Training in progress, step 500

.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 +1,70 @@
+---
+library_name: peft
+license: apache-2.0
+base_model: Qwen/Qwen2.5-7B
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: full_review
+  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. -->
+
+# full_review
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on the openreview_full_review dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.6107
+
+## 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: 5e-05
+- train_batch_size: 8
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 32
+- 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
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 3.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 1.6658        | 0.5089 | 600  | 1.6607          |
+| 1.5992        | 1.0178 | 1200 | 1.6405          |
+| 1.6182        | 1.5267 | 1800 | 1.6241          |
+| 1.5463        | 2.0356 | 2400 | 1.6182          |
+| 1.5356        | 2.5445 | 3000 | 1.6117          |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.46.1
+- Pytorch 2.5.1
+- Datasets 3.1.0
+- Tokenizers 0.20.3

adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "k_proj",
+    "down_proj",
+    "o_proj",
+    "gate_proj",
+    "up_proj",
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": true
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f509c157b1a48dbcfe3c6436b6488702ec21e641507c405cbe8ef94110a6dc36
+size 645976488

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 +1,13 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 1.6106535196304321,
+    "eval_runtime": 8.342,
+    "eval_samples_per_second": 45.672,
+    "eval_steps_per_second": 11.508,
+    "perplexity": 5.006081731731173,
+    "total_flos": 1.0283028491985224e+19,
+    "train_loss": 1.6124080855326697,
+    "train_runtime": 7405.1369,
+    "train_samples_per_second": 15.277,
+    "train_steps_per_second": 0.478
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 1.6106535196304321,
+    "eval_runtime": 8.342,
+    "eval_samples_per_second": 45.672,
+    "eval_steps_per_second": 11.508,
+    "perplexity": 5.006081731731173
+}

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 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 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": 1.0283028491985224e+19,
+    "train_loss": 1.6124080855326697,
+    "train_runtime": 7405.1369,
+    "train_samples_per_second": 15.277,
+    "train_steps_per_second": 0.478
+}

trainer_log.jsonl

@@ -0,0 +1,50 @@
+{"current_steps": 10, "total_steps": 1770, "loss": 2.0144, "lr": 2.824858757062147e-06, "epoch": 0.01694915254237288, "percentage": 0.56, "elapsed_time": "0:00:21", "remaining_time": "1:02:57"}
+{"current_steps": 20, "total_steps": 1770, "loss": 1.9709, "lr": 5.649717514124294e-06, "epoch": 0.03389830508474576, "percentage": 1.13, "elapsed_time": "0:00:41", "remaining_time": "1:01:01"}
+{"current_steps": 30, "total_steps": 1770, "loss": 1.8976, "lr": 8.47457627118644e-06, "epoch": 0.05084745762711865, "percentage": 1.69, "elapsed_time": "0:01:02", "remaining_time": "1:00:10"}
+{"current_steps": 40, "total_steps": 1770, "loss": 1.832, "lr": 1.1299435028248587e-05, "epoch": 0.06779661016949153, "percentage": 2.26, "elapsed_time": "0:01:22", "remaining_time": "0:59:36"}
+{"current_steps": 50, "total_steps": 1770, "loss": 1.7838, "lr": 1.4124293785310736e-05, "epoch": 0.0847457627118644, "percentage": 2.82, "elapsed_time": "0:01:43", "remaining_time": "0:59:08"}
+{"current_steps": 60, "total_steps": 1770, "loss": 1.7742, "lr": 1.694915254237288e-05, "epoch": 0.1016949152542373, "percentage": 3.39, "elapsed_time": "0:02:03", "remaining_time": "0:58:41"}
+{"current_steps": 70, "total_steps": 1770, "loss": 1.7601, "lr": 1.977401129943503e-05, "epoch": 0.11864406779661017, "percentage": 3.95, "elapsed_time": "0:02:23", "remaining_time": "0:58:15"}
+{"current_steps": 80, "total_steps": 1770, "loss": 1.7553, "lr": 2.2598870056497175e-05, "epoch": 0.13559322033898305, "percentage": 4.52, "elapsed_time": "0:02:44", "remaining_time": "0:57:51"}
+{"current_steps": 90, "total_steps": 1770, "loss": 1.7318, "lr": 2.5423728813559322e-05, "epoch": 0.15254237288135594, "percentage": 5.08, "elapsed_time": "0:03:04", "remaining_time": "0:57:28"}
+{"current_steps": 100, "total_steps": 1770, "loss": 1.7234, "lr": 2.8248587570621472e-05, "epoch": 0.1694915254237288, "percentage": 5.65, "elapsed_time": "0:03:25", "remaining_time": "0:57:05"}
+{"current_steps": 110, "total_steps": 1770, "loss": 1.731, "lr": 3.107344632768362e-05, "epoch": 0.1864406779661017, "percentage": 6.21, "elapsed_time": "0:03:45", "remaining_time": "0:56:43"}
+{"current_steps": 120, "total_steps": 1770, "loss": 1.7048, "lr": 3.389830508474576e-05, "epoch": 0.2033898305084746, "percentage": 6.78, "elapsed_time": "0:04:05", "remaining_time": "0:56:21"}
+{"current_steps": 130, "total_steps": 1770, "loss": 1.7091, "lr": 3.672316384180791e-05, "epoch": 0.22033898305084745, "percentage": 7.34, "elapsed_time": "0:04:26", "remaining_time": "0:55:59"}
+{"current_steps": 140, "total_steps": 1770, "loss": 1.7163, "lr": 3.954802259887006e-05, "epoch": 0.23728813559322035, "percentage": 7.91, "elapsed_time": "0:04:46", "remaining_time": "0:55:37"}
+{"current_steps": 150, "total_steps": 1770, "loss": 1.7085, "lr": 4.2372881355932206e-05, "epoch": 0.2542372881355932, "percentage": 8.47, "elapsed_time": "0:05:06", "remaining_time": "0:55:15"}
+{"current_steps": 160, "total_steps": 1770, "loss": 1.6966, "lr": 4.519774011299435e-05, "epoch": 0.2711864406779661, "percentage": 9.04, "elapsed_time": "0:05:27", "remaining_time": "0:54:53"}
+{"current_steps": 170, "total_steps": 1770, "loss": 1.7086, "lr": 4.80225988700565e-05, "epoch": 0.288135593220339, "percentage": 9.6, "elapsed_time": "0:05:47", "remaining_time": "0:54:32"}
+{"current_steps": 180, "total_steps": 1770, "loss": 1.7086, "lr": 4.999956245830044e-05, "epoch": 0.3050847457627119, "percentage": 10.17, "elapsed_time": "0:06:08", "remaining_time": "0:54:11"}
+{"current_steps": 190, "total_steps": 1770, "loss": 1.6885, "lr": 4.9991784365248394e-05, "epoch": 0.3220338983050847, "percentage": 10.73, "elapsed_time": "0:06:28", "remaining_time": "0:53:50"}
+{"current_steps": 200, "total_steps": 1770, "loss": 1.705, "lr": 4.997428660526443e-05, "epoch": 0.3389830508474576, "percentage": 11.3, "elapsed_time": "0:06:48", "remaining_time": "0:53:29"}
+{"current_steps": 210, "total_steps": 1770, "loss": 1.6889, "lr": 4.994707598348085e-05, "epoch": 0.3559322033898305, "percentage": 11.86, "elapsed_time": "0:07:09", "remaining_time": "0:53:08"}
+{"current_steps": 220, "total_steps": 1770, "loss": 1.6939, "lr": 4.991016308250258e-05, "epoch": 0.3728813559322034, "percentage": 12.43, "elapsed_time": "0:07:29", "remaining_time": "0:52:47"}
+{"current_steps": 230, "total_steps": 1770, "loss": 1.699, "lr": 4.9863562258291494e-05, "epoch": 0.3898305084745763, "percentage": 12.99, "elapsed_time": "0:07:49", "remaining_time": "0:52:26"}
+{"current_steps": 240, "total_steps": 1770, "loss": 1.6969, "lr": 4.980729163458312e-05, "epoch": 0.4067796610169492, "percentage": 13.56, "elapsed_time": "0:08:10", "remaining_time": "0:52:05"}
+{"current_steps": 250, "total_steps": 1770, "loss": 1.6931, "lr": 4.974137309583804e-05, "epoch": 0.423728813559322, "percentage": 14.12, "elapsed_time": "0:08:30", "remaining_time": "0:51:44"}
+{"current_steps": 260, "total_steps": 1770, "loss": 1.7061, "lr": 4.966583227873079e-05, "epoch": 0.4406779661016949, "percentage": 14.69, "elapsed_time": "0:08:51", "remaining_time": "0:51:24"}
+{"current_steps": 270, "total_steps": 1770, "loss": 1.6956, "lr": 4.9580698562179297e-05, "epoch": 0.4576271186440678, "percentage": 15.25, "elapsed_time": "0:09:11", "remaining_time": "0:51:03"}
+{"current_steps": 280, "total_steps": 1770, "loss": 1.7004, "lr": 4.948600505591905e-05, "epoch": 0.4745762711864407, "percentage": 15.82, "elapsed_time": "0:09:31", "remaining_time": "0:50:42"}
+{"current_steps": 290, "total_steps": 1770, "loss": 1.693, "lr": 4.938178858762622e-05, "epoch": 0.4915254237288136, "percentage": 16.38, "elapsed_time": "0:09:52", "remaining_time": "0:50:21"}
+{"current_steps": 300, "total_steps": 1770, "loss": 1.6819, "lr": 4.926808968859483e-05, "epoch": 0.5084745762711864, "percentage": 16.95, "elapsed_time": "0:10:12", "remaining_time": "0:50:01"}
+{"current_steps": 310, "total_steps": 1770, "loss": 1.6984, "lr": 4.9144952577973596e-05, "epoch": 0.5254237288135594, "percentage": 17.51, "elapsed_time": "0:10:32", "remaining_time": "0:49:40"}
+{"current_steps": 320, "total_steps": 1770, "loss": 1.6875, "lr": 4.90124251455684e-05, "epoch": 0.5423728813559322, "percentage": 18.08, "elapsed_time": "0:10:53", "remaining_time": "0:49:19"}
+{"current_steps": 330, "total_steps": 1770, "loss": 1.6758, "lr": 4.88705589332173e-05, "epoch": 0.559322033898305, "percentage": 18.64, "elapsed_time": "0:11:13", "remaining_time": "0:48:58"}
+{"current_steps": 340, "total_steps": 1770, "loss": 1.695, "lr": 4.871940911474513e-05, "epoch": 0.576271186440678, "percentage": 19.21, "elapsed_time": "0:11:33", "remaining_time": "0:48:38"}
+{"current_steps": 350, "total_steps": 1770, "loss": 1.6862, "lr": 4.8559034474505614e-05, "epoch": 0.5932203389830508, "percentage": 19.77, "elapsed_time": "0:11:54", "remaining_time": "0:48:17"}
+{"current_steps": 360, "total_steps": 1770, "loss": 1.6652, "lr": 4.838949738451929e-05, "epoch": 0.6101694915254238, "percentage": 20.34, "elapsed_time": "0:12:14", "remaining_time": "0:47:56"}
+{"current_steps": 370, "total_steps": 1770, "loss": 1.6898, "lr": 4.821086378021608e-05, "epoch": 0.6271186440677966, "percentage": 20.9, "elapsed_time": "0:12:34", "remaining_time": "0:47:35"}
+{"current_steps": 380, "total_steps": 1770, "loss": 1.6766, "lr": 4.802320313479214e-05, "epoch": 0.6440677966101694, "percentage": 21.47, "elapsed_time": "0:12:55", "remaining_time": "0:47:15"}
+{"current_steps": 390, "total_steps": 1770, "loss": 1.6838, "lr": 4.7826588432190614e-05, "epoch": 0.6610169491525424, "percentage": 22.03, "elapsed_time": "0:13:15", "remaining_time": "0:46:54"}
+{"current_steps": 400, "total_steps": 1770, "loss": 1.6702, "lr": 4.76210961387172e-05, "epoch": 0.6779661016949152, "percentage": 22.6, "elapsed_time": "0:13:35", "remaining_time": "0:46:33"}
+{"current_steps": 410, "total_steps": 1770, "loss": 1.6793, "lr": 4.7406806173301285e-05, "epoch": 0.6949152542372882, "percentage": 23.16, "elapsed_time": "0:13:56", "remaining_time": "0:46:13"}
+{"current_steps": 420, "total_steps": 1770, "loss": 1.6611, "lr": 4.7183801876414294e-05, "epoch": 0.711864406779661, "percentage": 23.73, "elapsed_time": "0:14:16", "remaining_time": "0:45:52"}
+{"current_steps": 430, "total_steps": 1770, "loss": 1.6896, "lr": 4.695216997765737e-05, "epoch": 0.7288135593220338, "percentage": 24.29, "elapsed_time": "0:14:36", "remaining_time": "0:45:32"}
+{"current_steps": 440, "total_steps": 1770, "loss": 1.6756, "lr": 4.6712000562031e-05, "epoch": 0.7457627118644068, "percentage": 24.86, "elapsed_time": "0:14:57", "remaining_time": "0:45:11"}
+{"current_steps": 450, "total_steps": 1770, "loss": 1.6658, "lr": 4.6463387034899645e-05, "epoch": 0.7627118644067796, "percentage": 25.42, "elapsed_time": "0:15:17", "remaining_time": "0:44:51"}
+{"current_steps": 460, "total_steps": 1770, "loss": 1.6841, "lr": 4.6206426085665046e-05, "epoch": 0.7796610169491526, "percentage": 25.99, "elapsed_time": "0:15:37", "remaining_time": "0:44:30"}
+{"current_steps": 470, "total_steps": 1770, "loss": 1.6947, "lr": 4.594121765016229e-05, "epoch": 0.7966101694915254, "percentage": 26.55, "elapsed_time": "0:15:58", "remaining_time": "0:44:10"}
+{"current_steps": 480, "total_steps": 1770, "loss": 1.6888, "lr": 4.5667864871793345e-05, "epoch": 0.8135593220338984, "percentage": 27.12, "elapsed_time": "0:16:18", "remaining_time": "0:43:49"}
+{"current_steps": 490, "total_steps": 1770, "loss": 1.6675, "lr": 4.538647406141308e-05, "epoch": 0.8305084745762712, "percentage": 27.68, "elapsed_time": "0:16:38", "remaining_time": "0:43:29"}
+{"current_steps": 500, "total_steps": 1770, "loss": 1.6828, "lr": 4.509715465598344e-05, "epoch": 0.847457627118644, "percentage": 28.25, "elapsed_time": "0:16:59", "remaining_time": "0:43:08"}

trainer_state.json

@@ -0,0 +1,2553 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 600,
+  "global_step": 3537,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008481764206955046,
+      "grad_norm": 0.9741957783699036,
+      "learning_rate": 1.4124293785310734e-06,
+      "loss": 2.0351,
+      "step": 10
+    },
+    {
+      "epoch": 0.016963528413910092,
+      "grad_norm": 0.6763953566551208,
+      "learning_rate": 2.824858757062147e-06,
+      "loss": 1.9766,
+      "step": 20
+    },
+    {
+      "epoch": 0.02544529262086514,
+      "grad_norm": 0.354457825422287,
+      "learning_rate": 4.23728813559322e-06,
+      "loss": 1.951,
+      "step": 30
+    },
+    {
+      "epoch": 0.033927056827820185,
+      "grad_norm": 0.31893497705459595,
+      "learning_rate": 5.649717514124294e-06,
+      "loss": 1.8941,
+      "step": 40
+    },
+    {
+      "epoch": 0.04240882103477523,
+      "grad_norm": 0.22654543817043304,
+      "learning_rate": 7.062146892655368e-06,
+      "loss": 1.8506,
+      "step": 50
+    },
+    {
+      "epoch": 0.05089058524173028,
+      "grad_norm": 0.2076452225446701,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 1.8052,
+      "step": 60
+    },
+    {
+      "epoch": 0.05937234944868533,
+      "grad_norm": 0.1928279548883438,
+      "learning_rate": 9.887005649717516e-06,
+      "loss": 1.8038,
+      "step": 70
+    },
+    {
+      "epoch": 0.06785411365564037,
+      "grad_norm": 0.18840375542640686,
+      "learning_rate": 1.1299435028248587e-05,
+      "loss": 1.7649,
+      "step": 80
+    },
+    {
+      "epoch": 0.07633587786259542,
+      "grad_norm": 0.19931818544864655,
+      "learning_rate": 1.2711864406779661e-05,
+      "loss": 1.7458,
+      "step": 90
+    },
+    {
+      "epoch": 0.08481764206955046,
+      "grad_norm": 0.20297980308532715,
+      "learning_rate": 1.4124293785310736e-05,
+      "loss": 1.75,
+      "step": 100
+    },
+    {
+      "epoch": 0.09329940627650551,
+      "grad_norm": 0.18958595395088196,
+      "learning_rate": 1.553672316384181e-05,
+      "loss": 1.7462,
+      "step": 110
+    },
+    {
+      "epoch": 0.10178117048346055,
+      "grad_norm": 0.20559008419513702,
+      "learning_rate": 1.694915254237288e-05,
+      "loss": 1.7414,
+      "step": 120
+    },
+    {
+      "epoch": 0.1102629346904156,
+      "grad_norm": 0.20397105813026428,
+      "learning_rate": 1.8361581920903956e-05,
+      "loss": 1.7446,
+      "step": 130
+    },
+    {
+      "epoch": 0.11874469889737066,
+      "grad_norm": 0.20660234987735748,
+      "learning_rate": 1.977401129943503e-05,
+      "loss": 1.7285,
+      "step": 140
+    },
+    {
+      "epoch": 0.1272264631043257,
+      "grad_norm": 0.2024795413017273,
+      "learning_rate": 2.1186440677966103e-05,
+      "loss": 1.7406,
+      "step": 150
+    },
+    {
+      "epoch": 0.13570822731128074,
+      "grad_norm": 0.20553280413150787,
+      "learning_rate": 2.2598870056497175e-05,
+      "loss": 1.7338,
+      "step": 160
+    },
+    {
+      "epoch": 0.1441899915182358,
+      "grad_norm": 0.2002541571855545,
+      "learning_rate": 2.401129943502825e-05,
+      "loss": 1.7048,
+      "step": 170
+    },
+    {
+      "epoch": 0.15267175572519084,
+      "grad_norm": 0.19995799660682678,
+      "learning_rate": 2.5423728813559322e-05,
+      "loss": 1.729,
+      "step": 180
+    },
+    {
+      "epoch": 0.16115351993214588,
+      "grad_norm": 0.19061961770057678,
+      "learning_rate": 2.68361581920904e-05,
+      "loss": 1.7157,
+      "step": 190
+    },
+    {
+      "epoch": 0.16963528413910092,
+      "grad_norm": 0.20361429452896118,
+      "learning_rate": 2.8248587570621472e-05,
+      "loss": 1.7065,
+      "step": 200
+    },
+    {
+      "epoch": 0.178117048346056,
+      "grad_norm": 0.19405007362365723,
+      "learning_rate": 2.9661016949152544e-05,
+      "loss": 1.7147,
+      "step": 210
+    },
+    {
+      "epoch": 0.18659881255301103,
+      "grad_norm": 0.21106290817260742,
+      "learning_rate": 3.107344632768362e-05,
+      "loss": 1.7245,
+      "step": 220
+    },
+    {
+      "epoch": 0.19508057675996607,
+      "grad_norm": 0.19132505357265472,
+      "learning_rate": 3.2485875706214694e-05,
+      "loss": 1.7045,
+      "step": 230
+    },
+    {
+      "epoch": 0.2035623409669211,
+      "grad_norm": 0.19091880321502686,
+      "learning_rate": 3.389830508474576e-05,
+      "loss": 1.6859,
+      "step": 240
+    },
+    {
+      "epoch": 0.21204410517387617,
+      "grad_norm": 0.19968806207180023,
+      "learning_rate": 3.531073446327684e-05,
+      "loss": 1.7053,
+      "step": 250
+    },
+    {
+      "epoch": 0.2205258693808312,
+      "grad_norm": 0.18798886239528656,
+      "learning_rate": 3.672316384180791e-05,
+      "loss": 1.6977,
+      "step": 260
+    },
+    {
+      "epoch": 0.22900763358778625,
+      "grad_norm": 0.18123075366020203,
+      "learning_rate": 3.813559322033898e-05,
+      "loss": 1.7306,
+      "step": 270
+    },
+    {
+      "epoch": 0.23748939779474132,
+      "grad_norm": 0.18372726440429688,
+      "learning_rate": 3.954802259887006e-05,
+      "loss": 1.688,
+      "step": 280
+    },
+    {
+      "epoch": 0.24597116200169636,
+      "grad_norm": 0.18568727374076843,
+      "learning_rate": 4.096045197740113e-05,
+      "loss": 1.6959,
+      "step": 290
+    },
+    {
+      "epoch": 0.2544529262086514,
+      "grad_norm": 0.18453723192214966,
+      "learning_rate": 4.2372881355932206e-05,
+      "loss": 1.7084,
+      "step": 300
+    },
+    {
+      "epoch": 0.26293469041560646,
+      "grad_norm": 0.1792721152305603,
+      "learning_rate": 4.378531073446328e-05,
+      "loss": 1.6758,
+      "step": 310
+    },
+    {
+      "epoch": 0.2714164546225615,
+      "grad_norm": 0.17322257161140442,
+      "learning_rate": 4.519774011299435e-05,
+      "loss": 1.7053,
+      "step": 320
+    },
+    {
+      "epoch": 0.27989821882951654,
+      "grad_norm": 0.18389643728733063,
+      "learning_rate": 4.6610169491525425e-05,
+      "loss": 1.7106,
+      "step": 330
+    },
+    {
+      "epoch": 0.2883799830364716,
+      "grad_norm": 0.18507392704486847,
+      "learning_rate": 4.80225988700565e-05,
+      "loss": 1.6952,
+      "step": 340
+    },
+    {
+      "epoch": 0.2968617472434266,
+      "grad_norm": 0.16895641386508942,
+      "learning_rate": 4.9435028248587575e-05,
+      "loss": 1.7141,
+      "step": 350
+    },
+    {
+      "epoch": 0.3053435114503817,
+      "grad_norm": 0.1711965948343277,
+      "learning_rate": 4.999956163314189e-05,
+      "loss": 1.6916,
+      "step": 360
+    },
+    {
+      "epoch": 0.3138252756573367,
+      "grad_norm": 0.1739186942577362,
+      "learning_rate": 4.9996882780236966e-05,
+      "loss": 1.6689,
+      "step": 370
+    },
+    {
+      "epoch": 0.32230703986429177,
+      "grad_norm": 0.17199921607971191,
+      "learning_rate": 4.99917688722129e-05,
+      "loss": 1.6957,
+      "step": 380
+    },
+    {
+      "epoch": 0.33078880407124683,
+      "grad_norm": 0.1741172820329666,
+      "learning_rate": 4.9984220407237743e-05,
+      "loss": 1.6925,
+      "step": 390
+    },
+    {
+      "epoch": 0.33927056827820185,
+      "grad_norm": 0.16634291410446167,
+      "learning_rate": 4.997423812064029e-05,
+      "loss": 1.7053,
+      "step": 400
+    },
+    {
+      "epoch": 0.3477523324851569,
+      "grad_norm": 0.16902585327625275,
+      "learning_rate": 4.996182298483854e-05,
+      "loss": 1.6835,
+      "step": 410
+    },
+    {
+      "epoch": 0.356234096692112,
+      "grad_norm": 0.16757601499557495,
+      "learning_rate": 4.9946976209244876e-05,
+      "loss": 1.6833,
+      "step": 420
+    },
+    {
+      "epoch": 0.364715860899067,
+      "grad_norm": 0.1582445502281189,
+      "learning_rate": 4.992969924014831e-05,
+      "loss": 1.6965,
+      "step": 430
+    },
+    {
+      "epoch": 0.37319762510602206,
+      "grad_norm": 0.1670326590538025,
+      "learning_rate": 4.990999376057358e-05,
+      "loss": 1.6797,
+      "step": 440
+    },
+    {
+      "epoch": 0.3816793893129771,
+      "grad_norm": 0.16612491011619568,
+      "learning_rate": 4.98878616901172e-05,
+      "loss": 1.7047,
+      "step": 450
+    },
+    {
+      "epoch": 0.39016115351993214,
+      "grad_norm": 0.16573746502399445,
+      "learning_rate": 4.986330518476046e-05,
+      "loss": 1.6825,
+      "step": 460
+    },
+    {
+      "epoch": 0.3986429177268872,
+      "grad_norm": 0.1650175005197525,
+      "learning_rate": 4.983632663665937e-05,
+      "loss": 1.6983,
+      "step": 470
+    },
+    {
+      "epoch": 0.4071246819338422,
+      "grad_norm": 0.16499648988246918,
+      "learning_rate": 4.980692867391172e-05,
+      "loss": 1.6838,
+      "step": 480
+    },
+    {
+      "epoch": 0.4156064461407973,
+      "grad_norm": 0.15969328582286835,
+      "learning_rate": 4.9775114160300974e-05,
+      "loss": 1.6812,
+      "step": 490
+    },
+    {
+      "epoch": 0.42408821034775235,
+      "grad_norm": 0.15962941944599152,
+      "learning_rate": 4.974088619501733e-05,
+      "loss": 1.6934,
+      "step": 500
+    },
+    {
+      "epoch": 0.43256997455470736,
+      "grad_norm": 0.1593630164861679,
+      "learning_rate": 4.970424811235582e-05,
+      "loss": 1.7084,
+      "step": 510
+    },
+    {
+      "epoch": 0.4410517387616624,
+      "grad_norm": 0.1627817004919052,
+      "learning_rate": 4.966520348139149e-05,
+      "loss": 1.6919,
+      "step": 520
+    },
+    {
+      "epoch": 0.4495335029686175,
+      "grad_norm": 0.16287963092327118,
+      "learning_rate": 4.9623756105631755e-05,
+      "loss": 1.6909,
+      "step": 530
+    },
+    {
+      "epoch": 0.4580152671755725,
+      "grad_norm": 0.15781639516353607,
+      "learning_rate": 4.9579910022645816e-05,
+      "loss": 1.6863,
+      "step": 540
+    },
+    {
+      "epoch": 0.46649703138252757,
+      "grad_norm": 0.1625601351261139,
+      "learning_rate": 4.9533669503671417e-05,
+      "loss": 1.6877,
+      "step": 550
+    },
+    {
+      "epoch": 0.47497879558948264,
+      "grad_norm": 0.1634383201599121,
+      "learning_rate": 4.94850390531987e-05,
+      "loss": 1.6949,
+      "step": 560
+    },
+    {
+      "epoch": 0.48346055979643765,
+      "grad_norm": 0.1620015949010849,
+      "learning_rate": 4.943402340853146e-05,
+      "loss": 1.6817,
+      "step": 570
+    },
+    {
+      "epoch": 0.4919423240033927,
+      "grad_norm": 0.16449208557605743,
+      "learning_rate": 4.938062753932562e-05,
+      "loss": 1.6908,
+      "step": 580
+    },
+    {
+      "epoch": 0.5004240882103478,
+      "grad_norm": 0.1599043905735016,
+      "learning_rate": 4.9324856647105144e-05,
+      "loss": 1.685,
+      "step": 590
+    },
+    {
+      "epoch": 0.5089058524173028,
+      "grad_norm": 0.1632467657327652,
+      "learning_rate": 4.9266716164755305e-05,
+      "loss": 1.6658,
+      "step": 600
+    },
+    {
+      "epoch": 0.5089058524173028,
+      "eval_loss": 1.6607000827789307,
+      "eval_runtime": 8.4197,
+      "eval_samples_per_second": 45.251,
+      "eval_steps_per_second": 11.402,
+      "step": 600
+    },
+    {
+      "epoch": 0.5173876166242578,
+      "grad_norm": 0.1606842428445816,
+      "learning_rate": 4.920621175599348e-05,
+      "loss": 1.6953,
+      "step": 610
+    },
+    {
+      "epoch": 0.5258693808312129,
+      "grad_norm": 0.16536976397037506,
+      "learning_rate": 4.9143349314817394e-05,
+      "loss": 1.6882,
+      "step": 620
+    },
+    {
+      "epoch": 0.5343511450381679,
+      "grad_norm": 0.16298532485961914,
+      "learning_rate": 4.9078134964930975e-05,
+      "loss": 1.696,
+      "step": 630
+    },
+    {
+      "epoch": 0.542832909245123,
+      "grad_norm": 0.15745271742343903,
+      "learning_rate": 4.90105750591478e-05,
+      "loss": 1.6657,
+      "step": 640
+    },
+    {
+      "epoch": 0.5513146734520781,
+      "grad_norm": 0.1618964523077011,
+      "learning_rate": 4.894067617877227e-05,
+      "loss": 1.6583,
+      "step": 650
+    },
+    {
+      "epoch": 0.5597964376590331,
+      "grad_norm": 0.17704178392887115,
+      "learning_rate": 4.886844513295846e-05,
+      "loss": 1.6808,
+      "step": 660
+    },
+    {
+      "epoch": 0.5682782018659881,
+      "grad_norm": 0.1587689071893692,
+      "learning_rate": 4.8793888958046855e-05,
+      "loss": 1.6988,
+      "step": 670
+    },
+    {
+      "epoch": 0.5767599660729432,
+      "grad_norm": 0.15959058701992035,
+      "learning_rate": 4.871701491687884e-05,
+      "loss": 1.6782,
+      "step": 680
+    },
+    {
+      "epoch": 0.5852417302798982,
+      "grad_norm": 0.16397462785243988,
+      "learning_rate": 4.863783049808929e-05,
+      "loss": 1.6755,
+      "step": 690
+    },
+    {
+      "epoch": 0.5937234944868532,
+      "grad_norm": 0.15620797872543335,
+      "learning_rate": 4.855634341537697e-05,
+      "loss": 1.6832,
+      "step": 700
+    },
+    {
+      "epoch": 0.6022052586938084,
+      "grad_norm": 0.16528008878231049,
+      "learning_rate": 4.847256160675319e-05,
+      "loss": 1.6524,
+      "step": 710
+    },
+    {
+      "epoch": 0.6106870229007634,
+      "grad_norm": 0.15817534923553467,
+      "learning_rate": 4.8386493233768465e-05,
+      "loss": 1.6641,
+      "step": 720
+    },
+    {
+      "epoch": 0.6191687871077184,
+      "grad_norm": 0.16023534536361694,
+      "learning_rate": 4.829814668071752e-05,
+      "loss": 1.6616,
+      "step": 730
+    },
+    {
+      "epoch": 0.6276505513146734,
+      "grad_norm": 0.15581044554710388,
+      "learning_rate": 4.820753055382253e-05,
+      "loss": 1.7041,
+      "step": 740
+    },
+    {
+      "epoch": 0.6361323155216285,
+      "grad_norm": 0.15969929099082947,
+      "learning_rate": 4.811465368039467e-05,
+      "loss": 1.6773,
+      "step": 750
+    },
+    {
+      "epoch": 0.6446140797285835,
+      "grad_norm": 0.16355305910110474,
+      "learning_rate": 4.801952510797433e-05,
+      "loss": 1.6619,
+      "step": 760
+    },
+    {
+      "epoch": 0.6530958439355385,
+      "grad_norm": 0.1603214144706726,
+      "learning_rate": 4.792215410344968e-05,
+      "loss": 1.6967,
+      "step": 770
+    },
+    {
+      "epoch": 0.6615776081424937,
+      "grad_norm": 0.16001801192760468,
+      "learning_rate": 4.7822550152153946e-05,
+      "loss": 1.657,
+      "step": 780
+    },
+    {
+      "epoch": 0.6700593723494487,
+      "grad_norm": 0.16406036913394928,
+      "learning_rate": 4.7720722956941434e-05,
+      "loss": 1.6645,
+      "step": 790
+    },
+    {
+      "epoch": 0.6785411365564037,
+      "grad_norm": 0.16242410242557526,
+      "learning_rate": 4.7616682437242295e-05,
+      "loss": 1.6618,
+      "step": 800
+    },
+    {
+      "epoch": 0.6870229007633588,
+      "grad_norm": 0.16283248364925385,
+      "learning_rate": 4.751043872809626e-05,
+      "loss": 1.6689,
+      "step": 810
+    },
+    {
+      "epoch": 0.6955046649703138,
+      "grad_norm": 0.16198423504829407,
+      "learning_rate": 4.740200217916534e-05,
+      "loss": 1.6751,
+      "step": 820
+    },
+    {
+      "epoch": 0.7039864291772688,
+      "grad_norm": 0.15994125604629517,
+      "learning_rate": 4.729138335372558e-05,
+      "loss": 1.6549,
+      "step": 830
+    },
+    {
+      "epoch": 0.712468193384224,
+      "grad_norm": 0.1621062308549881,
+      "learning_rate": 4.717859302763812e-05,
+      "loss": 1.6528,
+      "step": 840
+    },
+    {
+      "epoch": 0.720949957591179,
+      "grad_norm": 0.16383197903633118,
+      "learning_rate": 4.7063642188299384e-05,
+      "loss": 1.6689,
+      "step": 850
+    },
+    {
+      "epoch": 0.729431721798134,
+      "grad_norm": 0.16420704126358032,
+      "learning_rate": 4.694654203357082e-05,
+      "loss": 1.696,
+      "step": 860
+    },
+    {
+      "epoch": 0.7379134860050891,
+      "grad_norm": 0.16626745462417603,
+      "learning_rate": 4.682730397068803e-05,
+      "loss": 1.6834,
+      "step": 870
+    },
+    {
+      "epoch": 0.7463952502120441,
+      "grad_norm": 0.16311106085777283,
+      "learning_rate": 4.670593961514955e-05,
+      "loss": 1.653,
+      "step": 880
+    },
+    {
+      "epoch": 0.7548770144189991,
+      "grad_norm": 0.15614666044712067,
+      "learning_rate": 4.658246078958534e-05,
+      "loss": 1.6716,
+      "step": 890
+    },
+    {
+      "epoch": 0.7633587786259542,
+      "grad_norm": 0.16057974100112915,
+      "learning_rate": 4.645687952260508e-05,
+      "loss": 1.6456,
+      "step": 900
+    },
+    {
+      "epoch": 0.7718405428329093,
+      "grad_norm": 0.15990586578845978,
+      "learning_rate": 4.6329208047626425e-05,
+      "loss": 1.6783,
+      "step": 910
+    },
+    {
+      "epoch": 0.7803223070398643,
+      "grad_norm": 0.15808311104774475,
+      "learning_rate": 4.619945880168329e-05,
+      "loss": 1.6748,
+      "step": 920
+    },
+    {
+      "epoch": 0.7888040712468194,
+      "grad_norm": 0.16143541038036346,
+      "learning_rate": 4.6067644424214286e-05,
+      "loss": 1.6767,
+      "step": 930
+    },
+    {
+      "epoch": 0.7972858354537744,
+      "grad_norm": 0.16702161729335785,
+      "learning_rate": 4.5933777755831474e-05,
+      "loss": 1.6971,
+      "step": 940
+    },
+    {
+      "epoch": 0.8057675996607294,
+      "grad_norm": 0.15487882494926453,
+      "learning_rate": 4.579787183706952e-05,
+      "loss": 1.6897,
+      "step": 950
+    },
+    {
+      "epoch": 0.8142493638676844,
+      "grad_norm": 0.16486485302448273,
+      "learning_rate": 4.565993990711531e-05,
+      "loss": 1.6722,
+      "step": 960
+    },
+    {
+      "epoch": 0.8227311280746396,
+      "grad_norm": 0.16009177267551422,
+      "learning_rate": 4.551999540251834e-05,
+      "loss": 1.6464,
+      "step": 970
+    },
+    {
+      "epoch": 0.8312128922815946,
+      "grad_norm": 0.1609630137681961,
+      "learning_rate": 4.537805195588172e-05,
+      "loss": 1.6734,
+      "step": 980
+    },
+    {
+      "epoch": 0.8396946564885496,
+      "grad_norm": 0.16222642362117767,
+      "learning_rate": 4.523412339453422e-05,
+      "loss": 1.6824,
+      "step": 990
+    },
+    {
+      "epoch": 0.8481764206955047,
+      "grad_norm": 0.16156989336013794,
+      "learning_rate": 4.508822373918328e-05,
+      "loss": 1.6659,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8566581849024597,
+      "grad_norm": 0.16236943006515503,
+      "learning_rate": 4.494036720254915e-05,
+      "loss": 1.6707,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8651399491094147,
+      "grad_norm": 0.163671612739563,
+      "learning_rate": 4.479056818798044e-05,
+      "loss": 1.6561,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8736217133163698,
+      "grad_norm": 0.16348202526569366,
+      "learning_rate": 4.463884128805097e-05,
+      "loss": 1.6344,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8821034775233249,
+      "grad_norm": 0.16194137930870056,
+      "learning_rate": 4.448520128313826e-05,
+      "loss": 1.6583,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8905852417302799,
+      "grad_norm": 0.16495518386363983,
+      "learning_rate": 4.432966313998371e-05,
+      "loss": 1.6674,
+      "step": 1050
+    },
+    {
+      "epoch": 0.899067005937235,
+      "grad_norm": 0.15918731689453125,
+      "learning_rate": 4.4172242010234643e-05,
+      "loss": 1.6522,
+      "step": 1060
+    },
+    {
+      "epoch": 0.90754877014419,
+      "grad_norm": 0.16680502891540527,
+      "learning_rate": 4.401295322896827e-05,
+      "loss": 1.6563,
+      "step": 1070
+    },
+    {
+      "epoch": 0.916030534351145,
+      "grad_norm": 0.16706348955631256,
+      "learning_rate": 4.385181231319788e-05,
+      "loss": 1.6677,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9245122985581001,
+      "grad_norm": 0.16275151073932648,
+      "learning_rate": 4.368883496036125e-05,
+      "loss": 1.6784,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9329940627650551,
+      "grad_norm": 0.16376911103725433,
+      "learning_rate": 4.3524037046791455e-05,
+      "loss": 1.6565,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9414758269720102,
+      "grad_norm": 0.16600607335567474,
+      "learning_rate": 4.335743462617036e-05,
+      "loss": 1.6514,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9499575911789653,
+      "grad_norm": 0.1660010665655136,
+      "learning_rate": 4.318904392796467e-05,
+      "loss": 1.6546,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9584393553859203,
+      "grad_norm": 0.1595398634672165,
+      "learning_rate": 4.3018881355844996e-05,
+      "loss": 1.6445,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9669211195928753,
+      "grad_norm": 0.16763563454151154,
+      "learning_rate": 4.2846963486087926e-05,
+      "loss": 1.665,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9754028837998303,
+      "grad_norm": 0.16849684715270996,
+      "learning_rate": 4.2673307065961196e-05,
+      "loss": 1.6917,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9838846480067854,
+      "grad_norm": 0.16113999485969543,
+      "learning_rate": 4.249792901209233e-05,
+      "loss": 1.6593,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9923664122137404,
+      "grad_norm": 0.16176581382751465,
+      "learning_rate": 4.2320846408820644e-05,
+      "loss": 1.6515,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0008481764206956,
+      "grad_norm": 0.1654123216867447,
+      "learning_rate": 4.214207650653306e-05,
+      "loss": 1.6132,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0093299406276506,
+      "grad_norm": 0.17195528745651245,
+      "learning_rate": 4.196163671998364e-05,
+      "loss": 1.6052,
+      "step": 1190
+    },
+    {
+      "epoch": 1.0178117048346056,
+      "grad_norm": 0.1740170270204544,
+      "learning_rate": 4.1779544626597125e-05,
+      "loss": 1.5992,
+      "step": 1200
+    },
+    {
+      "epoch": 1.0178117048346056,
+      "eval_loss": 1.6405301094055176,
+      "eval_runtime": 8.3842,
+      "eval_samples_per_second": 45.442,
+      "eval_steps_per_second": 11.45,
+      "step": 1200
+    },
+    {
+      "epoch": 1.0262934690415606,
+      "grad_norm": 0.18201863765716553,
+      "learning_rate": 4.159581796475665e-05,
+      "loss": 1.6427,
+      "step": 1210
+    },
+    {
+      "epoch": 1.0347752332485156,
+      "grad_norm": 0.1829264909029007,
+      "learning_rate": 4.1410474632075794e-05,
+      "loss": 1.6227,
+      "step": 1220
+    },
+    {
+      "epoch": 1.0432569974554706,
+      "grad_norm": 0.17768743634223938,
+      "learning_rate": 4.122353268365507e-05,
+      "loss": 1.6279,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0517387616624259,
+      "grad_norm": 0.18258538842201233,
+      "learning_rate": 4.1035010330323086e-05,
+      "loss": 1.6102,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0602205258693809,
+      "grad_norm": 0.1885634958744049,
+      "learning_rate": 4.084492593686261e-05,
+      "loss": 1.6443,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0687022900763359,
+      "grad_norm": 0.18394209444522858,
+      "learning_rate": 4.0653298020221496e-05,
+      "loss": 1.6221,
+      "step": 1260
+    },
+    {
+      "epoch": 1.077184054283291,
+      "grad_norm": 0.18219374120235443,
+      "learning_rate": 4.046014524770893e-05,
+      "loss": 1.635,
+      "step": 1270
+    },
+    {
+      "epoch": 1.085665818490246,
+      "grad_norm": 0.18404074013233185,
+      "learning_rate": 4.026548643517694e-05,
+      "loss": 1.6255,
+      "step": 1280
+    },
+    {
+      "epoch": 1.094147582697201,
+      "grad_norm": 0.18222452700138092,
+      "learning_rate": 4.006934054518745e-05,
+      "loss": 1.5958,
+      "step": 1290
+    },
+    {
+      "epoch": 1.1026293469041561,
+      "grad_norm": 0.18307287991046906,
+      "learning_rate": 3.98717266851651e-05,
+      "loss": 1.6374,
+      "step": 1300
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.18946941196918488,
+      "learning_rate": 3.967266410553584e-05,
+      "loss": 1.6139,
+      "step": 1310
+    },
+    {
+      "epoch": 1.1195928753180662,
+      "grad_norm": 0.19172786176204681,
+      "learning_rate": 3.947217219785174e-05,
+      "loss": 1.6291,
+      "step": 1320
+    },
+    {
+      "epoch": 1.1280746395250212,
+      "grad_norm": 0.19109292328357697,
+      "learning_rate": 3.92702704929019e-05,
+      "loss": 1.6286,
+      "step": 1330
+    },
+    {
+      "epoch": 1.1365564037319762,
+      "grad_norm": 0.19251668453216553,
+      "learning_rate": 3.906697865880995e-05,
+      "loss": 1.6168,
+      "step": 1340
+    },
+    {
+      "epoch": 1.1450381679389312,
+      "grad_norm": 0.19415880739688873,
+      "learning_rate": 3.886231649911799e-05,
+      "loss": 1.6463,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1535199321458864,
+      "grad_norm": 0.19016750156879425,
+      "learning_rate": 3.865630395085754e-05,
+      "loss": 1.6448,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1620016963528415,
+      "grad_norm": 0.18788300454616547,
+      "learning_rate": 3.8448961082607346e-05,
+      "loss": 1.6236,
+      "step": 1370
+    },
+    {
+      "epoch": 1.1704834605597965,
+      "grad_norm": 0.19576455652713776,
+      "learning_rate": 3.82403080925384e-05,
+      "loss": 1.6289,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1789652247667515,
+      "grad_norm": 0.19166286289691925,
+      "learning_rate": 3.803036530644636e-05,
+      "loss": 1.6059,
+      "step": 1390
+    },
+    {
+      "epoch": 1.1874469889737065,
+      "grad_norm": 0.20226119458675385,
+      "learning_rate": 3.7819153175771524e-05,
+      "loss": 1.6339,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1959287531806615,
+      "grad_norm": 0.1938009411096573,
+      "learning_rate": 3.7606692275606596e-05,
+      "loss": 1.6112,
+      "step": 1410
+    },
+    {
+      "epoch": 1.2044105173876165,
+      "grad_norm": 0.19422097504138947,
+      "learning_rate": 3.739300330269232e-05,
+      "loss": 1.6095,
+      "step": 1420
+    },
+    {
+      "epoch": 1.2128922815945717,
+      "grad_norm": 0.19058625400066376,
+      "learning_rate": 3.7178107073401356e-05,
+      "loss": 1.602,
+      "step": 1430
+    },
+    {
+      "epoch": 1.2213740458015268,
+      "grad_norm": 0.19627830386161804,
+      "learning_rate": 3.696202452171047e-05,
+      "loss": 1.6087,
+      "step": 1440
+    },
+    {
+      "epoch": 1.2298558100084818,
+      "grad_norm": 0.19286489486694336,
+      "learning_rate": 3.674477669716125e-05,
+      "loss": 1.6143,
+      "step": 1450
+    },
+    {
+      "epoch": 1.2383375742154368,
+      "grad_norm": 0.19028903543949127,
+      "learning_rate": 3.6526384762809574e-05,
+      "loss": 1.6236,
+      "step": 1460
+    },
+    {
+      "epoch": 1.2468193384223918,
+      "grad_norm": 0.19716697931289673,
+      "learning_rate": 3.630686999316406e-05,
+      "loss": 1.6271,
+      "step": 1470
+    },
+    {
+      "epoch": 1.255301102629347,
+      "grad_norm": 0.19651146233081818,
+      "learning_rate": 3.608625377211357e-05,
+      "loss": 1.6396,
+      "step": 1480
+    },
+    {
+      "epoch": 1.263782866836302,
+      "grad_norm": 0.19856080412864685,
+      "learning_rate": 3.5864557590844164e-05,
+      "loss": 1.5935,
+      "step": 1490
+    },
+    {
+      "epoch": 1.272264631043257,
+      "grad_norm": 0.20016399025917053,
+      "learning_rate": 3.564180304574549e-05,
+      "loss": 1.6093,
+      "step": 1500
+    },
+    {
+      "epoch": 1.280746395250212,
+      "grad_norm": 0.19493816792964935,
+      "learning_rate": 3.541801183630706e-05,
+      "loss": 1.6601,
+      "step": 1510
+    },
+    {
+      "epoch": 1.289228159457167,
+      "grad_norm": 0.20034298300743103,
+      "learning_rate": 3.519320576300436e-05,
+      "loss": 1.6269,
+      "step": 1520
+    },
+    {
+      "epoch": 1.297709923664122,
+      "grad_norm": 0.19294461607933044,
+      "learning_rate": 3.4967406725175165e-05,
+      "loss": 1.6089,
+      "step": 1530
+    },
+    {
+      "epoch": 1.306191687871077,
+      "grad_norm": 0.20269368588924408,
+      "learning_rate": 3.4740636718886264e-05,
+      "loss": 1.6189,
+      "step": 1540
+    },
+    {
+      "epoch": 1.3146734520780323,
+      "grad_norm": 0.19257566332817078,
+      "learning_rate": 3.45129178347907e-05,
+      "loss": 1.6202,
+      "step": 1550
+    },
+    {
+      "epoch": 1.3231552162849873,
+      "grad_norm": 0.2031780183315277,
+      "learning_rate": 3.428427225597583e-05,
+      "loss": 1.6241,
+      "step": 1560
+    },
+    {
+      "epoch": 1.3316369804919423,
+      "grad_norm": 0.2029244303703308,
+      "learning_rate": 3.405472225580238e-05,
+      "loss": 1.6242,
+      "step": 1570
+    },
+    {
+      "epoch": 1.3401187446988974,
+      "grad_norm": 0.1954798549413681,
+      "learning_rate": 3.38242901957347e-05,
+      "loss": 1.6077,
+      "step": 1580
+    },
+    {
+      "epoch": 1.3486005089058524,
+      "grad_norm": 0.20626811683177948,
+      "learning_rate": 3.359299852316242e-05,
+      "loss": 1.6045,
+      "step": 1590
+    },
+    {
+      "epoch": 1.3570822731128076,
+      "grad_norm": 0.2025717794895172,
+      "learning_rate": 3.336086976921378e-05,
+      "loss": 1.6098,
+      "step": 1600
+    },
+    {
+      "epoch": 1.3655640373197624,
+      "grad_norm": 0.21111804246902466,
+      "learning_rate": 3.312792654656073e-05,
+      "loss": 1.6165,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3740458015267176,
+      "grad_norm": 0.19769315421581268,
+      "learning_rate": 3.28941915472162e-05,
+      "loss": 1.6013,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3825275657336726,
+      "grad_norm": 0.1956772804260254,
+      "learning_rate": 3.2659687540323516e-05,
+      "loss": 1.5936,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3910093299406276,
+      "grad_norm": 0.2080204337835312,
+      "learning_rate": 3.242443736993837e-05,
+      "loss": 1.5988,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3994910941475827,
+      "grad_norm": 0.20390017330646515,
+      "learning_rate": 3.218846395280349e-05,
+      "loss": 1.628,
+      "step": 1650
+    },
+    {
+      "epoch": 1.4079728583545377,
+      "grad_norm": 0.21278975903987885,
+      "learning_rate": 3.19517902761162e-05,
+      "loss": 1.6094,
+      "step": 1660
+    },
+    {
+      "epoch": 1.416454622561493,
+      "grad_norm": 0.19775566458702087,
+      "learning_rate": 3.171443939528919e-05,
+      "loss": 1.6263,
+      "step": 1670
+    },
+    {
+      "epoch": 1.424936386768448,
+      "grad_norm": 0.20688270032405853,
+      "learning_rate": 3.147643443170454e-05,
+      "loss": 1.5863,
+      "step": 1680
+    },
+    {
+      "epoch": 1.433418150975403,
+      "grad_norm": 0.20985737442970276,
+      "learning_rate": 3.123779857046134e-05,
+      "loss": 1.5944,
+      "step": 1690
+    },
+    {
+      "epoch": 1.441899915182358,
+      "grad_norm": 0.2099895030260086,
+      "learning_rate": 3.0998555058117216e-05,
+      "loss": 1.6376,
+      "step": 1700
+    },
+    {
+      "epoch": 1.450381679389313,
+      "grad_norm": 0.20941784977912903,
+      "learning_rate": 3.075872720042372e-05,
+      "loss": 1.6132,
+      "step": 1710
+    },
+    {
+      "epoch": 1.458863443596268,
+      "grad_norm": 0.20355844497680664,
+      "learning_rate": 3.0518338360056e-05,
+      "loss": 1.6218,
+      "step": 1720
+    },
+    {
+      "epoch": 1.467345207803223,
+      "grad_norm": 0.20220711827278137,
+      "learning_rate": 3.0277411954337005e-05,
+      "loss": 1.592,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4758269720101782,
+      "grad_norm": 0.20586727559566498,
+      "learning_rate": 3.0035971452956224e-05,
+      "loss": 1.615,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4843087362171332,
+      "grad_norm": 0.21427693963050842,
+      "learning_rate": 2.9794040375683485e-05,
+      "loss": 1.6025,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4927905004240882,
+      "grad_norm": 0.1974625289440155,
+      "learning_rate": 2.9551642290077707e-05,
+      "loss": 1.6188,
+      "step": 1760
+    },
+    {
+      "epoch": 1.5012722646310432,
+      "grad_norm": 0.2087961882352829,
+      "learning_rate": 2.9308800809191157e-05,
+      "loss": 1.6028,
+      "step": 1770
+    },
+    {
+      "epoch": 1.5097540288379983,
+      "grad_norm": 0.21272026002407074,
+      "learning_rate": 2.9065539589269147e-05,
+      "loss": 1.6252,
+      "step": 1780
+    },
+    {
+      "epoch": 1.5182357930449535,
+      "grad_norm": 0.20586243271827698,
+      "learning_rate": 2.8821882327445594e-05,
+      "loss": 1.6216,
+      "step": 1790
+    },
+    {
+      "epoch": 1.5267175572519083,
+      "grad_norm": 0.20619966089725494,
+      "learning_rate": 2.857785275943458e-05,
+      "loss": 1.6182,
+      "step": 1800
+    },
+    {
+      "epoch": 1.5267175572519083,
+      "eval_loss": 1.6240788698196411,
+      "eval_runtime": 8.3704,
+      "eval_samples_per_second": 45.518,
+      "eval_steps_per_second": 11.469,
+      "step": 1800
+    },
+    {
+      "epoch": 1.5351993214588635,
+      "grad_norm": 0.2109595090150833,
+      "learning_rate": 2.8333474657218174e-05,
+      "loss": 1.6002,
+      "step": 1810
+    },
+    {
+      "epoch": 1.5436810856658185,
+      "grad_norm": 0.2052215188741684,
+      "learning_rate": 2.8088771826730643e-05,
+      "loss": 1.5973,
+      "step": 1820
+    },
+    {
+      "epoch": 1.5521628498727735,
+      "grad_norm": 0.21743585169315338,
+      "learning_rate": 2.784376810553947e-05,
+      "loss": 1.6262,
+      "step": 1830
+    },
+    {
+      "epoch": 1.5606446140797285,
+      "grad_norm": 0.20770463347434998,
+      "learning_rate": 2.759848736052321e-05,
+      "loss": 1.5753,
+      "step": 1840
+    },
+    {
+      "epoch": 1.5691263782866836,
+      "grad_norm": 0.21669910848140717,
+      "learning_rate": 2.735295348554651e-05,
+      "loss": 1.6209,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5776081424936388,
+      "grad_norm": 0.21204906702041626,
+      "learning_rate": 2.7107190399132526e-05,
+      "loss": 1.6088,
+      "step": 1860
+    },
+    {
+      "epoch": 1.5860899067005936,
+      "grad_norm": 0.2132139503955841,
+      "learning_rate": 2.686122204213285e-05,
+      "loss": 1.6252,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5945716709075488,
+      "grad_norm": 0.22568291425704956,
+      "learning_rate": 2.6615072375395423e-05,
+      "loss": 1.6112,
+      "step": 1880
+    },
+    {
+      "epoch": 1.6030534351145038,
+      "grad_norm": 0.20681500434875488,
+      "learning_rate": 2.6368765377430344e-05,
+      "loss": 1.6146,
+      "step": 1890
+    },
+    {
+      "epoch": 1.6115351993214588,
+      "grad_norm": 0.20736300945281982,
+      "learning_rate": 2.6122325042074014e-05,
+      "loss": 1.5834,
+      "step": 1900
+    },
+    {
+      "epoch": 1.620016963528414,
+      "grad_norm": 0.20751895010471344,
+      "learning_rate": 2.5875775376151823e-05,
+      "loss": 1.6151,
+      "step": 1910
+    },
+    {
+      "epoch": 1.6284987277353689,
+      "grad_norm": 0.22000038623809814,
+      "learning_rate": 2.562914039713953e-05,
+      "loss": 1.5838,
+      "step": 1920
+    },
+    {
+      "epoch": 1.636980491942324,
+      "grad_norm": 0.20806461572647095,
+      "learning_rate": 2.5382444130823612e-05,
+      "loss": 1.6254,
+      "step": 1930
+    },
+    {
+      "epoch": 1.645462256149279,
+      "grad_norm": 0.21619880199432373,
+      "learning_rate": 2.5135710608960807e-05,
+      "loss": 1.5999,
+      "step": 1940
+    },
+    {
+      "epoch": 1.6539440203562341,
+      "grad_norm": 0.22222678363323212,
+      "learning_rate": 2.4888963866937067e-05,
+      "loss": 1.581,
+      "step": 1950
+    },
+    {
+      "epoch": 1.6624257845631891,
+      "grad_norm": 0.21882720291614532,
+      "learning_rate": 2.464222794142619e-05,
+      "loss": 1.6151,
+      "step": 1960
+    },
+    {
+      "epoch": 1.6709075487701441,
+      "grad_norm": 0.22589364647865295,
+      "learning_rate": 2.43955268680483e-05,
+      "loss": 1.6062,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6793893129770994,
+      "grad_norm": 0.21262851357460022,
+      "learning_rate": 2.4148884679028402e-05,
+      "loss": 1.6096,
+      "step": 1980
+    },
+    {
+      "epoch": 1.6878710771840542,
+      "grad_norm": 0.21310704946517944,
+      "learning_rate": 2.390232540085533e-05,
+      "loss": 1.5932,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6963528413910094,
+      "grad_norm": 0.21306438744068146,
+      "learning_rate": 2.3655873051941203e-05,
+      "loss": 1.5809,
+      "step": 2000
+    },
+    {
+      "epoch": 1.7048346055979644,
+      "grad_norm": 0.20825529098510742,
+      "learning_rate": 2.340955164028171e-05,
+      "loss": 1.5818,
+      "step": 2010
+    },
+    {
+      "epoch": 1.7133163698049194,
+      "grad_norm": 0.21667224168777466,
+      "learning_rate": 2.3163385161117367e-05,
+      "loss": 1.5911,
+      "step": 2020
+    },
+    {
+      "epoch": 1.7217981340118744,
+      "grad_norm": 0.2139502763748169,
+      "learning_rate": 2.291739759459605e-05,
+      "loss": 1.5816,
+      "step": 2030
+    },
+    {
+      "epoch": 1.7302798982188294,
+      "grad_norm": 0.2165052443742752,
+      "learning_rate": 2.267161290343694e-05,
+      "loss": 1.6033,
+      "step": 2040
+    },
+    {
+      "epoch": 1.7387616624257847,
+      "grad_norm": 0.2145167738199234,
+      "learning_rate": 2.2426055030596282e-05,
+      "loss": 1.6122,
+      "step": 2050
+    },
+    {
+      "epoch": 1.7472434266327395,
+      "grad_norm": 0.2100827991962433,
+      "learning_rate": 2.2180747896934956e-05,
+      "loss": 1.5995,
+      "step": 2060
+    },
+    {
+      "epoch": 1.7557251908396947,
+      "grad_norm": 0.22637659311294556,
+      "learning_rate": 2.193571539888824e-05,
+      "loss": 1.5938,
+      "step": 2070
+    },
+    {
+      "epoch": 1.7642069550466497,
+      "grad_norm": 0.2245241105556488,
+      "learning_rate": 2.169098140613797e-05,
+      "loss": 1.5792,
+      "step": 2080
+    },
+    {
+      "epoch": 1.7726887192536047,
+      "grad_norm": 0.2259100377559662,
+      "learning_rate": 2.1446569759287267e-05,
+      "loss": 1.5856,
+      "step": 2090
+    },
+    {
+      "epoch": 1.78117048346056,
+      "grad_norm": 0.22213774919509888,
+      "learning_rate": 2.120250426753815e-05,
+      "loss": 1.5875,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7896522476675147,
+      "grad_norm": 0.22346043586730957,
+      "learning_rate": 2.095880870637216e-05,
+      "loss": 1.5909,
+      "step": 2110
+    },
+    {
+      "epoch": 1.79813401187447,
+      "grad_norm": 0.2293236404657364,
+      "learning_rate": 2.071550681523429e-05,
+      "loss": 1.5812,
+      "step": 2120
+    },
+    {
+      "epoch": 1.806615776081425,
+      "grad_norm": 0.22071833908557892,
+      "learning_rate": 2.0472622295220413e-05,
+      "loss": 1.5909,
+      "step": 2130
+    },
+    {
+      "epoch": 1.81509754028838,
+      "grad_norm": 0.22376208007335663,
+      "learning_rate": 2.023017880676849e-05,
+      "loss": 1.5852,
+      "step": 2140
+    },
+    {
+      "epoch": 1.823579304495335,
+      "grad_norm": 0.23016902804374695,
+      "learning_rate": 1.9988199967353642e-05,
+      "loss": 1.5997,
+      "step": 2150
+    },
+    {
+      "epoch": 1.83206106870229,
+      "grad_norm": 0.22549933195114136,
+      "learning_rate": 1.974670934918753e-05,
+      "loss": 1.6062,
+      "step": 2160
+    },
+    {
+      "epoch": 1.8405428329092453,
+      "grad_norm": 0.2221376597881317,
+      "learning_rate": 1.9505730476922064e-05,
+      "loss": 1.6112,
+      "step": 2170
+    },
+    {
+      "epoch": 1.8490245971162,
+      "grad_norm": 0.22362485527992249,
+      "learning_rate": 1.9265286825357742e-05,
+      "loss": 1.6001,
+      "step": 2180
+    },
+    {
+      "epoch": 1.8575063613231553,
+      "grad_norm": 0.22703245282173157,
+      "learning_rate": 1.9025401817156905e-05,
+      "loss": 1.6122,
+      "step": 2190
+    },
+    {
+      "epoch": 1.8659881255301103,
+      "grad_norm": 0.2253347635269165,
+      "learning_rate": 1.8786098820561972e-05,
+      "loss": 1.5837,
+      "step": 2200
+    },
+    {
+      "epoch": 1.8744698897370653,
+      "grad_norm": 0.2247924655675888,
+      "learning_rate": 1.8547401147119127e-05,
+      "loss": 1.6056,
+      "step": 2210
+    },
+    {
+      "epoch": 1.8829516539440203,
+      "grad_norm": 0.2206358015537262,
+      "learning_rate": 1.8309332049407363e-05,
+      "loss": 1.5898,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8914334181509753,
+      "grad_norm": 0.23167766630649567,
+      "learning_rate": 1.8071914718773393e-05,
+      "loss": 1.5803,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8999151823579306,
+      "grad_norm": 0.22543945908546448,
+      "learning_rate": 1.783517228307245e-05,
+      "loss": 1.5908,
+      "step": 2240
+    },
+    {
+      "epoch": 1.9083969465648853,
+      "grad_norm": 0.22845198214054108,
+      "learning_rate": 1.7599127804415317e-05,
+      "loss": 1.6114,
+      "step": 2250
+    },
+    {
+      "epoch": 1.9168787107718406,
+      "grad_norm": 0.22125060856342316,
+      "learning_rate": 1.7363804276921748e-05,
+      "loss": 1.5835,
+      "step": 2260
+    },
+    {
+      "epoch": 1.9253604749787956,
+      "grad_norm": 0.22639022767543793,
+      "learning_rate": 1.7129224624480506e-05,
+      "loss": 1.5937,
+      "step": 2270
+    },
+    {
+      "epoch": 1.9338422391857506,
+      "grad_norm": 0.23626133799552917,
+      "learning_rate": 1.689541169851625e-05,
+      "loss": 1.6138,
+      "step": 2280
+    },
+    {
+      "epoch": 1.9423240033927058,
+      "grad_norm": 0.2249774932861328,
+      "learning_rate": 1.6662388275763484e-05,
+      "loss": 1.5972,
+      "step": 2290
+    },
+    {
+      "epoch": 1.9508057675996606,
+      "grad_norm": 0.22697053849697113,
+      "learning_rate": 1.643017705604774e-05,
+      "loss": 1.6007,
+      "step": 2300
+    },
+    {
+      "epoch": 1.9592875318066159,
+      "grad_norm": 0.23137332499027252,
+      "learning_rate": 1.6198800660074353e-05,
+      "loss": 1.5868,
+      "step": 2310
+    },
+    {
+      "epoch": 1.9677692960135709,
+      "grad_norm": 0.2348482310771942,
+      "learning_rate": 1.5968281627224823e-05,
+      "loss": 1.5554,
+      "step": 2320
+    },
+    {
+      "epoch": 1.9762510602205259,
+      "grad_norm": 0.2407807558774948,
+      "learning_rate": 1.573864241336119e-05,
+      "loss": 1.6222,
+      "step": 2330
+    },
+    {
+      "epoch": 1.984732824427481,
+      "grad_norm": 0.22666969895362854,
+      "learning_rate": 1.550990538863848e-05,
+      "loss": 1.5684,
+      "step": 2340
+    },
+    {
+      "epoch": 1.993214588634436,
+      "grad_norm": 0.23586216568946838,
+      "learning_rate": 1.528209283532559e-05,
+      "loss": 1.6103,
+      "step": 2350
+    },
+    {
+      "epoch": 2.001696352841391,
+      "grad_norm": 0.2294708639383316,
+      "learning_rate": 1.5055226945634571e-05,
+      "loss": 1.5866,
+      "step": 2360
+    },
+    {
+      "epoch": 2.010178117048346,
+      "grad_norm": 0.2632901072502136,
+      "learning_rate": 1.4829329819558905e-05,
+      "loss": 1.524,
+      "step": 2370
+    },
+    {
+      "epoch": 2.018659881255301,
+      "grad_norm": 0.26669132709503174,
+      "learning_rate": 1.4604423462720574e-05,
+      "loss": 1.5222,
+      "step": 2380
+    },
+    {
+      "epoch": 2.027141645462256,
+      "grad_norm": 0.245851069688797,
+      "learning_rate": 1.438052978422641e-05,
+      "loss": 1.538,
+      "step": 2390
+    },
+    {
+      "epoch": 2.035623409669211,
+      "grad_norm": 0.2613518238067627,
+      "learning_rate": 1.4157670594533879e-05,
+      "loss": 1.5463,
+      "step": 2400
+    },
+    {
+      "epoch": 2.035623409669211,
+      "eval_loss": 1.6182470321655273,
+      "eval_runtime": 8.3631,
+      "eval_samples_per_second": 45.557,
+      "eval_steps_per_second": 11.479,
+      "step": 2400
+    },
+    {
+      "epoch": 2.0441051738761664,
+      "grad_norm": 0.25607672333717346,
+      "learning_rate": 1.393586760332633e-05,
+      "loss": 1.5402,
+      "step": 2410
+    },
+    {
+      "epoch": 2.052586938083121,
+      "grad_norm": 0.24335207045078278,
+      "learning_rate": 1.3715142417398275e-05,
+      "loss": 1.5312,
+      "step": 2420
+    },
+    {
+      "epoch": 2.0610687022900764,
+      "grad_norm": 0.25492334365844727,
+      "learning_rate": 1.3495516538550495e-05,
+      "loss": 1.5251,
+      "step": 2430
+    },
+    {
+      "epoch": 2.0695504664970312,
+      "grad_norm": 0.26885706186294556,
+      "learning_rate": 1.3277011361495498e-05,
+      "loss": 1.5536,
+      "step": 2440
+    },
+    {
+      "epoch": 2.0780322307039865,
+      "grad_norm": 0.2573489844799042,
+      "learning_rate": 1.3059648171773361e-05,
+      "loss": 1.5608,
+      "step": 2450
+    },
+    {
+      "epoch": 2.0865139949109412,
+      "grad_norm": 0.26441746950149536,
+      "learning_rate": 1.2843448143678175e-05,
+      "loss": 1.5314,
+      "step": 2460
+    },
+    {
+      "epoch": 2.0949957591178965,
+      "grad_norm": 0.2639044523239136,
+      "learning_rate": 1.2628432338195423e-05,
+      "loss": 1.5769,
+      "step": 2470
+    },
+    {
+      "epoch": 2.1034775233248517,
+      "grad_norm": 0.26254355907440186,
+      "learning_rate": 1.2414621700950307e-05,
+      "loss": 1.5392,
+      "step": 2480
+    },
+    {
+      "epoch": 2.1119592875318065,
+      "grad_norm": 0.2559792697429657,
+      "learning_rate": 1.2202037060167338e-05,
+      "loss": 1.5482,
+      "step": 2490
+    },
+    {
+      "epoch": 2.1204410517387617,
+      "grad_norm": 0.2633630037307739,
+      "learning_rate": 1.1990699124641385e-05,
+      "loss": 1.5305,
+      "step": 2500
+    },
+    {
+      "epoch": 2.1289228159457165,
+      "grad_norm": 0.26149672269821167,
+      "learning_rate": 1.1780628481720333e-05,
+      "loss": 1.5205,
+      "step": 2510
+    },
+    {
+      "epoch": 2.1374045801526718,
+      "grad_norm": 0.2706843912601471,
+      "learning_rate": 1.1571845595299571e-05,
+      "loss": 1.5377,
+      "step": 2520
+    },
+    {
+      "epoch": 2.145886344359627,
+      "grad_norm": 0.2620759606361389,
+      "learning_rate": 1.1364370803828536e-05,
+      "loss": 1.5274,
+      "step": 2530
+    },
+    {
+      "epoch": 2.154368108566582,
+      "grad_norm": 0.2704004645347595,
+      "learning_rate": 1.1158224318329439e-05,
+      "loss": 1.5422,
+      "step": 2540
+    },
+    {
+      "epoch": 2.162849872773537,
+      "grad_norm": 0.2624742090702057,
+      "learning_rate": 1.0953426220428417e-05,
+      "loss": 1.5189,
+      "step": 2550
+    },
+    {
+      "epoch": 2.171331636980492,
+      "grad_norm": 0.2630760669708252,
+      "learning_rate": 1.0749996460399334e-05,
+      "loss": 1.545,
+      "step": 2560
+    },
+    {
+      "epoch": 2.179813401187447,
+      "grad_norm": 0.2676001489162445,
+      "learning_rate": 1.0547954855220258e-05,
+      "loss": 1.5395,
+      "step": 2570
+    },
+    {
+      "epoch": 2.188295165394402,
+      "grad_norm": 0.2664811611175537,
+      "learning_rate": 1.0347321086643098e-05,
+      "loss": 1.5427,
+      "step": 2580
+    },
+    {
+      "epoch": 2.196776929601357,
+      "grad_norm": 0.28437891602516174,
+      "learning_rate": 1.0148114699276254e-05,
+      "loss": 1.5424,
+      "step": 2590
+    },
+    {
+      "epoch": 2.2052586938083123,
+      "grad_norm": 0.27211999893188477,
+      "learning_rate": 9.950355098680716e-06,
+      "loss": 1.5274,
+      "step": 2600
+    },
+    {
+      "epoch": 2.213740458015267,
+      "grad_norm": 0.2823927104473114,
+      "learning_rate": 9.754061549479684e-06,
+      "loss": 1.5466,
+      "step": 2610
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.26873740553855896,
+      "learning_rate": 9.559253173481894e-06,
+      "loss": 1.5228,
+      "step": 2620
+    },
+    {
+      "epoch": 2.230703986429177,
+      "grad_norm": 0.27045539021492004,
+      "learning_rate": 9.365948947818917e-06,
+      "loss": 1.5392,
+      "step": 2630
+    },
+    {
+      "epoch": 2.2391857506361323,
+      "grad_norm": 0.2740824222564697,
+      "learning_rate": 9.174167703096487e-06,
+      "loss": 1.5312,
+      "step": 2640
+    },
+    {
+      "epoch": 2.247667514843087,
+      "grad_norm": 0.2652507424354553,
+      "learning_rate": 8.983928121560142e-06,
+      "loss": 1.5506,
+      "step": 2650
+    },
+    {
+      "epoch": 2.2561492790500424,
+      "grad_norm": 0.27051135897636414,
+      "learning_rate": 8.795248735275295e-06,
+      "loss": 1.5392,
+      "step": 2660
+    },
+    {
+      "epoch": 2.2646310432569976,
+      "grad_norm": 0.26868537068367004,
+      "learning_rate": 8.608147924321954e-06,
+      "loss": 1.5313,
+      "step": 2670
+    },
+    {
+      "epoch": 2.2731128074639524,
+      "grad_norm": 0.27669626474380493,
+      "learning_rate": 8.422643915004242e-06,
+      "loss": 1.5379,
+      "step": 2680
+    },
+    {
+      "epoch": 2.2815945716709076,
+      "grad_norm": 0.2646745443344116,
+      "learning_rate": 8.238754778074869e-06,
+      "loss": 1.5187,
+      "step": 2690
+    },
+    {
+      "epoch": 2.2900763358778624,
+      "grad_norm": 0.2804393470287323,
+      "learning_rate": 8.056498426974817e-06,
+      "loss": 1.5436,
+      "step": 2700
+    },
+    {
+      "epoch": 2.2985581000848176,
+      "grad_norm": 0.2765372097492218,
+      "learning_rate": 7.875892616088293e-06,
+      "loss": 1.5349,
+      "step": 2710
+    },
+    {
+      "epoch": 2.307039864291773,
+      "grad_norm": 0.26837730407714844,
+      "learning_rate": 7.696954939013182e-06,
+      "loss": 1.5157,
+      "step": 2720
+    },
+    {
+      "epoch": 2.3155216284987277,
+      "grad_norm": 0.27085524797439575,
+      "learning_rate": 7.5197028268472134e-06,
+      "loss": 1.5115,
+      "step": 2730
+    },
+    {
+      "epoch": 2.324003392705683,
+      "grad_norm": 0.2781482934951782,
+      "learning_rate": 7.344153546489904e-06,
+      "loss": 1.5391,
+      "step": 2740
+    },
+    {
+      "epoch": 2.3324851569126377,
+      "grad_norm": 0.2616721987724304,
+      "learning_rate": 7.17032419896051e-06,
+      "loss": 1.5305,
+      "step": 2750
+    },
+    {
+      "epoch": 2.340966921119593,
+      "grad_norm": 0.2856769859790802,
+      "learning_rate": 6.998231717732148e-06,
+      "loss": 1.5011,
+      "step": 2760
+    },
+    {
+      "epoch": 2.349448685326548,
+      "grad_norm": 0.27664417028427124,
+      "learning_rate": 6.827892867082236e-06,
+      "loss": 1.503,
+      "step": 2770
+    },
+    {
+      "epoch": 2.357930449533503,
+      "grad_norm": 0.2847400903701782,
+      "learning_rate": 6.659324240459389e-06,
+      "loss": 1.5295,
+      "step": 2780
+    },
+    {
+      "epoch": 2.366412213740458,
+      "grad_norm": 0.2821629047393799,
+      "learning_rate": 6.492542258867013e-06,
+      "loss": 1.539,
+      "step": 2790
+    },
+    {
+      "epoch": 2.374893977947413,
+      "grad_norm": 0.28155988454818726,
+      "learning_rate": 6.327563169263651e-06,
+      "loss": 1.5195,
+      "step": 2800
+    },
+    {
+      "epoch": 2.383375742154368,
+      "grad_norm": 0.275939017534256,
+      "learning_rate": 6.164403042980277e-06,
+      "loss": 1.5154,
+      "step": 2810
+    },
+    {
+      "epoch": 2.391857506361323,
+      "grad_norm": 0.28075718879699707,
+      "learning_rate": 6.00307777415475e-06,
+      "loss": 1.5294,
+      "step": 2820
+    },
+    {
+      "epoch": 2.4003392705682782,
+      "grad_norm": 0.2765721082687378,
+      "learning_rate": 5.843603078183465e-06,
+      "loss": 1.5282,
+      "step": 2830
+    },
+    {
+      "epoch": 2.408821034775233,
+      "grad_norm": 0.27522534132003784,
+      "learning_rate": 5.685994490190474e-06,
+      "loss": 1.5159,
+      "step": 2840
+    },
+    {
+      "epoch": 2.4173027989821882,
+      "grad_norm": 0.276583731174469,
+      "learning_rate": 5.530267363514124e-06,
+      "loss": 1.5402,
+      "step": 2850
+    },
+    {
+      "epoch": 2.4257845631891435,
+      "grad_norm": 0.27267441153526306,
+      "learning_rate": 5.3764368682114315e-06,
+      "loss": 1.5255,
+      "step": 2860
+    },
+    {
+      "epoch": 2.4342663273960983,
+      "grad_norm": 0.27260467410087585,
+      "learning_rate": 5.2245179895803044e-06,
+      "loss": 1.5335,
+      "step": 2870
+    },
+    {
+      "epoch": 2.4427480916030535,
+      "grad_norm": 0.2702862620353699,
+      "learning_rate": 5.074525526699736e-06,
+      "loss": 1.5305,
+      "step": 2880
+    },
+    {
+      "epoch": 2.4512298558100083,
+      "grad_norm": 0.2800707519054413,
+      "learning_rate": 4.926474090988193e-06,
+      "loss": 1.5533,
+      "step": 2890
+    },
+    {
+      "epoch": 2.4597116200169635,
+      "grad_norm": 0.2815302908420563,
+      "learning_rate": 4.780378104780234e-06,
+      "loss": 1.5158,
+      "step": 2900
+    },
+    {
+      "epoch": 2.4681933842239188,
+      "grad_norm": 0.2695189416408539,
+      "learning_rate": 4.636251799921579e-06,
+      "loss": 1.525,
+      "step": 2910
+    },
+    {
+      "epoch": 2.4766751484308736,
+      "grad_norm": 0.27255234122276306,
+      "learning_rate": 4.4941092163827135e-06,
+      "loss": 1.514,
+      "step": 2920
+    },
+    {
+      "epoch": 2.485156912637829,
+      "grad_norm": 0.27873003482818604,
+      "learning_rate": 4.353964200891192e-06,
+      "loss": 1.5393,
+      "step": 2930
+    },
+    {
+      "epoch": 2.4936386768447836,
+      "grad_norm": 0.2896173298358917,
+      "learning_rate": 4.215830405582777e-06,
+      "loss": 1.5401,
+      "step": 2940
+    },
+    {
+      "epoch": 2.502120441051739,
+      "grad_norm": 0.2764180302619934,
+      "learning_rate": 4.079721286671528e-06,
+      "loss": 1.5417,
+      "step": 2950
+    },
+    {
+      "epoch": 2.510602205258694,
+      "grad_norm": 0.27540919184684753,
+      "learning_rate": 3.945650103138962e-06,
+      "loss": 1.5291,
+      "step": 2960
+    },
+    {
+      "epoch": 2.519083969465649,
+      "grad_norm": 0.2762376070022583,
+      "learning_rate": 3.813629915442435e-06,
+      "loss": 1.5305,
+      "step": 2970
+    },
+    {
+      "epoch": 2.527565733672604,
+      "grad_norm": 0.2932542562484741,
+      "learning_rate": 3.6836735842429027e-06,
+      "loss": 1.5355,
+      "step": 2980
+    },
+    {
+      "epoch": 2.536047497879559,
+      "grad_norm": 0.28812217712402344,
+      "learning_rate": 3.5557937691520382e-06,
+      "loss": 1.5618,
+      "step": 2990
+    },
+    {
+      "epoch": 2.544529262086514,
+      "grad_norm": 0.28096404671669006,
+      "learning_rate": 3.4300029274990734e-06,
+      "loss": 1.5356,
+      "step": 3000
+    },
+    {
+      "epoch": 2.544529262086514,
+      "eval_loss": 1.6116546392440796,
+      "eval_runtime": 8.3692,
+      "eval_samples_per_second": 45.524,
+      "eval_steps_per_second": 11.471,
+      "step": 3000
+    },
+    {
+      "epoch": 2.553011026293469,
+      "grad_norm": 0.2974391579627991,
+      "learning_rate": 3.3063133131172403e-06,
+      "loss": 1.5331,
+      "step": 3010
+    },
+    {
+      "epoch": 2.561492790500424,
+      "grad_norm": 0.2741416394710541,
+      "learning_rate": 3.1847369751500766e-06,
+      "loss": 1.5251,
+      "step": 3020
+    },
+    {
+      "epoch": 2.569974554707379,
+      "grad_norm": 0.27761682868003845,
+      "learning_rate": 3.065285756877684e-06,
+      "loss": 1.5296,
+      "step": 3030
+    },
+    {
+      "epoch": 2.578456318914334,
+      "grad_norm": 0.2878273129463196,
+      "learning_rate": 2.947971294562976e-06,
+      "loss": 1.5432,
+      "step": 3040
+    },
+    {
+      "epoch": 2.5869380831212894,
+      "grad_norm": 0.2833128869533539,
+      "learning_rate": 2.8328050163182017e-06,
+      "loss": 1.5209,
+      "step": 3050
+    },
+    {
+      "epoch": 2.595419847328244,
+      "grad_norm": 0.3064183294773102,
+      "learning_rate": 2.7197981409916416e-06,
+      "loss": 1.5096,
+      "step": 3060
+    },
+    {
+      "epoch": 2.6039016115351994,
+      "grad_norm": 0.2813494801521301,
+      "learning_rate": 2.6089616770747394e-06,
+      "loss": 1.5133,
+      "step": 3070
+    },
+    {
+      "epoch": 2.612383375742154,
+      "grad_norm": 0.28296294808387756,
+      "learning_rate": 2.500306421629728e-06,
+      "loss": 1.5477,
+      "step": 3080
+    },
+    {
+      "epoch": 2.6208651399491094,
+      "grad_norm": 0.2628287672996521,
+      "learning_rate": 2.393842959237813e-06,
+      "loss": 1.5082,
+      "step": 3090
+    },
+    {
+      "epoch": 2.6293469041560646,
+      "grad_norm": 0.2780253291130066,
+      "learning_rate": 2.289581660968121e-06,
+      "loss": 1.5568,
+      "step": 3100
+    },
+    {
+      "epoch": 2.6378286683630194,
+      "grad_norm": 0.27985766530036926,
+      "learning_rate": 2.187532683367385e-06,
+      "loss": 1.5065,
+      "step": 3110
+    },
+    {
+      "epoch": 2.6463104325699747,
+      "grad_norm": 0.28796765208244324,
+      "learning_rate": 2.0877059674705464e-06,
+      "loss": 1.5217,
+      "step": 3120
+    },
+    {
+      "epoch": 2.6547921967769295,
+      "grad_norm": 0.28680670261383057,
+      "learning_rate": 1.9901112378323846e-06,
+      "loss": 1.5044,
+      "step": 3130
+    },
+    {
+      "epoch": 2.6632739609838847,
+      "grad_norm": 0.2811967730522156,
+      "learning_rate": 1.8947580015801575e-06,
+      "loss": 1.536,
+      "step": 3140
+    },
+    {
+      "epoch": 2.67175572519084,
+      "grad_norm": 0.2895421087741852,
+      "learning_rate": 1.8016555474875213e-06,
+      "loss": 1.5325,
+      "step": 3150
+    },
+    {
+      "epoch": 2.6802374893977947,
+      "grad_norm": 0.284627228975296,
+      "learning_rate": 1.710812945069637e-06,
+      "loss": 1.5302,
+      "step": 3160
+    },
+    {
+      "epoch": 2.68871925360475,
+      "grad_norm": 0.27630600333213806,
+      "learning_rate": 1.622239043699686e-06,
+      "loss": 1.5353,
+      "step": 3170
+    },
+    {
+      "epoch": 2.6972010178117047,
+      "grad_norm": 0.28358718752861023,
+      "learning_rate": 1.535942471746807e-06,
+      "loss": 1.5171,
+      "step": 3180
+    },
+    {
+      "epoch": 2.70568278201866,
+      "grad_norm": 0.2867433428764343,
+      "learning_rate": 1.4519316357355733e-06,
+      "loss": 1.5308,
+      "step": 3190
+    },
+    {
+      "epoch": 2.714164546225615,
+      "grad_norm": 0.29057636857032776,
+      "learning_rate": 1.3702147195270687e-06,
+      "loss": 1.5366,
+      "step": 3200
+    },
+    {
+      "epoch": 2.72264631043257,
+      "grad_norm": 0.28410962224006653,
+      "learning_rate": 1.2907996835216734e-06,
+      "loss": 1.5148,
+      "step": 3210
+    },
+    {
+      "epoch": 2.731128074639525,
+      "grad_norm": 0.2737291157245636,
+      "learning_rate": 1.2136942638835995e-06,
+      "loss": 1.5228,
+      "step": 3220
+    },
+    {
+      "epoch": 2.73960983884648,
+      "grad_norm": 0.28176769614219666,
+      "learning_rate": 1.1389059717872741e-06,
+      "loss": 1.5197,
+      "step": 3230
+    },
+    {
+      "epoch": 2.7480916030534353,
+      "grad_norm": 0.27323269844055176,
+      "learning_rate": 1.06644209268566e-06,
+      "loss": 1.5125,
+      "step": 3240
+    },
+    {
+      "epoch": 2.75657336726039,
+      "grad_norm": 0.2849061191082001,
+      "learning_rate": 9.963096856005216e-07,
+      "loss": 1.5549,
+      "step": 3250
+    },
+    {
+      "epoch": 2.7650551314673453,
+      "grad_norm": 0.27499040961265564,
+      "learning_rate": 9.285155824347969e-07,
+      "loss": 1.5183,
+      "step": 3260
+    },
+    {
+      "epoch": 2.7735368956743,
+      "grad_norm": 0.2779834270477295,
+      "learning_rate": 8.630663873070627e-07,
+      "loss": 1.5323,
+      "step": 3270
+    },
+    {
+      "epoch": 2.7820186598812553,
+      "grad_norm": 0.2841133177280426,
+      "learning_rate": 7.999684759082083e-07,
+      "loss": 1.5202,
+      "step": 3280
+    },
+    {
+      "epoch": 2.7905004240882105,
+      "grad_norm": 0.2832767367362976,
+      "learning_rate": 7.392279948803371e-07,
+      "loss": 1.541,
+      "step": 3290
+    },
+    {
+      "epoch": 2.7989821882951653,
+      "grad_norm": 0.29252272844314575,
+      "learning_rate": 6.808508612179937e-07,
+      "loss": 1.5437,
+      "step": 3300
+    },
+    {
+      "epoch": 2.8074639525021206,
+      "grad_norm": 0.28466904163360596,
+      "learning_rate": 6.248427616917907e-07,
+      "loss": 1.5489,
+      "step": 3310
+    },
+    {
+      "epoch": 2.8159457167090753,
+      "grad_norm": 0.2813858091831207,
+      "learning_rate": 5.712091522943996e-07,
+      "loss": 1.5159,
+      "step": 3320
+    },
+    {
+      "epoch": 2.8244274809160306,
+      "grad_norm": 0.2804953157901764,
+      "learning_rate": 5.199552577090927e-07,
+      "loss": 1.5359,
+      "step": 3330
+    },
+    {
+      "epoch": 2.832909245122986,
+      "grad_norm": 0.2917366027832031,
+      "learning_rate": 4.710860708007553e-07,
+      "loss": 1.5281,
+      "step": 3340
+    },
+    {
+      "epoch": 2.8413910093299406,
+      "grad_norm": 0.2769799530506134,
+      "learning_rate": 4.2460635212952515e-07,
+      "loss": 1.511,
+      "step": 3350
+    },
+    {
+      "epoch": 2.849872773536896,
+      "grad_norm": 0.29529932141304016,
+      "learning_rate": 3.805206294870439e-07,
+      "loss": 1.5074,
+      "step": 3360
+    },
+    {
+      "epoch": 2.8583545377438506,
+      "grad_norm": 0.2867056131362915,
+      "learning_rate": 3.388331974553821e-07,
+      "loss": 1.5001,
+      "step": 3370
+    },
+    {
+      "epoch": 2.866836301950806,
+      "grad_norm": 0.2916123569011688,
+      "learning_rate": 2.995481169886849e-07,
+      "loss": 1.5362,
+      "step": 3380
+    },
+    {
+      "epoch": 2.875318066157761,
+      "grad_norm": 0.2851467430591583,
+      "learning_rate": 2.6266921501758024e-07,
+      "loss": 1.5205,
+      "step": 3390
+    },
+    {
+      "epoch": 2.883799830364716,
+      "grad_norm": 0.29022783041000366,
+      "learning_rate": 2.2820008407637983e-07,
+      "loss": 1.5117,
+      "step": 3400
+    },
+    {
+      "epoch": 2.8922815945716707,
+      "grad_norm": 0.2804555296897888,
+      "learning_rate": 1.9614408195310907e-07,
+      "loss": 1.5184,
+      "step": 3410
+    },
+    {
+      "epoch": 2.900763358778626,
+      "grad_norm": 0.2887100279331207,
+      "learning_rate": 1.6650433136241872e-07,
+      "loss": 1.5141,
+      "step": 3420
+    },
+    {
+      "epoch": 2.909245122985581,
+      "grad_norm": 0.27130427956581116,
+      "learning_rate": 1.3928371964138653e-07,
+      "loss": 1.5328,
+      "step": 3430
+    },
+    {
+      "epoch": 2.917726887192536,
+      "grad_norm": 0.2768704891204834,
+      "learning_rate": 1.1448489846824228e-07,
+      "loss": 1.5251,
+      "step": 3440
+    },
+    {
+      "epoch": 2.926208651399491,
+      "grad_norm": 0.28594911098480225,
+      "learning_rate": 9.211028360406326e-08,
+      "loss": 1.5303,
+      "step": 3450
+    },
+    {
+      "epoch": 2.934690415606446,
+      "grad_norm": 0.2850119471549988,
+      "learning_rate": 7.216205465743753e-08,
+      "loss": 1.5128,
+      "step": 3460
+    },
+    {
+      "epoch": 2.943172179813401,
+      "grad_norm": 0.28492864966392517,
+      "learning_rate": 5.464215487214763e-08,
+      "loss": 1.5382,
+      "step": 3470
+    },
+    {
+      "epoch": 2.9516539440203564,
+      "grad_norm": 0.2814091444015503,
+      "learning_rate": 3.95522909378665e-08,
+      "loss": 1.5374,
+      "step": 3480
+    },
+    {
+      "epoch": 2.960135708227311,
+      "grad_norm": 0.287667453289032,
+      "learning_rate": 2.6893932823898738e-08,
+      "loss": 1.5138,
+      "step": 3490
+    },
+    {
+      "epoch": 2.9686174724342664,
+      "grad_norm": 0.2752053737640381,
+      "learning_rate": 1.666831363598964e-08,
+      "loss": 1.5293,
+      "step": 3500
+    },
+    {
+      "epoch": 2.9770992366412212,
+      "grad_norm": 0.2834467887878418,
+      "learning_rate": 8.876429496193494e-09,
+      "loss": 1.545,
+      "step": 3510
+    },
+    {
+      "epoch": 2.9855810008481765,
+      "grad_norm": 0.28982028365135193,
+      "learning_rate": 3.519039445851191e-09,
+      "loss": 1.5387,
+      "step": 3520
+    },
+    {
+      "epoch": 2.9940627650551317,
+      "grad_norm": 0.2901974320411682,
+      "learning_rate": 5.966653716410476e-10,
+      "loss": 1.533,
+      "step": 3530
+    },
+    {
+      "epoch": 3.0,
+      "step": 3537,
+      "total_flos": 1.0283028491985224e+19,
+      "train_loss": 1.6124080855326697,
+      "train_runtime": 7405.1369,
+      "train_samples_per_second": 15.277,
+      "train_steps_per_second": 0.478
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 3537,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0283028491985224e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e01bd501f6e727644895f2601acfc8091dbed9f3c8163e45e210fe7b9bf724db
+size 7032