Add files using upload-large-folder tool

.gitattributes

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

README.md

@@ -0,0 +1,61 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen2.5-Math-1.5B-Instruct
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: Speaking_way_Qwen__QwQ-32B-Preview_cc_cc
+  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. -->
+
+# Speaking_way_Qwen__QwQ-32B-Preview_cc_cc
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-Math-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B-Instruct) on the Speaking_way_Qwen__QwQ-32B-Preview_cc_cc dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.4302
+
+## 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: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 4
+- total_eval_batch_size: 4
+- 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
+- num_epochs: 2
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.46.1
+- Pytorch 2.6.0+cu124
+- Datasets 3.1.0
+- Tokenizers 0.20.3

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 0.430238276720047,
+    "eval_runtime": 0.29,
+    "eval_samples_per_second": 10.345,
+    "eval_steps_per_second": 3.448,
+    "total_flos": 3190623830016.0,
+    "train_loss": 0.43240329596613136,
+    "train_runtime": 137.9931,
+    "train_samples_per_second": 3.247,
+    "train_steps_per_second": 0.812
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-Math-1.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 4096,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.1",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 0.430238276720047,
+    "eval_runtime": 0.29,
+    "eval_samples_per_second": 10.345,
+    "eval_steps_per_second": 3.448
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "transformers_version": "4.46.1"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77d54457f8e9fe76cafe3c4b7fb4bcbe57db5e21f29b2c176e6bab7bab57d9b3
+size 3087467144

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": "<|im_end|>",
+    "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        {{- 'Please reason step by step, and put your final answer within \\\\boxed{}.' }}\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\\nPlease reason step by step, and put your final answer within \\\\boxed{}.<|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": "<|im_end|>",
+  "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": 2.0,
+    "total_flos": 3190623830016.0,
+    "train_loss": 0.43240329596613136,
+    "train_runtime": 137.9931,
+    "train_samples_per_second": 3.247,
+    "train_steps_per_second": 0.812
+}

trainer_log.jsonl

@@ -0,0 +1,113 @@
+{"current_steps": 1, "total_steps": 112, "loss": 0.6974, "lr": 9.998033131915266e-06, "epoch": 0.017857142857142856, "percentage": 0.89, "elapsed_time": "0:00:02", "remaining_time": "0:04:15"}
+{"current_steps": 2, "total_steps": 112, "loss": 0.8358, "lr": 9.992134075089085e-06, "epoch": 0.03571428571428571, "percentage": 1.79, "elapsed_time": "0:00:05", "remaining_time": "0:05:24"}
+{"current_steps": 3, "total_steps": 112, "loss": 0.6795, "lr": 9.982307470588097e-06, "epoch": 0.05357142857142857, "percentage": 2.68, "elapsed_time": "0:00:06", "remaining_time": "0:04:12"}
+{"current_steps": 4, "total_steps": 112, "loss": 0.6181, "lr": 9.968561049466214e-06, "epoch": 0.07142857142857142, "percentage": 3.57, "elapsed_time": "0:00:08", "remaining_time": "0:03:36"}
+{"current_steps": 5, "total_steps": 112, "loss": 0.5699, "lr": 9.950905626682229e-06, "epoch": 0.08928571428571429, "percentage": 4.46, "elapsed_time": "0:00:09", "remaining_time": "0:03:14"}
+{"current_steps": 6, "total_steps": 112, "loss": 0.721, "lr": 9.92935509259118e-06, "epoch": 0.10714285714285714, "percentage": 5.36, "elapsed_time": "0:00:10", "remaining_time": "0:02:59"}
+{"current_steps": 7, "total_steps": 112, "loss": 0.6084, "lr": 9.903926402016153e-06, "epoch": 0.125, "percentage": 6.25, "elapsed_time": "0:00:11", "remaining_time": "0:02:48"}
+{"current_steps": 8, "total_steps": 112, "loss": 0.5652, "lr": 9.874639560909118e-06, "epoch": 0.14285714285714285, "percentage": 7.14, "elapsed_time": "0:00:12", "remaining_time": "0:02:39"}
+{"current_steps": 9, "total_steps": 112, "loss": 0.5108, "lr": 9.841517610611309e-06, "epoch": 0.16071428571428573, "percentage": 8.04, "elapsed_time": "0:00:13", "remaining_time": "0:02:32"}
+{"current_steps": 10, "total_steps": 112, "loss": 0.6436, "lr": 9.804586609725499e-06, "epoch": 0.17857142857142858, "percentage": 8.93, "elapsed_time": "0:00:14", "remaining_time": "0:02:26"}
+{"current_steps": 11, "total_steps": 112, "loss": 0.4997, "lr": 9.763875613614482e-06, "epoch": 0.19642857142857142, "percentage": 9.82, "elapsed_time": "0:00:15", "remaining_time": "0:02:21"}
+{"current_steps": 12, "total_steps": 112, "loss": 0.6113, "lr": 9.719416651541839e-06, "epoch": 0.21428571428571427, "percentage": 10.71, "elapsed_time": "0:00:16", "remaining_time": "0:02:17"}
+{"current_steps": 13, "total_steps": 112, "loss": 0.5743, "lr": 9.671244701472999e-06, "epoch": 0.23214285714285715, "percentage": 11.61, "elapsed_time": "0:00:17", "remaining_time": "0:02:14"}
+{"current_steps": 14, "total_steps": 112, "loss": 0.6872, "lr": 9.619397662556434e-06, "epoch": 0.25, "percentage": 12.5, "elapsed_time": "0:00:18", "remaining_time": "0:02:10"}
+{"current_steps": 15, "total_steps": 112, "loss": 0.5076, "lr": 9.563916325306595e-06, "epoch": 0.26785714285714285, "percentage": 13.39, "elapsed_time": "0:00:19", "remaining_time": "0:02:07"}
+{"current_steps": 16, "total_steps": 112, "loss": 0.4452, "lr": 9.504844339512096e-06, "epoch": 0.2857142857142857, "percentage": 14.29, "elapsed_time": "0:00:20", "remaining_time": "0:02:04"}
+{"current_steps": 17, "total_steps": 112, "loss": 0.5083, "lr": 9.442228179894362e-06, "epoch": 0.30357142857142855, "percentage": 15.18, "elapsed_time": "0:00:21", "remaining_time": "0:02:02"}
+{"current_steps": 18, "total_steps": 112, "loss": 0.5823, "lr": 9.376117109543769e-06, "epoch": 0.32142857142857145, "percentage": 16.07, "elapsed_time": "0:00:22", "remaining_time": "0:02:00"}
+{"current_steps": 19, "total_steps": 112, "loss": 0.4695, "lr": 9.306563141162046e-06, "epoch": 0.3392857142857143, "percentage": 16.96, "elapsed_time": "0:00:24", "remaining_time": "0:01:57"}
+{"current_steps": 20, "total_steps": 112, "loss": 0.593, "lr": 9.233620996141421e-06, "epoch": 0.35714285714285715, "percentage": 17.86, "elapsed_time": "0:00:25", "remaining_time": "0:01:55"}
+{"current_steps": 21, "total_steps": 112, "loss": 0.4043, "lr": 9.157348061512728e-06, "epoch": 0.375, "percentage": 18.75, "elapsed_time": "0:00:26", "remaining_time": "0:01:53"}
+{"current_steps": 22, "total_steps": 112, "loss": 0.4668, "lr": 9.077804344796302e-06, "epoch": 0.39285714285714285, "percentage": 19.64, "elapsed_time": "0:00:27", "remaining_time": "0:01:51"}
+{"current_steps": 23, "total_steps": 112, "loss": 0.6571, "lr": 8.995052426791247e-06, "epoch": 0.4107142857142857, "percentage": 20.54, "elapsed_time": "0:00:28", "remaining_time": "0:01:49"}
+{"current_steps": 24, "total_steps": 112, "loss": 0.4553, "lr": 8.90915741234015e-06, "epoch": 0.42857142857142855, "percentage": 21.43, "elapsed_time": "0:00:29", "remaining_time": "0:01:47"}
+{"current_steps": 25, "total_steps": 112, "loss": 0.4036, "lr": 8.820186879108038e-06, "epoch": 0.44642857142857145, "percentage": 22.32, "elapsed_time": "0:00:30", "remaining_time": "0:01:45"}
+{"current_steps": 26, "total_steps": 112, "loss": 0.4632, "lr": 8.728210824415829e-06, "epoch": 0.4642857142857143, "percentage": 23.21, "elapsed_time": "0:00:31", "remaining_time": "0:01:44"}
+{"current_steps": 27, "total_steps": 112, "loss": 0.4398, "lr": 8.633301610170136e-06, "epoch": 0.48214285714285715, "percentage": 24.11, "elapsed_time": "0:00:32", "remaining_time": "0:01:42"}
+{"current_steps": 28, "total_steps": 112, "loss": 0.4645, "lr": 8.535533905932739e-06, "epoch": 0.5, "percentage": 25.0, "elapsed_time": "0:00:33", "remaining_time": "0:01:40"}
+{"current_steps": 29, "total_steps": 112, "loss": 0.4404, "lr": 8.43498463017451e-06, "epoch": 0.5178571428571429, "percentage": 25.89, "elapsed_time": "0:00:34", "remaining_time": "0:01:39"}
+{"current_steps": 30, "total_steps": 112, "loss": 0.4694, "lr": 8.331732889760021e-06, "epoch": 0.5357142857142857, "percentage": 26.79, "elapsed_time": "0:00:35", "remaining_time": "0:01:37"}
+{"current_steps": 31, "total_steps": 112, "loss": 0.4527, "lr": 8.22585991771044e-06, "epoch": 0.5535714285714286, "percentage": 27.68, "elapsed_time": "0:00:36", "remaining_time": "0:01:36"}
+{"current_steps": 32, "total_steps": 112, "loss": 0.528, "lr": 8.117449009293668e-06, "epoch": 0.5714285714285714, "percentage": 28.57, "elapsed_time": "0:00:37", "remaining_time": "0:01:34"}
+{"current_steps": 33, "total_steps": 112, "loss": 0.5084, "lr": 8.00658545649203e-06, "epoch": 0.5892857142857143, "percentage": 29.46, "elapsed_time": "0:00:39", "remaining_time": "0:01:33"}
+{"current_steps": 34, "total_steps": 112, "loss": 0.4599, "lr": 7.89335648089903e-06, "epoch": 0.6071428571428571, "percentage": 30.36, "elapsed_time": "0:00:40", "remaining_time": "0:01:32"}
+{"current_steps": 35, "total_steps": 112, "loss": 0.3561, "lr": 7.777851165098012e-06, "epoch": 0.625, "percentage": 31.25, "elapsed_time": "0:00:41", "remaining_time": "0:01:30"}
+{"current_steps": 36, "total_steps": 112, "loss": 0.4894, "lr": 7.660160382576683e-06, "epoch": 0.6428571428571429, "percentage": 32.14, "elapsed_time": "0:00:42", "remaining_time": "0:01:29"}
+{"current_steps": 37, "total_steps": 112, "loss": 0.5342, "lr": 7.540376726232648e-06, "epoch": 0.6607142857142857, "percentage": 33.04, "elapsed_time": "0:00:43", "remaining_time": "0:01:27"}
+{"current_steps": 38, "total_steps": 112, "loss": 0.5164, "lr": 7.4185944355261996e-06, "epoch": 0.6785714285714286, "percentage": 33.93, "elapsed_time": "0:00:44", "remaining_time": "0:01:26"}
+{"current_steps": 39, "total_steps": 112, "loss": 0.4159, "lr": 7.294909322337689e-06, "epoch": 0.6964285714285714, "percentage": 34.82, "elapsed_time": "0:00:45", "remaining_time": "0:01:25"}
+{"current_steps": 40, "total_steps": 112, "loss": 0.3821, "lr": 7.169418695587791e-06, "epoch": 0.7142857142857143, "percentage": 35.71, "elapsed_time": "0:00:46", "remaining_time": "0:01:23"}
+{"current_steps": 41, "total_steps": 112, "loss": 0.3631, "lr": 7.042221284679982e-06, "epoch": 0.7321428571428571, "percentage": 36.61, "elapsed_time": "0:00:47", "remaining_time": "0:01:22"}
+{"current_steps": 42, "total_steps": 112, "loss": 0.3928, "lr": 6.913417161825449e-06, "epoch": 0.75, "percentage": 37.5, "elapsed_time": "0:00:48", "remaining_time": "0:01:21"}
+{"current_steps": 43, "total_steps": 112, "loss": 0.4428, "lr": 6.783107663311566e-06, "epoch": 0.7678571428571429, "percentage": 38.39, "elapsed_time": "0:00:49", "remaining_time": "0:01:20"}
+{"current_steps": 44, "total_steps": 112, "loss": 0.3863, "lr": 6.651395309775837e-06, "epoch": 0.7857142857142857, "percentage": 39.29, "elapsed_time": "0:00:50", "remaining_time": "0:01:18"}
+{"current_steps": 45, "total_steps": 112, "loss": 0.3592, "lr": 6.518383725548074e-06, "epoch": 0.8035714285714286, "percentage": 40.18, "elapsed_time": "0:00:52", "remaining_time": "0:01:17"}
+{"current_steps": 46, "total_steps": 112, "loss": 0.44, "lr": 6.384177557124247e-06, "epoch": 0.8214285714285714, "percentage": 41.07, "elapsed_time": "0:00:53", "remaining_time": "0:01:16"}
+{"current_steps": 47, "total_steps": 112, "loss": 0.4012, "lr": 6.248882390836135e-06, "epoch": 0.8392857142857143, "percentage": 41.96, "elapsed_time": "0:00:54", "remaining_time": "0:01:15"}
+{"current_steps": 48, "total_steps": 112, "loss": 0.369, "lr": 6.112604669781572e-06, "epoch": 0.8571428571428571, "percentage": 42.86, "elapsed_time": "0:00:55", "remaining_time": "0:01:13"}
+{"current_steps": 49, "total_steps": 112, "loss": 0.4084, "lr": 5.975451610080643e-06, "epoch": 0.875, "percentage": 43.75, "elapsed_time": "0:00:56", "remaining_time": "0:01:12"}
+{"current_steps": 50, "total_steps": 112, "loss": 0.37, "lr": 5.837531116523683e-06, "epoch": 0.8928571428571429, "percentage": 44.64, "elapsed_time": "0:00:57", "remaining_time": "0:01:11"}
+{"current_steps": 51, "total_steps": 112, "loss": 0.4036, "lr": 5.698951697677498e-06, "epoch": 0.9107142857142857, "percentage": 45.54, "elapsed_time": "0:00:58", "remaining_time": "0:01:10"}
+{"current_steps": 52, "total_steps": 112, "loss": 0.367, "lr": 5.559822380516539e-06, "epoch": 0.9285714285714286, "percentage": 46.43, "elapsed_time": "0:00:59", "remaining_time": "0:01:08"}
+{"current_steps": 53, "total_steps": 112, "loss": 0.4085, "lr": 5.420252624646238e-06, "epoch": 0.9464285714285714, "percentage": 47.32, "elapsed_time": "0:01:00", "remaining_time": "0:01:07"}
+{"current_steps": 54, "total_steps": 112, "loss": 0.4651, "lr": 5.2803522361859596e-06, "epoch": 0.9642857142857143, "percentage": 48.21, "elapsed_time": "0:01:01", "remaining_time": "0:01:06"}
+{"current_steps": 55, "total_steps": 112, "loss": 0.395, "lr": 5.140231281379345e-06, "epoch": 0.9821428571428571, "percentage": 49.11, "elapsed_time": "0:01:02", "remaining_time": "0:01:05"}
+{"current_steps": 56, "total_steps": 112, "loss": 0.4398, "lr": 5e-06, "epoch": 1.0, "percentage": 50.0, "elapsed_time": "0:01:04", "remaining_time": "0:01:04"}
+{"current_steps": 57, "total_steps": 112, "loss": 0.3122, "lr": 4.859768718620656e-06, "epoch": 1.0178571428571428, "percentage": 50.89, "elapsed_time": "0:01:05", "remaining_time": "0:01:02"}
+{"current_steps": 58, "total_steps": 112, "loss": 0.4237, "lr": 4.719647763814041e-06, "epoch": 1.0357142857142858, "percentage": 51.79, "elapsed_time": "0:01:06", "remaining_time": "0:01:01"}
+{"current_steps": 59, "total_steps": 112, "loss": 0.4167, "lr": 4.579747375353763e-06, "epoch": 1.0535714285714286, "percentage": 52.68, "elapsed_time": "0:01:07", "remaining_time": "0:01:00"}
+{"current_steps": 60, "total_steps": 112, "loss": 0.381, "lr": 4.4401776194834615e-06, "epoch": 1.0714285714285714, "percentage": 53.57, "elapsed_time": "0:01:08", "remaining_time": "0:00:59"}
+{"current_steps": 61, "total_steps": 112, "loss": 0.4389, "lr": 4.3010483023225045e-06, "epoch": 1.0892857142857142, "percentage": 54.46, "elapsed_time": "0:01:09", "remaining_time": "0:00:58"}
+{"current_steps": 62, "total_steps": 112, "loss": 0.4297, "lr": 4.162468883476319e-06, "epoch": 1.1071428571428572, "percentage": 55.36, "elapsed_time": "0:01:10", "remaining_time": "0:00:56"}
+{"current_steps": 63, "total_steps": 112, "loss": 0.3709, "lr": 4.02454838991936e-06, "epoch": 1.125, "percentage": 56.25, "elapsed_time": "0:01:11", "remaining_time": "0:00:55"}
+{"current_steps": 64, "total_steps": 112, "loss": 0.3859, "lr": 3.887395330218429e-06, "epoch": 1.1428571428571428, "percentage": 57.14, "elapsed_time": "0:01:12", "remaining_time": "0:00:54"}
+{"current_steps": 65, "total_steps": 112, "loss": 0.3352, "lr": 3.751117609163865e-06, "epoch": 1.1607142857142858, "percentage": 58.04, "elapsed_time": "0:01:13", "remaining_time": "0:00:53"}
+{"current_steps": 66, "total_steps": 112, "loss": 0.3092, "lr": 3.6158224428757538e-06, "epoch": 1.1785714285714286, "percentage": 58.93, "elapsed_time": "0:01:15", "remaining_time": "0:00:52"}
+{"current_steps": 67, "total_steps": 112, "loss": 0.5043, "lr": 3.4816162744519266e-06, "epoch": 1.1964285714285714, "percentage": 59.82, "elapsed_time": "0:01:16", "remaining_time": "0:00:51"}
+{"current_steps": 68, "total_steps": 112, "loss": 0.3273, "lr": 3.3486046902241663e-06, "epoch": 1.2142857142857142, "percentage": 60.71, "elapsed_time": "0:01:17", "remaining_time": "0:00:50"}
+{"current_steps": 69, "total_steps": 112, "loss": 0.3542, "lr": 3.216892336688435e-06, "epoch": 1.2321428571428572, "percentage": 61.61, "elapsed_time": "0:01:18", "remaining_time": "0:00:48"}
+{"current_steps": 70, "total_steps": 112, "loss": 0.3339, "lr": 3.0865828381745515e-06, "epoch": 1.25, "percentage": 62.5, "elapsed_time": "0:01:19", "remaining_time": "0:00:47"}
+{"current_steps": 71, "total_steps": 112, "loss": 0.2963, "lr": 2.95777871532002e-06, "epoch": 1.2678571428571428, "percentage": 63.39, "elapsed_time": "0:01:20", "remaining_time": "0:00:46"}
+{"current_steps": 72, "total_steps": 112, "loss": 0.413, "lr": 2.83058130441221e-06, "epoch": 1.2857142857142856, "percentage": 64.29, "elapsed_time": "0:01:21", "remaining_time": "0:00:45"}
+{"current_steps": 73, "total_steps": 112, "loss": 0.3826, "lr": 2.705090677662311e-06, "epoch": 1.3035714285714286, "percentage": 65.18, "elapsed_time": "0:01:22", "remaining_time": "0:00:44"}
+{"current_steps": 74, "total_steps": 112, "loss": 0.3197, "lr": 2.5814055644738013e-06, "epoch": 1.3214285714285714, "percentage": 66.07, "elapsed_time": "0:01:23", "remaining_time": "0:00:43"}
+{"current_steps": 75, "total_steps": 112, "loss": 0.2913, "lr": 2.4596232737673544e-06, "epoch": 1.3392857142857144, "percentage": 66.96, "elapsed_time": "0:01:24", "remaining_time": "0:00:41"}
+{"current_steps": 76, "total_steps": 112, "loss": 0.3952, "lr": 2.339839617423318e-06, "epoch": 1.3571428571428572, "percentage": 67.86, "elapsed_time": "0:01:26", "remaining_time": "0:00:40"}
+{"current_steps": 77, "total_steps": 112, "loss": 0.3802, "lr": 2.2221488349019903e-06, "epoch": 1.375, "percentage": 68.75, "elapsed_time": "0:01:27", "remaining_time": "0:00:39"}
+{"current_steps": 78, "total_steps": 112, "loss": 0.3438, "lr": 2.1066435191009717e-06, "epoch": 1.3928571428571428, "percentage": 69.64, "elapsed_time": "0:01:28", "remaining_time": "0:00:38"}
+{"current_steps": 79, "total_steps": 112, "loss": 0.3833, "lr": 1.9934145435079705e-06, "epoch": 1.4107142857142856, "percentage": 70.54, "elapsed_time": "0:01:29", "remaining_time": "0:00:37"}
+{"current_steps": 80, "total_steps": 112, "loss": 0.2786, "lr": 1.8825509907063328e-06, "epoch": 1.4285714285714286, "percentage": 71.43, "elapsed_time": "0:01:30", "remaining_time": "0:00:36"}
+{"current_steps": 81, "total_steps": 112, "loss": 0.3328, "lr": 1.7741400822895633e-06, "epoch": 1.4464285714285714, "percentage": 72.32, "elapsed_time": "0:01:31", "remaining_time": "0:00:35"}
+{"current_steps": 82, "total_steps": 112, "loss": 0.3151, "lr": 1.6682671102399806e-06, "epoch": 1.4642857142857144, "percentage": 73.21, "elapsed_time": "0:01:32", "remaining_time": "0:00:33"}
+{"current_steps": 83, "total_steps": 112, "loss": 0.3865, "lr": 1.5650153698254916e-06, "epoch": 1.4821428571428572, "percentage": 74.11, "elapsed_time": "0:01:33", "remaining_time": "0:00:32"}
+{"current_steps": 84, "total_steps": 112, "loss": 0.3846, "lr": 1.4644660940672628e-06, "epoch": 1.5, "percentage": 75.0, "elapsed_time": "0:01:34", "remaining_time": "0:00:31"}
+{"current_steps": 85, "total_steps": 112, "loss": 0.3494, "lr": 1.3666983898298659e-06, "epoch": 1.5178571428571428, "percentage": 75.89, "elapsed_time": "0:01:35", "remaining_time": "0:00:30"}
+{"current_steps": 86, "total_steps": 112, "loss": 0.372, "lr": 1.2717891755841722e-06, "epoch": 1.5357142857142856, "percentage": 76.79, "elapsed_time": "0:01:37", "remaining_time": "0:00:29"}
+{"current_steps": 87, "total_steps": 112, "loss": 0.3104, "lr": 1.1798131208919628e-06, "epoch": 1.5535714285714286, "percentage": 77.68, "elapsed_time": "0:01:38", "remaining_time": "0:00:28"}
+{"current_steps": 88, "total_steps": 112, "loss": 0.3788, "lr": 1.0908425876598512e-06, "epoch": 1.5714285714285714, "percentage": 78.57, "elapsed_time": "0:01:39", "remaining_time": "0:00:27"}
+{"current_steps": 89, "total_steps": 112, "loss": 0.3517, "lr": 1.004947573208756e-06, "epoch": 1.5892857142857144, "percentage": 79.46, "elapsed_time": "0:01:40", "remaining_time": "0:00:25"}
+{"current_steps": 90, "total_steps": 112, "loss": 0.3701, "lr": 9.221956552036992e-07, "epoch": 1.6071428571428572, "percentage": 80.36, "elapsed_time": "0:01:41", "remaining_time": "0:00:24"}
+{"current_steps": 91, "total_steps": 112, "loss": 0.3948, "lr": 8.426519384872733e-07, "epoch": 1.625, "percentage": 81.25, "elapsed_time": "0:01:42", "remaining_time": "0:00:23"}
+{"current_steps": 92, "total_steps": 112, "loss": 0.3536, "lr": 7.663790038585794e-07, "epoch": 1.6428571428571428, "percentage": 82.14, "elapsed_time": "0:01:43", "remaining_time": "0:00:22"}
+{"current_steps": 93, "total_steps": 112, "loss": 0.3571, "lr": 6.934368588379553e-07, "epoch": 1.6607142857142856, "percentage": 83.04, "elapsed_time": "0:01:44", "remaining_time": "0:00:21"}
+{"current_steps": 94, "total_steps": 112, "loss": 0.4095, "lr": 6.238828904562316e-07, "epoch": 1.6785714285714286, "percentage": 83.93, "elapsed_time": "0:01:45", "remaining_time": "0:00:20"}
+{"current_steps": 95, "total_steps": 112, "loss": 0.4505, "lr": 5.577718201056392e-07, "epoch": 1.6964285714285714, "percentage": 84.82, "elapsed_time": "0:01:46", "remaining_time": "0:00:19"}
+{"current_steps": 96, "total_steps": 112, "loss": 0.3527, "lr": 4.951556604879049e-07, "epoch": 1.7142857142857144, "percentage": 85.71, "elapsed_time": "0:01:48", "remaining_time": "0:00:18"}
+{"current_steps": 97, "total_steps": 112, "loss": 0.3811, "lr": 4.3608367469340553e-07, "epoch": 1.7321428571428572, "percentage": 86.61, "elapsed_time": "0:01:49", "remaining_time": "0:00:16"}
+{"current_steps": 98, "total_steps": 112, "loss": 0.3452, "lr": 3.8060233744356634e-07, "epoch": 1.75, "percentage": 87.5, "elapsed_time": "0:01:50", "remaining_time": "0:00:15"}
+{"current_steps": 99, "total_steps": 112, "loss": 0.3598, "lr": 3.287552985270015e-07, "epoch": 1.7678571428571428, "percentage": 88.39, "elapsed_time": "0:01:51", "remaining_time": "0:00:14"}
+{"current_steps": 100, "total_steps": 112, "loss": 0.4311, "lr": 2.8058334845816214e-07, "epoch": 1.7857142857142856, "percentage": 89.29, "elapsed_time": "0:01:52", "remaining_time": "0:00:13"}
+{"current_steps": 101, "total_steps": 112, "loss": 0.4095, "lr": 2.3612438638551837e-07, "epoch": 1.8035714285714286, "percentage": 90.18, "elapsed_time": "0:01:53", "remaining_time": "0:00:12"}
+{"current_steps": 102, "total_steps": 112, "loss": 0.3693, "lr": 1.9541339027450256e-07, "epoch": 1.8214285714285714, "percentage": 91.07, "elapsed_time": "0:01:54", "remaining_time": "0:00:11"}
+{"current_steps": 103, "total_steps": 112, "loss": 0.3714, "lr": 1.5848238938869332e-07, "epoch": 1.8392857142857144, "percentage": 91.96, "elapsed_time": "0:01:55", "remaining_time": "0:00:10"}
+{"current_steps": 104, "total_steps": 112, "loss": 0.3524, "lr": 1.253604390908819e-07, "epoch": 1.8571428571428572, "percentage": 92.86, "elapsed_time": "0:01:56", "remaining_time": "0:00:08"}
+{"current_steps": 105, "total_steps": 112, "loss": 0.4088, "lr": 9.607359798384785e-08, "epoch": 1.875, "percentage": 93.75, "elapsed_time": "0:01:57", "remaining_time": "0:00:07"}
+{"current_steps": 106, "total_steps": 112, "loss": 0.2922, "lr": 7.064490740882057e-08, "epoch": 1.8928571428571428, "percentage": 94.64, "elapsed_time": "0:01:59", "remaining_time": "0:00:06"}
+{"current_steps": 107, "total_steps": 112, "loss": 0.3933, "lr": 4.909437331777178e-08, "epoch": 1.9107142857142856, "percentage": 95.54, "elapsed_time": "0:02:00", "remaining_time": "0:00:05"}
+{"current_steps": 108, "total_steps": 112, "loss": 0.4854, "lr": 3.143895053378698e-08, "epoch": 1.9285714285714286, "percentage": 96.43, "elapsed_time": "0:02:01", "remaining_time": "0:00:04"}
+{"current_steps": 109, "total_steps": 112, "loss": 0.389, "lr": 1.769252941190458e-08, "epoch": 1.9464285714285714, "percentage": 97.32, "elapsed_time": "0:02:02", "remaining_time": "0:00:03"}
+{"current_steps": 110, "total_steps": 112, "loss": 0.3559, "lr": 7.865924910916977e-09, "epoch": 1.9642857142857144, "percentage": 98.21, "elapsed_time": "0:02:03", "remaining_time": "0:00:02"}
+{"current_steps": 111, "total_steps": 112, "loss": 0.3549, "lr": 1.9668680847356735e-09, "epoch": 1.9821428571428572, "percentage": 99.11, "elapsed_time": "0:02:04", "remaining_time": "0:00:01"}
+{"current_steps": 112, "total_steps": 112, "loss": 0.4064, "lr": 0.0, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "0:02:05", "remaining_time": "0:00:00"}
+{"current_steps": 112, "total_steps": 112, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "0:02:17", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,826 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 112,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.017857142857142856,
+      "grad_norm": 1.3127956310764763,
+      "learning_rate": 9.998033131915266e-06,
+      "loss": 0.6974,
+      "step": 1
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 1.6235178804357906,
+      "learning_rate": 9.992134075089085e-06,
+      "loss": 0.8358,
+      "step": 2
+    },
+    {
+      "epoch": 0.05357142857142857,
+      "grad_norm": 2.174957867334859,
+      "learning_rate": 9.982307470588097e-06,
+      "loss": 0.6795,
+      "step": 3
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.0346935109585302,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.6181,
+      "step": 4
+    },
+    {
+      "epoch": 0.08928571428571429,
+      "grad_norm": 0.7666095700709454,
+      "learning_rate": 9.950905626682229e-06,
+      "loss": 0.5699,
+      "step": 5
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 1.056810418580731,
+      "learning_rate": 9.92935509259118e-06,
+      "loss": 0.721,
+      "step": 6
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 1.1811684804284484,
+      "learning_rate": 9.903926402016153e-06,
+      "loss": 0.6084,
+      "step": 7
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.7990534841171754,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.5652,
+      "step": 8
+    },
+    {
+      "epoch": 0.16071428571428573,
+      "grad_norm": 0.8027486215505741,
+      "learning_rate": 9.841517610611309e-06,
+      "loss": 0.5108,
+      "step": 9
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 1.0218975470411973,
+      "learning_rate": 9.804586609725499e-06,
+      "loss": 0.6436,
+      "step": 10
+    },
+    {
+      "epoch": 0.19642857142857142,
+      "grad_norm": 0.7369129809098821,
+      "learning_rate": 9.763875613614482e-06,
+      "loss": 0.4997,
+      "step": 11
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.2070251338796105,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.6113,
+      "step": 12
+    },
+    {
+      "epoch": 0.23214285714285715,
+      "grad_norm": 1.4072933439802777,
+      "learning_rate": 9.671244701472999e-06,
+      "loss": 0.5743,
+      "step": 13
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7626018676782569,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.6872,
+      "step": 14
+    },
+    {
+      "epoch": 0.26785714285714285,
+      "grad_norm": 0.7422325448624677,
+      "learning_rate": 9.563916325306595e-06,
+      "loss": 0.5076,
+      "step": 15
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.6377052718389377,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.4452,
+      "step": 16
+    },
+    {
+      "epoch": 0.30357142857142855,
+      "grad_norm": 0.7561204248290452,
+      "learning_rate": 9.442228179894362e-06,
+      "loss": 0.5083,
+      "step": 17
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 1.2650799618566455,
+      "learning_rate": 9.376117109543769e-06,
+      "loss": 0.5823,
+      "step": 18
+    },
+    {
+      "epoch": 0.3392857142857143,
+      "grad_norm": 0.581342829090626,
+      "learning_rate": 9.306563141162046e-06,
+      "loss": 0.4695,
+      "step": 19
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.7813559756892376,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.593,
+      "step": 20
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.5147151070065507,
+      "learning_rate": 9.157348061512728e-06,
+      "loss": 0.4043,
+      "step": 21
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 0.6193835941442335,
+      "learning_rate": 9.077804344796302e-06,
+      "loss": 0.4668,
+      "step": 22
+    },
+    {
+      "epoch": 0.4107142857142857,
+      "grad_norm": 1.6783107665848302,
+      "learning_rate": 8.995052426791247e-06,
+      "loss": 0.6571,
+      "step": 23
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.5555819389623258,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.4553,
+      "step": 24
+    },
+    {
+      "epoch": 0.44642857142857145,
+      "grad_norm": 0.5718575691046155,
+      "learning_rate": 8.820186879108038e-06,
+      "loss": 0.4036,
+      "step": 25
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 0.8949856246701675,
+      "learning_rate": 8.728210824415829e-06,
+      "loss": 0.4632,
+      "step": 26
+    },
+    {
+      "epoch": 0.48214285714285715,
+      "grad_norm": 0.5215068458859718,
+      "learning_rate": 8.633301610170136e-06,
+      "loss": 0.4398,
+      "step": 27
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8402876806141819,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.4645,
+      "step": 28
+    },
+    {
+      "epoch": 0.5178571428571429,
+      "grad_norm": 0.5479176399993333,
+      "learning_rate": 8.43498463017451e-06,
+      "loss": 0.4404,
+      "step": 29
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 0.5051393454153713,
+      "learning_rate": 8.331732889760021e-06,
+      "loss": 0.4694,
+      "step": 30
+    },
+    {
+      "epoch": 0.5535714285714286,
+      "grad_norm": 0.5042489206195668,
+      "learning_rate": 8.22585991771044e-06,
+      "loss": 0.4527,
+      "step": 31
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.7581308052833844,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.528,
+      "step": 32
+    },
+    {
+      "epoch": 0.5892857142857143,
+      "grad_norm": 0.8309207449398666,
+      "learning_rate": 8.00658545649203e-06,
+      "loss": 0.5084,
+      "step": 33
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 0.4894933847266823,
+      "learning_rate": 7.89335648089903e-06,
+      "loss": 0.4599,
+      "step": 34
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.4544927514370947,
+      "learning_rate": 7.777851165098012e-06,
+      "loss": 0.3561,
+      "step": 35
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.5368608960905278,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.4894,
+      "step": 36
+    },
+    {
+      "epoch": 0.6607142857142857,
+      "grad_norm": 0.7489932429378126,
+      "learning_rate": 7.540376726232648e-06,
+      "loss": 0.5342,
+      "step": 37
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 0.7110021752951674,
+      "learning_rate": 7.4185944355261996e-06,
+      "loss": 0.5164,
+      "step": 38
+    },
+    {
+      "epoch": 0.6964285714285714,
+      "grad_norm": 0.46135875629116857,
+      "learning_rate": 7.294909322337689e-06,
+      "loss": 0.4159,
+      "step": 39
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.46634313324048204,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.3821,
+      "step": 40
+    },
+    {
+      "epoch": 0.7321428571428571,
+      "grad_norm": 0.442108131123772,
+      "learning_rate": 7.042221284679982e-06,
+      "loss": 0.3631,
+      "step": 41
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6202570606265427,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.3928,
+      "step": 42
+    },
+    {
+      "epoch": 0.7678571428571429,
+      "grad_norm": 0.4617402615471091,
+      "learning_rate": 6.783107663311566e-06,
+      "loss": 0.4428,
+      "step": 43
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.4711187506365836,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.3863,
+      "step": 44
+    },
+    {
+      "epoch": 0.8035714285714286,
+      "grad_norm": 0.457670345971296,
+      "learning_rate": 6.518383725548074e-06,
+      "loss": 0.3592,
+      "step": 45
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 0.5047833734012847,
+      "learning_rate": 6.384177557124247e-06,
+      "loss": 0.44,
+      "step": 46
+    },
+    {
+      "epoch": 0.8392857142857143,
+      "grad_norm": 0.5943584851168806,
+      "learning_rate": 6.248882390836135e-06,
+      "loss": 0.4012,
+      "step": 47
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.6166445509506383,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.369,
+      "step": 48
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.5685679275581732,
+      "learning_rate": 5.975451610080643e-06,
+      "loss": 0.4084,
+      "step": 49
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 0.950425254537538,
+      "learning_rate": 5.837531116523683e-06,
+      "loss": 0.37,
+      "step": 50
+    },
+    {
+      "epoch": 0.9107142857142857,
+      "grad_norm": 0.9078747142445511,
+      "learning_rate": 5.698951697677498e-06,
+      "loss": 0.4036,
+      "step": 51
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.5303910886714994,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 0.367,
+      "step": 52
+    },
+    {
+      "epoch": 0.9464285714285714,
+      "grad_norm": 0.5049393170633624,
+      "learning_rate": 5.420252624646238e-06,
+      "loss": 0.4085,
+      "step": 53
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 0.7767996696471057,
+      "learning_rate": 5.2803522361859596e-06,
+      "loss": 0.4651,
+      "step": 54
+    },
+    {
+      "epoch": 0.9821428571428571,
+      "grad_norm": 0.4397977349242177,
+      "learning_rate": 5.140231281379345e-06,
+      "loss": 0.395,
+      "step": 55
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7956588101595404,
+      "learning_rate": 5e-06,
+      "loss": 0.4398,
+      "step": 56
+    },
+    {
+      "epoch": 1.0178571428571428,
+      "grad_norm": 0.45928806489610763,
+      "learning_rate": 4.859768718620656e-06,
+      "loss": 0.3122,
+      "step": 57
+    },
+    {
+      "epoch": 1.0357142857142858,
+      "grad_norm": 0.827318205786827,
+      "learning_rate": 4.719647763814041e-06,
+      "loss": 0.4237,
+      "step": 58
+    },
+    {
+      "epoch": 1.0535714285714286,
+      "grad_norm": 0.4705518513429398,
+      "learning_rate": 4.579747375353763e-06,
+      "loss": 0.4167,
+      "step": 59
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.44935873798777193,
+      "learning_rate": 4.4401776194834615e-06,
+      "loss": 0.381,
+      "step": 60
+    },
+    {
+      "epoch": 1.0892857142857142,
+      "grad_norm": 1.0790511738287123,
+      "learning_rate": 4.3010483023225045e-06,
+      "loss": 0.4389,
+      "step": 61
+    },
+    {
+      "epoch": 1.1071428571428572,
+      "grad_norm": 0.5167341420269965,
+      "learning_rate": 4.162468883476319e-06,
+      "loss": 0.4297,
+      "step": 62
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.4751425972361663,
+      "learning_rate": 4.02454838991936e-06,
+      "loss": 0.3709,
+      "step": 63
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.46933112747456135,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.3859,
+      "step": 64
+    },
+    {
+      "epoch": 1.1607142857142858,
+      "grad_norm": 0.4535370691471418,
+      "learning_rate": 3.751117609163865e-06,
+      "loss": 0.3352,
+      "step": 65
+    },
+    {
+      "epoch": 1.1785714285714286,
+      "grad_norm": 0.3951477368192192,
+      "learning_rate": 3.6158224428757538e-06,
+      "loss": 0.3092,
+      "step": 66
+    },
+    {
+      "epoch": 1.1964285714285714,
+      "grad_norm": 0.7886622825309841,
+      "learning_rate": 3.4816162744519266e-06,
+      "loss": 0.5043,
+      "step": 67
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.6035836233234189,
+      "learning_rate": 3.3486046902241663e-06,
+      "loss": 0.3273,
+      "step": 68
+    },
+    {
+      "epoch": 1.2321428571428572,
+      "grad_norm": 0.4453444432493159,
+      "learning_rate": 3.216892336688435e-06,
+      "loss": 0.3542,
+      "step": 69
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4616692143123522,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.3339,
+      "step": 70
+    },
+    {
+      "epoch": 1.2678571428571428,
+      "grad_norm": 0.6212474624908485,
+      "learning_rate": 2.95777871532002e-06,
+      "loss": 0.2963,
+      "step": 71
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.4998040866415646,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.413,
+      "step": 72
+    },
+    {
+      "epoch": 1.3035714285714286,
+      "grad_norm": 0.49877033009506233,
+      "learning_rate": 2.705090677662311e-06,
+      "loss": 0.3826,
+      "step": 73
+    },
+    {
+      "epoch": 1.3214285714285714,
+      "grad_norm": 0.5053625656331764,
+      "learning_rate": 2.5814055644738013e-06,
+      "loss": 0.3197,
+      "step": 74
+    },
+    {
+      "epoch": 1.3392857142857144,
+      "grad_norm": 0.6618000990443785,
+      "learning_rate": 2.4596232737673544e-06,
+      "loss": 0.2913,
+      "step": 75
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.5246995960203857,
+      "learning_rate": 2.339839617423318e-06,
+      "loss": 0.3952,
+      "step": 76
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.9440702443020683,
+      "learning_rate": 2.2221488349019903e-06,
+      "loss": 0.3802,
+      "step": 77
+    },
+    {
+      "epoch": 1.3928571428571428,
+      "grad_norm": 0.4294098434698266,
+      "learning_rate": 2.1066435191009717e-06,
+      "loss": 0.3438,
+      "step": 78
+    },
+    {
+      "epoch": 1.4107142857142856,
+      "grad_norm": 0.4622956798352517,
+      "learning_rate": 1.9934145435079705e-06,
+      "loss": 0.3833,
+      "step": 79
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.45317159382122363,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.2786,
+      "step": 80
+    },
+    {
+      "epoch": 1.4464285714285714,
+      "grad_norm": 0.5753084317133276,
+      "learning_rate": 1.7741400822895633e-06,
+      "loss": 0.3328,
+      "step": 81
+    },
+    {
+      "epoch": 1.4642857142857144,
+      "grad_norm": 0.4504291255760464,
+      "learning_rate": 1.6682671102399806e-06,
+      "loss": 0.3151,
+      "step": 82
+    },
+    {
+      "epoch": 1.4821428571428572,
+      "grad_norm": 0.4975145319469907,
+      "learning_rate": 1.5650153698254916e-06,
+      "loss": 0.3865,
+      "step": 83
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4765892756109124,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.3846,
+      "step": 84
+    },
+    {
+      "epoch": 1.5178571428571428,
+      "grad_norm": 0.4373909001618784,
+      "learning_rate": 1.3666983898298659e-06,
+      "loss": 0.3494,
+      "step": 85
+    },
+    {
+      "epoch": 1.5357142857142856,
+      "grad_norm": 0.4529685208275607,
+      "learning_rate": 1.2717891755841722e-06,
+      "loss": 0.372,
+      "step": 86
+    },
+    {
+      "epoch": 1.5535714285714286,
+      "grad_norm": 0.4273734734099652,
+      "learning_rate": 1.1798131208919628e-06,
+      "loss": 0.3104,
+      "step": 87
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.4532632023869939,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.3788,
+      "step": 88
+    },
+    {
+      "epoch": 1.5892857142857144,
+      "grad_norm": 0.43287416507346416,
+      "learning_rate": 1.004947573208756e-06,
+      "loss": 0.3517,
+      "step": 89
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 0.4625067047681074,
+      "learning_rate": 9.221956552036992e-07,
+      "loss": 0.3701,
+      "step": 90
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.6769722699829624,
+      "learning_rate": 8.426519384872733e-07,
+      "loss": 0.3948,
+      "step": 91
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.4296113071894928,
+      "learning_rate": 7.663790038585794e-07,
+      "loss": 0.3536,
+      "step": 92
+    },
+    {
+      "epoch": 1.6607142857142856,
+      "grad_norm": 0.46596528542425175,
+      "learning_rate": 6.934368588379553e-07,
+      "loss": 0.3571,
+      "step": 93
+    },
+    {
+      "epoch": 1.6785714285714286,
+      "grad_norm": 0.4729100333646761,
+      "learning_rate": 6.238828904562316e-07,
+      "loss": 0.4095,
+      "step": 94
+    },
+    {
+      "epoch": 1.6964285714285714,
+      "grad_norm": 0.5079122264353754,
+      "learning_rate": 5.577718201056392e-07,
+      "loss": 0.4505,
+      "step": 95
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.4430532987642348,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.3527,
+      "step": 96
+    },
+    {
+      "epoch": 1.7321428571428572,
+      "grad_norm": 0.45397870266443635,
+      "learning_rate": 4.3608367469340553e-07,
+      "loss": 0.3811,
+      "step": 97
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.46268426701682147,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.3452,
+      "step": 98
+    },
+    {
+      "epoch": 1.7678571428571428,
+      "grad_norm": 0.48417556947646817,
+      "learning_rate": 3.287552985270015e-07,
+      "loss": 0.3598,
+      "step": 99
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.5123072804560189,
+      "learning_rate": 2.8058334845816214e-07,
+      "loss": 0.4311,
+      "step": 100
+    },
+    {
+      "epoch": 1.8035714285714286,
+      "grad_norm": 0.4834267967160676,
+      "learning_rate": 2.3612438638551837e-07,
+      "loss": 0.4095,
+      "step": 101
+    },
+    {
+      "epoch": 1.8214285714285714,
+      "grad_norm": 0.7562919146465696,
+      "learning_rate": 1.9541339027450256e-07,
+      "loss": 0.3693,
+      "step": 102
+    },
+    {
+      "epoch": 1.8392857142857144,
+      "grad_norm": 0.4719509441422737,
+      "learning_rate": 1.5848238938869332e-07,
+      "loss": 0.3714,
+      "step": 103
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.5016599958234566,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.3524,
+      "step": 104
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.4943834992783747,
+      "learning_rate": 9.607359798384785e-08,
+      "loss": 0.4088,
+      "step": 105
+    },
+    {
+      "epoch": 1.8928571428571428,
+      "grad_norm": 0.40867879752100184,
+      "learning_rate": 7.064490740882057e-08,
+      "loss": 0.2922,
+      "step": 106
+    },
+    {
+      "epoch": 1.9107142857142856,
+      "grad_norm": 0.4800434565787566,
+      "learning_rate": 4.909437331777178e-08,
+      "loss": 0.3933,
+      "step": 107
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.5483442024534406,
+      "learning_rate": 3.143895053378698e-08,
+      "loss": 0.4854,
+      "step": 108
+    },
+    {
+      "epoch": 1.9464285714285714,
+      "grad_norm": 0.6164063051281681,
+      "learning_rate": 1.769252941190458e-08,
+      "loss": 0.389,
+      "step": 109
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 0.4275474287127897,
+      "learning_rate": 7.865924910916977e-09,
+      "loss": 0.3559,
+      "step": 110
+    },
+    {
+      "epoch": 1.9821428571428572,
+      "grad_norm": 0.5161206468414776,
+      "learning_rate": 1.9668680847356735e-09,
+      "loss": 0.3549,
+      "step": 111
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7252772591275618,
+      "learning_rate": 0.0,
+      "loss": 0.4064,
+      "step": 112
+    },
+    {
+      "epoch": 2.0,
+      "step": 112,
+      "total_flos": 3190623830016.0,
+      "train_loss": 0.43240329596613136,
+      "train_runtime": 137.9931,
+      "train_samples_per_second": 3.247,
+      "train_steps_per_second": 0.812
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 112,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 70000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3190623830016.0,
+  "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:c8e2c0ab3c23881d10317b6f2dd6c5004ac0794c7c0620f65ae175c4d0aaa1f4
+size 7224