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.o_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.q_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.q_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.v_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.48.self_attn.v_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.input_layernorm.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.mlp.down_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.mlp.gate_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.mlp.up_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.post_attention_layernorm.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.k_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.k_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.o_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.q_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.q_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.v_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.49.self_attn.v_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.50.input_layernorm.weight": "model-00020-of-00031.safetensors",
+    "model.layers.50.mlp.down_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.50.mlp.gate_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.50.mlp.up_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.50.post_attention_layernorm.weight": "model-00020-of-00031.safetensors",
+    "model.layers.50.self_attn.k_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.50.self_attn.k_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.50.self_attn.o_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.50.self_attn.q_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.50.self_attn.q_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.50.self_attn.v_proj.bias": "model-00019-of-00031.safetensors",
+    "model.layers.50.self_attn.v_proj.weight": "model-00019-of-00031.safetensors",
+    "model.layers.51.input_layernorm.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.mlp.down_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.mlp.gate_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.mlp.up_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.post_attention_layernorm.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.k_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.k_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.o_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.q_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.q_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.v_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.51.self_attn.v_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.input_layernorm.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.mlp.down_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.mlp.gate_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.mlp.up_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.post_attention_layernorm.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.k_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.k_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.o_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.q_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.q_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.v_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.52.self_attn.v_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.53.input_layernorm.weight": "model-00021-of-00031.safetensors",
+    "model.layers.53.mlp.down_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.53.mlp.gate_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.53.mlp.up_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.53.post_attention_layernorm.weight": "model-00021-of-00031.safetensors",
+    "model.layers.53.self_attn.k_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.53.self_attn.k_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.53.self_attn.o_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.53.self_attn.q_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.53.self_attn.q_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.53.self_attn.v_proj.bias": "model-00020-of-00031.safetensors",
+    "model.layers.53.self_attn.v_proj.weight": "model-00020-of-00031.safetensors",
+    "model.layers.54.input_layernorm.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.mlp.down_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.mlp.gate_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.mlp.up_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.post_attention_layernorm.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.k_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.k_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.o_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.q_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.q_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.v_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.54.self_attn.v_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.input_layernorm.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.mlp.down_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.mlp.gate_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.mlp.up_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.post_attention_layernorm.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.k_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.k_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.o_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.q_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.q_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.v_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.55.self_attn.v_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.56.input_layernorm.weight": "model-00022-of-00031.safetensors",
+    "model.layers.56.mlp.down_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.56.mlp.gate_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.56.mlp.up_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.56.post_attention_layernorm.weight": "model-00022-of-00031.safetensors",
+    "model.layers.56.self_attn.k_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.56.self_attn.k_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.56.self_attn.o_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.56.self_attn.q_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.56.self_attn.q_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.56.self_attn.v_proj.bias": "model-00021-of-00031.safetensors",
+    "model.layers.56.self_attn.v_proj.weight": "model-00021-of-00031.safetensors",
+    "model.layers.57.input_layernorm.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.mlp.down_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.mlp.gate_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.mlp.up_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.post_attention_layernorm.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.k_proj.bias": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.k_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.o_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.q_proj.bias": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.q_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.v_proj.bias": "model-00022-of-00031.safetensors",
+    "model.layers.57.self_attn.v_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.input_layernorm.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.mlp.down_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.mlp.gate_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.mlp.up_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.post_attention_layernorm.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.k_proj.bias": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.k_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.o_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.q_proj.bias": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.q_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.v_proj.bias": "model-00022-of-00031.safetensors",
+    "model.layers.58.self_attn.v_proj.weight": "model-00022-of-00031.safetensors",
+    "model.layers.59.input_layernorm.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.mlp.down_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.mlp.gate_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.mlp.up_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.post_attention_layernorm.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.k_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.k_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.o_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.q_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.q_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.v_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.59.self_attn.v_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00004-of-00031.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00004-of-00031.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00003-of-00031.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00003-of-00031.safetensors",
+    "model.layers.60.input_layernorm.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.mlp.down_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.mlp.gate_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.mlp.up_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.post_attention_layernorm.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.k_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.k_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.o_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.q_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.q_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.v_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.60.self_attn.v_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.61.input_layernorm.weight": "model-00024-of-00031.safetensors",
+    "model.layers.61.mlp.down_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.61.mlp.gate_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.61.mlp.up_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.61.post_attention_layernorm.weight": "model-00024-of-00031.safetensors",
+    "model.layers.61.self_attn.k_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.61.self_attn.k_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.61.self_attn.o_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.61.self_attn.q_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.61.self_attn.q_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.61.self_attn.v_proj.bias": "model-00023-of-00031.safetensors",
+    "model.layers.61.self_attn.v_proj.weight": "model-00023-of-00031.safetensors",
+    "model.layers.62.input_layernorm.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.mlp.down_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.mlp.gate_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.mlp.up_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.post_attention_layernorm.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.k_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.k_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.o_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.q_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.q_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.v_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.62.self_attn.v_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.input_layernorm.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.mlp.down_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.mlp.gate_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.mlp.up_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.post_attention_layernorm.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.k_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.k_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.o_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.q_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.q_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.v_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.63.self_attn.v_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.64.input_layernorm.weight": "model-00025-of-00031.safetensors",
+    "model.layers.64.mlp.down_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.64.mlp.gate_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.64.mlp.up_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.64.post_attention_layernorm.weight": "model-00025-of-00031.safetensors",
+    "model.layers.64.self_attn.k_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.64.self_attn.k_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.64.self_attn.o_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.64.self_attn.q_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.64.self_attn.q_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.64.self_attn.v_proj.bias": "model-00024-of-00031.safetensors",
+    "model.layers.64.self_attn.v_proj.weight": "model-00024-of-00031.safetensors",
+    "model.layers.65.input_layernorm.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.mlp.down_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.mlp.gate_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.mlp.up_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.post_attention_layernorm.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.k_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.k_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.o_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.q_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.q_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.v_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.65.self_attn.v_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.input_layernorm.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.mlp.down_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.mlp.gate_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.mlp.up_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.post_attention_layernorm.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.k_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.k_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.o_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.q_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.q_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.v_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.66.self_attn.v_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.67.input_layernorm.weight": "model-00026-of-00031.safetensors",
+    "model.layers.67.mlp.down_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.67.mlp.gate_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.67.mlp.up_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.67.post_attention_layernorm.weight": "model-00026-of-00031.safetensors",
+    "model.layers.67.self_attn.k_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.67.self_attn.k_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.67.self_attn.o_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.67.self_attn.q_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.67.self_attn.q_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.67.self_attn.v_proj.bias": "model-00025-of-00031.safetensors",
+    "model.layers.67.self_attn.v_proj.weight": "model-00025-of-00031.safetensors",
+    "model.layers.68.input_layernorm.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.mlp.down_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.mlp.gate_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.mlp.up_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.post_attention_layernorm.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.k_proj.bias": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.k_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.o_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.q_proj.bias": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.q_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.v_proj.bias": "model-00026-of-00031.safetensors",
+    "model.layers.68.self_attn.v_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.input_layernorm.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.mlp.down_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.mlp.gate_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.mlp.up_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.post_attention_layernorm.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.k_proj.bias": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.k_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.o_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.q_proj.bias": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.q_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.v_proj.bias": "model-00026-of-00031.safetensors",
+    "model.layers.69.self_attn.v_proj.weight": "model-00026-of-00031.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.70.input_layernorm.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.mlp.down_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.mlp.gate_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.mlp.up_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.post_attention_layernorm.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.k_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.k_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.o_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.q_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.q_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.v_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.70.self_attn.v_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.input_layernorm.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.mlp.down_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.mlp.gate_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.mlp.up_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.post_attention_layernorm.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.k_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.k_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.o_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.q_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.q_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.v_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.71.self_attn.v_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.72.input_layernorm.weight": "model-00028-of-00031.safetensors",
+    "model.layers.72.mlp.down_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.72.mlp.gate_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.72.mlp.up_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.72.post_attention_layernorm.weight": "model-00028-of-00031.safetensors",
+    "model.layers.72.self_attn.k_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.72.self_attn.k_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.72.self_attn.o_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.72.self_attn.q_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.72.self_attn.q_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.72.self_attn.v_proj.bias": "model-00027-of-00031.safetensors",
+    "model.layers.72.self_attn.v_proj.weight": "model-00027-of-00031.safetensors",
+    "model.layers.73.input_layernorm.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.mlp.down_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.mlp.gate_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.mlp.up_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.post_attention_layernorm.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.k_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.k_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.o_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.q_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.q_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.v_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.73.self_attn.v_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.input_layernorm.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.mlp.down_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.mlp.gate_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.mlp.up_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.post_attention_layernorm.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.k_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.k_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.o_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.q_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.q_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.v_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.74.self_attn.v_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.75.input_layernorm.weight": "model-00029-of-00031.safetensors",
+    "model.layers.75.mlp.down_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.75.mlp.gate_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.75.mlp.up_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.75.post_attention_layernorm.weight": "model-00029-of-00031.safetensors",
+    "model.layers.75.self_attn.k_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.75.self_attn.k_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.75.self_attn.o_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.75.self_attn.q_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.75.self_attn.q_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.75.self_attn.v_proj.bias": "model-00028-of-00031.safetensors",
+    "model.layers.75.self_attn.v_proj.weight": "model-00028-of-00031.safetensors",
+    "model.layers.76.input_layernorm.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.mlp.down_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.mlp.gate_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.mlp.up_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.post_attention_layernorm.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.k_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.k_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.o_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.q_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.q_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.v_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.76.self_attn.v_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.input_layernorm.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.mlp.down_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.mlp.gate_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.mlp.up_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.post_attention_layernorm.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.k_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.k_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.o_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.q_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.q_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.v_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.77.self_attn.v_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.78.input_layernorm.weight": "model-00030-of-00031.safetensors",
+    "model.layers.78.mlp.down_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.78.mlp.gate_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.78.mlp.up_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.78.post_attention_layernorm.weight": "model-00030-of-00031.safetensors",
+    "model.layers.78.self_attn.k_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.78.self_attn.k_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.78.self_attn.o_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.78.self_attn.q_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.78.self_attn.q_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.78.self_attn.v_proj.bias": "model-00029-of-00031.safetensors",
+    "model.layers.78.self_attn.v_proj.weight": "model-00029-of-00031.safetensors",
+    "model.layers.79.input_layernorm.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.mlp.down_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.mlp.gate_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.mlp.up_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.post_attention_layernorm.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.k_proj.bias": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.k_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.o_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.q_proj.bias": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.q_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.v_proj.bias": "model-00030-of-00031.safetensors",
+    "model.layers.79.self_attn.v_proj.weight": "model-00030-of-00031.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00005-of-00031.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00005-of-00031.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00005-of-00031.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00004-of-00031.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00004-of-00031.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00004-of-00031.safetensors",
+    "model.norm.weight": "model-00030-of-00031.safetensors"
+  }
+}

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
+  }
+}

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": 3.0,
+    "total_flos": 2.982866014384947e+16,
+    "train_loss": 0.3923433954339652,
+    "train_runtime": 48299.7361,
+    "train_samples_per_second": 7.078,
+    "train_steps_per_second": 0.007
+}

trainer_log.jsonl

@@ -0,0 +1,337 @@
+{"current_steps": 1, "total_steps": 336, "loss": 0.8158, "lr": 2.9411764705882356e-07, "epoch": 0.008928571428571428, "percentage": 0.3, "elapsed_time": "0:02:11", "remaining_time": "12:15:56"}
+{"current_steps": 2, "total_steps": 336, "loss": 0.8361, "lr": 5.882352941176471e-07, "epoch": 0.017857142857142856, "percentage": 0.6, "elapsed_time": "0:04:19", "remaining_time": "12:03:31"}
+{"current_steps": 3, "total_steps": 336, "loss": 0.8422, "lr": 8.823529411764707e-07, "epoch": 0.026785714285714284, "percentage": 0.89, "elapsed_time": "0:06:23", "remaining_time": "11:49:11"}
+{"current_steps": 4, "total_steps": 336, "loss": 0.8212, "lr": 1.1764705882352942e-06, "epoch": 0.03571428571428571, "percentage": 1.19, "elapsed_time": "0:08:18", "remaining_time": "11:30:06"}
+{"current_steps": 5, "total_steps": 336, "loss": 0.8039, "lr": 1.4705882352941177e-06, "epoch": 0.044642857142857144, "percentage": 1.49, "elapsed_time": "0:10:12", "remaining_time": "11:15:58"}
+{"current_steps": 6, "total_steps": 336, "loss": 0.7598, "lr": 1.7647058823529414e-06, "epoch": 0.05357142857142857, "percentage": 1.79, "elapsed_time": "0:12:13", "remaining_time": "11:12:44"}
+{"current_steps": 7, "total_steps": 336, "loss": 0.7132, "lr": 2.058823529411765e-06, "epoch": 0.0625, "percentage": 2.08, "elapsed_time": "0:14:13", "remaining_time": "11:08:44"}
+{"current_steps": 8, "total_steps": 336, "loss": 0.7006, "lr": 2.3529411764705885e-06, "epoch": 0.07142857142857142, "percentage": 2.38, "elapsed_time": "0:16:10", "remaining_time": "11:02:52"}
+{"current_steps": 9, "total_steps": 336, "loss": 0.6781, "lr": 2.647058823529412e-06, "epoch": 0.08035714285714286, "percentage": 2.68, "elapsed_time": "0:18:07", "remaining_time": "10:58:48"}
+{"current_steps": 10, "total_steps": 336, "loss": 0.6716, "lr": 2.9411764705882355e-06, "epoch": 0.08928571428571429, "percentage": 2.98, "elapsed_time": "0:20:05", "remaining_time": "10:55:14"}
+{"current_steps": 11, "total_steps": 336, "loss": 0.6524, "lr": 3.2352941176470594e-06, "epoch": 0.09821428571428571, "percentage": 3.27, "elapsed_time": "0:22:03", "remaining_time": "10:51:45"}
+{"current_steps": 12, "total_steps": 336, "loss": 0.6202, "lr": 3.529411764705883e-06, "epoch": 0.10714285714285714, "percentage": 3.57, "elapsed_time": "0:23:58", "remaining_time": "10:47:13"}
+{"current_steps": 13, "total_steps": 336, "loss": 0.6104, "lr": 3.8235294117647055e-06, "epoch": 0.11607142857142858, "percentage": 3.87, "elapsed_time": "0:25:52", "remaining_time": "10:42:53"}
+{"current_steps": 14, "total_steps": 336, "loss": 0.6036, "lr": 4.11764705882353e-06, "epoch": 0.125, "percentage": 4.17, "elapsed_time": "0:27:47", "remaining_time": "10:39:19"}
+{"current_steps": 15, "total_steps": 336, "loss": 0.5893, "lr": 4.411764705882353e-06, "epoch": 0.13392857142857142, "percentage": 4.46, "elapsed_time": "0:29:42", "remaining_time": "10:35:36"}
+{"current_steps": 16, "total_steps": 336, "loss": 0.5752, "lr": 4.705882352941177e-06, "epoch": 0.14285714285714285, "percentage": 4.76, "elapsed_time": "0:31:35", "remaining_time": "10:31:56"}
+{"current_steps": 17, "total_steps": 336, "loss": 0.5527, "lr": 5e-06, "epoch": 0.15178571428571427, "percentage": 5.06, "elapsed_time": "0:33:32", "remaining_time": "10:29:17"}
+{"current_steps": 18, "total_steps": 336, "loss": 0.5438, "lr": 5.294117647058824e-06, "epoch": 0.16071428571428573, "percentage": 5.36, "elapsed_time": "0:35:27", "remaining_time": "10:26:28"}
+{"current_steps": 19, "total_steps": 336, "loss": 0.5295, "lr": 5.588235294117647e-06, "epoch": 0.16964285714285715, "percentage": 5.65, "elapsed_time": "0:37:22", "remaining_time": "10:23:28"}
+{"current_steps": 20, "total_steps": 336, "loss": 0.5227, "lr": 5.882352941176471e-06, "epoch": 0.17857142857142858, "percentage": 5.95, "elapsed_time": "0:39:21", "remaining_time": "10:21:59"}
+{"current_steps": 21, "total_steps": 336, "loss": 0.5146, "lr": 6.176470588235295e-06, "epoch": 0.1875, "percentage": 6.25, "elapsed_time": "0:41:18", "remaining_time": "10:19:43"}
+{"current_steps": 22, "total_steps": 336, "loss": 0.503, "lr": 6.470588235294119e-06, "epoch": 0.19642857142857142, "percentage": 6.55, "elapsed_time": "0:43:15", "remaining_time": "10:17:24"}
+{"current_steps": 23, "total_steps": 336, "loss": 0.499, "lr": 6.764705882352942e-06, "epoch": 0.20535714285714285, "percentage": 6.85, "elapsed_time": "0:45:09", "remaining_time": "10:14:28"}
+{"current_steps": 24, "total_steps": 336, "loss": 0.4901, "lr": 7.058823529411766e-06, "epoch": 0.21428571428571427, "percentage": 7.14, "elapsed_time": "0:47:06", "remaining_time": "10:12:28"}
+{"current_steps": 25, "total_steps": 336, "loss": 0.4882, "lr": 7.352941176470589e-06, "epoch": 0.22321428571428573, "percentage": 7.44, "elapsed_time": "0:49:04", "remaining_time": "10:10:24"}
+{"current_steps": 26, "total_steps": 336, "loss": 0.4789, "lr": 7.647058823529411e-06, "epoch": 0.23214285714285715, "percentage": 7.74, "elapsed_time": "0:50:59", "remaining_time": "10:08:01"}
+{"current_steps": 27, "total_steps": 336, "loss": 0.4792, "lr": 7.941176470588236e-06, "epoch": 0.24107142857142858, "percentage": 8.04, "elapsed_time": "0:53:02", "remaining_time": "10:06:56"}
+{"current_steps": 28, "total_steps": 336, "loss": 0.4665, "lr": 8.23529411764706e-06, "epoch": 0.25, "percentage": 8.33, "elapsed_time": "0:54:56", "remaining_time": "10:04:23"}
+{"current_steps": 29, "total_steps": 336, "loss": 0.4666, "lr": 8.529411764705883e-06, "epoch": 0.25892857142857145, "percentage": 8.63, "elapsed_time": "0:56:52", "remaining_time": "10:02:06"}
+{"current_steps": 30, "total_steps": 336, "loss": 0.4644, "lr": 8.823529411764707e-06, "epoch": 0.26785714285714285, "percentage": 8.93, "elapsed_time": "0:58:49", "remaining_time": "9:59:59"}
+{"current_steps": 31, "total_steps": 336, "loss": 0.4552, "lr": 9.11764705882353e-06, "epoch": 0.2767857142857143, "percentage": 9.23, "elapsed_time": "1:00:48", "remaining_time": "9:58:14"}
+{"current_steps": 32, "total_steps": 336, "loss": 0.4611, "lr": 9.411764705882354e-06, "epoch": 0.2857142857142857, "percentage": 9.52, "elapsed_time": "1:02:44", "remaining_time": "9:56:01"}
+{"current_steps": 33, "total_steps": 336, "loss": 0.449, "lr": 9.705882352941177e-06, "epoch": 0.29464285714285715, "percentage": 9.82, "elapsed_time": "1:04:42", "remaining_time": "9:54:05"}
+{"current_steps": 34, "total_steps": 336, "loss": 0.4543, "lr": 1e-05, "epoch": 0.30357142857142855, "percentage": 10.12, "elapsed_time": "1:06:44", "remaining_time": "9:52:48"}
+{"current_steps": 35, "total_steps": 336, "loss": 0.4449, "lr": 9.999729465939036e-06, "epoch": 0.3125, "percentage": 10.42, "elapsed_time": "1:08:47", "remaining_time": "9:51:36"}
+{"current_steps": 36, "total_steps": 336, "loss": 0.4503, "lr": 9.998917893031615e-06, "epoch": 0.32142857142857145, "percentage": 10.71, "elapsed_time": "1:10:45", "remaining_time": "9:49:39"}
+{"current_steps": 37, "total_steps": 336, "loss": 0.4421, "lr": 9.997565369100983e-06, "epoch": 0.33035714285714285, "percentage": 11.01, "elapsed_time": "1:12:46", "remaining_time": "9:48:07"}
+{"current_steps": 38, "total_steps": 336, "loss": 0.4453, "lr": 9.995672040508656e-06, "epoch": 0.3392857142857143, "percentage": 11.31, "elapsed_time": "1:14:47", "remaining_time": "9:46:34"}
+{"current_steps": 39, "total_steps": 336, "loss": 0.4388, "lr": 9.993238112138584e-06, "epoch": 0.3482142857142857, "percentage": 11.61, "elapsed_time": "1:16:51", "remaining_time": "9:45:17"}
+{"current_steps": 40, "total_steps": 336, "loss": 0.4407, "lr": 9.990263847374976e-06, "epoch": 0.35714285714285715, "percentage": 11.9, "elapsed_time": "1:18:51", "remaining_time": "9:43:30"}
+{"current_steps": 41, "total_steps": 336, "loss": 0.4419, "lr": 9.986749568073804e-06, "epoch": 0.36607142857142855, "percentage": 12.2, "elapsed_time": "1:20:52", "remaining_time": "9:41:57"}
+{"current_steps": 42, "total_steps": 336, "loss": 0.438, "lr": 9.982695654527966e-06, "epoch": 0.375, "percentage": 12.5, "elapsed_time": "1:22:57", "remaining_time": "9:40:44"}
+{"current_steps": 43, "total_steps": 336, "loss": 0.4299, "lr": 9.97810254542614e-06, "epoch": 0.38392857142857145, "percentage": 12.8, "elapsed_time": "1:24:58", "remaining_time": "9:39:00"}
+{"current_steps": 44, "total_steps": 336, "loss": 0.4248, "lr": 9.972970737805312e-06, "epoch": 0.39285714285714285, "percentage": 13.1, "elapsed_time": "1:26:58", "remaining_time": "9:37:14"}
+{"current_steps": 45, "total_steps": 336, "loss": 0.4333, "lr": 9.96730078699698e-06, "epoch": 0.4017857142857143, "percentage": 13.39, "elapsed_time": "1:29:00", "remaining_time": "9:35:34"}
+{"current_steps": 46, "total_steps": 336, "loss": 0.4266, "lr": 9.961093306567076e-06, "epoch": 0.4107142857142857, "percentage": 13.69, "elapsed_time": "1:31:05", "remaining_time": "9:34:16"}
+{"current_steps": 47, "total_steps": 336, "loss": 0.4253, "lr": 9.954348968249552e-06, "epoch": 0.41964285714285715, "percentage": 13.99, "elapsed_time": "1:33:09", "remaining_time": "9:32:49"}
+{"current_steps": 48, "total_steps": 336, "loss": 0.4229, "lr": 9.947068501873702e-06, "epoch": 0.42857142857142855, "percentage": 14.29, "elapsed_time": "1:35:14", "remaining_time": "9:31:25"}
+{"current_steps": 49, "total_steps": 336, "loss": 0.423, "lr": 9.93925269528518e-06, "epoch": 0.4375, "percentage": 14.58, "elapsed_time": "1:37:21", "remaining_time": "9:30:16"}
+{"current_steps": 50, "total_steps": 336, "loss": 0.4136, "lr": 9.930902394260746e-06, "epoch": 0.44642857142857145, "percentage": 14.88, "elapsed_time": "1:39:27", "remaining_time": "9:28:54"}
+{"current_steps": 51, "total_steps": 336, "loss": 0.4238, "lr": 9.922018502416736e-06, "epoch": 0.45535714285714285, "percentage": 15.18, "elapsed_time": "1:46:01", "remaining_time": "9:52:29"}
+{"current_steps": 52, "total_steps": 336, "loss": 0.4162, "lr": 9.912601981111287e-06, "epoch": 0.4642857142857143, "percentage": 15.48, "elapsed_time": "1:48:05", "remaining_time": "9:50:22"}
+{"current_steps": 53, "total_steps": 336, "loss": 0.4115, "lr": 9.902653849340296e-06, "epoch": 0.4732142857142857, "percentage": 15.77, "elapsed_time": "1:50:08", "remaining_time": "9:48:08"}
+{"current_steps": 54, "total_steps": 336, "loss": 0.4144, "lr": 9.892175183627161e-06, "epoch": 0.48214285714285715, "percentage": 16.07, "elapsed_time": "1:52:13", "remaining_time": "9:46:04"}
+{"current_steps": 55, "total_steps": 336, "loss": 0.4163, "lr": 9.881167117906276e-06, "epoch": 0.49107142857142855, "percentage": 16.37, "elapsed_time": "1:54:22", "remaining_time": "9:44:19"}
+{"current_steps": 56, "total_steps": 336, "loss": 0.412, "lr": 9.869630843400331e-06, "epoch": 0.5, "percentage": 16.67, "elapsed_time": "1:56:30", "remaining_time": "9:42:34"}
+{"current_steps": 57, "total_steps": 336, "loss": 0.4196, "lr": 9.8575676084914e-06, "epoch": 0.5089285714285714, "percentage": 16.96, "elapsed_time": "1:58:35", "remaining_time": "9:40:30"}
+{"current_steps": 58, "total_steps": 336, "loss": 0.4135, "lr": 9.844978718585855e-06, "epoch": 0.5178571428571429, "percentage": 17.26, "elapsed_time": "2:00:42", "remaining_time": "9:38:33"}
+{"current_steps": 59, "total_steps": 336, "loss": 0.4135, "lr": 9.831865535973103e-06, "epoch": 0.5267857142857143, "percentage": 17.56, "elapsed_time": "2:02:47", "remaining_time": "9:36:29"}
+{"current_steps": 60, "total_steps": 336, "loss": 0.4085, "lr": 9.81822947967816e-06, "epoch": 0.5357142857142857, "percentage": 17.86, "elapsed_time": "2:04:54", "remaining_time": "9:34:34"}
+{"current_steps": 61, "total_steps": 336, "loss": 0.4112, "lr": 9.804072025308096e-06, "epoch": 0.5446428571428571, "percentage": 18.15, "elapsed_time": "2:06:59", "remaining_time": "9:32:32"}
+{"current_steps": 62, "total_steps": 336, "loss": 0.4086, "lr": 9.789394704892364e-06, "epoch": 0.5535714285714286, "percentage": 18.45, "elapsed_time": "2:09:09", "remaining_time": "9:30:48"}
+{"current_steps": 63, "total_steps": 336, "loss": 0.4077, "lr": 9.774199106717004e-06, "epoch": 0.5625, "percentage": 18.75, "elapsed_time": "2:11:18", "remaining_time": "9:29:01"}
+{"current_steps": 64, "total_steps": 336, "loss": 0.4072, "lr": 9.758486875152766e-06, "epoch": 0.5714285714285714, "percentage": 19.05, "elapsed_time": "2:13:22", "remaining_time": "9:26:48"}
+{"current_steps": 65, "total_steps": 336, "loss": 0.4015, "lr": 9.742259710477178e-06, "epoch": 0.5803571428571429, "percentage": 19.35, "elapsed_time": "2:15:27", "remaining_time": "9:24:45"}
+{"current_steps": 66, "total_steps": 336, "loss": 0.4035, "lr": 9.725519368690539e-06, "epoch": 0.5892857142857143, "percentage": 19.64, "elapsed_time": "2:17:36", "remaining_time": "9:22:56"}
+{"current_steps": 67, "total_steps": 336, "loss": 0.4061, "lr": 9.70826766132591e-06, "epoch": 0.5982142857142857, "percentage": 19.94, "elapsed_time": "2:19:40", "remaining_time": "9:20:46"}
+{"current_steps": 68, "total_steps": 336, "loss": 0.4086, "lr": 9.690506455253073e-06, "epoch": 0.6071428571428571, "percentage": 20.24, "elapsed_time": "2:21:47", "remaining_time": "9:18:48"}
+{"current_steps": 69, "total_steps": 336, "loss": 0.4017, "lr": 9.672237672476506e-06, "epoch": 0.6160714285714286, "percentage": 20.54, "elapsed_time": "2:23:55", "remaining_time": "9:16:56"}
+{"current_steps": 70, "total_steps": 336, "loss": 0.4017, "lr": 9.65346328992741e-06, "epoch": 0.625, "percentage": 20.83, "elapsed_time": "2:26:07", "remaining_time": "9:15:16"}
+{"current_steps": 71, "total_steps": 336, "loss": 0.4043, "lr": 9.634185339249766e-06, "epoch": 0.6339285714285714, "percentage": 21.13, "elapsed_time": "2:28:18", "remaining_time": "9:13:32"}
+{"current_steps": 72, "total_steps": 336, "loss": 0.396, "lr": 9.614405906580486e-06, "epoch": 0.6428571428571429, "percentage": 21.43, "elapsed_time": "2:30:24", "remaining_time": "9:11:30"}
+{"current_steps": 73, "total_steps": 336, "loss": 0.4062, "lr": 9.594127132323669e-06, "epoch": 0.6517857142857143, "percentage": 21.73, "elapsed_time": "2:32:32", "remaining_time": "9:09:33"}
+{"current_steps": 74, "total_steps": 336, "loss": 0.3977, "lr": 9.573351210918976e-06, "epoch": 0.6607142857142857, "percentage": 22.02, "elapsed_time": "2:34:40", "remaining_time": "9:07:37"}
+{"current_steps": 75, "total_steps": 336, "loss": 0.3993, "lr": 9.55208039060416e-06, "epoch": 0.6696428571428571, "percentage": 22.32, "elapsed_time": "2:36:49", "remaining_time": "9:05:43"}
+{"current_steps": 76, "total_steps": 336, "loss": 0.4126, "lr": 9.53031697317178e-06, "epoch": 0.6785714285714286, "percentage": 22.62, "elapsed_time": "2:38:57", "remaining_time": "9:03:48"}
+{"current_steps": 77, "total_steps": 336, "loss": 0.4068, "lr": 9.50806331372012e-06, "epoch": 0.6875, "percentage": 22.92, "elapsed_time": "2:41:11", "remaining_time": "9:02:09"}
+{"current_steps": 78, "total_steps": 336, "loss": 0.4014, "lr": 9.485321820398321e-06, "epoch": 0.6964285714285714, "percentage": 23.21, "elapsed_time": "2:43:25", "remaining_time": "9:00:32"}
+{"current_steps": 79, "total_steps": 336, "loss": 0.3995, "lr": 9.4620949541458e-06, "epoch": 0.7053571428571429, "percentage": 23.51, "elapsed_time": "2:45:32", "remaining_time": "8:58:30"}
+{"current_steps": 80, "total_steps": 336, "loss": 0.4055, "lr": 9.43838522842594e-06, "epoch": 0.7142857142857143, "percentage": 23.81, "elapsed_time": "2:47:40", "remaining_time": "8:56:33"}
+{"current_steps": 81, "total_steps": 336, "loss": 0.3976, "lr": 9.41419520895409e-06, "epoch": 0.7232142857142857, "percentage": 24.11, "elapsed_time": "2:49:51", "remaining_time": "8:54:43"}
+{"current_steps": 82, "total_steps": 336, "loss": 0.3922, "lr": 9.389527513419935e-06, "epoch": 0.7321428571428571, "percentage": 24.4, "elapsed_time": "2:52:05", "remaining_time": "8:53:02"}
+{"current_steps": 83, "total_steps": 336, "loss": 0.4018, "lr": 9.364384811204212e-06, "epoch": 0.7410714285714286, "percentage": 24.7, "elapsed_time": "2:54:18", "remaining_time": "8:51:19"}
+{"current_steps": 84, "total_steps": 336, "loss": 0.399, "lr": 9.338769823089853e-06, "epoch": 0.75, "percentage": 25.0, "elapsed_time": "2:56:35", "remaining_time": "8:49:45"}
+{"current_steps": 85, "total_steps": 336, "loss": 0.3948, "lr": 9.312685320967566e-06, "epoch": 0.7589285714285714, "percentage": 25.3, "elapsed_time": "2:58:51", "remaining_time": "8:48:08"}
+{"current_steps": 86, "total_steps": 336, "loss": 0.4007, "lr": 9.286134127535859e-06, "epoch": 0.7678571428571429, "percentage": 25.6, "elapsed_time": "3:01:03", "remaining_time": "8:46:20"}
+{"current_steps": 87, "total_steps": 336, "loss": 0.3944, "lr": 9.25911911599561e-06, "epoch": 0.7767857142857143, "percentage": 25.89, "elapsed_time": "3:03:13", "remaining_time": "8:44:24"}
+{"current_steps": 88, "total_steps": 336, "loss": 0.4006, "lr": 9.231643209739128e-06, "epoch": 0.7857142857142857, "percentage": 26.19, "elapsed_time": "3:05:24", "remaining_time": "8:42:30"}
+{"current_steps": 89, "total_steps": 336, "loss": 0.3894, "lr": 9.203709382033814e-06, "epoch": 0.7946428571428571, "percentage": 26.49, "elapsed_time": "3:07:39", "remaining_time": "8:40:47"}
+{"current_steps": 90, "total_steps": 336, "loss": 0.3965, "lr": 9.175320655700407e-06, "epoch": 0.8035714285714286, "percentage": 26.79, "elapsed_time": "3:09:53", "remaining_time": "8:39:02"}
+{"current_steps": 91, "total_steps": 336, "loss": 0.4021, "lr": 9.146480102785871e-06, "epoch": 0.8125, "percentage": 27.08, "elapsed_time": "3:12:11", "remaining_time": "8:37:25"}
+{"current_steps": 92, "total_steps": 336, "loss": 0.3979, "lr": 9.117190844230971e-06, "epoch": 0.8214285714285714, "percentage": 27.38, "elapsed_time": "3:14:29", "remaining_time": "8:35:49"}
+{"current_steps": 93, "total_steps": 336, "loss": 0.3961, "lr": 9.08745604953253e-06, "epoch": 0.8303571428571429, "percentage": 27.68, "elapsed_time": "3:16:45", "remaining_time": "8:34:07"}
+{"current_steps": 94, "total_steps": 336, "loss": 0.3947, "lr": 9.057278936400453e-06, "epoch": 0.8392857142857143, "percentage": 27.98, "elapsed_time": "3:19:00", "remaining_time": "8:32:19"}
+{"current_steps": 95, "total_steps": 336, "loss": 0.3993, "lr": 9.026662770409524e-06, "epoch": 0.8482142857142857, "percentage": 28.27, "elapsed_time": "3:21:16", "remaining_time": "8:30:37"}
+{"current_steps": 96, "total_steps": 336, "loss": 0.3997, "lr": 8.99561086464603e-06, "epoch": 0.8571428571428571, "percentage": 28.57, "elapsed_time": "3:23:33", "remaining_time": "8:28:53"}
+{"current_steps": 97, "total_steps": 336, "loss": 0.3985, "lr": 8.964126579349237e-06, "epoch": 0.8660714285714286, "percentage": 28.87, "elapsed_time": "3:25:50", "remaining_time": "8:27:10"}
+{"current_steps": 98, "total_steps": 336, "loss": 0.3857, "lr": 8.932213321547769e-06, "epoch": 0.875, "percentage": 29.17, "elapsed_time": "3:28:09", "remaining_time": "8:25:31"}
+{"current_steps": 99, "total_steps": 336, "loss": 0.3985, "lr": 8.899874544690921e-06, "epoch": 0.8839285714285714, "percentage": 29.46, "elapsed_time": "3:30:29", "remaining_time": "8:23:53"}
+{"current_steps": 100, "total_steps": 336, "loss": 0.3912, "lr": 8.86711374827494e-06, "epoch": 0.8928571428571429, "percentage": 29.76, "elapsed_time": "3:32:50", "remaining_time": "8:22:19"}
+{"current_steps": 101, "total_steps": 336, "loss": 0.3972, "lr": 8.833934477464348e-06, "epoch": 0.9017857142857143, "percentage": 30.06, "elapsed_time": "3:39:32", "remaining_time": "8:30:48"}
+{"current_steps": 102, "total_steps": 336, "loss": 0.3889, "lr": 8.800340322708291e-06, "epoch": 0.9107142857142857, "percentage": 30.36, "elapsed_time": "3:41:49", "remaining_time": "8:28:54"}
+{"current_steps": 103, "total_steps": 336, "loss": 0.3991, "lr": 8.766334919352018e-06, "epoch": 0.9196428571428571, "percentage": 30.65, "elapsed_time": "3:44:04", "remaining_time": "8:26:53"}
+{"current_steps": 104, "total_steps": 336, "loss": 0.3879, "lr": 8.73192194724347e-06, "epoch": 0.9285714285714286, "percentage": 30.95, "elapsed_time": "3:46:24", "remaining_time": "8:25:02"}
+{"current_steps": 105, "total_steps": 336, "loss": 0.3962, "lr": 8.697105130335084e-06, "epoch": 0.9375, "percentage": 31.25, "elapsed_time": "3:48:43", "remaining_time": "8:23:11"}
+{"current_steps": 106, "total_steps": 336, "loss": 0.3909, "lr": 8.661888236280813e-06, "epoch": 0.9464285714285714, "percentage": 31.55, "elapsed_time": "3:51:05", "remaining_time": "8:21:24"}
+{"current_steps": 107, "total_steps": 336, "loss": 0.3899, "lr": 8.626275076028397e-06, "epoch": 0.9553571428571429, "percentage": 31.85, "elapsed_time": "3:53:27", "remaining_time": "8:19:38"}
+{"current_steps": 108, "total_steps": 336, "loss": 0.3846, "lr": 8.590269503406986e-06, "epoch": 0.9642857142857143, "percentage": 32.14, "elapsed_time": "3:55:43", "remaining_time": "8:17:38"}
+{"current_steps": 109, "total_steps": 336, "loss": 0.3922, "lr": 8.553875414710088e-06, "epoch": 0.9732142857142857, "percentage": 32.44, "elapsed_time": "3:58:04", "remaining_time": "8:15:48"}
+{"current_steps": 110, "total_steps": 336, "loss": 0.3909, "lr": 8.517096748273951e-06, "epoch": 0.9821428571428571, "percentage": 32.74, "elapsed_time": "4:00:25", "remaining_time": "8:13:58"}
+{"current_steps": 111, "total_steps": 336, "loss": 0.3893, "lr": 8.479937484051368e-06, "epoch": 0.9910714285714286, "percentage": 33.04, "elapsed_time": "4:02:42", "remaining_time": "8:11:59"}
+{"current_steps": 112, "total_steps": 336, "loss": 0.3836, "lr": 8.442401643181e-06, "epoch": 1.0, "percentage": 33.33, "elapsed_time": "4:04:59", "remaining_time": "8:09:59"}
+{"current_steps": 113, "total_steps": 336, "loss": 0.3733, "lr": 8.404493287552232e-06, "epoch": 1.0089285714285714, "percentage": 33.63, "elapsed_time": "4:07:21", "remaining_time": "8:08:09"}
+{"current_steps": 114, "total_steps": 336, "loss": 0.3687, "lr": 8.366216519365623e-06, "epoch": 1.0178571428571428, "percentage": 33.93, "elapsed_time": "4:09:43", "remaining_time": "8:06:17"}
+{"current_steps": 115, "total_steps": 336, "loss": 0.3734, "lr": 8.327575480688985e-06, "epoch": 1.0267857142857142, "percentage": 34.23, "elapsed_time": "4:12:02", "remaining_time": "8:04:22"}
+{"current_steps": 116, "total_steps": 336, "loss": 0.3782, "lr": 8.288574353009164e-06, "epoch": 1.0357142857142858, "percentage": 34.52, "elapsed_time": "4:14:23", "remaining_time": "8:02:27"}
+{"current_steps": 117, "total_steps": 336, "loss": 0.3745, "lr": 8.249217356779544e-06, "epoch": 1.0446428571428572, "percentage": 34.82, "elapsed_time": "4:16:48", "remaining_time": "8:00:42"}
+{"current_steps": 118, "total_steps": 336, "loss": 0.3731, "lr": 8.20950875096333e-06, "epoch": 1.0535714285714286, "percentage": 35.12, "elapsed_time": "4:19:23", "remaining_time": "7:59:12"}
+{"current_steps": 119, "total_steps": 336, "loss": 0.3711, "lr": 8.169452832572676e-06, "epoch": 1.0625, "percentage": 35.42, "elapsed_time": "4:21:42", "remaining_time": "7:57:14"}
+{"current_steps": 120, "total_steps": 336, "loss": 0.3701, "lr": 8.129053936203688e-06, "epoch": 1.0714285714285714, "percentage": 35.71, "elapsed_time": "4:24:07", "remaining_time": "7:55:25"}
+{"current_steps": 121, "total_steps": 336, "loss": 0.3729, "lr": 8.088316433567369e-06, "epoch": 1.0803571428571428, "percentage": 36.01, "elapsed_time": "4:26:33", "remaining_time": "7:53:37"}
+{"current_steps": 122, "total_steps": 336, "loss": 0.3725, "lr": 8.04724473301652e-06, "epoch": 1.0892857142857142, "percentage": 36.31, "elapsed_time": "4:28:56", "remaining_time": "7:51:45"}
+{"current_steps": 123, "total_steps": 336, "loss": 0.3689, "lr": 8.005843279068724e-06, "epoch": 1.0982142857142858, "percentage": 36.61, "elapsed_time": "4:31:17", "remaining_time": "7:49:47"}
+{"current_steps": 124, "total_steps": 336, "loss": 0.377, "lr": 7.964116551925365e-06, "epoch": 1.1071428571428572, "percentage": 36.9, "elapsed_time": "4:33:36", "remaining_time": "7:47:46"}
+{"current_steps": 125, "total_steps": 336, "loss": 0.3684, "lr": 7.92206906698682e-06, "epoch": 1.1160714285714286, "percentage": 37.2, "elapsed_time": "4:35:58", "remaining_time": "7:45:50"}
+{"current_steps": 126, "total_steps": 336, "loss": 0.3744, "lr": 7.879705374363831e-06, "epoch": 1.125, "percentage": 37.5, "elapsed_time": "4:38:17", "remaining_time": "7:43:48"}
+{"current_steps": 127, "total_steps": 336, "loss": 0.3754, "lr": 7.837030058385117e-06, "epoch": 1.1339285714285714, "percentage": 37.8, "elapsed_time": "4:40:39", "remaining_time": "7:41:52"}
+{"current_steps": 128, "total_steps": 336, "loss": 0.3712, "lr": 7.794047737101298e-06, "epoch": 1.1428571428571428, "percentage": 38.1, "elapsed_time": "4:43:04", "remaining_time": "7:39:59"}
+{"current_steps": 129, "total_steps": 336, "loss": 0.3721, "lr": 7.750763061785139e-06, "epoch": 1.1517857142857142, "percentage": 38.39, "elapsed_time": "4:45:28", "remaining_time": "7:38:05"}
+{"current_steps": 130, "total_steps": 336, "loss": 0.3754, "lr": 7.707180716428237e-06, "epoch": 1.1607142857142858, "percentage": 38.69, "elapsed_time": "4:47:48", "remaining_time": "7:36:04"}
+{"current_steps": 131, "total_steps": 336, "loss": 0.3675, "lr": 7.663305417234146e-06, "epoch": 1.1696428571428572, "percentage": 38.99, "elapsed_time": "4:50:08", "remaining_time": "7:34:02"}
+{"current_steps": 132, "total_steps": 336, "loss": 0.3786, "lr": 7.619141912108008e-06, "epoch": 1.1785714285714286, "percentage": 39.29, "elapsed_time": "4:52:45", "remaining_time": "7:32:27"}
+{"current_steps": 133, "total_steps": 336, "loss": 0.3731, "lr": 7.57469498014278e-06, "epoch": 1.1875, "percentage": 39.58, "elapsed_time": "4:55:06", "remaining_time": "7:30:25"}
+{"current_steps": 134, "total_steps": 336, "loss": 0.3735, "lr": 7.529969431102063e-06, "epoch": 1.1964285714285714, "percentage": 39.88, "elapsed_time": "4:57:28", "remaining_time": "7:28:25"}
+{"current_steps": 135, "total_steps": 336, "loss": 0.3689, "lr": 7.484970104899624e-06, "epoch": 1.2053571428571428, "percentage": 40.18, "elapsed_time": "4:59:51", "remaining_time": "7:26:27"}
+{"current_steps": 136, "total_steps": 336, "loss": 0.3704, "lr": 7.4397018710756415e-06, "epoch": 1.2142857142857142, "percentage": 40.48, "elapsed_time": "5:02:16", "remaining_time": "7:24:30"}
+{"current_steps": 137, "total_steps": 336, "loss": 0.3702, "lr": 7.394169628269771e-06, "epoch": 1.2232142857142858, "percentage": 40.77, "elapsed_time": "5:04:38", "remaining_time": "7:22:29"}
+{"current_steps": 138, "total_steps": 336, "loss": 0.3688, "lr": 7.34837830369103e-06, "epoch": 1.2321428571428572, "percentage": 41.07, "elapsed_time": "5:06:57", "remaining_time": "7:20:25"}
+{"current_steps": 139, "total_steps": 336, "loss": 0.3731, "lr": 7.302332852584619e-06, "epoch": 1.2410714285714286, "percentage": 41.37, "elapsed_time": "5:09:21", "remaining_time": "7:18:26"}
+{"current_steps": 140, "total_steps": 336, "loss": 0.3691, "lr": 7.2560382576956875e-06, "epoch": 1.25, "percentage": 41.67, "elapsed_time": "5:11:50", "remaining_time": "7:16:35"}
+{"current_steps": 141, "total_steps": 336, "loss": 0.3664, "lr": 7.209499528730138e-06, "epoch": 1.2589285714285714, "percentage": 41.96, "elapsed_time": "5:14:24", "remaining_time": "7:14:49"}
+{"current_steps": 142, "total_steps": 336, "loss": 0.3702, "lr": 7.162721701812506e-06, "epoch": 1.2678571428571428, "percentage": 42.26, "elapsed_time": "5:16:49", "remaining_time": "7:12:50"}
+{"current_steps": 143, "total_steps": 336, "loss": 0.3722, "lr": 7.115709838940983e-06, "epoch": 1.2767857142857144, "percentage": 42.56, "elapsed_time": "5:19:17", "remaining_time": "7:10:55"}
+{"current_steps": 144, "total_steps": 336, "loss": 0.3741, "lr": 7.068469027439642e-06, "epoch": 1.2857142857142856, "percentage": 42.86, "elapsed_time": "5:21:46", "remaining_time": "7:09:01"}
+{"current_steps": 145, "total_steps": 336, "loss": 0.3675, "lr": 7.02100437940791e-06, "epoch": 1.2946428571428572, "percentage": 43.15, "elapsed_time": "5:24:10", "remaining_time": "7:07:00"}
+{"current_steps": 146, "total_steps": 336, "loss": 0.3695, "lr": 6.9733210311673826e-06, "epoch": 1.3035714285714286, "percentage": 43.45, "elapsed_time": "5:26:30", "remaining_time": "7:04:54"}
+{"current_steps": 147, "total_steps": 336, "loss": 0.3665, "lr": 6.925424142705997e-06, "epoch": 1.3125, "percentage": 43.75, "elapsed_time": "5:28:53", "remaining_time": "7:02:51"}
+{"current_steps": 148, "total_steps": 336, "loss": 0.3702, "lr": 6.8773188971196515e-06, "epoch": 1.3214285714285714, "percentage": 44.05, "elapsed_time": "5:31:13", "remaining_time": "7:00:45"}
+{"current_steps": 149, "total_steps": 336, "loss": 0.3739, "lr": 6.829010500051319e-06, "epoch": 1.3303571428571428, "percentage": 44.35, "elapsed_time": "5:33:37", "remaining_time": "6:58:42"}
+{"current_steps": 150, "total_steps": 336, "loss": 0.3689, "lr": 6.780504179127735e-06, "epoch": 1.3392857142857144, "percentage": 44.64, "elapsed_time": "5:36:03", "remaining_time": "6:56:42"}
+{"current_steps": 151, "total_steps": 336, "loss": 0.3686, "lr": 6.731805183393696e-06, "epoch": 1.3482142857142856, "percentage": 44.94, "elapsed_time": "5:42:55", "remaining_time": "7:00:08"}
+{"current_steps": 152, "total_steps": 336, "loss": 0.3685, "lr": 6.682918782744033e-06, "epoch": 1.3571428571428572, "percentage": 45.24, "elapsed_time": "5:45:16", "remaining_time": "6:57:58"}
+{"current_steps": 153, "total_steps": 336, "loss": 0.3626, "lr": 6.63385026735334e-06, "epoch": 1.3660714285714286, "percentage": 45.54, "elapsed_time": "5:47:38", "remaining_time": "6:55:48"}
+{"current_steps": 154, "total_steps": 336, "loss": 0.3635, "lr": 6.584604947103515e-06, "epoch": 1.375, "percentage": 45.83, "elapsed_time": "5:49:58", "remaining_time": "6:53:36"}
+{"current_steps": 155, "total_steps": 336, "loss": 0.3699, "lr": 6.535188151009143e-06, "epoch": 1.3839285714285714, "percentage": 46.13, "elapsed_time": "5:52:19", "remaining_time": "6:51:25"}
+{"current_steps": 156, "total_steps": 336, "loss": 0.362, "lr": 6.4856052266408375e-06, "epoch": 1.3928571428571428, "percentage": 46.43, "elapsed_time": "5:54:41", "remaining_time": "6:49:15"}
+{"current_steps": 157, "total_steps": 336, "loss": 0.3652, "lr": 6.4358615395465506e-06, "epoch": 1.4017857142857144, "percentage": 46.73, "elapsed_time": "5:57:05", "remaining_time": "6:47:07"}
+{"current_steps": 158, "total_steps": 336, "loss": 0.3703, "lr": 6.385962472670953e-06, "epoch": 1.4107142857142856, "percentage": 47.02, "elapsed_time": "5:59:28", "remaining_time": "6:44:59"}
+{"current_steps": 159, "total_steps": 336, "loss": 0.3624, "lr": 6.335913425772926e-06, "epoch": 1.4196428571428572, "percentage": 47.32, "elapsed_time": "6:01:50", "remaining_time": "6:42:48"}
+{"current_steps": 160, "total_steps": 336, "loss": 0.3663, "lr": 6.28571981484123e-06, "epoch": 1.4285714285714286, "percentage": 47.62, "elapsed_time": "6:04:10", "remaining_time": "6:40:35"}
+{"current_steps": 161, "total_steps": 336, "loss": 0.3708, "lr": 6.235387071508427e-06, "epoch": 1.4375, "percentage": 47.92, "elapsed_time": "6:06:31", "remaining_time": "6:38:24"}
+{"current_steps": 162, "total_steps": 336, "loss": 0.3703, "lr": 6.184920642463095e-06, "epoch": 1.4464285714285714, "percentage": 48.21, "elapsed_time": "6:08:53", "remaining_time": "6:36:12"}
+{"current_steps": 163, "total_steps": 336, "loss": 0.3731, "lr": 6.1343259888604335e-06, "epoch": 1.4553571428571428, "percentage": 48.51, "elapsed_time": "6:11:16", "remaining_time": "6:34:03"}
+{"current_steps": 164, "total_steps": 336, "loss": 0.3677, "lr": 6.083608585731283e-06, "epoch": 1.4642857142857144, "percentage": 48.81, "elapsed_time": "6:13:41", "remaining_time": "6:31:55"}
+{"current_steps": 165, "total_steps": 336, "loss": 0.3708, "lr": 6.032773921389655e-06, "epoch": 1.4732142857142856, "percentage": 49.11, "elapsed_time": "6:16:05", "remaining_time": "6:29:46"}
+{"current_steps": 166, "total_steps": 336, "loss": 0.3665, "lr": 5.9818274968388225e-06, "epoch": 1.4821428571428572, "percentage": 49.4, "elapsed_time": "6:18:29", "remaining_time": "6:27:36"}
+{"current_steps": 167, "total_steps": 336, "loss": 0.3625, "lr": 5.930774825176034e-06, "epoch": 1.4910714285714286, "percentage": 49.7, "elapsed_time": "6:20:49", "remaining_time": "6:25:23"}
+{"current_steps": 168, "total_steps": 336, "loss": 0.3699, "lr": 5.879621430995927e-06, "epoch": 1.5, "percentage": 50.0, "elapsed_time": "6:23:06", "remaining_time": "6:23:06"}
+{"current_steps": 169, "total_steps": 336, "loss": 0.3678, "lr": 5.8283728497926865e-06, "epoch": 1.5089285714285714, "percentage": 50.3, "elapsed_time": "6:25:25", "remaining_time": "6:20:51"}
+{"current_steps": 170, "total_steps": 336, "loss": 0.3677, "lr": 5.777034627361025e-06, "epoch": 1.5178571428571428, "percentage": 50.6, "elapsed_time": "6:27:46", "remaining_time": "6:18:38"}
+{"current_steps": 171, "total_steps": 336, "loss": 0.3697, "lr": 5.725612319196065e-06, "epoch": 1.5267857142857144, "percentage": 50.89, "elapsed_time": "6:30:04", "remaining_time": "6:16:23"}
+{"current_steps": 172, "total_steps": 336, "loss": 0.3696, "lr": 5.674111489892144e-06, "epoch": 1.5357142857142856, "percentage": 51.19, "elapsed_time": "6:32:25", "remaining_time": "6:14:10"}
+{"current_steps": 173, "total_steps": 336, "loss": 0.3744, "lr": 5.622537712540664e-06, "epoch": 1.5446428571428572, "percentage": 51.49, "elapsed_time": "6:34:48", "remaining_time": "6:11:59"}
+{"current_steps": 174, "total_steps": 336, "loss": 0.3668, "lr": 5.570896568126994e-06, "epoch": 1.5535714285714286, "percentage": 51.79, "elapsed_time": "6:37:15", "remaining_time": "6:09:51"}
+{"current_steps": 175, "total_steps": 336, "loss": 0.3649, "lr": 5.5191936449265345e-06, "epoch": 1.5625, "percentage": 52.08, "elapsed_time": "6:39:30", "remaining_time": "6:07:33"}
+{"current_steps": 176, "total_steps": 336, "loss": 0.3683, "lr": 5.4674345379e-06, "epoch": 1.5714285714285714, "percentage": 52.38, "elapsed_time": "6:41:47", "remaining_time": "6:05:16"}
+{"current_steps": 177, "total_steps": 336, "loss": 0.3658, "lr": 5.415624848087959e-06, "epoch": 1.5803571428571428, "percentage": 52.68, "elapsed_time": "6:44:08", "remaining_time": "6:03:02"}
+{"current_steps": 178, "total_steps": 336, "loss": 0.3663, "lr": 5.36377018200472e-06, "epoch": 1.5892857142857144, "percentage": 52.98, "elapsed_time": "6:46:27", "remaining_time": "6:00:47"}
+{"current_steps": 179, "total_steps": 336, "loss": 0.3618, "lr": 5.311876151031642e-06, "epoch": 1.5982142857142856, "percentage": 53.27, "elapsed_time": "6:48:48", "remaining_time": "5:58:34"}
+{"current_steps": 180, "total_steps": 336, "loss": 0.3707, "lr": 5.259948370809902e-06, "epoch": 1.6071428571428572, "percentage": 53.57, "elapsed_time": "6:51:11", "remaining_time": "5:56:22"}
+{"current_steps": 181, "total_steps": 336, "loss": 0.3647, "lr": 5.207992460632805e-06, "epoch": 1.6160714285714286, "percentage": 53.87, "elapsed_time": "6:53:35", "remaining_time": "5:54:10"}
+{"current_steps": 182, "total_steps": 336, "loss": 0.3646, "lr": 5.156014042837696e-06, "epoch": 1.625, "percentage": 54.17, "elapsed_time": "6:55:58", "remaining_time": "5:51:58"}
+{"current_steps": 183, "total_steps": 336, "loss": 0.3603, "lr": 5.104018742197557e-06, "epoch": 1.6339285714285714, "percentage": 54.46, "elapsed_time": "6:58:16", "remaining_time": "5:49:42"}
+{"current_steps": 184, "total_steps": 336, "loss": 0.3596, "lr": 5.052012185312322e-06, "epoch": 1.6428571428571428, "percentage": 54.76, "elapsed_time": "7:00:36", "remaining_time": "5:47:27"}
+{"current_steps": 185, "total_steps": 336, "loss": 0.3646, "lr": 5e-06, "epoch": 1.6517857142857144, "percentage": 55.06, "elapsed_time": "7:02:56", "remaining_time": "5:45:12"}
+{"current_steps": 186, "total_steps": 336, "loss": 0.3616, "lr": 4.94798781468768e-06, "epoch": 1.6607142857142856, "percentage": 55.36, "elapsed_time": "7:05:17", "remaining_time": "5:42:58"}
+{"current_steps": 187, "total_steps": 336, "loss": 0.366, "lr": 4.895981257802444e-06, "epoch": 1.6696428571428572, "percentage": 55.65, "elapsed_time": "7:07:39", "remaining_time": "5:40:45"}
+{"current_steps": 188, "total_steps": 336, "loss": 0.3613, "lr": 4.843985957162304e-06, "epoch": 1.6785714285714286, "percentage": 55.95, "elapsed_time": "7:10:01", "remaining_time": "5:38:31"}
+{"current_steps": 189, "total_steps": 336, "loss": 0.3614, "lr": 4.792007539367198e-06, "epoch": 1.6875, "percentage": 56.25, "elapsed_time": "7:12:25", "remaining_time": "5:36:19"}
+{"current_steps": 190, "total_steps": 336, "loss": 0.3665, "lr": 4.740051629190099e-06, "epoch": 1.6964285714285714, "percentage": 56.55, "elapsed_time": "7:14:48", "remaining_time": "5:34:06"}
+{"current_steps": 191, "total_steps": 336, "loss": 0.3605, "lr": 4.6881238489683596e-06, "epoch": 1.7053571428571428, "percentage": 56.85, "elapsed_time": "7:17:07", "remaining_time": "5:31:50"}
+{"current_steps": 192, "total_steps": 336, "loss": 0.3687, "lr": 4.636229817995281e-06, "epoch": 1.7142857142857144, "percentage": 57.14, "elapsed_time": "7:19:26", "remaining_time": "5:29:34"}
+{"current_steps": 193, "total_steps": 336, "loss": 0.37, "lr": 4.584375151912043e-06, "epoch": 1.7232142857142856, "percentage": 57.44, "elapsed_time": "7:21:47", "remaining_time": "5:27:20"}
+{"current_steps": 194, "total_steps": 336, "loss": 0.3699, "lr": 4.532565462099999e-06, "epoch": 1.7321428571428572, "percentage": 57.74, "elapsed_time": "7:24:07", "remaining_time": "5:25:05"}
+{"current_steps": 195, "total_steps": 336, "loss": 0.3634, "lr": 4.480806355073467e-06, "epoch": 1.7410714285714286, "percentage": 58.04, "elapsed_time": "7:26:30", "remaining_time": "5:22:51"}
+{"current_steps": 196, "total_steps": 336, "loss": 0.3647, "lr": 4.429103431873009e-06, "epoch": 1.75, "percentage": 58.33, "elapsed_time": "7:28:53", "remaining_time": "5:20:38"}
+{"current_steps": 197, "total_steps": 336, "loss": 0.3685, "lr": 4.377462287459338e-06, "epoch": 1.7589285714285714, "percentage": 58.63, "elapsed_time": "7:31:15", "remaining_time": "5:18:23"}
+{"current_steps": 198, "total_steps": 336, "loss": 0.359, "lr": 4.3258885101078565e-06, "epoch": 1.7678571428571428, "percentage": 58.93, "elapsed_time": "7:33:37", "remaining_time": "5:16:09"}
+{"current_steps": 199, "total_steps": 336, "loss": 0.3589, "lr": 4.274387680803936e-06, "epoch": 1.7767857142857144, "percentage": 59.23, "elapsed_time": "7:35:55", "remaining_time": "5:13:52"}
+{"current_steps": 200, "total_steps": 336, "loss": 0.3657, "lr": 4.2229653726389765e-06, "epoch": 1.7857142857142856, "percentage": 59.52, "elapsed_time": "7:38:17", "remaining_time": "5:11:38"}
+{"current_steps": 201, "total_steps": 336, "loss": 0.3637, "lr": 4.171627150207314e-06, "epoch": 1.7946428571428572, "percentage": 59.82, "elapsed_time": "7:44:55", "remaining_time": "5:12:16"}
+{"current_steps": 202, "total_steps": 336, "loss": 0.3655, "lr": 4.120378569004074e-06, "epoch": 1.8035714285714286, "percentage": 60.12, "elapsed_time": "7:47:18", "remaining_time": "5:09:59"}
+{"current_steps": 203, "total_steps": 336, "loss": 0.3662, "lr": 4.069225174823968e-06, "epoch": 1.8125, "percentage": 60.42, "elapsed_time": "7:49:42", "remaining_time": "5:07:44"}
+{"current_steps": 204, "total_steps": 336, "loss": 0.3612, "lr": 4.018172503161179e-06, "epoch": 1.8214285714285714, "percentage": 60.71, "elapsed_time": "7:52:05", "remaining_time": "5:05:28"}
+{"current_steps": 205, "total_steps": 336, "loss": 0.3594, "lr": 3.967226078610346e-06, "epoch": 1.8303571428571428, "percentage": 61.01, "elapsed_time": "7:54:25", "remaining_time": "5:03:10"}
+{"current_steps": 206, "total_steps": 336, "loss": 0.3685, "lr": 3.9163914142687185e-06, "epoch": 1.8392857142857144, "percentage": 61.31, "elapsed_time": "7:56:45", "remaining_time": "5:00:52"}
+{"current_steps": 207, "total_steps": 336, "loss": 0.3582, "lr": 3.865674011139567e-06, "epoch": 1.8482142857142856, "percentage": 61.61, "elapsed_time": "7:59:07", "remaining_time": "4:58:34"}
+{"current_steps": 208, "total_steps": 336, "loss": 0.3641, "lr": 3.815079357536907e-06, "epoch": 1.8571428571428572, "percentage": 61.9, "elapsed_time": "8:01:30", "remaining_time": "4:56:18"}
+{"current_steps": 209, "total_steps": 336, "loss": 0.3665, "lr": 3.7646129284915754e-06, "epoch": 1.8660714285714286, "percentage": 62.2, "elapsed_time": "8:03:52", "remaining_time": "4:54:01"}
+{"current_steps": 210, "total_steps": 336, "loss": 0.3612, "lr": 3.714280185158771e-06, "epoch": 1.875, "percentage": 62.5, "elapsed_time": "8:06:19", "remaining_time": "4:51:47"}
+{"current_steps": 211, "total_steps": 336, "loss": 0.3629, "lr": 3.6640865742270756e-06, "epoch": 1.8839285714285714, "percentage": 62.8, "elapsed_time": "8:08:40", "remaining_time": "4:49:30"}
+{"current_steps": 212, "total_steps": 336, "loss": 0.3646, "lr": 3.614037527329048e-06, "epoch": 1.8928571428571428, "percentage": 63.1, "elapsed_time": "8:11:04", "remaining_time": "4:47:13"}
+{"current_steps": 213, "total_steps": 336, "loss": 0.3607, "lr": 3.5641384604534503e-06, "epoch": 1.9017857142857144, "percentage": 63.39, "elapsed_time": "8:13:23", "remaining_time": "4:44:55"}
+{"current_steps": 214, "total_steps": 336, "loss": 0.3613, "lr": 3.5143947733591633e-06, "epoch": 1.9107142857142856, "percentage": 63.69, "elapsed_time": "8:15:43", "remaining_time": "4:42:36"}
+{"current_steps": 215, "total_steps": 336, "loss": 0.365, "lr": 3.4648118489908588e-06, "epoch": 1.9196428571428572, "percentage": 63.99, "elapsed_time": "8:18:09", "remaining_time": "4:40:21"}
+{"current_steps": 216, "total_steps": 336, "loss": 0.3644, "lr": 3.4153950528964867e-06, "epoch": 1.9285714285714286, "percentage": 64.29, "elapsed_time": "8:20:28", "remaining_time": "4:38:02"}
+{"current_steps": 217, "total_steps": 336, "loss": 0.3662, "lr": 3.366149732646661e-06, "epoch": 1.9375, "percentage": 64.58, "elapsed_time": "8:22:52", "remaining_time": "4:35:46"}
+{"current_steps": 218, "total_steps": 336, "loss": 0.3598, "lr": 3.3170812172559695e-06, "epoch": 1.9464285714285714, "percentage": 64.88, "elapsed_time": "8:25:19", "remaining_time": "4:33:31"}
+{"current_steps": 219, "total_steps": 336, "loss": 0.3625, "lr": 3.268194816606305e-06, "epoch": 1.9553571428571428, "percentage": 65.18, "elapsed_time": "8:27:42", "remaining_time": "4:31:14"}
+{"current_steps": 220, "total_steps": 336, "loss": 0.364, "lr": 3.2194958208722656e-06, "epoch": 1.9642857142857144, "percentage": 65.48, "elapsed_time": "8:30:03", "remaining_time": "4:28:56"}
+{"current_steps": 221, "total_steps": 336, "loss": 0.3618, "lr": 3.1709894999486828e-06, "epoch": 1.9732142857142856, "percentage": 65.77, "elapsed_time": "8:32:23", "remaining_time": "4:26:38"}
+{"current_steps": 222, "total_steps": 336, "loss": 0.3627, "lr": 3.1226811028803514e-06, "epoch": 1.9821428571428572, "percentage": 66.07, "elapsed_time": "8:34:43", "remaining_time": "4:24:19"}
+{"current_steps": 223, "total_steps": 336, "loss": 0.3661, "lr": 3.0745758572940044e-06, "epoch": 1.9910714285714286, "percentage": 66.37, "elapsed_time": "8:37:06", "remaining_time": "4:22:01"}
+{"current_steps": 224, "total_steps": 336, "loss": 0.3518, "lr": 3.0266789688326187e-06, "epoch": 2.0, "percentage": 66.67, "elapsed_time": "8:39:28", "remaining_time": "4:19:44"}
+{"current_steps": 225, "total_steps": 336, "loss": 0.3488, "lr": 2.978995620592092e-06, "epoch": 2.0089285714285716, "percentage": 66.96, "elapsed_time": "8:41:53", "remaining_time": "4:17:28"}
+{"current_steps": 226, "total_steps": 336, "loss": 0.3498, "lr": 2.9315309725603596e-06, "epoch": 2.017857142857143, "percentage": 67.26, "elapsed_time": "8:44:20", "remaining_time": "4:15:12"}
+{"current_steps": 227, "total_steps": 336, "loss": 0.3411, "lr": 2.884290161059017e-06, "epoch": 2.0267857142857144, "percentage": 67.56, "elapsed_time": "8:46:44", "remaining_time": "4:12:55"}
+{"current_steps": 228, "total_steps": 336, "loss": 0.354, "lr": 2.8372782981874964e-06, "epoch": 2.0357142857142856, "percentage": 67.86, "elapsed_time": "8:49:08", "remaining_time": "4:10:38"}
+{"current_steps": 229, "total_steps": 336, "loss": 0.3454, "lr": 2.7905004712698646e-06, "epoch": 2.044642857142857, "percentage": 68.15, "elapsed_time": "8:51:32", "remaining_time": "4:08:21"}
+{"current_steps": 230, "total_steps": 336, "loss": 0.3528, "lr": 2.7439617423043146e-06, "epoch": 2.0535714285714284, "percentage": 68.45, "elapsed_time": "8:53:57", "remaining_time": "4:06:04"}
+{"current_steps": 231, "total_steps": 336, "loss": 0.3489, "lr": 2.697667147415383e-06, "epoch": 2.0625, "percentage": 68.75, "elapsed_time": "8:56:21", "remaining_time": "4:03:48"}
+{"current_steps": 232, "total_steps": 336, "loss": 0.3436, "lr": 2.6516216963089698e-06, "epoch": 2.0714285714285716, "percentage": 69.05, "elapsed_time": "8:58:49", "remaining_time": "4:01:32"}
+{"current_steps": 233, "total_steps": 336, "loss": 0.3432, "lr": 2.605830371730229e-06, "epoch": 2.080357142857143, "percentage": 69.35, "elapsed_time": "9:01:16", "remaining_time": "3:59:16"}
+{"current_steps": 234, "total_steps": 336, "loss": 0.3424, "lr": 2.560298128924358e-06, "epoch": 2.0892857142857144, "percentage": 69.64, "elapsed_time": "9:03:43", "remaining_time": "3:57:00"}
+{"current_steps": 235, "total_steps": 336, "loss": 0.3421, "lr": 2.5150298951003783e-06, "epoch": 2.0982142857142856, "percentage": 69.94, "elapsed_time": "9:06:07", "remaining_time": "3:54:43"}
+{"current_steps": 236, "total_steps": 336, "loss": 0.3458, "lr": 2.470030568897938e-06, "epoch": 2.107142857142857, "percentage": 70.24, "elapsed_time": "9:08:30", "remaining_time": "3:52:25"}
+{"current_steps": 237, "total_steps": 336, "loss": 0.349, "lr": 2.425305019857222e-06, "epoch": 2.1160714285714284, "percentage": 70.54, "elapsed_time": "9:10:54", "remaining_time": "3:50:07"}
+{"current_steps": 238, "total_steps": 336, "loss": 0.3507, "lr": 2.3808580878919948e-06, "epoch": 2.125, "percentage": 70.83, "elapsed_time": "9:13:19", "remaining_time": "3:47:50"}
+{"current_steps": 239, "total_steps": 336, "loss": 0.3429, "lr": 2.336694582765857e-06, "epoch": 2.1339285714285716, "percentage": 71.13, "elapsed_time": "9:15:46", "remaining_time": "3:45:33"}
+{"current_steps": 240, "total_steps": 336, "loss": 0.3459, "lr": 2.2928192835717642e-06, "epoch": 2.142857142857143, "percentage": 71.43, "elapsed_time": "9:18:12", "remaining_time": "3:43:17"}
+{"current_steps": 241, "total_steps": 336, "loss": 0.3471, "lr": 2.2492369382148634e-06, "epoch": 2.1517857142857144, "percentage": 71.73, "elapsed_time": "9:20:41", "remaining_time": "3:41:01"}
+{"current_steps": 242, "total_steps": 336, "loss": 0.3456, "lr": 2.205952262898704e-06, "epoch": 2.1607142857142856, "percentage": 72.02, "elapsed_time": "9:23:07", "remaining_time": "3:38:43"}
+{"current_steps": 243, "total_steps": 336, "loss": 0.3459, "lr": 2.1629699416148832e-06, "epoch": 2.169642857142857, "percentage": 72.32, "elapsed_time": "9:25:30", "remaining_time": "3:36:25"}
+{"current_steps": 244, "total_steps": 336, "loss": 0.3513, "lr": 2.120294625636171e-06, "epoch": 2.1785714285714284, "percentage": 72.62, "elapsed_time": "9:27:55", "remaining_time": "3:34:08"}
+{"current_steps": 245, "total_steps": 336, "loss": 0.3423, "lr": 2.077930933013182e-06, "epoch": 2.1875, "percentage": 72.92, "elapsed_time": "9:30:19", "remaining_time": "3:31:49"}
+{"current_steps": 246, "total_steps": 336, "loss": 0.3511, "lr": 2.0358834480746363e-06, "epoch": 2.1964285714285716, "percentage": 73.21, "elapsed_time": "9:32:43", "remaining_time": "3:29:32"}
+{"current_steps": 247, "total_steps": 336, "loss": 0.3412, "lr": 1.9941567209312768e-06, "epoch": 2.205357142857143, "percentage": 73.51, "elapsed_time": "9:35:11", "remaining_time": "3:27:15"}
+{"current_steps": 248, "total_steps": 336, "loss": 0.3439, "lr": 1.9527552669834797e-06, "epoch": 2.2142857142857144, "percentage": 73.81, "elapsed_time": "9:37:39", "remaining_time": "3:24:58"}
+{"current_steps": 249, "total_steps": 336, "loss": 0.3429, "lr": 1.911683566432633e-06, "epoch": 2.2232142857142856, "percentage": 74.11, "elapsed_time": "9:40:08", "remaining_time": "3:22:42"}
+{"current_steps": 250, "total_steps": 336, "loss": 0.3399, "lr": 1.8709460637963123e-06, "epoch": 2.232142857142857, "percentage": 74.4, "elapsed_time": "9:42:32", "remaining_time": "3:20:23"}
+{"current_steps": 251, "total_steps": 336, "loss": 0.3506, "lr": 1.8305471674273262e-06, "epoch": 2.2410714285714284, "percentage": 74.7, "elapsed_time": "9:49:13", "remaining_time": "3:19:32"}
+{"current_steps": 252, "total_steps": 336, "loss": 0.3421, "lr": 1.7904912490366723e-06, "epoch": 2.25, "percentage": 75.0, "elapsed_time": "9:51:36", "remaining_time": "3:17:12"}
+{"current_steps": 253, "total_steps": 336, "loss": 0.3516, "lr": 1.7507826432204571e-06, "epoch": 2.2589285714285716, "percentage": 75.3, "elapsed_time": "9:54:00", "remaining_time": "3:14:52"}
+{"current_steps": 254, "total_steps": 336, "loss": 0.3476, "lr": 1.711425646990838e-06, "epoch": 2.267857142857143, "percentage": 75.6, "elapsed_time": "9:56:28", "remaining_time": "3:12:33"}
+{"current_steps": 255, "total_steps": 336, "loss": 0.3456, "lr": 1.6724245193110178e-06, "epoch": 2.2767857142857144, "percentage": 75.89, "elapsed_time": "9:58:55", "remaining_time": "3:10:14"}
+{"current_steps": 256, "total_steps": 336, "loss": 0.3546, "lr": 1.6337834806343783e-06, "epoch": 2.2857142857142856, "percentage": 76.19, "elapsed_time": "10:01:22", "remaining_time": "3:07:55"}
+{"current_steps": 257, "total_steps": 336, "loss": 0.3518, "lr": 1.595506712447768e-06, "epoch": 2.294642857142857, "percentage": 76.49, "elapsed_time": "10:03:48", "remaining_time": "3:05:36"}
+{"current_steps": 258, "total_steps": 336, "loss": 0.3488, "lr": 1.557598356819e-06, "epoch": 2.3035714285714284, "percentage": 76.79, "elapsed_time": "10:06:11", "remaining_time": "3:03:15"}
+{"current_steps": 259, "total_steps": 336, "loss": 0.3471, "lr": 1.5200625159486322e-06, "epoch": 2.3125, "percentage": 77.08, "elapsed_time": "10:08:30", "remaining_time": "3:00:54"}
+{"current_steps": 260, "total_steps": 336, "loss": 0.3419, "lr": 1.482903251726049e-06, "epoch": 2.3214285714285716, "percentage": 77.38, "elapsed_time": "10:10:56", "remaining_time": "2:58:34"}
+{"current_steps": 261, "total_steps": 336, "loss": 0.3458, "lr": 1.446124585289913e-06, "epoch": 2.330357142857143, "percentage": 77.68, "elapsed_time": "10:13:20", "remaining_time": "2:56:14"}
+{"current_steps": 262, "total_steps": 336, "loss": 0.3493, "lr": 1.409730496593016e-06, "epoch": 2.3392857142857144, "percentage": 77.98, "elapsed_time": "10:15:47", "remaining_time": "2:53:55"}
+{"current_steps": 263, "total_steps": 336, "loss": 0.3503, "lr": 1.3737249239716043e-06, "epoch": 2.3482142857142856, "percentage": 78.27, "elapsed_time": "10:18:16", "remaining_time": "2:51:36"}
+{"current_steps": 264, "total_steps": 336, "loss": 0.3493, "lr": 1.3381117637191887e-06, "epoch": 2.357142857142857, "percentage": 78.57, "elapsed_time": "10:20:45", "remaining_time": "2:49:17"}
+{"current_steps": 265, "total_steps": 336, "loss": 0.3467, "lr": 1.3028948696649164e-06, "epoch": 2.3660714285714284, "percentage": 78.87, "elapsed_time": "10:23:11", "remaining_time": "2:46:57"}
+{"current_steps": 266, "total_steps": 336, "loss": 0.3443, "lr": 1.2680780527565313e-06, "epoch": 2.375, "percentage": 79.17, "elapsed_time": "10:25:35", "remaining_time": "2:44:37"}
+{"current_steps": 267, "total_steps": 336, "loss": 0.3461, "lr": 1.2336650806479826e-06, "epoch": 2.3839285714285716, "percentage": 79.46, "elapsed_time": "10:28:00", "remaining_time": "2:42:17"}
+{"current_steps": 268, "total_steps": 336, "loss": 0.3543, "lr": 1.1996596772917091e-06, "epoch": 2.392857142857143, "percentage": 79.76, "elapsed_time": "10:30:26", "remaining_time": "2:39:57"}
+{"current_steps": 269, "total_steps": 336, "loss": 0.3471, "lr": 1.1660655225356533e-06, "epoch": 2.4017857142857144, "percentage": 80.06, "elapsed_time": "10:32:53", "remaining_time": "2:37:38"}
+{"current_steps": 270, "total_steps": 336, "loss": 0.3474, "lr": 1.132886251725061e-06, "epoch": 2.4107142857142856, "percentage": 80.36, "elapsed_time": "10:35:21", "remaining_time": "2:35:18"}
+{"current_steps": 271, "total_steps": 336, "loss": 0.3451, "lr": 1.1001254553090812e-06, "epoch": 2.419642857142857, "percentage": 80.65, "elapsed_time": "10:37:50", "remaining_time": "2:32:59"}
+{"current_steps": 272, "total_steps": 336, "loss": 0.3474, "lr": 1.0677866784522317e-06, "epoch": 2.4285714285714284, "percentage": 80.95, "elapsed_time": "10:40:19", "remaining_time": "2:30:39"}
+{"current_steps": 273, "total_steps": 336, "loss": 0.3459, "lr": 1.0358734206507642e-06, "epoch": 2.4375, "percentage": 81.25, "elapsed_time": "10:42:45", "remaining_time": "2:28:19"}
+{"current_steps": 274, "total_steps": 336, "loss": 0.3387, "lr": 1.004389135353972e-06, "epoch": 2.4464285714285716, "percentage": 81.55, "elapsed_time": "10:45:08", "remaining_time": "2:25:58"}
+{"current_steps": 275, "total_steps": 336, "loss": 0.347, "lr": 9.733372295904776e-07, "epoch": 2.455357142857143, "percentage": 81.85, "elapsed_time": "10:47:32", "remaining_time": "2:23:38"}
+{"current_steps": 276, "total_steps": 336, "loss": 0.3467, "lr": 9.427210635995482e-07, "epoch": 2.4642857142857144, "percentage": 82.14, "elapsed_time": "10:49:55", "remaining_time": "2:21:17"}
+{"current_steps": 277, "total_steps": 336, "loss": 0.3524, "lr": 9.125439504674699e-07, "epoch": 2.4732142857142856, "percentage": 82.44, "elapsed_time": "10:52:23", "remaining_time": "2:18:57"}
+{"current_steps": 278, "total_steps": 336, "loss": 0.351, "lr": 8.828091557690288e-07, "epoch": 2.482142857142857, "percentage": 82.74, "elapsed_time": "10:54:50", "remaining_time": "2:16:37"}
+{"current_steps": 279, "total_steps": 336, "loss": 0.3455, "lr": 8.535198972141295e-07, "epoch": 2.4910714285714284, "percentage": 83.04, "elapsed_time": "10:57:20", "remaining_time": "2:14:17"}
+{"current_steps": 280, "total_steps": 336, "loss": 0.3441, "lr": 8.246793442995954e-07, "epoch": 2.5, "percentage": 83.33, "elapsed_time": "10:59:49", "remaining_time": "2:11:57"}
+{"current_steps": 281, "total_steps": 336, "loss": 0.338, "lr": 7.962906179661872e-07, "epoch": 2.508928571428571, "percentage": 83.63, "elapsed_time": "11:02:13", "remaining_time": "2:09:37"}
+{"current_steps": 282, "total_steps": 336, "loss": 0.3423, "lr": 7.68356790260873e-07, "epoch": 2.517857142857143, "percentage": 83.93, "elapsed_time": "11:04:36", "remaining_time": "2:07:16"}
+{"current_steps": 283, "total_steps": 336, "loss": 0.3437, "lr": 7.408808840043913e-07, "epoch": 2.5267857142857144, "percentage": 84.23, "elapsed_time": "11:07:00", "remaining_time": "2:04:55"}
+{"current_steps": 284, "total_steps": 336, "loss": 0.3533, "lr": 7.138658724641417e-07, "epoch": 2.5357142857142856, "percentage": 84.52, "elapsed_time": "11:09:27", "remaining_time": "2:02:34"}
+{"current_steps": 285, "total_steps": 336, "loss": 0.3473, "lr": 6.873146790324359e-07, "epoch": 2.544642857142857, "percentage": 84.82, "elapsed_time": "11:11:54", "remaining_time": "2:00:14"}
+{"current_steps": 286, "total_steps": 336, "loss": 0.3466, "lr": 6.612301769101464e-07, "epoch": 2.553571428571429, "percentage": 85.12, "elapsed_time": "11:14:23", "remaining_time": "1:57:54"}
+{"current_steps": 287, "total_steps": 336, "loss": 0.3526, "lr": 6.356151887957901e-07, "epoch": 2.5625, "percentage": 85.42, "elapsed_time": "11:16:51", "remaining_time": "1:55:33"}
+{"current_steps": 288, "total_steps": 336, "loss": 0.3436, "lr": 6.104724865800665e-07, "epoch": 2.571428571428571, "percentage": 85.71, "elapsed_time": "11:19:18", "remaining_time": "1:53:13"}
+{"current_steps": 289, "total_steps": 336, "loss": 0.3435, "lr": 5.858047910459109e-07, "epoch": 2.580357142857143, "percentage": 86.01, "elapsed_time": "11:21:42", "remaining_time": "1:50:51"}
+{"current_steps": 290, "total_steps": 336, "loss": 0.3425, "lr": 5.616147715740611e-07, "epoch": 2.5892857142857144, "percentage": 86.31, "elapsed_time": "11:24:07", "remaining_time": "1:48:30"}
+{"current_steps": 291, "total_steps": 336, "loss": 0.3437, "lr": 5.379050458541995e-07, "epoch": 2.5982142857142856, "percentage": 86.61, "elapsed_time": "11:26:35", "remaining_time": "1:46:10"}
+{"current_steps": 292, "total_steps": 336, "loss": 0.3477, "lr": 5.146781796016798e-07, "epoch": 2.607142857142857, "percentage": 86.9, "elapsed_time": "11:29:15", "remaining_time": "1:43:51"}
+{"current_steps": 293, "total_steps": 336, "loss": 0.3452, "lr": 4.919366862798807e-07, "epoch": 2.616071428571429, "percentage": 87.2, "elapsed_time": "11:31:42", "remaining_time": "1:41:30"}
+{"current_steps": 294, "total_steps": 336, "loss": 0.3487, "lr": 4.696830268282204e-07, "epoch": 2.625, "percentage": 87.5, "elapsed_time": "11:34:11", "remaining_time": "1:39:10"}
+{"current_steps": 295, "total_steps": 336, "loss": 0.3458, "lr": 4.479196093958421e-07, "epoch": 2.633928571428571, "percentage": 87.8, "elapsed_time": "11:36:41", "remaining_time": "1:36:49"}
+{"current_steps": 296, "total_steps": 336, "loss": 0.3467, "lr": 4.2664878908102556e-07, "epoch": 2.642857142857143, "percentage": 88.1, "elapsed_time": "11:39:07", "remaining_time": "1:34:28"}
+{"current_steps": 297, "total_steps": 336, "loss": 0.3517, "lr": 4.058728676763313e-07, "epoch": 2.6517857142857144, "percentage": 88.39, "elapsed_time": "11:41:34", "remaining_time": "1:32:07"}
+{"current_steps": 298, "total_steps": 336, "loss": 0.3459, "lr": 3.855940934195146e-07, "epoch": 2.6607142857142856, "percentage": 88.69, "elapsed_time": "11:43:58", "remaining_time": "1:29:46"}
+{"current_steps": 299, "total_steps": 336, "loss": 0.3506, "lr": 3.6581466075023443e-07, "epoch": 2.669642857142857, "percentage": 88.99, "elapsed_time": "11:46:25", "remaining_time": "1:27:25"}
+{"current_steps": 300, "total_steps": 336, "loss": 0.3445, "lr": 3.4653671007259084e-07, "epoch": 2.678571428571429, "percentage": 89.29, "elapsed_time": "11:48:55", "remaining_time": "1:25:04"}
+{"current_steps": 301, "total_steps": 336, "loss": 0.3417, "lr": 3.2776232752349536e-07, "epoch": 2.6875, "percentage": 89.58, "elapsed_time": "11:55:44", "remaining_time": "1:23:13"}
+{"current_steps": 302, "total_steps": 336, "loss": 0.3439, "lr": 3.0949354474692937e-07, "epoch": 2.696428571428571, "percentage": 89.88, "elapsed_time": "11:58:09", "remaining_time": "1:20:51"}
+{"current_steps": 303, "total_steps": 336, "loss": 0.3439, "lr": 2.9173233867409057e-07, "epoch": 2.705357142857143, "percentage": 90.18, "elapsed_time": "12:00:34", "remaining_time": "1:18:28"}
+{"current_steps": 304, "total_steps": 336, "loss": 0.3495, "lr": 2.7448063130946224e-07, "epoch": 2.7142857142857144, "percentage": 90.48, "elapsed_time": "12:03:00", "remaining_time": "1:16:06"}
+{"current_steps": 305, "total_steps": 336, "loss": 0.3417, "lr": 2.577402895228243e-07, "epoch": 2.7232142857142856, "percentage": 90.77, "elapsed_time": "12:05:26", "remaining_time": "1:13:43"}
+{"current_steps": 306, "total_steps": 336, "loss": 0.343, "lr": 2.4151312484723465e-07, "epoch": 2.732142857142857, "percentage": 91.07, "elapsed_time": "12:07:51", "remaining_time": "1:11:21"}
+{"current_steps": 307, "total_steps": 336, "loss": 0.3462, "lr": 2.2580089328299747e-07, "epoch": 2.741071428571429, "percentage": 91.37, "elapsed_time": "12:10:16", "remaining_time": "1:08:59"}
+{"current_steps": 308, "total_steps": 336, "loss": 0.3443, "lr": 2.106052951076365e-07, "epoch": 2.75, "percentage": 91.67, "elapsed_time": "12:12:42", "remaining_time": "1:06:36"}
+{"current_steps": 309, "total_steps": 336, "loss": 0.345, "lr": 1.959279746919057e-07, "epoch": 2.758928571428571, "percentage": 91.96, "elapsed_time": "12:15:05", "remaining_time": "1:04:13"}
+{"current_steps": 310, "total_steps": 336, "loss": 0.3458, "lr": 1.8177052032184285e-07, "epoch": 2.767857142857143, "percentage": 92.26, "elapsed_time": "12:17:28", "remaining_time": "1:01:51"}
+{"current_steps": 311, "total_steps": 336, "loss": 0.3482, "lr": 1.6813446402689783e-07, "epoch": 2.7767857142857144, "percentage": 92.56, "elapsed_time": "12:19:52", "remaining_time": "0:59:28"}
+{"current_steps": 312, "total_steps": 336, "loss": 0.3412, "lr": 1.5502128141414496e-07, "epoch": 2.7857142857142856, "percentage": 92.86, "elapsed_time": "12:22:14", "remaining_time": "0:57:05"}
+{"current_steps": 313, "total_steps": 336, "loss": 0.3438, "lr": 1.4243239150860123e-07, "epoch": 2.794642857142857, "percentage": 93.15, "elapsed_time": "12:24:39", "remaining_time": "0:54:43"}
+{"current_steps": 314, "total_steps": 336, "loss": 0.3497, "lr": 1.303691565996712e-07, "epoch": 2.803571428571429, "percentage": 93.45, "elapsed_time": "12:27:05", "remaining_time": "0:52:20"}
+{"current_steps": 315, "total_steps": 336, "loss": 0.3449, "lr": 1.1883288209372512e-07, "epoch": 2.8125, "percentage": 93.75, "elapsed_time": "12:29:34", "remaining_time": "0:49:58"}
+{"current_steps": 316, "total_steps": 336, "loss": 0.3448, "lr": 1.0782481637284014e-07, "epoch": 2.821428571428571, "percentage": 94.05, "elapsed_time": "12:32:01", "remaining_time": "0:47:35"}
+{"current_steps": 317, "total_steps": 336, "loss": 0.3474, "lr": 9.734615065970454e-08, "epoch": 2.830357142857143, "percentage": 94.35, "elapsed_time": "12:34:29", "remaining_time": "0:45:13"}
+{"current_steps": 318, "total_steps": 336, "loss": 0.3484, "lr": 8.739801888871468e-08, "epoch": 2.8392857142857144, "percentage": 94.64, "elapsed_time": "12:36:54", "remaining_time": "0:42:50"}
+{"current_steps": 319, "total_steps": 336, "loss": 0.3508, "lr": 7.79814975832649e-08, "epoch": 2.8482142857142856, "percentage": 94.94, "elapsed_time": "12:39:19", "remaining_time": "0:40:27"}
+{"current_steps": 320, "total_steps": 336, "loss": 0.3438, "lr": 6.909760573925561e-08, "epoch": 2.857142857142857, "percentage": 95.24, "elapsed_time": "12:41:44", "remaining_time": "0:38:05"}
+{"current_steps": 321, "total_steps": 336, "loss": 0.3466, "lr": 6.074730471482049e-08, "epoch": 2.866071428571429, "percentage": 95.54, "elapsed_time": "12:44:10", "remaining_time": "0:35:42"}
+{"current_steps": 322, "total_steps": 336, "loss": 0.3441, "lr": 5.2931498126298495e-08, "epoch": 2.875, "percentage": 95.83, "elapsed_time": "12:46:36", "remaining_time": "0:33:19"}
+{"current_steps": 323, "total_steps": 336, "loss": 0.3475, "lr": 4.565103175044883e-08, "epoch": 2.883928571428571, "percentage": 96.13, "elapsed_time": "12:49:00", "remaining_time": "0:30:57"}
+{"current_steps": 324, "total_steps": 336, "loss": 0.3478, "lr": 3.890669343292464e-08, "epoch": 2.892857142857143, "percentage": 96.43, "elapsed_time": "12:51:27", "remaining_time": "0:28:34"}
+{"current_steps": 325, "total_steps": 336, "loss": 0.3431, "lr": 3.269921300301959e-08, "epoch": 2.9017857142857144, "percentage": 96.73, "elapsed_time": "12:53:53", "remaining_time": "0:26:11"}
+{"current_steps": 326, "total_steps": 336, "loss": 0.3441, "lr": 2.702926219468882e-08, "epoch": 2.9107142857142856, "percentage": 97.02, "elapsed_time": "12:56:19", "remaining_time": "0:23:48"}
+{"current_steps": 327, "total_steps": 336, "loss": 0.3538, "lr": 2.1897454573860388e-08, "epoch": 2.919642857142857, "percentage": 97.32, "elapsed_time": "12:58:43", "remaining_time": "0:21:25"}
+{"current_steps": 328, "total_steps": 336, "loss": 0.3431, "lr": 1.7304345472035634e-08, "epoch": 2.928571428571429, "percentage": 97.62, "elapsed_time": "13:01:08", "remaining_time": "0:19:03"}
+{"current_steps": 329, "total_steps": 336, "loss": 0.3412, "lr": 1.3250431926197793e-08, "epoch": 2.9375, "percentage": 97.92, "elapsed_time": "13:03:35", "remaining_time": "0:16:40"}
+{"current_steps": 330, "total_steps": 336, "loss": 0.346, "lr": 9.73615262502503e-09, "epoch": 2.946428571428571, "percentage": 98.21, "elapsed_time": "13:06:02", "remaining_time": "0:14:17"}
+{"current_steps": 331, "total_steps": 336, "loss": 0.3491, "lr": 6.761887861417293e-09, "epoch": 2.955357142857143, "percentage": 98.51, "elapsed_time": "13:08:28", "remaining_time": "0:11:54"}
+{"current_steps": 332, "total_steps": 336, "loss": 0.3433, "lr": 4.327959491344791e-09, "epoch": 2.9642857142857144, "percentage": 98.81, "elapsed_time": "13:10:56", "remaining_time": "0:09:31"}
+{"current_steps": 333, "total_steps": 336, "loss": 0.3467, "lr": 2.4346308990175205e-09, "epoch": 2.9732142857142856, "percentage": 99.11, "elapsed_time": "13:13:25", "remaining_time": "0:07:08"}
+{"current_steps": 334, "total_steps": 336, "loss": 0.343, "lr": 1.082106968385288e-09, "epoch": 2.982142857142857, "percentage": 99.4, "elapsed_time": "13:15:46", "remaining_time": "0:04:45"}
+{"current_steps": 335, "total_steps": 336, "loss": 0.3447, "lr": 2.7053406096433366e-10, "epoch": 2.991071428571429, "percentage": 99.7, "elapsed_time": "13:18:10", "remaining_time": "0:02:22"}
+{"current_steps": 336, "total_steps": 336, "loss": 0.3407, "lr": 0.0, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "13:20:36", "remaining_time": "0:00:00"}
+{"current_steps": 336, "total_steps": 336, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "13:24:55", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2394 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 336,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008928571428571428,
+      "grad_norm": 6.749195575714111,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 0.8158,
+      "step": 1
+    },
+    {
+      "epoch": 0.017857142857142856,
+      "grad_norm": 6.9488749504089355,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.8361,
+      "step": 2
+    },
+    {
+      "epoch": 0.026785714285714284,
+      "grad_norm": 7.039111137390137,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 0.8422,
+      "step": 3
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 6.606884002685547,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.8212,
+      "step": 4
+    },
+    {
+      "epoch": 0.044642857142857144,
+      "grad_norm": 5.828409194946289,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 0.8039,
+      "step": 5
+    },
+    {
+      "epoch": 0.05357142857142857,
+      "grad_norm": 3.9500532150268555,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.7598,
+      "step": 6
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 2.794076681137085,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 0.7132,
+      "step": 7
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 2.651289463043213,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.7006,
+      "step": 8
+    },
+    {
+      "epoch": 0.08035714285714286,
+      "grad_norm": 1.9226497411727905,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 0.6781,
+      "step": 9
+    },
+    {
+      "epoch": 0.08928571428571429,
+      "grad_norm": 1.8058496713638306,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.6716,
+      "step": 10
+    },
+    {
+      "epoch": 0.09821428571428571,
+      "grad_norm": 1.5553733110427856,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 0.6524,
+      "step": 11
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 2.1922216415405273,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.6202,
+      "step": 12
+    },
+    {
+      "epoch": 0.11607142857142858,
+      "grad_norm": 2.5917115211486816,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 0.6104,
+      "step": 13
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 2.4120588302612305,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.6036,
+      "step": 14
+    },
+    {
+      "epoch": 0.13392857142857142,
+      "grad_norm": 2.0048017501831055,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.5893,
+      "step": 15
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.3664730787277222,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.5752,
+      "step": 16
+    },
+    {
+      "epoch": 0.15178571428571427,
+      "grad_norm": 1.232675552368164,
+      "learning_rate": 5e-06,
+      "loss": 0.5527,
+      "step": 17
+    },
+    {
+      "epoch": 0.16071428571428573,
+      "grad_norm": 1.2626478672027588,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.5438,
+      "step": 18
+    },
+    {
+      "epoch": 0.16964285714285715,
+      "grad_norm": 1.1723309755325317,
+      "learning_rate": 5.588235294117647e-06,
+      "loss": 0.5295,
+      "step": 19
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 0.9530285000801086,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.5227,
+      "step": 20
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.8222931027412415,
+      "learning_rate": 6.176470588235295e-06,
+      "loss": 0.5146,
+      "step": 21
+    },
+    {
+      "epoch": 0.19642857142857142,
+      "grad_norm": 0.8412429094314575,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.503,
+      "step": 22
+    },
+    {
+      "epoch": 0.20535714285714285,
+      "grad_norm": 0.8323157429695129,
+      "learning_rate": 6.764705882352942e-06,
+      "loss": 0.499,
+      "step": 23
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.6052507758140564,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.4901,
+      "step": 24
+    },
+    {
+      "epoch": 0.22321428571428573,
+      "grad_norm": 0.5770189762115479,
+      "learning_rate": 7.352941176470589e-06,
+      "loss": 0.4882,
+      "step": 25
+    },
+    {
+      "epoch": 0.23214285714285715,
+      "grad_norm": 0.6112756729125977,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.4789,
+      "step": 26
+    },
+    {
+      "epoch": 0.24107142857142858,
+      "grad_norm": 0.6391553282737732,
+      "learning_rate": 7.941176470588236e-06,
+      "loss": 0.4792,
+      "step": 27
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48618483543395996,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.4665,
+      "step": 28
+    },
+    {
+      "epoch": 0.25892857142857145,
+      "grad_norm": 0.4175974726676941,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 0.4666,
+      "step": 29
+    },
+    {
+      "epoch": 0.26785714285714285,
+      "grad_norm": 0.5063344836235046,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.4644,
+      "step": 30
+    },
+    {
+      "epoch": 0.2767857142857143,
+      "grad_norm": 0.4095499813556671,
+      "learning_rate": 9.11764705882353e-06,
+      "loss": 0.4552,
+      "step": 31
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.44540026783943176,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.4611,
+      "step": 32
+    },
+    {
+      "epoch": 0.29464285714285715,
+      "grad_norm": 0.40555301308631897,
+      "learning_rate": 9.705882352941177e-06,
+      "loss": 0.449,
+      "step": 33
+    },
+    {
+      "epoch": 0.30357142857142855,
+      "grad_norm": 0.3954616189002991,
+      "learning_rate": 1e-05,
+      "loss": 0.4543,
+      "step": 34
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.4039325416088104,
+      "learning_rate": 9.999729465939036e-06,
+      "loss": 0.4449,
+      "step": 35
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 0.4171324372291565,
+      "learning_rate": 9.998917893031615e-06,
+      "loss": 0.4503,
+      "step": 36
+    },
+    {
+      "epoch": 0.33035714285714285,
+      "grad_norm": 0.323896586894989,
+      "learning_rate": 9.997565369100983e-06,
+      "loss": 0.4421,
+      "step": 37
+    },
+    {
+      "epoch": 0.3392857142857143,
+      "grad_norm": 0.3594343364238739,
+      "learning_rate": 9.995672040508656e-06,
+      "loss": 0.4453,
+      "step": 38
+    },
+    {
+      "epoch": 0.3482142857142857,
+      "grad_norm": 0.33965522050857544,
+      "learning_rate": 9.993238112138584e-06,
+      "loss": 0.4388,
+      "step": 39
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.33565545082092285,
+      "learning_rate": 9.990263847374976e-06,
+      "loss": 0.4407,
+      "step": 40
+    },
+    {
+      "epoch": 0.36607142857142855,
+      "grad_norm": 0.312674343585968,
+      "learning_rate": 9.986749568073804e-06,
+      "loss": 0.4419,
+      "step": 41
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.33573076128959656,
+      "learning_rate": 9.982695654527966e-06,
+      "loss": 0.438,
+      "step": 42
+    },
+    {
+      "epoch": 0.38392857142857145,
+      "grad_norm": 0.3331693708896637,
+      "learning_rate": 9.97810254542614e-06,
+      "loss": 0.4299,
+      "step": 43
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 0.3128435015678406,
+      "learning_rate": 9.972970737805312e-06,
+      "loss": 0.4248,
+      "step": 44
+    },
+    {
+      "epoch": 0.4017857142857143,
+      "grad_norm": 0.2929725646972656,
+      "learning_rate": 9.96730078699698e-06,
+      "loss": 0.4333,
+      "step": 45
+    },
+    {
+      "epoch": 0.4107142857142857,
+      "grad_norm": 0.27605733275413513,
+      "learning_rate": 9.961093306567076e-06,
+      "loss": 0.4266,
+      "step": 46
+    },
+    {
+      "epoch": 0.41964285714285715,
+      "grad_norm": 0.3021415174007416,
+      "learning_rate": 9.954348968249552e-06,
+      "loss": 0.4253,
+      "step": 47
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.26961633563041687,
+      "learning_rate": 9.947068501873702e-06,
+      "loss": 0.4229,
+      "step": 48
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.27512669563293457,
+      "learning_rate": 9.93925269528518e-06,
+      "loss": 0.423,
+      "step": 49
+    },
+    {
+      "epoch": 0.44642857142857145,
+      "grad_norm": 0.25938841700553894,
+      "learning_rate": 9.930902394260746e-06,
+      "loss": 0.4136,
+      "step": 50
+    },
+    {
+      "epoch": 0.45535714285714285,
+      "grad_norm": 0.24410875141620636,
+      "learning_rate": 9.922018502416736e-06,
+      "loss": 0.4238,
+      "step": 51
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 0.24541054666042328,
+      "learning_rate": 9.912601981111287e-06,
+      "loss": 0.4162,
+      "step": 52
+    },
+    {
+      "epoch": 0.4732142857142857,
+      "grad_norm": 0.23068013787269592,
+      "learning_rate": 9.902653849340296e-06,
+      "loss": 0.4115,
+      "step": 53
+    },
+    {
+      "epoch": 0.48214285714285715,
+      "grad_norm": 0.2482675462961197,
+      "learning_rate": 9.892175183627161e-06,
+      "loss": 0.4144,
+      "step": 54
+    },
+    {
+      "epoch": 0.49107142857142855,
+      "grad_norm": 0.23585791885852814,
+      "learning_rate": 9.881167117906276e-06,
+      "loss": 0.4163,
+      "step": 55
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2257727086544037,
+      "learning_rate": 9.869630843400331e-06,
+      "loss": 0.412,
+      "step": 56
+    },
+    {
+      "epoch": 0.5089285714285714,
+      "grad_norm": 0.2443915456533432,
+      "learning_rate": 9.8575676084914e-06,
+      "loss": 0.4196,
+      "step": 57
+    },
+    {
+      "epoch": 0.5178571428571429,
+      "grad_norm": 0.2699064314365387,
+      "learning_rate": 9.844978718585855e-06,
+      "loss": 0.4135,
+      "step": 58
+    },
+    {
+      "epoch": 0.5267857142857143,
+      "grad_norm": 0.27562785148620605,
+      "learning_rate": 9.831865535973103e-06,
+      "loss": 0.4135,
+      "step": 59
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 0.28133267164230347,
+      "learning_rate": 9.81822947967816e-06,
+      "loss": 0.4085,
+      "step": 60
+    },
+    {
+      "epoch": 0.5446428571428571,
+      "grad_norm": 0.2471129447221756,
+      "learning_rate": 9.804072025308096e-06,
+      "loss": 0.4112,
+      "step": 61
+    },
+    {
+      "epoch": 0.5535714285714286,
+      "grad_norm": 0.3286897540092468,
+      "learning_rate": 9.789394704892364e-06,
+      "loss": 0.4086,
+      "step": 62
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.27339598536491394,
+      "learning_rate": 9.774199106717004e-06,
+      "loss": 0.4077,
+      "step": 63
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.24371126294136047,
+      "learning_rate": 9.758486875152766e-06,
+      "loss": 0.4072,
+      "step": 64
+    },
+    {
+      "epoch": 0.5803571428571429,
+      "grad_norm": 0.26882895827293396,
+      "learning_rate": 9.742259710477178e-06,
+      "loss": 0.4015,
+      "step": 65
+    },
+    {
+      "epoch": 0.5892857142857143,
+      "grad_norm": 0.3289386034011841,
+      "learning_rate": 9.725519368690539e-06,
+      "loss": 0.4035,
+      "step": 66
+    },
+    {
+      "epoch": 0.5982142857142857,
+      "grad_norm": 0.33041179180145264,
+      "learning_rate": 9.70826766132591e-06,
+      "loss": 0.4061,
+      "step": 67
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 0.2799140512943268,
+      "learning_rate": 9.690506455253073e-06,
+      "loss": 0.4086,
+      "step": 68
+    },
+    {
+      "epoch": 0.6160714285714286,
+      "grad_norm": 0.33769428730010986,
+      "learning_rate": 9.672237672476506e-06,
+      "loss": 0.4017,
+      "step": 69
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.3362935483455658,
+      "learning_rate": 9.65346328992741e-06,
+      "loss": 0.4017,
+      "step": 70
+    },
+    {
+      "epoch": 0.6339285714285714,
+      "grad_norm": 0.3309652507305145,
+      "learning_rate": 9.634185339249766e-06,
+      "loss": 0.4043,
+      "step": 71
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.3004245162010193,
+      "learning_rate": 9.614405906580486e-06,
+      "loss": 0.396,
+      "step": 72
+    },
+    {
+      "epoch": 0.6517857142857143,
+      "grad_norm": 0.31212133169174194,
+      "learning_rate": 9.594127132323669e-06,
+      "loss": 0.4062,
+      "step": 73
+    },
+    {
+      "epoch": 0.6607142857142857,
+      "grad_norm": 0.2722640335559845,
+      "learning_rate": 9.573351210918976e-06,
+      "loss": 0.3977,
+      "step": 74
+    },
+    {
+      "epoch": 0.6696428571428571,
+      "grad_norm": 0.3749098777770996,
+      "learning_rate": 9.55208039060416e-06,
+      "loss": 0.3993,
+      "step": 75
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 0.3566979467868805,
+      "learning_rate": 9.53031697317178e-06,
+      "loss": 0.4126,
+      "step": 76
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.32874590158462524,
+      "learning_rate": 9.50806331372012e-06,
+      "loss": 0.4068,
+      "step": 77
+    },
+    {
+      "epoch": 0.6964285714285714,
+      "grad_norm": 0.29310765862464905,
+      "learning_rate": 9.485321820398321e-06,
+      "loss": 0.4014,
+      "step": 78
+    },
+    {
+      "epoch": 0.7053571428571429,
+      "grad_norm": 0.31753551959991455,
+      "learning_rate": 9.4620949541458e-06,
+      "loss": 0.3995,
+      "step": 79
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.30886319279670715,
+      "learning_rate": 9.43838522842594e-06,
+      "loss": 0.4055,
+      "step": 80
+    },
+    {
+      "epoch": 0.7232142857142857,
+      "grad_norm": 0.3552889823913574,
+      "learning_rate": 9.41419520895409e-06,
+      "loss": 0.3976,
+      "step": 81
+    },
+    {
+      "epoch": 0.7321428571428571,
+      "grad_norm": 0.3814694285392761,
+      "learning_rate": 9.389527513419935e-06,
+      "loss": 0.3922,
+      "step": 82
+    },
+    {
+      "epoch": 0.7410714285714286,
+      "grad_norm": 0.4074476361274719,
+      "learning_rate": 9.364384811204212e-06,
+      "loss": 0.4018,
+      "step": 83
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38614606857299805,
+      "learning_rate": 9.338769823089853e-06,
+      "loss": 0.399,
+      "step": 84
+    },
+    {
+      "epoch": 0.7589285714285714,
+      "grad_norm": 0.3035358488559723,
+      "learning_rate": 9.312685320967566e-06,
+      "loss": 0.3948,
+      "step": 85
+    },
+    {
+      "epoch": 0.7678571428571429,
+      "grad_norm": 0.3059258759021759,
+      "learning_rate": 9.286134127535859e-06,
+      "loss": 0.4007,
+      "step": 86
+    },
+    {
+      "epoch": 0.7767857142857143,
+      "grad_norm": 0.39105063676834106,
+      "learning_rate": 9.25911911599561e-06,
+      "loss": 0.3944,
+      "step": 87
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.40922218561172485,
+      "learning_rate": 9.231643209739128e-06,
+      "loss": 0.4006,
+      "step": 88
+    },
+    {
+      "epoch": 0.7946428571428571,
+      "grad_norm": 0.35727784037590027,
+      "learning_rate": 9.203709382033814e-06,
+      "loss": 0.3894,
+      "step": 89
+    },
+    {
+      "epoch": 0.8035714285714286,
+      "grad_norm": 0.26759040355682373,
+      "learning_rate": 9.175320655700407e-06,
+      "loss": 0.3965,
+      "step": 90
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.2492348849773407,
+      "learning_rate": 9.146480102785871e-06,
+      "loss": 0.4021,
+      "step": 91
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 0.29503515362739563,
+      "learning_rate": 9.117190844230971e-06,
+      "loss": 0.3979,
+      "step": 92
+    },
+    {
+      "epoch": 0.8303571428571429,
+      "grad_norm": 0.28773820400238037,
+      "learning_rate": 9.08745604953253e-06,
+      "loss": 0.3961,
+      "step": 93
+    },
+    {
+      "epoch": 0.8392857142857143,
+      "grad_norm": 0.2832280397415161,
+      "learning_rate": 9.057278936400453e-06,
+      "loss": 0.3947,
+      "step": 94
+    },
+    {
+      "epoch": 0.8482142857142857,
+      "grad_norm": 0.26580309867858887,
+      "learning_rate": 9.026662770409524e-06,
+      "loss": 0.3993,
+      "step": 95
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.2617496848106384,
+      "learning_rate": 8.99561086464603e-06,
+      "loss": 0.3997,
+      "step": 96
+    },
+    {
+      "epoch": 0.8660714285714286,
+      "grad_norm": 0.3017551302909851,
+      "learning_rate": 8.964126579349237e-06,
+      "loss": 0.3985,
+      "step": 97
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.2986689507961273,
+      "learning_rate": 8.932213321547769e-06,
+      "loss": 0.3857,
+      "step": 98
+    },
+    {
+      "epoch": 0.8839285714285714,
+      "grad_norm": 0.3144254982471466,
+      "learning_rate": 8.899874544690921e-06,
+      "loss": 0.3985,
+      "step": 99
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 0.25709250569343567,
+      "learning_rate": 8.86711374827494e-06,
+      "loss": 0.3912,
+      "step": 100
+    },
+    {
+      "epoch": 0.9017857142857143,
+      "grad_norm": 0.35801225900650024,
+      "learning_rate": 8.833934477464348e-06,
+      "loss": 0.3972,
+      "step": 101
+    },
+    {
+      "epoch": 0.9107142857142857,
+      "grad_norm": 0.3704852759838104,
+      "learning_rate": 8.800340322708291e-06,
+      "loss": 0.3889,
+      "step": 102
+    },
+    {
+      "epoch": 0.9196428571428571,
+      "grad_norm": 0.3720284700393677,
+      "learning_rate": 8.766334919352018e-06,
+      "loss": 0.3991,
+      "step": 103
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.3510439991950989,
+      "learning_rate": 8.73192194724347e-06,
+      "loss": 0.3879,
+      "step": 104
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.29455068707466125,
+      "learning_rate": 8.697105130335084e-06,
+      "loss": 0.3962,
+      "step": 105
+    },
+    {
+      "epoch": 0.9464285714285714,
+      "grad_norm": 0.25835826992988586,
+      "learning_rate": 8.661888236280813e-06,
+      "loss": 0.3909,
+      "step": 106
+    },
+    {
+      "epoch": 0.9553571428571429,
+      "grad_norm": 0.29132160544395447,
+      "learning_rate": 8.626275076028397e-06,
+      "loss": 0.3899,
+      "step": 107
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 0.3577162027359009,
+      "learning_rate": 8.590269503406986e-06,
+      "loss": 0.3846,
+      "step": 108
+    },
+    {
+      "epoch": 0.9732142857142857,
+      "grad_norm": 0.28617140650749207,
+      "learning_rate": 8.553875414710088e-06,
+      "loss": 0.3922,
+      "step": 109
+    },
+    {
+      "epoch": 0.9821428571428571,
+      "grad_norm": 0.34976303577423096,
+      "learning_rate": 8.517096748273951e-06,
+      "loss": 0.3909,
+      "step": 110
+    },
+    {
+      "epoch": 0.9910714285714286,
+      "grad_norm": 0.307277113199234,
+      "learning_rate": 8.479937484051368e-06,
+      "loss": 0.3893,
+      "step": 111
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33036214113235474,
+      "learning_rate": 8.442401643181e-06,
+      "loss": 0.3836,
+      "step": 112
+    },
+    {
+      "epoch": 1.0089285714285714,
+      "grad_norm": 0.3078720271587372,
+      "learning_rate": 8.404493287552232e-06,
+      "loss": 0.3733,
+      "step": 113
+    },
+    {
+      "epoch": 1.0178571428571428,
+      "grad_norm": 0.297649085521698,
+      "learning_rate": 8.366216519365623e-06,
+      "loss": 0.3687,
+      "step": 114
+    },
+    {
+      "epoch": 1.0267857142857142,
+      "grad_norm": 0.2785271406173706,
+      "learning_rate": 8.327575480688985e-06,
+      "loss": 0.3734,
+      "step": 115
+    },
+    {
+      "epoch": 1.0357142857142858,
+      "grad_norm": 0.35386645793914795,
+      "learning_rate": 8.288574353009164e-06,
+      "loss": 0.3782,
+      "step": 116
+    },
+    {
+      "epoch": 1.0446428571428572,
+      "grad_norm": 0.33655157685279846,
+      "learning_rate": 8.249217356779544e-06,
+      "loss": 0.3745,
+      "step": 117
+    },
+    {
+      "epoch": 1.0535714285714286,
+      "grad_norm": 0.24362798035144806,
+      "learning_rate": 8.20950875096333e-06,
+      "loss": 0.3731,
+      "step": 118
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 0.2674640119075775,
+      "learning_rate": 8.169452832572676e-06,
+      "loss": 0.3711,
+      "step": 119
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.26036393642425537,
+      "learning_rate": 8.129053936203688e-06,
+      "loss": 0.3701,
+      "step": 120
+    },
+    {
+      "epoch": 1.0803571428571428,
+      "grad_norm": 0.24254180490970612,
+      "learning_rate": 8.088316433567369e-06,
+      "loss": 0.3729,
+      "step": 121
+    },
+    {
+      "epoch": 1.0892857142857142,
+      "grad_norm": 0.24308055639266968,
+      "learning_rate": 8.04724473301652e-06,
+      "loss": 0.3725,
+      "step": 122
+    },
+    {
+      "epoch": 1.0982142857142858,
+      "grad_norm": 0.2283226102590561,
+      "learning_rate": 8.005843279068724e-06,
+      "loss": 0.3689,
+      "step": 123
+    },
+    {
+      "epoch": 1.1071428571428572,
+      "grad_norm": 0.2410878688097,
+      "learning_rate": 7.964116551925365e-06,
+      "loss": 0.377,
+      "step": 124
+    },
+    {
+      "epoch": 1.1160714285714286,
+      "grad_norm": 0.25953209400177,
+      "learning_rate": 7.92206906698682e-06,
+      "loss": 0.3684,
+      "step": 125
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.2646959125995636,
+      "learning_rate": 7.879705374363831e-06,
+      "loss": 0.3744,
+      "step": 126
+    },
+    {
+      "epoch": 1.1339285714285714,
+      "grad_norm": 0.23834243416786194,
+      "learning_rate": 7.837030058385117e-06,
+      "loss": 0.3754,
+      "step": 127
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.25867897272109985,
+      "learning_rate": 7.794047737101298e-06,
+      "loss": 0.3712,
+      "step": 128
+    },
+    {
+      "epoch": 1.1517857142857142,
+      "grad_norm": 0.26810479164123535,
+      "learning_rate": 7.750763061785139e-06,
+      "loss": 0.3721,
+      "step": 129
+    },
+    {
+      "epoch": 1.1607142857142858,
+      "grad_norm": 0.23617470264434814,
+      "learning_rate": 7.707180716428237e-06,
+      "loss": 0.3754,
+      "step": 130
+    },
+    {
+      "epoch": 1.1696428571428572,
+      "grad_norm": 0.24265629053115845,
+      "learning_rate": 7.663305417234146e-06,
+      "loss": 0.3675,
+      "step": 131
+    },
+    {
+      "epoch": 1.1785714285714286,
+      "grad_norm": 0.2951458692550659,
+      "learning_rate": 7.619141912108008e-06,
+      "loss": 0.3786,
+      "step": 132
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 0.24244728684425354,
+      "learning_rate": 7.57469498014278e-06,
+      "loss": 0.3731,
+      "step": 133
+    },
+    {
+      "epoch": 1.1964285714285714,
+      "grad_norm": 0.21746976673603058,
+      "learning_rate": 7.529969431102063e-06,
+      "loss": 0.3735,
+      "step": 134
+    },
+    {
+      "epoch": 1.2053571428571428,
+      "grad_norm": 0.25599175691604614,
+      "learning_rate": 7.484970104899624e-06,
+      "loss": 0.3689,
+      "step": 135
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.2244032621383667,
+      "learning_rate": 7.4397018710756415e-06,
+      "loss": 0.3704,
+      "step": 136
+    },
+    {
+      "epoch": 1.2232142857142858,
+      "grad_norm": 0.25709277391433716,
+      "learning_rate": 7.394169628269771e-06,
+      "loss": 0.3702,
+      "step": 137
+    },
+    {
+      "epoch": 1.2321428571428572,
+      "grad_norm": 0.2774679958820343,
+      "learning_rate": 7.34837830369103e-06,
+      "loss": 0.3688,
+      "step": 138
+    },
+    {
+      "epoch": 1.2410714285714286,
+      "grad_norm": 0.28378257155418396,
+      "learning_rate": 7.302332852584619e-06,
+      "loss": 0.3731,
+      "step": 139
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.24090154469013214,
+      "learning_rate": 7.2560382576956875e-06,
+      "loss": 0.3691,
+      "step": 140
+    },
+    {
+      "epoch": 1.2589285714285714,
+      "grad_norm": 0.2390367090702057,
+      "learning_rate": 7.209499528730138e-06,
+      "loss": 0.3664,
+      "step": 141
+    },
+    {
+      "epoch": 1.2678571428571428,
+      "grad_norm": 0.2720012664794922,
+      "learning_rate": 7.162721701812506e-06,
+      "loss": 0.3702,
+      "step": 142
+    },
+    {
+      "epoch": 1.2767857142857144,
+      "grad_norm": 0.2866012156009674,
+      "learning_rate": 7.115709838940983e-06,
+      "loss": 0.3722,
+      "step": 143
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.23692139983177185,
+      "learning_rate": 7.068469027439642e-06,
+      "loss": 0.3741,
+      "step": 144
+    },
+    {
+      "epoch": 1.2946428571428572,
+      "grad_norm": 0.21032477915287018,
+      "learning_rate": 7.02100437940791e-06,
+      "loss": 0.3675,
+      "step": 145
+    },
+    {
+      "epoch": 1.3035714285714286,
+      "grad_norm": 0.2565901577472687,
+      "learning_rate": 6.9733210311673826e-06,
+      "loss": 0.3695,
+      "step": 146
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 0.3054284453392029,
+      "learning_rate": 6.925424142705997e-06,
+      "loss": 0.3665,
+      "step": 147
+    },
+    {
+      "epoch": 1.3214285714285714,
+      "grad_norm": 0.2568247318267822,
+      "learning_rate": 6.8773188971196515e-06,
+      "loss": 0.3702,
+      "step": 148
+    },
+    {
+      "epoch": 1.3303571428571428,
+      "grad_norm": 0.2898830473423004,
+      "learning_rate": 6.829010500051319e-06,
+      "loss": 0.3739,
+      "step": 149
+    },
+    {
+      "epoch": 1.3392857142857144,
+      "grad_norm": 0.22603343427181244,
+      "learning_rate": 6.780504179127735e-06,
+      "loss": 0.3689,
+      "step": 150
+    },
+    {
+      "epoch": 1.3482142857142856,
+      "grad_norm": 0.26400893926620483,
+      "learning_rate": 6.731805183393696e-06,
+      "loss": 0.3686,
+      "step": 151
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.21483619511127472,
+      "learning_rate": 6.682918782744033e-06,
+      "loss": 0.3685,
+      "step": 152
+    },
+    {
+      "epoch": 1.3660714285714286,
+      "grad_norm": 0.28114837408065796,
+      "learning_rate": 6.63385026735334e-06,
+      "loss": 0.3626,
+      "step": 153
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.22158163785934448,
+      "learning_rate": 6.584604947103515e-06,
+      "loss": 0.3635,
+      "step": 154
+    },
+    {
+      "epoch": 1.3839285714285714,
+      "grad_norm": 0.2537171542644501,
+      "learning_rate": 6.535188151009143e-06,
+      "loss": 0.3699,
+      "step": 155
+    },
+    {
+      "epoch": 1.3928571428571428,
+      "grad_norm": 0.24269886314868927,
+      "learning_rate": 6.4856052266408375e-06,
+      "loss": 0.362,
+      "step": 156
+    },
+    {
+      "epoch": 1.4017857142857144,
+      "grad_norm": 0.22118201851844788,
+      "learning_rate": 6.4358615395465506e-06,
+      "loss": 0.3652,
+      "step": 157
+    },
+    {
+      "epoch": 1.4107142857142856,
+      "grad_norm": 0.2311353087425232,
+      "learning_rate": 6.385962472670953e-06,
+      "loss": 0.3703,
+      "step": 158
+    },
+    {
+      "epoch": 1.4196428571428572,
+      "grad_norm": 0.21577298641204834,
+      "learning_rate": 6.335913425772926e-06,
+      "loss": 0.3624,
+      "step": 159
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.23977944254875183,
+      "learning_rate": 6.28571981484123e-06,
+      "loss": 0.3663,
+      "step": 160
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 0.2253430187702179,
+      "learning_rate": 6.235387071508427e-06,
+      "loss": 0.3708,
+      "step": 161
+    },
+    {
+      "epoch": 1.4464285714285714,
+      "grad_norm": 0.20987780392169952,
+      "learning_rate": 6.184920642463095e-06,
+      "loss": 0.3703,
+      "step": 162
+    },
+    {
+      "epoch": 1.4553571428571428,
+      "grad_norm": 0.23987387120723724,
+      "learning_rate": 6.1343259888604335e-06,
+      "loss": 0.3731,
+      "step": 163
+    },
+    {
+      "epoch": 1.4642857142857144,
+      "grad_norm": 0.22025255858898163,
+      "learning_rate": 6.083608585731283e-06,
+      "loss": 0.3677,
+      "step": 164
+    },
+    {
+      "epoch": 1.4732142857142856,
+      "grad_norm": 0.19166462123394012,
+      "learning_rate": 6.032773921389655e-06,
+      "loss": 0.3708,
+      "step": 165
+    },
+    {
+      "epoch": 1.4821428571428572,
+      "grad_norm": 0.2034110724925995,
+      "learning_rate": 5.9818274968388225e-06,
+      "loss": 0.3665,
+      "step": 166
+    },
+    {
+      "epoch": 1.4910714285714286,
+      "grad_norm": 0.21674256026744843,
+      "learning_rate": 5.930774825176034e-06,
+      "loss": 0.3625,
+      "step": 167
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.22111833095550537,
+      "learning_rate": 5.879621430995927e-06,
+      "loss": 0.3699,
+      "step": 168
+    },
+    {
+      "epoch": 1.5089285714285714,
+      "grad_norm": 0.20402273535728455,
+      "learning_rate": 5.8283728497926865e-06,
+      "loss": 0.3678,
+      "step": 169
+    },
+    {
+      "epoch": 1.5178571428571428,
+      "grad_norm": 0.2157091498374939,
+      "learning_rate": 5.777034627361025e-06,
+      "loss": 0.3677,
+      "step": 170
+    },
+    {
+      "epoch": 1.5267857142857144,
+      "grad_norm": 0.20562361180782318,
+      "learning_rate": 5.725612319196065e-06,
+      "loss": 0.3697,
+      "step": 171
+    },
+    {
+      "epoch": 1.5357142857142856,
+      "grad_norm": 0.1970471292734146,
+      "learning_rate": 5.674111489892144e-06,
+      "loss": 0.3696,
+      "step": 172
+    },
+    {
+      "epoch": 1.5446428571428572,
+      "grad_norm": 0.19568419456481934,
+      "learning_rate": 5.622537712540664e-06,
+      "loss": 0.3744,
+      "step": 173
+    },
+    {
+      "epoch": 1.5535714285714286,
+      "grad_norm": 0.20620059967041016,
+      "learning_rate": 5.570896568126994e-06,
+      "loss": 0.3668,
+      "step": 174
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.22930361330509186,
+      "learning_rate": 5.5191936449265345e-06,
+      "loss": 0.3649,
+      "step": 175
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.19315338134765625,
+      "learning_rate": 5.4674345379e-06,
+      "loss": 0.3683,
+      "step": 176
+    },
+    {
+      "epoch": 1.5803571428571428,
+      "grad_norm": 0.20728451013565063,
+      "learning_rate": 5.415624848087959e-06,
+      "loss": 0.3658,
+      "step": 177
+    },
+    {
+      "epoch": 1.5892857142857144,
+      "grad_norm": 0.1867711842060089,
+      "learning_rate": 5.36377018200472e-06,
+      "loss": 0.3663,
+      "step": 178
+    },
+    {
+      "epoch": 1.5982142857142856,
+      "grad_norm": 0.18480746448040009,
+      "learning_rate": 5.311876151031642e-06,
+      "loss": 0.3618,
+      "step": 179
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 0.20057101547718048,
+      "learning_rate": 5.259948370809902e-06,
+      "loss": 0.3707,
+      "step": 180
+    },
+    {
+      "epoch": 1.6160714285714286,
+      "grad_norm": 0.18401624262332916,
+      "learning_rate": 5.207992460632805e-06,
+      "loss": 0.3647,
+      "step": 181
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.17545711994171143,
+      "learning_rate": 5.156014042837696e-06,
+      "loss": 0.3646,
+      "step": 182
+    },
+    {
+      "epoch": 1.6339285714285714,
+      "grad_norm": 0.20088276267051697,
+      "learning_rate": 5.104018742197557e-06,
+      "loss": 0.3603,
+      "step": 183
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.1916016936302185,
+      "learning_rate": 5.052012185312322e-06,
+      "loss": 0.3596,
+      "step": 184
+    },
+    {
+      "epoch": 1.6517857142857144,
+      "grad_norm": 0.18089112639427185,
+      "learning_rate": 5e-06,
+      "loss": 0.3646,
+      "step": 185
+    },
+    {
+      "epoch": 1.6607142857142856,
+      "grad_norm": 0.21182607114315033,
+      "learning_rate": 4.94798781468768e-06,
+      "loss": 0.3616,
+      "step": 186
+    },
+    {
+      "epoch": 1.6696428571428572,
+      "grad_norm": 0.1948297768831253,
+      "learning_rate": 4.895981257802444e-06,
+      "loss": 0.366,
+      "step": 187
+    },
+    {
+      "epoch": 1.6785714285714286,
+      "grad_norm": 0.18150609731674194,
+      "learning_rate": 4.843985957162304e-06,
+      "loss": 0.3613,
+      "step": 188
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 0.22999927401542664,
+      "learning_rate": 4.792007539367198e-06,
+      "loss": 0.3614,
+      "step": 189
+    },
+    {
+      "epoch": 1.6964285714285714,
+      "grad_norm": 0.18993772566318512,
+      "learning_rate": 4.740051629190099e-06,
+      "loss": 0.3665,
+      "step": 190
+    },
+    {
+      "epoch": 1.7053571428571428,
+      "grad_norm": 0.20203571021556854,
+      "learning_rate": 4.6881238489683596e-06,
+      "loss": 0.3605,
+      "step": 191
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.1820489913225174,
+      "learning_rate": 4.636229817995281e-06,
+      "loss": 0.3687,
+      "step": 192
+    },
+    {
+      "epoch": 1.7232142857142856,
+      "grad_norm": 0.1792946755886078,
+      "learning_rate": 4.584375151912043e-06,
+      "loss": 0.37,
+      "step": 193
+    },
+    {
+      "epoch": 1.7321428571428572,
+      "grad_norm": 0.1797734498977661,
+      "learning_rate": 4.532565462099999e-06,
+      "loss": 0.3699,
+      "step": 194
+    },
+    {
+      "epoch": 1.7410714285714286,
+      "grad_norm": 0.17329907417297363,
+      "learning_rate": 4.480806355073467e-06,
+      "loss": 0.3634,
+      "step": 195
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.17865252494812012,
+      "learning_rate": 4.429103431873009e-06,
+      "loss": 0.3647,
+      "step": 196
+    },
+    {
+      "epoch": 1.7589285714285714,
+      "grad_norm": 0.18890346586704254,
+      "learning_rate": 4.377462287459338e-06,
+      "loss": 0.3685,
+      "step": 197
+    },
+    {
+      "epoch": 1.7678571428571428,
+      "grad_norm": 0.16928838193416595,
+      "learning_rate": 4.3258885101078565e-06,
+      "loss": 0.359,
+      "step": 198
+    },
+    {
+      "epoch": 1.7767857142857144,
+      "grad_norm": 0.16775333881378174,
+      "learning_rate": 4.274387680803936e-06,
+      "loss": 0.3589,
+      "step": 199
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.1853839010000229,
+      "learning_rate": 4.2229653726389765e-06,
+      "loss": 0.3657,
+      "step": 200
+    },
+    {
+      "epoch": 1.7946428571428572,
+      "grad_norm": 0.17588171362876892,
+      "learning_rate": 4.171627150207314e-06,
+      "loss": 0.3637,
+      "step": 201
+    },
+    {
+      "epoch": 1.8035714285714286,
+      "grad_norm": 0.1664077490568161,
+      "learning_rate": 4.120378569004074e-06,
+      "loss": 0.3655,
+      "step": 202
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 0.17974969744682312,
+      "learning_rate": 4.069225174823968e-06,
+      "loss": 0.3662,
+      "step": 203
+    },
+    {
+      "epoch": 1.8214285714285714,
+      "grad_norm": 0.17224469780921936,
+      "learning_rate": 4.018172503161179e-06,
+      "loss": 0.3612,
+      "step": 204
+    },
+    {
+      "epoch": 1.8303571428571428,
+      "grad_norm": 0.1866443157196045,
+      "learning_rate": 3.967226078610346e-06,
+      "loss": 0.3594,
+      "step": 205
+    },
+    {
+      "epoch": 1.8392857142857144,
+      "grad_norm": 0.17376399040222168,
+      "learning_rate": 3.9163914142687185e-06,
+      "loss": 0.3685,
+      "step": 206
+    },
+    {
+      "epoch": 1.8482142857142856,
+      "grad_norm": 0.17465896904468536,
+      "learning_rate": 3.865674011139567e-06,
+      "loss": 0.3582,
+      "step": 207
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.15824931859970093,
+      "learning_rate": 3.815079357536907e-06,
+      "loss": 0.3641,
+      "step": 208
+    },
+    {
+      "epoch": 1.8660714285714286,
+      "grad_norm": 0.1589619368314743,
+      "learning_rate": 3.7646129284915754e-06,
+      "loss": 0.3665,
+      "step": 209
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.16915646195411682,
+      "learning_rate": 3.714280185158771e-06,
+      "loss": 0.3612,
+      "step": 210
+    },
+    {
+      "epoch": 1.8839285714285714,
+      "grad_norm": 0.16680151224136353,
+      "learning_rate": 3.6640865742270756e-06,
+      "loss": 0.3629,
+      "step": 211
+    },
+    {
+      "epoch": 1.8928571428571428,
+      "grad_norm": 0.17910495400428772,
+      "learning_rate": 3.614037527329048e-06,
+      "loss": 0.3646,
+      "step": 212
+    },
+    {
+      "epoch": 1.9017857142857144,
+      "grad_norm": 0.14773690700531006,
+      "learning_rate": 3.5641384604534503e-06,
+      "loss": 0.3607,
+      "step": 213
+    },
+    {
+      "epoch": 1.9107142857142856,
+      "grad_norm": 0.20091548562049866,
+      "learning_rate": 3.5143947733591633e-06,
+      "loss": 0.3613,
+      "step": 214
+    },
+    {
+      "epoch": 1.9196428571428572,
+      "grad_norm": 0.17596937716007233,
+      "learning_rate": 3.4648118489908588e-06,
+      "loss": 0.365,
+      "step": 215
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.18936946988105774,
+      "learning_rate": 3.4153950528964867e-06,
+      "loss": 0.3644,
+      "step": 216
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 0.16055169701576233,
+      "learning_rate": 3.366149732646661e-06,
+      "loss": 0.3662,
+      "step": 217
+    },
+    {
+      "epoch": 1.9464285714285714,
+      "grad_norm": 0.18520261347293854,
+      "learning_rate": 3.3170812172559695e-06,
+      "loss": 0.3598,
+      "step": 218
+    },
+    {
+      "epoch": 1.9553571428571428,
+      "grad_norm": 0.16777203977108002,
+      "learning_rate": 3.268194816606305e-06,
+      "loss": 0.3625,
+      "step": 219
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 0.19867220520973206,
+      "learning_rate": 3.2194958208722656e-06,
+      "loss": 0.364,
+      "step": 220
+    },
+    {
+      "epoch": 1.9732142857142856,
+      "grad_norm": 0.15577928721904755,
+      "learning_rate": 3.1709894999486828e-06,
+      "loss": 0.3618,
+      "step": 221
+    },
+    {
+      "epoch": 1.9821428571428572,
+      "grad_norm": 0.16875170171260834,
+      "learning_rate": 3.1226811028803514e-06,
+      "loss": 0.3627,
+      "step": 222
+    },
+    {
+      "epoch": 1.9910714285714286,
+      "grad_norm": 0.17682914435863495,
+      "learning_rate": 3.0745758572940044e-06,
+      "loss": 0.3661,
+      "step": 223
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.17448174953460693,
+      "learning_rate": 3.0266789688326187e-06,
+      "loss": 0.3518,
+      "step": 224
+    },
+    {
+      "epoch": 2.0089285714285716,
+      "grad_norm": 0.16910579800605774,
+      "learning_rate": 2.978995620592092e-06,
+      "loss": 0.3488,
+      "step": 225
+    },
+    {
+      "epoch": 2.017857142857143,
+      "grad_norm": 0.1821155995130539,
+      "learning_rate": 2.9315309725603596e-06,
+      "loss": 0.3498,
+      "step": 226
+    },
+    {
+      "epoch": 2.0267857142857144,
+      "grad_norm": 0.17017966508865356,
+      "learning_rate": 2.884290161059017e-06,
+      "loss": 0.3411,
+      "step": 227
+    },
+    {
+      "epoch": 2.0357142857142856,
+      "grad_norm": 0.1491294950246811,
+      "learning_rate": 2.8372782981874964e-06,
+      "loss": 0.354,
+      "step": 228
+    },
+    {
+      "epoch": 2.044642857142857,
+      "grad_norm": 0.172164186835289,
+      "learning_rate": 2.7905004712698646e-06,
+      "loss": 0.3454,
+      "step": 229
+    },
+    {
+      "epoch": 2.0535714285714284,
+      "grad_norm": 0.180723175406456,
+      "learning_rate": 2.7439617423043146e-06,
+      "loss": 0.3528,
+      "step": 230
+    },
+    {
+      "epoch": 2.0625,
+      "grad_norm": 0.1604950726032257,
+      "learning_rate": 2.697667147415383e-06,
+      "loss": 0.3489,
+      "step": 231
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.16484476625919342,
+      "learning_rate": 2.6516216963089698e-06,
+      "loss": 0.3436,
+      "step": 232
+    },
+    {
+      "epoch": 2.080357142857143,
+      "grad_norm": 0.1700424700975418,
+      "learning_rate": 2.605830371730229e-06,
+      "loss": 0.3432,
+      "step": 233
+    },
+    {
+      "epoch": 2.0892857142857144,
+      "grad_norm": 0.16941745579242706,
+      "learning_rate": 2.560298128924358e-06,
+      "loss": 0.3424,
+      "step": 234
+    },
+    {
+      "epoch": 2.0982142857142856,
+      "grad_norm": 0.187941312789917,
+      "learning_rate": 2.5150298951003783e-06,
+      "loss": 0.3421,
+      "step": 235
+    },
+    {
+      "epoch": 2.107142857142857,
+      "grad_norm": 0.1790265589952469,
+      "learning_rate": 2.470030568897938e-06,
+      "loss": 0.3458,
+      "step": 236
+    },
+    {
+      "epoch": 2.1160714285714284,
+      "grad_norm": 0.16201697289943695,
+      "learning_rate": 2.425305019857222e-06,
+      "loss": 0.349,
+      "step": 237
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.18233942985534668,
+      "learning_rate": 2.3808580878919948e-06,
+      "loss": 0.3507,
+      "step": 238
+    },
+    {
+      "epoch": 2.1339285714285716,
+      "grad_norm": 0.16857150197029114,
+      "learning_rate": 2.336694582765857e-06,
+      "loss": 0.3429,
+      "step": 239
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.1519509106874466,
+      "learning_rate": 2.2928192835717642e-06,
+      "loss": 0.3459,
+      "step": 240
+    },
+    {
+      "epoch": 2.1517857142857144,
+      "grad_norm": 0.15660250186920166,
+      "learning_rate": 2.2492369382148634e-06,
+      "loss": 0.3471,
+      "step": 241
+    },
+    {
+      "epoch": 2.1607142857142856,
+      "grad_norm": 0.1472284495830536,
+      "learning_rate": 2.205952262898704e-06,
+      "loss": 0.3456,
+      "step": 242
+    },
+    {
+      "epoch": 2.169642857142857,
+      "grad_norm": 0.15189002454280853,
+      "learning_rate": 2.1629699416148832e-06,
+      "loss": 0.3459,
+      "step": 243
+    },
+    {
+      "epoch": 2.1785714285714284,
+      "grad_norm": 0.13611957430839539,
+      "learning_rate": 2.120294625636171e-06,
+      "loss": 0.3513,
+      "step": 244
+    },
+    {
+      "epoch": 2.1875,
+      "grad_norm": 0.14140184223651886,
+      "learning_rate": 2.077930933013182e-06,
+      "loss": 0.3423,
+      "step": 245
+    },
+    {
+      "epoch": 2.1964285714285716,
+      "grad_norm": 0.1375838965177536,
+      "learning_rate": 2.0358834480746363e-06,
+      "loss": 0.3511,
+      "step": 246
+    },
+    {
+      "epoch": 2.205357142857143,
+      "grad_norm": 0.1536509245634079,
+      "learning_rate": 1.9941567209312768e-06,
+      "loss": 0.3412,
+      "step": 247
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.15353716909885406,
+      "learning_rate": 1.9527552669834797e-06,
+      "loss": 0.3439,
+      "step": 248
+    },
+    {
+      "epoch": 2.2232142857142856,
+      "grad_norm": 0.1261204034090042,
+      "learning_rate": 1.911683566432633e-06,
+      "loss": 0.3429,
+      "step": 249
+    },
+    {
+      "epoch": 2.232142857142857,
+      "grad_norm": 0.1508454829454422,
+      "learning_rate": 1.8709460637963123e-06,
+      "loss": 0.3399,
+      "step": 250
+    },
+    {
+      "epoch": 2.2410714285714284,
+      "grad_norm": 0.13777896761894226,
+      "learning_rate": 1.8305471674273262e-06,
+      "loss": 0.3506,
+      "step": 251
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.1431542932987213,
+      "learning_rate": 1.7904912490366723e-06,
+      "loss": 0.3421,
+      "step": 252
+    },
+    {
+      "epoch": 2.2589285714285716,
+      "grad_norm": 0.15625214576721191,
+      "learning_rate": 1.7507826432204571e-06,
+      "loss": 0.3516,
+      "step": 253
+    },
+    {
+      "epoch": 2.267857142857143,
+      "grad_norm": 0.13694243133068085,
+      "learning_rate": 1.711425646990838e-06,
+      "loss": 0.3476,
+      "step": 254
+    },
+    {
+      "epoch": 2.2767857142857144,
+      "grad_norm": 0.13206931948661804,
+      "learning_rate": 1.6724245193110178e-06,
+      "loss": 0.3456,
+      "step": 255
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.1381482183933258,
+      "learning_rate": 1.6337834806343783e-06,
+      "loss": 0.3546,
+      "step": 256
+    },
+    {
+      "epoch": 2.294642857142857,
+      "grad_norm": 0.14691098034381866,
+      "learning_rate": 1.595506712447768e-06,
+      "loss": 0.3518,
+      "step": 257
+    },
+    {
+      "epoch": 2.3035714285714284,
+      "grad_norm": 0.13349969685077667,
+      "learning_rate": 1.557598356819e-06,
+      "loss": 0.3488,
+      "step": 258
+    },
+    {
+      "epoch": 2.3125,
+      "grad_norm": 0.14272207021713257,
+      "learning_rate": 1.5200625159486322e-06,
+      "loss": 0.3471,
+      "step": 259
+    },
+    {
+      "epoch": 2.3214285714285716,
+      "grad_norm": 0.13418231904506683,
+      "learning_rate": 1.482903251726049e-06,
+      "loss": 0.3419,
+      "step": 260
+    },
+    {
+      "epoch": 2.330357142857143,
+      "grad_norm": 0.13040195405483246,
+      "learning_rate": 1.446124585289913e-06,
+      "loss": 0.3458,
+      "step": 261
+    },
+    {
+      "epoch": 2.3392857142857144,
+      "grad_norm": 0.12616047263145447,
+      "learning_rate": 1.409730496593016e-06,
+      "loss": 0.3493,
+      "step": 262
+    },
+    {
+      "epoch": 2.3482142857142856,
+      "grad_norm": 0.20311793684959412,
+      "learning_rate": 1.3737249239716043e-06,
+      "loss": 0.3503,
+      "step": 263
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.160719633102417,
+      "learning_rate": 1.3381117637191887e-06,
+      "loss": 0.3493,
+      "step": 264
+    },
+    {
+      "epoch": 2.3660714285714284,
+      "grad_norm": 0.1257162094116211,
+      "learning_rate": 1.3028948696649164e-06,
+      "loss": 0.3467,
+      "step": 265
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.15282198786735535,
+      "learning_rate": 1.2680780527565313e-06,
+      "loss": 0.3443,
+      "step": 266
+    },
+    {
+      "epoch": 2.3839285714285716,
+      "grad_norm": 0.1382765769958496,
+      "learning_rate": 1.2336650806479826e-06,
+      "loss": 0.3461,
+      "step": 267
+    },
+    {
+      "epoch": 2.392857142857143,
+      "grad_norm": 0.13694579899311066,
+      "learning_rate": 1.1996596772917091e-06,
+      "loss": 0.3543,
+      "step": 268
+    },
+    {
+      "epoch": 2.4017857142857144,
+      "grad_norm": 0.13849100470542908,
+      "learning_rate": 1.1660655225356533e-06,
+      "loss": 0.3471,
+      "step": 269
+    },
+    {
+      "epoch": 2.4107142857142856,
+      "grad_norm": 0.13622619211673737,
+      "learning_rate": 1.132886251725061e-06,
+      "loss": 0.3474,
+      "step": 270
+    },
+    {
+      "epoch": 2.419642857142857,
+      "grad_norm": 0.13429802656173706,
+      "learning_rate": 1.1001254553090812e-06,
+      "loss": 0.3451,
+      "step": 271
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.1249421089887619,
+      "learning_rate": 1.0677866784522317e-06,
+      "loss": 0.3474,
+      "step": 272
+    },
+    {
+      "epoch": 2.4375,
+      "grad_norm": 0.1234789565205574,
+      "learning_rate": 1.0358734206507642e-06,
+      "loss": 0.3459,
+      "step": 273
+    },
+    {
+      "epoch": 2.4464285714285716,
+      "grad_norm": 0.12073200196027756,
+      "learning_rate": 1.004389135353972e-06,
+      "loss": 0.3387,
+      "step": 274
+    },
+    {
+      "epoch": 2.455357142857143,
+      "grad_norm": 0.1367228776216507,
+      "learning_rate": 9.733372295904776e-07,
+      "loss": 0.347,
+      "step": 275
+    },
+    {
+      "epoch": 2.4642857142857144,
+      "grad_norm": 0.12393829226493835,
+      "learning_rate": 9.427210635995482e-07,
+      "loss": 0.3467,
+      "step": 276
+    },
+    {
+      "epoch": 2.4732142857142856,
+      "grad_norm": 0.1407790631055832,
+      "learning_rate": 9.125439504674699e-07,
+      "loss": 0.3524,
+      "step": 277
+    },
+    {
+      "epoch": 2.482142857142857,
+      "grad_norm": 0.1231057420372963,
+      "learning_rate": 8.828091557690288e-07,
+      "loss": 0.351,
+      "step": 278
+    },
+    {
+      "epoch": 2.4910714285714284,
+      "grad_norm": 0.1206686794757843,
+      "learning_rate": 8.535198972141295e-07,
+      "loss": 0.3455,
+      "step": 279
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.12657472491264343,
+      "learning_rate": 8.246793442995954e-07,
+      "loss": 0.3441,
+      "step": 280
+    },
+    {
+      "epoch": 2.508928571428571,
+      "grad_norm": 0.13261254131793976,
+      "learning_rate": 7.962906179661872e-07,
+      "loss": 0.338,
+      "step": 281
+    },
+    {
+      "epoch": 2.517857142857143,
+      "grad_norm": 0.13191427290439606,
+      "learning_rate": 7.68356790260873e-07,
+      "loss": 0.3423,
+      "step": 282
+    },
+    {
+      "epoch": 2.5267857142857144,
+      "grad_norm": 0.13889721035957336,
+      "learning_rate": 7.408808840043913e-07,
+      "loss": 0.3437,
+      "step": 283
+    },
+    {
+      "epoch": 2.5357142857142856,
+      "grad_norm": 0.13105542957782745,
+      "learning_rate": 7.138658724641417e-07,
+      "loss": 0.3533,
+      "step": 284
+    },
+    {
+      "epoch": 2.544642857142857,
+      "grad_norm": 0.12581577897071838,
+      "learning_rate": 6.873146790324359e-07,
+      "loss": 0.3473,
+      "step": 285
+    },
+    {
+      "epoch": 2.553571428571429,
+      "grad_norm": 0.12375447899103165,
+      "learning_rate": 6.612301769101464e-07,
+      "loss": 0.3466,
+      "step": 286
+    },
+    {
+      "epoch": 2.5625,
+      "grad_norm": 0.11879701167345047,
+      "learning_rate": 6.356151887957901e-07,
+      "loss": 0.3526,
+      "step": 287
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.1344556361436844,
+      "learning_rate": 6.104724865800665e-07,
+      "loss": 0.3436,
+      "step": 288
+    },
+    {
+      "epoch": 2.580357142857143,
+      "grad_norm": 0.1175764948129654,
+      "learning_rate": 5.858047910459109e-07,
+      "loss": 0.3435,
+      "step": 289
+    },
+    {
+      "epoch": 2.5892857142857144,
+      "grad_norm": 0.12031050771474838,
+      "learning_rate": 5.616147715740611e-07,
+      "loss": 0.3425,
+      "step": 290
+    },
+    {
+      "epoch": 2.5982142857142856,
+      "grad_norm": 0.1191452294588089,
+      "learning_rate": 5.379050458541995e-07,
+      "loss": 0.3437,
+      "step": 291
+    },
+    {
+      "epoch": 2.607142857142857,
+      "grad_norm": 0.11280887573957443,
+      "learning_rate": 5.146781796016798e-07,
+      "loss": 0.3477,
+      "step": 292
+    },
+    {
+      "epoch": 2.616071428571429,
+      "grad_norm": 0.12038522213697433,
+      "learning_rate": 4.919366862798807e-07,
+      "loss": 0.3452,
+      "step": 293
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.1229255199432373,
+      "learning_rate": 4.696830268282204e-07,
+      "loss": 0.3487,
+      "step": 294
+    },
+    {
+      "epoch": 2.633928571428571,
+      "grad_norm": 0.12023057788610458,
+      "learning_rate": 4.479196093958421e-07,
+      "loss": 0.3458,
+      "step": 295
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.11441769450902939,
+      "learning_rate": 4.2664878908102556e-07,
+      "loss": 0.3467,
+      "step": 296
+    },
+    {
+      "epoch": 2.6517857142857144,
+      "grad_norm": 0.12089654058218002,
+      "learning_rate": 4.058728676763313e-07,
+      "loss": 0.3517,
+      "step": 297
+    },
+    {
+      "epoch": 2.6607142857142856,
+      "grad_norm": 0.11661281436681747,
+      "learning_rate": 3.855940934195146e-07,
+      "loss": 0.3459,
+      "step": 298
+    },
+    {
+      "epoch": 2.669642857142857,
+      "grad_norm": 0.11176279187202454,
+      "learning_rate": 3.6581466075023443e-07,
+      "loss": 0.3506,
+      "step": 299
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 0.1134781613945961,
+      "learning_rate": 3.4653671007259084e-07,
+      "loss": 0.3445,
+      "step": 300
+    },
+    {
+      "epoch": 2.6875,
+      "grad_norm": 0.11702830344438553,
+      "learning_rate": 3.2776232752349536e-07,
+      "loss": 0.3417,
+      "step": 301
+    },
+    {
+      "epoch": 2.696428571428571,
+      "grad_norm": 0.11138033121824265,
+      "learning_rate": 3.0949354474692937e-07,
+      "loss": 0.3439,
+      "step": 302
+    },
+    {
+      "epoch": 2.705357142857143,
+      "grad_norm": 0.11514540761709213,
+      "learning_rate": 2.9173233867409057e-07,
+      "loss": 0.3439,
+      "step": 303
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.11949031054973602,
+      "learning_rate": 2.7448063130946224e-07,
+      "loss": 0.3495,
+      "step": 304
+    },
+    {
+      "epoch": 2.7232142857142856,
+      "grad_norm": 0.10531047731637955,
+      "learning_rate": 2.577402895228243e-07,
+      "loss": 0.3417,
+      "step": 305
+    },
+    {
+      "epoch": 2.732142857142857,
+      "grad_norm": 0.11205192655324936,
+      "learning_rate": 2.4151312484723465e-07,
+      "loss": 0.343,
+      "step": 306
+    },
+    {
+      "epoch": 2.741071428571429,
+      "grad_norm": 0.10984615981578827,
+      "learning_rate": 2.2580089328299747e-07,
+      "loss": 0.3462,
+      "step": 307
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.10972468554973602,
+      "learning_rate": 2.106052951076365e-07,
+      "loss": 0.3443,
+      "step": 308
+    },
+    {
+      "epoch": 2.758928571428571,
+      "grad_norm": 0.14334134757518768,
+      "learning_rate": 1.959279746919057e-07,
+      "loss": 0.345,
+      "step": 309
+    },
+    {
+      "epoch": 2.767857142857143,
+      "grad_norm": 0.11364768445491791,
+      "learning_rate": 1.8177052032184285e-07,
+      "loss": 0.3458,
+      "step": 310
+    },
+    {
+      "epoch": 2.7767857142857144,
+      "grad_norm": 0.11008045077323914,
+      "learning_rate": 1.6813446402689783e-07,
+      "loss": 0.3482,
+      "step": 311
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.1126861646771431,
+      "learning_rate": 1.5502128141414496e-07,
+      "loss": 0.3412,
+      "step": 312
+    },
+    {
+      "epoch": 2.794642857142857,
+      "grad_norm": 0.11044777929782867,
+      "learning_rate": 1.4243239150860123e-07,
+      "loss": 0.3438,
+      "step": 313
+    },
+    {
+      "epoch": 2.803571428571429,
+      "grad_norm": 0.12912730872631073,
+      "learning_rate": 1.303691565996712e-07,
+      "loss": 0.3497,
+      "step": 314
+    },
+    {
+      "epoch": 2.8125,
+      "grad_norm": 0.11161229759454727,
+      "learning_rate": 1.1883288209372512e-07,
+      "loss": 0.3449,
+      "step": 315
+    },
+    {
+      "epoch": 2.821428571428571,
+      "grad_norm": 0.11409316956996918,
+      "learning_rate": 1.0782481637284014e-07,
+      "loss": 0.3448,
+      "step": 316
+    },
+    {
+      "epoch": 2.830357142857143,
+      "grad_norm": 0.11476591974496841,
+      "learning_rate": 9.734615065970454e-08,
+      "loss": 0.3474,
+      "step": 317
+    },
+    {
+      "epoch": 2.8392857142857144,
+      "grad_norm": 0.11300153285264969,
+      "learning_rate": 8.739801888871468e-08,
+      "loss": 0.3484,
+      "step": 318
+    },
+    {
+      "epoch": 2.8482142857142856,
+      "grad_norm": 0.10717196017503738,
+      "learning_rate": 7.79814975832649e-08,
+      "loss": 0.3508,
+      "step": 319
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.10945117473602295,
+      "learning_rate": 6.909760573925561e-08,
+      "loss": 0.3438,
+      "step": 320
+    },
+    {
+      "epoch": 2.866071428571429,
+      "grad_norm": 0.11592280864715576,
+      "learning_rate": 6.074730471482049e-08,
+      "loss": 0.3466,
+      "step": 321
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.10622705519199371,
+      "learning_rate": 5.2931498126298495e-08,
+      "loss": 0.3441,
+      "step": 322
+    },
+    {
+      "epoch": 2.883928571428571,
+      "grad_norm": 0.11332610249519348,
+      "learning_rate": 4.565103175044883e-08,
+      "loss": 0.3475,
+      "step": 323
+    },
+    {
+      "epoch": 2.892857142857143,
+      "grad_norm": 0.12698130309581757,
+      "learning_rate": 3.890669343292464e-08,
+      "loss": 0.3478,
+      "step": 324
+    },
+    {
+      "epoch": 2.9017857142857144,
+      "grad_norm": 0.11187270283699036,
+      "learning_rate": 3.269921300301959e-08,
+      "loss": 0.3431,
+      "step": 325
+    },
+    {
+      "epoch": 2.9107142857142856,
+      "grad_norm": 0.1112656369805336,
+      "learning_rate": 2.702926219468882e-08,
+      "loss": 0.3441,
+      "step": 326
+    },
+    {
+      "epoch": 2.919642857142857,
+      "grad_norm": 0.11196763813495636,
+      "learning_rate": 2.1897454573860388e-08,
+      "loss": 0.3538,
+      "step": 327
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.11176672577857971,
+      "learning_rate": 1.7304345472035634e-08,
+      "loss": 0.3431,
+      "step": 328
+    },
+    {
+      "epoch": 2.9375,
+      "grad_norm": 0.11075292527675629,
+      "learning_rate": 1.3250431926197793e-08,
+      "loss": 0.3412,
+      "step": 329
+    },
+    {
+      "epoch": 2.946428571428571,
+      "grad_norm": 0.10811440646648407,
+      "learning_rate": 9.73615262502503e-09,
+      "loss": 0.346,
+      "step": 330
+    },
+    {
+      "epoch": 2.955357142857143,
+      "grad_norm": 0.1118391752243042,
+      "learning_rate": 6.761887861417293e-09,
+      "loss": 0.3491,
+      "step": 331
+    },
+    {
+      "epoch": 2.9642857142857144,
+      "grad_norm": 0.10914882272481918,
+      "learning_rate": 4.327959491344791e-09,
+      "loss": 0.3433,
+      "step": 332
+    },
+    {
+      "epoch": 2.9732142857142856,
+      "grad_norm": 0.11297401040792465,
+      "learning_rate": 2.4346308990175205e-09,
+      "loss": 0.3467,
+      "step": 333
+    },
+    {
+      "epoch": 2.982142857142857,
+      "grad_norm": 0.1087309867143631,
+      "learning_rate": 1.082106968385288e-09,
+      "loss": 0.343,
+      "step": 334
+    },
+    {
+      "epoch": 2.991071428571429,
+      "grad_norm": 0.1101929172873497,
+      "learning_rate": 2.7053406096433366e-10,
+      "loss": 0.3447,
+      "step": 335
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.13185350596904755,
+      "learning_rate": 0.0,
+      "loss": 0.3407,
+      "step": 336
+    },
+    {
+      "epoch": 3.0,
+      "step": 336,
+      "total_flos": 2.982866014384947e+16,
+      "train_loss": 0.3923433954339652,
+      "train_runtime": 48299.7361,
+      "train_samples_per_second": 7.078,
+      "train_steps_per_second": 0.007
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 336,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.982866014384947e+16,
+  "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:eb0244194182469e91b3173374e869ca007a72c3ea3f197b9369a89cb93be320
+size 7608