Upload folder using huggingface_hub

.gitattributes

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

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,63 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen3.5-9B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: qwen35_caption_galore
+  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. -->
+
+# qwen35_caption_galore
+
+This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 5e-6), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 5e-6), merger_lr),
+    }
+- train_batch_size: 3
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 20
+- total_train_batch_size: 60
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine_with_min_lr
+- lr_scheduler_warmup_steps: 0.05
+- num_epochs: 4
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.5.3
+- Pytorch 2.11.0+cu128
+- Datasets 4.0.0
+- Tokenizers 0.22.2

README.md

@@ -0,0 +1,63 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen3.5-9B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: qwen35_caption_galore
+  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. -->
+
+# qwen35_caption_galore
+
+This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 5e-6), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 5e-6), merger_lr),
+    }
+- train_batch_size: 3
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 20
+- total_train_batch_size: 60
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine_with_min_lr
+- lr_scheduler_warmup_steps: 0.05
+- num_epochs: 4
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.5.3
+- Pytorch 2.11.0+cu128
+- Datasets 4.0.0
+- Tokenizers 0.22.2

all_results.json

@@ -0,0 +1,9 @@
+{
+    "effective_tokens_per_sec": 7301.591607227352,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.888545433905317,
+    "train_runtime": 9327.0059,
+    "train_samples_per_second": 6.459,
+    "train_steps_per_second": 0.108
+}

chat_template.jinja

@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+    {%- if content is string %}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping %}
+        {%- for item in content %}
+            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain images.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Picture ' ~ image_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+            {%- elif 'video' in item or item.type == 'video' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain videos.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Video ' ~ video_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+            {%- elif 'text' in item %}
+                {{- item.text }}
+            {%- else %}
+                {{- raise_exception('Unexpected item type in content.') }}
+            {%- endif %}
+        {%- endfor %}
+    {%- elif content is none or content is undefined %}
+        {{- '' }}
+    {%- else %}
+        {{- raise_exception('Unexpected content type.') }}
+    {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+    {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+    {{- '<|im_start|>system\n' }}
+    {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>" }}
+    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {%- if content %}
+            {{- '\n\n' + content }}
+        {%- endif %}
+    {%- endif %}
+    {{- '<|im_end|>\n' }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" %}
+        {%- set content = render_content(message.content, false)|trim %}
+        {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
+            {%- set ns.multi_step_tool = false %}
+            {%- set ns.last_query_index = index %}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+    {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+    {%- set content = render_content(message.content, true)|trim %}
+    {%- if message.role == "system" %}
+        {%- if not loop.first %}
+            {{- raise_exception('System message must be at the beginning.') }}
+        {%- endif %}
+    {%- elif message.role == "user" %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- set reasoning_content = reasoning_content|trim %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if tool_call.function is defined %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {%- if loop.first %}
+                    {%- if content|trim %}
+                        {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- else %}
+                        {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- endif %}
+                {%- else %}
+                    {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                {%- endif %}
+                {%- if tool_call.arguments is defined %}
+                    {%- for args_name, args_value in tool_call.arguments|items %}
+                        {{- '<parameter=' + args_name + '>\n' }}
+                        {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
+                        {{- args_value }}
+                        {{- '\n</parameter>\n' }}
+                    {%- endfor %}
+                {%- endif %}
+                {{- '</function>\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.previtem and loop.previtem.role != "tool" %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if not loop.last and loop.nextitem.role != "tool" %}
+            {{- '<|im_end|>\n' }}
+        {%- elif loop.last %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- else %}
+        {{- raise_exception('Unexpected message role.') }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- else %}
+        {{- '<think>\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,113 @@
+{
+  "architectures": [
+    "Qwen3_5ForConditionalGeneration"
+  ],
+  "dtype": "bfloat16",
+  "eos_token_id": 248046,
+  "hidden_size": 4096,
+  "image_token_id": 248056,
+  "model_type": "qwen3_5",
+  "pad_token_id": 248044,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "attn_output_gate": true,
+    "bos_token_id": null,
+    "dtype": "bfloat16",
+    "eos_token_id": 248044,
+    "full_attention_interval": 4,
+    "head_dim": 256,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "initializer_range": 0.02,
+    "intermediate_size": 12288,
+    "layer_types": [
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention"
+    ],
+    "linear_conv_kernel_dim": 4,
+    "linear_key_head_dim": 128,
+    "linear_num_key_heads": 16,
+    "linear_num_value_heads": 32,
+    "linear_value_head_dim": 128,
+    "mamba_ssm_dtype": "float32",
+    "max_position_embeddings": 262144,
+    "mlp_only_layers": [],
+    "model_type": "qwen3_5_text",
+    "mtp_num_hidden_layers": 1,
+    "mtp_use_dedicated_embeddings": false,
+    "num_attention_heads": 16,
+    "num_hidden_layers": 32,
+    "num_key_value_heads": 4,
+    "pad_token_id": null,
+    "partial_rotary_factor": 0.25,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        11,
+        11,
+        10
+      ],
+      "partial_rotary_factor": 0.25,
+      "rope_theta": 10000000,
+      "rope_type": "default"
+    },
+    "tie_word_embeddings": false,
+    "use_cache": false,
+    "vocab_size": 248320
+  },
+  "tie_word_embeddings": false,
+  "transformers_version": "5.5.3",
+  "use_cache": false,
+  "video_token_id": 248057,
+  "vision_config": {
+    "deepstack_visual_indexes": [],
+    "depth": 27,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "in_channels": 3,
+    "initializer_range": 0.02,
+    "intermediate_size": 4304,
+    "model_type": "qwen3_5",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 4096,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 248054,
+  "vision_start_token_id": 248053
+}

generation_config.json

@@ -0,0 +1,10 @@
+{
+  "_from_model_config": true,
+  "eos_token_id": [
+    248046,
+    248044
+  ],
+  "pad_token_id": 248044,
+  "transformers_version": "5.5.3",
+  "use_cache": true
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b95ae7daf5c4aacb7636eeee4289b1571f774e360f5a3e3576e1c2a5b0d62bae
+size 20859273368

processor_config.json

@@ -0,0 +1,60 @@
+{
+  "image_processor": {
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "Qwen2VLImageProcessor",
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "merge_size": 2,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+      "longest_edge": 16777216,
+      "shortest_edge": 65536
+    },
+    "temporal_patch_size": 2
+  },
+  "processor_class": "Qwen3VLProcessor",
+  "video_processor": {
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "do_sample_frames": true,
+    "fps": 2,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "max_frames": 768,
+    "merge_size": 2,
+    "min_frames": 4,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "return_metadata": false,
+    "size": {
+      "longest_edge": 25165824,
+      "shortest_edge": 4096
+    },
+    "temporal_patch_size": 2,
+    "video_processor_type": "Qwen3VLVideoProcessor"
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
+size 19989343

tokenizer_config.json

@@ -0,0 +1,33 @@
+{
+  "add_prefix_space": false,
+  "audio_bos_token": "<|audio_start|>",
+  "audio_eos_token": "<|audio_end|>",
+  "audio_token": "<|audio_pad|>",
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "image_token": "<|image_pad|>",
+  "is_local": true,
+  "model_max_length": 262144,
+  "model_specific_special_tokens": {
+    "audio_bos_token": "<|audio_start|>",
+    "audio_eos_token": "<|audio_end|>",
+    "audio_token": "<|audio_pad|>",
+    "image_token": "<|image_pad|>",
+    "video_token": "<|video_pad|>",
+    "vision_bos_token": "<|vision_start|>",
+    "vision_eos_token": "<|vision_end|>"
+  },
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": null,
+  "video_token": "<|video_pad|>",
+  "vision_bos_token": "<|vision_start|>",
+  "vision_eos_token": "<|vision_end|>"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "effective_tokens_per_sec": 7301.591607227352,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.888545433905317,
+    "train_runtime": 9327.0059,
+    "train_samples_per_second": 6.459,
+    "train_steps_per_second": 0.108
+}

trainer_state.json

@@ -0,0 +1,3557 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 1004,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00796812749003984,
+      "grad_norm": 0.9559739828109741,
+      "learning_rate": 9.803921568627452e-08,
+      "loss": 2.094315528869629,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.7918368577957153,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 2.0106358528137207,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 3.103429079055786,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 2.1994524002075195,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.6494978070259094,
+      "learning_rate": 6.862745098039217e-07,
+      "loss": 1.9166667461395264,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 1.7068390846252441,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 1.5706766843795776,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 1.4902676343917847,
+      "learning_rate": 1.0784313725490197e-06,
+      "loss": 1.816737174987793,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 0.6236617565155029,
+      "learning_rate": 1.2745098039215686e-06,
+      "loss": 2.7344751358032227,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.360865831375122,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 4.102913856506348,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 0.964033305644989,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.7476555109024048,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.5470930337905884,
+      "learning_rate": 1.8627450980392158e-06,
+      "loss": 1.6582436561584473,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.5301034450531006,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 1.5105293989181519,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 0.6716785430908203,
+      "learning_rate": 2.254901960784314e-06,
+      "loss": 1.4427540302276611,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.3430458903312683,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 1.5457741022109985,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.5500040054321289,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 1.642572283744812,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 1.0999526977539062,
+      "learning_rate": 2.843137254901961e-06,
+      "loss": 1.4051127433776855,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 0.5437412261962891,
+      "learning_rate": 3.03921568627451e-06,
+      "loss": 1.216549038887024,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 0.6122533082962036,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 1.3082672357559204,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.3810500502586365,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 1.4653483629226685,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 0.8308849930763245,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 1.2415324449539185,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.9473059177398682,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 1.3710896968841553,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 0.8783873915672302,
+      "learning_rate": 4.019607843137255e-06,
+      "loss": 0.8346668481826782,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.48223596811294556,
+      "learning_rate": 4.215686274509805e-06,
+      "loss": 1.0442719459533691,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 0.6098504662513733,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.9686055183410645,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 1.1614059209823608,
+      "learning_rate": 4.607843137254902e-06,
+      "loss": 1.2143298387527466,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.39265233278274536,
+      "learning_rate": 4.803921568627452e-06,
+      "loss": 1.3708750009536743,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.8406171798706055,
+      "learning_rate": 5e-06,
+      "loss": 1.3206267356872559,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.7920119762420654,
+      "learning_rate": 4.999951098195543e-06,
+      "loss": 1.127969741821289,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.6165183782577515,
+      "learning_rate": 4.9998043949078495e-06,
+      "loss": 1.5454459190368652,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 1.2174687385559082,
+      "learning_rate": 4.9995598965138565e-06,
+      "loss": 1.369892954826355,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.4563896656036377,
+      "learning_rate": 4.9992176136414875e-06,
+      "loss": 0.7839393019676208,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.4270339608192444,
+      "learning_rate": 4.998777561169186e-06,
+      "loss": 1.3749765157699585,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 0.7023680210113525,
+      "learning_rate": 4.998239758225272e-06,
+      "loss": 1.129468321800232,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.8103557229042053,
+      "learning_rate": 4.99760422818711e-06,
+      "loss": 1.2961775064468384,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 0.7736934423446655,
+      "learning_rate": 4.9968709986800925e-06,
+      "loss": 0.8295523524284363,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.8463128209114075,
+      "learning_rate": 4.996040101576436e-06,
+      "loss": 1.552251935005188,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 0.48394837975502014,
+      "learning_rate": 4.995111572993804e-06,
+      "loss": 0.8132213354110718,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 3.6982743740081787,
+      "learning_rate": 4.994085453293725e-06,
+      "loss": 1.137037992477417,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.5522757172584534,
+      "learning_rate": 4.992961787079848e-06,
+      "loss": 1.4612939357757568,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 2.682084560394287,
+      "learning_rate": 4.991740623196001e-06,
+      "loss": 0.8877551555633545,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.279124915599823,
+      "learning_rate": 4.990422014724068e-06,
+      "loss": 0.8413684368133545,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.3474184274673462,
+      "learning_rate": 4.989006018981675e-06,
+      "loss": 1.320245385169983,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.7194346785545349,
+      "learning_rate": 4.987492697519712e-06,
+      "loss": 0.8597245812416077,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.3525860905647278,
+      "learning_rate": 4.985882116119643e-06,
+      "loss": 1.3105067014694214,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 2.683076858520508,
+      "learning_rate": 4.9841743447906576e-06,
+      "loss": 0.8781213760375977,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.49996867775917053,
+      "learning_rate": 4.982369457766621e-06,
+      "loss": 1.325507640838623,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.5643993616104126,
+      "learning_rate": 4.980467533502853e-06,
+      "loss": 1.1099426746368408,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.39738336205482483,
+      "learning_rate": 4.978468654672709e-06,
+      "loss": 1.247117519378662,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 1.4850482940673828,
+      "learning_rate": 4.976372908163999e-06,
+      "loss": 1.0908377170562744,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.6579986214637756,
+      "learning_rate": 4.974180385075195e-06,
+      "loss": 1.2663590908050537,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 1.2138111591339111,
+      "learning_rate": 4.971891180711488e-06,
+      "loss": 0.8543062806129456,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.43239423632621765,
+      "learning_rate": 4.9695053945806295e-06,
+      "loss": 1.2789909839630127,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.5163455009460449,
+      "learning_rate": 4.9670231303886194e-06,
+      "loss": 0.7750987410545349,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 1.9173283576965332,
+      "learning_rate": 4.9644444960351865e-06,
+      "loss": 0.7063390612602234,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.4428473711013794,
+      "learning_rate": 4.9617696036091095e-06,
+      "loss": 0.9797186255455017,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.45927050709724426,
+      "learning_rate": 4.958998569383335e-06,
+      "loss": 1.2907061576843262,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.6998914480209351,
+      "learning_rate": 4.956131513809928e-06,
+      "loss": 0.8183092474937439,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.6853180527687073,
+      "learning_rate": 4.953168561514838e-06,
+      "loss": 1.128021001815796,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.396808385848999,
+      "learning_rate": 4.950109841292475e-06,
+      "loss": 1.2935739755630493,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.3683837652206421,
+      "learning_rate": 4.946955486100119e-06,
+      "loss": 1.351388931274414,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.5228427052497864,
+      "learning_rate": 4.943705633052134e-06,
+      "loss": 1.2385647296905518,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.3592073619365692,
+      "learning_rate": 4.940360423414013e-06,
+      "loss": 0.9587579369544983,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 0.8788257241249084,
+      "learning_rate": 4.936920002596233e-06,
+      "loss": 1.0263392925262451,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 0.7780712842941284,
+      "learning_rate": 4.9333845201479375e-06,
+      "loss": 0.9243409633636475,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.6171746850013733,
+      "learning_rate": 4.9297541297504345e-06,
+      "loss": 1.270082950592041,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.3748152554035187,
+      "learning_rate": 4.9260289892105164e-06,
+      "loss": 0.944374144077301,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 1.0059514045715332,
+      "learning_rate": 4.922209260453599e-06,
+      "loss": 0.7532623410224915,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.4450763165950775,
+      "learning_rate": 4.9182951095166875e-06,
+      "loss": 0.9002277255058289,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.762507975101471,
+      "learning_rate": 4.914286706541153e-06,
+      "loss": 1.270380973815918,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.47578826546669006,
+      "learning_rate": 4.910184225765341e-06,
+      "loss": 0.9316639304161072,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.1545168161392212,
+      "learning_rate": 4.905987845516996e-06,
+      "loss": 1.0807777643203735,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.604080319404602,
+      "learning_rate": 4.90169774820551e-06,
+      "loss": 1.016838788986206,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.3630507290363312,
+      "learning_rate": 4.897314120313996e-06,
+      "loss": 1.2297433614730835,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.541435718536377,
+      "learning_rate": 4.892837152391174e-06,
+      "loss": 1.3433854579925537,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.48429420590400696,
+      "learning_rate": 4.888267039043101e-06,
+      "loss": 0.9228965640068054,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.7947145104408264,
+      "learning_rate": 4.883603978924698e-06,
+      "loss": 0.9337972402572632,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 0.6225354075431824,
+      "learning_rate": 4.878848174731126e-06,
+      "loss": 0.9235936999320984,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.77800452709198,
+      "learning_rate": 4.873999833188967e-06,
+      "loss": 1.1111277341842651,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 8.632575988769531,
+      "learning_rate": 4.8690591650472455e-06,
+      "loss": 1.0318198204040527,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 0.5781624913215637,
+      "learning_rate": 4.864026385068259e-06,
+      "loss": 1.1300413608551025,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.352184534072876,
+      "learning_rate": 4.858901712018251e-06,
+      "loss": 1.0068553686141968,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.22476018965244293,
+      "learning_rate": 4.8536853686578965e-06,
+      "loss": 0.7624706625938416,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.9084291458129883,
+      "learning_rate": 4.84837758173262e-06,
+      "loss": 1.1348692178726196,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.5189490914344788,
+      "learning_rate": 4.8429785819627395e-06,
+      "loss": 0.8424814939498901,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.45786410570144653,
+      "learning_rate": 4.83748860403344e-06,
+      "loss": 1.2375285625457764,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 1.0779087543487549,
+      "learning_rate": 4.831907886584566e-06,
+      "loss": 0.8901735544204712,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.5550018548965454,
+      "learning_rate": 4.8262366722002535e-06,
+      "loss": 0.8349870443344116,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.24916230142116547,
+      "learning_rate": 4.820475207398386e-06,
+      "loss": 1.2311174869537354,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 1.2258037328720093,
+      "learning_rate": 4.8146237426198735e-06,
+      "loss": 0.7291948199272156,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 1.2274919748306274,
+      "learning_rate": 4.808682532217772e-06,
+      "loss": 0.907733678817749,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 0.8285712003707886,
+      "learning_rate": 4.802651834446221e-06,
+      "loss": 1.05784273147583,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.29364529252052307,
+      "learning_rate": 4.796531911449226e-06,
+      "loss": 0.49492236971855164,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.29135143756866455,
+      "learning_rate": 4.790323029249252e-06,
+      "loss": 1.2699949741363525,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.2689671814441681,
+      "learning_rate": 4.784025457735675e-06,
+      "loss": 1.2555361986160278,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.329478919506073,
+      "learning_rate": 4.777639470653035e-06,
+      "loss": 1.0946720838546753,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 1.0551133155822754,
+      "learning_rate": 4.771165345589148e-06,
+      "loss": 0.8700244426727295,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 1.9477455615997314,
+      "learning_rate": 4.7646033639630355e-06,
+      "loss": 1.1975181102752686,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.32198113203048706,
+      "learning_rate": 4.757953811012688e-06,
+      "loss": 0.8865339756011963,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 0.7740492224693298,
+      "learning_rate": 4.7512169757826735e-06,
+      "loss": 0.5861401557922363,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 2.783029556274414,
+      "learning_rate": 4.74439315111157e-06,
+      "loss": 0.7686101794242859,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 2.4759202003479004,
+      "learning_rate": 4.737482633619234e-06,
+      "loss": 1.0390477180480957,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.1833098977804184,
+      "learning_rate": 4.730485723693912e-06,
+      "loss": 0.9645057916641235,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.7762478590011597,
+      "learning_rate": 4.723402725479177e-06,
+      "loss": 1.513939619064331,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.3035716414451599,
+      "learning_rate": 4.71623394686071e-06,
+      "loss": 1.2250189781188965,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.348950058221817,
+      "learning_rate": 4.708979699452923e-06,
+      "loss": 0.5983444452285767,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.26674261689186096,
+      "learning_rate": 4.701640298585405e-06,
+      "loss": 1.2691799402236938,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.2950589954853058,
+      "learning_rate": 4.694216063289219e-06,
+      "loss": 1.2244665622711182,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.4690109193325043,
+      "learning_rate": 4.686707316283033e-06,
+      "loss": 0.9395098090171814,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.792752742767334,
+      "learning_rate": 4.679114383959095e-06,
+      "loss": 1.4253650903701782,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 1.2172280550003052,
+      "learning_rate": 4.6714375963690415e-06,
+      "loss": 0.8411707282066345,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.2963148355484009,
+      "learning_rate": 4.6636772872095545e-06,
+      "loss": 1.2389003038406372,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.4124774932861328,
+      "learning_rate": 4.6558337938078505e-06,
+      "loss": 1.485334873199463,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.44793665409088135,
+      "learning_rate": 4.647907457107025e-06,
+      "loss": 0.6865295767784119,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.3873876631259918,
+      "learning_rate": 4.639898621651226e-06,
+      "loss": 0.9913421869277954,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 0.525278627872467,
+      "learning_rate": 4.631807635570681e-06,
+      "loss": 1.0780147314071655,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.6837154626846313,
+      "learning_rate": 4.623634850566563e-06,
+      "loss": 1.181473731994629,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.4443308115005493,
+      "learning_rate": 4.615380621895701e-06,
+      "loss": 1.3275084495544434,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.31839582324028015,
+      "learning_rate": 4.607045308355141e-06,
+      "loss": 0.8152991533279419,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.28175586462020874,
+      "learning_rate": 4.598629272266549e-06,
+      "loss": 0.8213876485824585,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 0.5143842101097107,
+      "learning_rate": 4.590132879460458e-06,
+      "loss": 0.45163553953170776,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 0.9065913558006287,
+      "learning_rate": 4.5815564992603705e-06,
+      "loss": 0.8838123679161072,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.4655047357082367,
+      "learning_rate": 4.572900504466702e-06,
+      "loss": 1.1349900960922241,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.29092493653297424,
+      "learning_rate": 4.564165271340577e-06,
+      "loss": 1.2160018682479858,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 2.0495662689208984,
+      "learning_rate": 4.555351179587472e-06,
+      "loss": 0.9666154384613037,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.24971216917037964,
+      "learning_rate": 4.546458612340714e-06,
+      "loss": 0.9788669347763062,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 2.5510385036468506,
+      "learning_rate": 4.537487956144821e-06,
+      "loss": 0.9759644269943237,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.5234197974205017,
+      "learning_rate": 4.528439600938706e-06,
+      "loss": 0.935538113117218,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.26992306113243103,
+      "learning_rate": 4.519313940038721e-06,
+      "loss": 0.7100508213043213,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.540461540222168,
+      "learning_rate": 4.510111370121563e-06,
+      "loss": 1.173165202140808,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.3485596776008606,
+      "learning_rate": 4.500832291207033e-06,
+      "loss": 1.1555724143981934,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 1.3314050436019897,
+      "learning_rate": 4.491477106640646e-06,
+      "loss": 1.069810390472412,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.3811550438404083,
+      "learning_rate": 4.482046223076092e-06,
+      "loss": 1.0255157947540283,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.5846409201622009,
+      "learning_rate": 4.4725400504575725e-06,
+      "loss": 1.0753873586654663,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 0.8359457850456238,
+      "learning_rate": 4.462959002001971e-06,
+      "loss": 0.6470549702644348,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.2318486124277115,
+      "learning_rate": 4.453303494180895e-06,
+      "loss": 0.9214377999305725,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.3804703652858734,
+      "learning_rate": 4.443573946702573e-06,
+      "loss": 0.5500368475914001,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.30357447266578674,
+      "learning_rate": 4.433770782493606e-06,
+      "loss": 1.2407550811767578,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 0.8035040497779846,
+      "learning_rate": 4.42389442768059e-06,
+      "loss": 0.5976806282997131,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.3053663372993469,
+      "learning_rate": 4.413945311571586e-06,
+      "loss": 0.9198079109191895,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.4868881404399872,
+      "learning_rate": 4.40392386663747e-06,
+      "loss": 0.8197845220565796,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.3025258779525757,
+      "learning_rate": 4.393830528493118e-06,
+      "loss": 1.036171555519104,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.3065246641635895,
+      "learning_rate": 4.3836657358784875e-06,
+      "loss": 1.1696443557739258,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.39688268303871155,
+      "learning_rate": 4.373429930639534e-06,
+      "loss": 1.192521095275879,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.2639002799987793,
+      "learning_rate": 4.3631235577090085e-06,
+      "loss": 0.27921262383461,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.18626777827739716,
+      "learning_rate": 4.35274706508712e-06,
+      "loss": 0.7266547679901123,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 4.527546405792236,
+      "learning_rate": 4.342300903822057e-06,
+      "loss": 0.9568579196929932,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.2621271312236786,
+      "learning_rate": 4.331785527990384e-06,
+      "loss": 0.48270687460899353,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.29555654525756836,
+      "learning_rate": 4.3212013946773045e-06,
+      "loss": 1.1897711753845215,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.25580552220344543,
+      "learning_rate": 4.31054896395679e-06,
+      "loss": 1.2302889823913574,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.3190980553627014,
+      "learning_rate": 4.299828698871581e-06,
+      "loss": 1.353623867034912,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.550437331199646,
+      "learning_rate": 4.289041065413062e-06,
+      "loss": 0.6007822155952454,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.3428783416748047,
+      "learning_rate": 4.278186532501003e-06,
+      "loss": 0.9104017615318298,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.42507946491241455,
+      "learning_rate": 4.267265571963179e-06,
+      "loss": 0.5576983690261841,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.2930675446987152,
+      "learning_rate": 4.256278658514858e-06,
+      "loss": 1.1836248636245728,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 0.9647249579429626,
+      "learning_rate": 4.245226269738165e-06,
+      "loss": 1.0202300548553467,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.7804911732673645,
+      "learning_rate": 4.23410888606133e-06,
+      "loss": 1.1470812559127808,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.5448903441429138,
+      "learning_rate": 4.222926990737793e-06,
+      "loss": 1.053821086883545,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 0.6600915193557739,
+      "learning_rate": 4.211681069825207e-06,
+      "loss": 0.8763217926025391,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.643913745880127,
+      "learning_rate": 4.200371612164305e-06,
+      "loss": 0.7322060465812683,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.8507025241851807,
+      "learning_rate": 4.1889991093576555e-06,
+      "loss": 0.6315903663635254,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.5612606406211853,
+      "learning_rate": 4.177564055748284e-06,
+      "loss": 0.5154302716255188,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.4319795072078705,
+      "learning_rate": 4.166066948398199e-06,
+      "loss": 0.9969758987426758,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 1.2382147312164307,
+      "learning_rate": 4.154508287066772e-06,
+      "loss": 0.9225918650627136,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 1.0497831106185913,
+      "learning_rate": 4.142888574189025e-06,
+      "loss": 1.2936601638793945,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.3049759268760681,
+      "learning_rate": 4.131208314853778e-06,
+      "loss": 0.6818143725395203,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.34813225269317627,
+      "learning_rate": 4.119468016781707e-06,
+      "loss": 0.9008133411407471,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.4372757077217102,
+      "learning_rate": 4.107668190303266e-06,
+      "loss": 0.9360118508338928,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 1.0530791282653809,
+      "learning_rate": 4.095809348336506e-06,
+      "loss": 0.6509593725204468,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.8471210598945618,
+      "learning_rate": 4.083892006364777e-06,
+      "loss": 0.8736714124679565,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 0.5901526212692261,
+      "learning_rate": 4.071916682414327e-06,
+      "loss": 1.1537514925003052,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.34044766426086426,
+      "learning_rate": 4.059883897031777e-06,
+      "loss": 1.143200397491455,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.2364099770784378,
+      "learning_rate": 4.0477941732615e-06,
+      "loss": 1.1832212209701538,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.39826545119285583,
+      "learning_rate": 4.035648036622879e-06,
+      "loss": 1.1535452604293823,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 0.4491971731185913,
+      "learning_rate": 4.023446015087469e-06,
+      "loss": 0.7076215744018555,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.3007599711418152,
+      "learning_rate": 4.0111886390560446e-06,
+      "loss": 1.1757667064666748,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.24296985566616058,
+      "learning_rate": 3.998876441335544e-06,
+      "loss": 0.685849130153656,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.34658387303352356,
+      "learning_rate": 3.98650995711591e-06,
+      "loss": 1.1981104612350464,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.43268612027168274,
+      "learning_rate": 3.974089723946826e-06,
+      "loss": 1.1309443712234497,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.9115822315216064,
+      "learning_rate": 3.9616162817143465e-06,
+      "loss": 0.9174013733863831,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.36179691553115845,
+      "learning_rate": 3.949090172617437e-06,
+      "loss": 1.1577059030532837,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.48457685112953186,
+      "learning_rate": 3.936511941144397e-06,
+      "loss": 0.6025640964508057,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.25623995065689087,
+      "learning_rate": 3.9238821340491936e-06,
+      "loss": 0.7281185984611511,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.3597607910633087,
+      "learning_rate": 3.911201300327703e-06,
+      "loss": 1.0819116830825806,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.22393296658992767,
+      "learning_rate": 3.898469991193835e-06,
+      "loss": 1.1407146453857422,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.29782551527023315,
+      "learning_rate": 3.885688760055581e-06,
+      "loss": 0.6114221215248108,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.464743971824646,
+      "learning_rate": 3.872858162490956e-06,
+      "loss": 0.6897773146629333,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.5882709622383118,
+      "learning_rate": 3.859978756223845e-06,
+      "loss": 1.220623254776001,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.4881476163864136,
+      "learning_rate": 3.847051101099768e-06,
+      "loss": 1.1533688306808472,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.25827768445014954,
+      "learning_rate": 3.834075759061535e-06,
+      "loss": 1.1448545455932617,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 0.9065626859664917,
+      "learning_rate": 3.8210532941248255e-06,
+      "loss": 0.9347513914108276,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.3092288374900818,
+      "learning_rate": 3.8079842723536688e-06,
+      "loss": 1.13028883934021,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.3807051479816437,
+      "learning_rate": 3.7948692618358424e-06,
+      "loss": 0.7535625100135803,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.6893432140350342,
+      "learning_rate": 3.781708832658169e-06,
+      "loss": 0.7608675360679626,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 0.5436515808105469,
+      "learning_rate": 3.7685035568817492e-06,
+      "loss": 1.038297414779663,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 0.42842552065849304,
+      "learning_rate": 3.7552540085170826e-06,
+      "loss": 1.154944896697998,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 0.5742346048355103,
+      "learning_rate": 3.7419607634991235e-06,
+      "loss": 0.5783440470695496,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.3722529709339142,
+      "learning_rate": 3.7286243996622432e-06,
+      "loss": 0.5923794507980347,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 1.5448358058929443,
+      "learning_rate": 3.715245496715116e-06,
+      "loss": 1.1293904781341553,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.48172831535339355,
+      "learning_rate": 3.7018246362155134e-06,
+      "loss": 1.0739167928695679,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.37734299898147583,
+      "learning_rate": 3.6883624015450344e-06,
+      "loss": 1.1407253742218018,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.3849587142467499,
+      "learning_rate": 3.6748593778837367e-06,
+      "loss": 0.6090486645698547,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.3636282682418823,
+      "learning_rate": 3.661316152184708e-06,
+      "loss": 1.125139594078064,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.2964485287666321,
+      "learning_rate": 3.6477333131485483e-06,
+      "loss": 1.2747722864151,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 1.8840792179107666,
+      "learning_rate": 3.634111451197781e-06,
+      "loss": 0.7516986131668091,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.33791375160217285,
+      "learning_rate": 3.6204511584511865e-06,
+      "loss": 0.5845279693603516,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.4675881266593933,
+      "learning_rate": 3.606753028698067e-06,
+      "loss": 0.7826358675956726,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.44286930561065674,
+      "learning_rate": 3.593017657372436e-06,
+      "loss": 0.5201093554496765,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.2707214951515198,
+      "learning_rate": 3.5792456415271288e-06,
+      "loss": 0.851132869720459,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.2824823558330536,
+      "learning_rate": 3.565437579807858e-06,
+      "loss": 0.7483075857162476,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.5290979743003845,
+      "learning_rate": 3.551594072427188e-06,
+      "loss": 1.004777193069458,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.24640345573425293,
+      "learning_rate": 3.537715721138444e-06,
+      "loss": 1.1453437805175781,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.4027877151966095,
+      "learning_rate": 3.523803129209554e-06,
+      "loss": 1.034494400024414,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 1.150445818901062,
+      "learning_rate": 3.5098569013968314e-06,
+      "loss": 0.5956304669380188,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.40078258514404297,
+      "learning_rate": 3.49587764391868e-06,
+      "loss": 1.2166188955307007,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.314955472946167,
+      "learning_rate": 3.4818659644292473e-06,
+      "loss": 1.1487468481063843,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.5719289183616638,
+      "learning_rate": 3.4678224719920085e-06,
+      "loss": 0.7450273633003235,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 0.4806522727012634,
+      "learning_rate": 3.4537477770532933e-06,
+      "loss": 0.9380613565444946,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.7582850456237793,
+      "learning_rate": 3.4396424914157527e-06,
+      "loss": 0.6869011521339417,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 0.7151138782501221,
+      "learning_rate": 3.4255072282117574e-06,
+      "loss": 0.3773868978023529,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.4724314212799072,
+      "learning_rate": 3.411342601876754e-06,
+      "loss": 0.7629727125167847,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 5.844729900360107,
+      "learning_rate": 3.3971492281225554e-06,
+      "loss": 0.719204843044281,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.6904979348182678,
+      "learning_rate": 3.382927723910572e-06,
+      "loss": 1.4966500997543335,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.46376219391822815,
+      "learning_rate": 3.3686787074249966e-06,
+      "loss": 0.8824137449264526,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.3775095045566559,
+      "learning_rate": 3.354402798045935e-06,
+      "loss": 1.0556694269180298,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.8376180529594421,
+      "learning_rate": 3.340100616322479e-06,
+      "loss": 1.1164145469665527,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.2619885504245758,
+      "learning_rate": 3.3257727839457356e-06,
+      "loss": 0.8621953725814819,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.1829259842634201,
+      "learning_rate": 3.311419923721798e-06,
+      "loss": 0.5333537459373474,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 2.3281333446502686,
+      "learning_rate": 3.29704265954468e-06,
+      "loss": 0.9579188823699951,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 0.503736674785614,
+      "learning_rate": 3.282641616369193e-06,
+      "loss": 0.6507978439331055,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.579704761505127,
+      "learning_rate": 3.2682174201837773e-06,
+      "loss": 1.060945987701416,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.38510552048683167,
+      "learning_rate": 3.2537706979832955e-06,
+      "loss": 0.7646783590316772,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.38651424646377563,
+      "learning_rate": 3.23930207774178e-06,
+      "loss": 0.7947995066642761,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.23826530575752258,
+      "learning_rate": 3.2248121883851296e-06,
+      "loss": 0.11674472689628601,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.29104629158973694,
+      "learning_rate": 3.210301659763777e-06,
+      "loss": 1.1686278581619263,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 0.6639391183853149,
+      "learning_rate": 3.1957711226253073e-06,
+      "loss": 0.5482738018035889,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.46378275752067566,
+      "learning_rate": 3.181221208587041e-06,
+      "loss": 1.1552459001541138,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.4766862392425537,
+      "learning_rate": 3.16665255010858e-06,
+      "loss": 1.0234571695327759,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 0.6543024778366089,
+      "learning_rate": 3.152065780464311e-06,
+      "loss": 1.1956716775894165,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.27256646752357483,
+      "learning_rate": 3.1374615337158874e-06,
+      "loss": 0.9821120500564575,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.3706796169281006,
+      "learning_rate": 3.1228404446846577e-06,
+      "loss": 0.8647980093955994,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 0.4896516501903534,
+      "learning_rate": 3.1082031489240775e-06,
+      "loss": 0.6228200197219849,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.6764768362045288,
+      "learning_rate": 3.093550282692081e-06,
+      "loss": 0.7682955861091614,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.3454812467098236,
+      "learning_rate": 3.0788824829234206e-06,
+      "loss": 1.1249513626098633,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.5250853896141052,
+      "learning_rate": 3.064200387201987e-06,
+      "loss": 1.1730482578277588,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.21981775760650635,
+      "learning_rate": 3.0495046337330917e-06,
+      "loss": 1.1183021068572998,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.3227817714214325,
+      "learning_rate": 3.034795861315723e-06,
+      "loss": 1.1290582418441772,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.27887821197509766,
+      "learning_rate": 3.020074709314781e-06,
+      "loss": 0.795586347579956,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 3.773707151412964,
+      "learning_rate": 3.0053418176332844e-06,
+      "loss": 0.6430038213729858,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.2873072326183319,
+      "learning_rate": 2.9905978266845565e-06,
+      "loss": 1.0878937244415283,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 2.1222825050354004,
+      "learning_rate": 2.975843377364385e-06,
+      "loss": 0.8686354160308838,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.28823626041412354,
+      "learning_rate": 2.961079111023167e-06,
+      "loss": 1.1724278926849365,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.35042816400527954,
+      "learning_rate": 2.9463056694380263e-06,
+      "loss": 0.6018155217170715,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.20316095650196075,
+      "learning_rate": 2.9315236947849213e-06,
+      "loss": 0.6505711078643799,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.5584422945976257,
+      "learning_rate": 2.916733829610726e-06,
+      "loss": 0.6233350038528442,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.10072977095842361,
+      "learning_rate": 2.901936716805302e-06,
+      "loss": 0.5613674521446228,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.4069754183292389,
+      "learning_rate": 2.887132999573554e-06,
+      "loss": 0.8825414776802063,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.40306705236434937,
+      "learning_rate": 2.8723233214074675e-06,
+      "loss": 0.7677954435348511,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 0.5223978161811829,
+      "learning_rate": 2.857508326058142e-06,
+      "loss": 0.29404082894325256,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.5514689683914185,
+      "learning_rate": 2.842688657507805e-06,
+      "loss": 0.9230751991271973,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 0.5080574154853821,
+      "learning_rate": 2.8278649599418194e-06,
+      "loss": 0.5748634338378906,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.5595948100090027,
+      "learning_rate": 2.813037877720683e-06,
+      "loss": 0.872228741645813,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.33069849014282227,
+      "learning_rate": 2.798208055352017e-06,
+      "loss": 1.0266979932785034,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.6691774129867554,
+      "learning_rate": 2.7833761374625546e-06,
+      "loss": 0.9545827507972717,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.7659925222396851,
+      "learning_rate": 2.7685427687701172e-06,
+      "loss": 0.45812368392944336,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.34194955229759216,
+      "learning_rate": 2.753708594055588e-06,
+      "loss": 0.7933403253555298,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.18707995116710663,
+      "learning_rate": 2.7388742581348893e-06,
+      "loss": 0.6936202645301819,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.24901489913463593,
+      "learning_rate": 2.7240404058309487e-06,
+      "loss": 0.23291048407554626,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.3907551169395447,
+      "learning_rate": 2.709207681945673e-06,
+      "loss": 0.7976186275482178,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 1.126518726348877,
+      "learning_rate": 2.6943767312319167e-06,
+      "loss": 0.6725423336029053,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 0.49583446979522705,
+      "learning_rate": 2.679548198365462e-06,
+      "loss": 0.6779805421829224,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.9810777902603149,
+      "learning_rate": 2.6647227279169873e-06,
+      "loss": 0.9634326100349426,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.693539023399353,
+      "learning_rate": 2.6499009643240527e-06,
+      "loss": 0.9543125629425049,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.46630221605300903,
+      "learning_rate": 2.6350835518630907e-06,
+      "loss": 0.8142679333686829,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.6899120807647705,
+      "learning_rate": 2.620271134621397e-06,
+      "loss": 0.7164907455444336,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.38001900911331177,
+      "learning_rate": 2.605464356469131e-06,
+      "loss": 1.0059504508972168,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.3405975103378296,
+      "learning_rate": 2.5906638610313336e-06,
+      "loss": 1.0714585781097412,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.8568170666694641,
+      "learning_rate": 2.575870291659946e-06,
+      "loss": 0.7432149648666382,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 1.3127059936523438,
+      "learning_rate": 2.5610842914058446e-06,
+      "loss": 0.458040326833725,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.3774378299713135,
+      "learning_rate": 2.5463065029908906e-06,
+      "loss": 0.6683325171470642,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.5897888541221619,
+      "learning_rate": 2.5315375687799896e-06,
+      "loss": 0.866702675819397,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.4325333833694458,
+      "learning_rate": 2.5167781307531714e-06,
+      "loss": 0.9445369839668274,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.3828630745410919,
+      "learning_rate": 2.5020288304776822e-06,
+      "loss": 1.061798095703125,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 2.932656764984131,
+      "learning_rate": 2.487290309080097e-06,
+      "loss": 0.8735225200653076,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.3896097242832184,
+      "learning_rate": 2.472563207218454e-06,
+      "loss": 0.6837027072906494,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.253146231174469,
+      "learning_rate": 2.457848165054402e-06,
+      "loss": 0.8419521450996399,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.29789260029792786,
+      "learning_rate": 2.4431458222253755e-06,
+      "loss": 1.0405691862106323,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 1.392357349395752,
+      "learning_rate": 2.4284568178167923e-06,
+      "loss": 0.6272971034049988,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.28172406554222107,
+      "learning_rate": 2.413781790334272e-06,
+      "loss": 0.7470545768737793,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.29609525203704834,
+      "learning_rate": 2.3991213776758776e-06,
+      "loss": 0.25965481996536255,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.3925531208515167,
+      "learning_rate": 2.3844762171043975e-06,
+      "loss": 0.643407940864563,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.3190953731536865,
+      "learning_rate": 2.3698469452196328e-06,
+      "loss": 0.6364563703536987,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 0.5373165011405945,
+      "learning_rate": 2.355234197930735e-06,
+      "loss": 0.6536794900894165,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.6590198874473572,
+      "learning_rate": 2.340638610428554e-06,
+      "loss": 0.7957019209861755,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.3126554489135742,
+      "learning_rate": 2.3260608171580412e-06,
+      "loss": 0.6391379237174988,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.5144351720809937,
+      "learning_rate": 2.311501451790658e-06,
+      "loss": 0.8941548466682434,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.3901940584182739,
+      "learning_rate": 2.2969611471968386e-06,
+      "loss": 1.0340503454208374,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.2865762412548065,
+      "learning_rate": 2.2824405354184776e-06,
+      "loss": 1.0633257627487183,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.3029857873916626,
+      "learning_rate": 2.267940247641458e-06,
+      "loss": 0.7299340963363647,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.23658427596092224,
+      "learning_rate": 2.2534609141682135e-06,
+      "loss": 1.0125393867492676,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.28016626834869385,
+      "learning_rate": 2.2390031643903286e-06,
+      "loss": 0.7341445088386536,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 2.312579870223999,
+      "learning_rate": 2.224567626761185e-06,
+      "loss": 0.9467394948005676,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.41042718291282654,
+      "learning_rate": 2.2101549287686384e-06,
+      "loss": 1.046839952468872,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.19729259610176086,
+      "learning_rate": 2.195765696907745e-06,
+      "loss": 0.5880594253540039,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 2.2867958545684814,
+      "learning_rate": 2.1814005566535303e-06,
+      "loss": 0.22595608234405518,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.2568455636501312,
+      "learning_rate": 2.167060132433798e-06,
+      "loss": 0.6130040884017944,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.21212895214557648,
+      "learning_rate": 2.1527450476019895e-06,
+      "loss": 0.9618557691574097,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.2596210539340973,
+      "learning_rate": 2.138455924410085e-06,
+      "loss": 0.7713019847869873,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.32125142216682434,
+      "learning_rate": 2.1241933839815606e-06,
+      "loss": 0.5870032906532288,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.4021800458431244,
+      "learning_rate": 2.109958046284382e-06,
+      "loss": 0.9384274482727051,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.5727471113204956,
+      "learning_rate": 2.0957505301040604e-06,
+      "loss": 0.7279608249664307,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.6546733975410461,
+      "learning_rate": 2.081571453016756e-06,
+      "loss": 0.8627814650535583,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 8.880243301391602,
+      "learning_rate": 2.067421431362427e-06,
+      "loss": 0.8626641631126404,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.558358371257782,
+      "learning_rate": 2.053301080218044e-06,
+      "loss": 0.810906708240509,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 1.0155774354934692,
+      "learning_rate": 2.039211013370853e-06,
+      "loss": 0.671716034412384,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.5573283433914185,
+      "learning_rate": 2.025151843291693e-06,
+      "loss": 1.0284332036972046,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.4150453507900238,
+      "learning_rate": 2.0111241811083733e-06,
+      "loss": 0.5160080194473267,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 0.9103673100471497,
+      "learning_rate": 1.9971286365791092e-06,
+      "loss": 0.4943941533565521,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.4069664478302002,
+      "learning_rate": 1.9831658180660197e-06,
+      "loss": 1.0363861322402954,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 2.3980255126953125,
+      "learning_rate": 1.9692363325086774e-06,
+      "loss": 0.9750573635101318,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.7514590620994568,
+      "learning_rate": 1.95534078539773e-06,
+      "loss": 0.9939650297164917,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.21718713641166687,
+      "learning_rate": 1.9414797807485796e-06,
+      "loss": 0.7119104266166687,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.38140931725502014,
+      "learning_rate": 1.9276539210751295e-06,
+      "loss": 1.012832760810852,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.35234305262565613,
+      "learning_rate": 1.9138638073635886e-06,
+      "loss": 1.1279911994934082,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.4425659775733948,
+      "learning_rate": 1.90011003904635e-06,
+      "loss": 0.5873862504959106,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.558262050151825,
+      "learning_rate": 1.886393213975939e-06,
+      "loss": 1.0476487874984741,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.508423924446106,
+      "learning_rate": 1.8727139283990203e-06,
+      "loss": 0.8155238628387451,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.6962931156158447,
+      "learning_rate": 1.8590727769304789e-06,
+      "loss": 0.8759396076202393,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.34868013858795166,
+      "learning_rate": 1.8454703525275814e-06,
+      "loss": 0.6775892972946167,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.4656111001968384,
+      "learning_rate": 1.8319072464641912e-06,
+      "loss": 0.36424151062965393,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.6093098521232605,
+      "learning_rate": 1.8183840483050741e-06,
+      "loss": 0.5552644729614258,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.693278431892395,
+      "learning_rate": 1.8049013458802667e-06,
+      "loss": 0.9318357110023499,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 6.787789344787598,
+      "learning_rate": 1.7914597252595267e-06,
+      "loss": 1.0147026777267456,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.49275878071784973,
+      "learning_rate": 1.7780597707268582e-06,
+      "loss": 0.5891736745834351,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 2.4305739402770996,
+      "learning_rate": 1.7647020647551105e-06,
+      "loss": 0.9491831660270691,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.14006826281547546,
+      "learning_rate": 1.7513871879806605e-06,
+      "loss": 0.5863680839538574,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 1.8849611282348633,
+      "learning_rate": 1.7381157191781766e-06,
+      "loss": 0.7551034688949585,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.2770822048187256,
+      "learning_rate": 1.7248882352354571e-06,
+      "loss": 0.42556312680244446,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.28983649611473083,
+      "learning_rate": 1.7117053111283528e-06,
+      "loss": 1.0456041097640991,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 1.258420467376709,
+      "learning_rate": 1.6985675198957752e-06,
+      "loss": 0.7082871198654175,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.3922531008720398,
+      "learning_rate": 1.6854754326147916e-06,
+      "loss": 0.6282980442047119,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.35938650369644165,
+      "learning_rate": 1.6724296183757925e-06,
+      "loss": 1.0413717031478882,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.3763040602207184,
+      "learning_rate": 1.6594306442577629e-06,
+      "loss": 1.0706406831741333,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.39896076917648315,
+      "learning_rate": 1.6464790753036298e-06,
+      "loss": 0.7209835052490234,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 1.1747177839279175,
+      "learning_rate": 1.6335754744956935e-06,
+      "loss": 0.4854739010334015,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.34756121039390564,
+      "learning_rate": 1.6207204027311685e-06,
+      "loss": 1.055281162261963,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 5.085588455200195,
+      "learning_rate": 1.6079144187977925e-06,
+      "loss": 0.5201956033706665,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.30022647976875305,
+      "learning_rate": 1.595158079349542e-06,
+      "loss": 1.1175742149353027,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.7968214750289917,
+      "learning_rate": 1.5824519388824311e-06,
+      "loss": 0.8962827920913696,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.9491831660270691,
+      "learning_rate": 1.5697965497104117e-06,
+      "loss": 0.6911844611167908,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.35476648807525635,
+      "learning_rate": 1.557192461941367e-06,
+      "loss": 0.9178895354270935,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.4761117398738861,
+      "learning_rate": 1.5446402234531934e-06,
+      "loss": 1.0637744665145874,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.31293293833732605,
+      "learning_rate": 1.5321403798699924e-06,
+      "loss": 0.6461257934570312,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.5470765233039856,
+      "learning_rate": 1.5196934745383456e-06,
+      "loss": 1.2460167407989502,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.21187008917331696,
+      "learning_rate": 1.5073000485037044e-06,
+      "loss": 1.264428734779358,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.5381740927696228,
+      "learning_rate": 1.4949606404868638e-06,
+      "loss": 1.1104791164398193,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.6435518264770508,
+      "learning_rate": 1.4826757868605519e-06,
+      "loss": 0.5485662221908569,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.28011929988861084,
+      "learning_rate": 1.4704460216261125e-06,
+      "loss": 1.0793448686599731,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.3088745176792145,
+      "learning_rate": 1.458271876390287e-06,
+      "loss": 0.623894989490509,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.3380941152572632,
+      "learning_rate": 1.446153880342117e-06,
+      "loss": 1.0458285808563232,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 0.733073890209198,
+      "learning_rate": 1.4340925602299337e-06,
+      "loss": 0.6764969825744629,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.2842513918876648,
+      "learning_rate": 1.4220884403384656e-06,
+      "loss": 0.5900357961654663,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.4980064034461975,
+      "learning_rate": 1.4101420424660433e-06,
+      "loss": 0.7422754764556885,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.5263570547103882,
+      "learning_rate": 1.3982538859019219e-06,
+      "loss": 0.621476411819458,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.34509947896003723,
+      "learning_rate": 1.3864244874037096e-06,
+      "loss": 1.008946418762207,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.2824869751930237,
+      "learning_rate": 1.3746543611749007e-06,
+      "loss": 0.5831305384635925,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.35344237089157104,
+      "learning_rate": 1.3629440188425285e-06,
+      "loss": 0.9950648546218872,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.4040015935897827,
+      "learning_rate": 1.351293969434921e-06,
+      "loss": 0.4139801263809204,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.2865569591522217,
+      "learning_rate": 1.3397047193595809e-06,
+      "loss": 1.0677540302276611,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 1.1370466947555542,
+      "learning_rate": 1.3281767723811651e-06,
+      "loss": 0.43931376934051514,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 1.879893183708191,
+      "learning_rate": 1.3167106295995923e-06,
+      "loss": 0.8960666060447693,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.44126853346824646,
+      "learning_rate": 1.3053067894282624e-06,
+      "loss": 1.1758267879486084,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 11.338512420654297,
+      "learning_rate": 1.2939657475723844e-06,
+      "loss": 0.48371633887290955,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.27216267585754395,
+      "learning_rate": 1.2826879970074346e-06,
+      "loss": 0.9077948927879333,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.2819850444793701,
+      "learning_rate": 1.271474027957727e-06,
+      "loss": 0.8929274082183838,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.36070162057876587,
+      "learning_rate": 1.2603243278751043e-06,
+      "loss": 0.5501183271408081,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.3890959620475769,
+      "learning_rate": 1.249239381417746e-06,
+      "loss": 0.1415092945098877,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.25319811701774597,
+      "learning_rate": 1.238219670429104e-06,
+      "loss": 1.0322911739349365,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.24590496718883514,
+      "learning_rate": 1.2272656739169589e-06,
+      "loss": 0.5773763656616211,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.40709877014160156,
+      "learning_rate": 1.2163778680325948e-06,
+      "loss": 0.6369284987449646,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.22158946096897125,
+      "learning_rate": 1.2055567260501055e-06,
+      "loss": 0.3463606834411621,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.3506253957748413,
+      "learning_rate": 1.1948027183458182e-06,
+      "loss": 0.7111614346504211,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 0.6726877689361572,
+      "learning_rate": 1.18411631237785e-06,
+      "loss": 0.477986216545105,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.4653856158256531,
+      "learning_rate": 1.173497972665788e-06,
+      "loss": 0.9941408634185791,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.5946168899536133,
+      "learning_rate": 1.1629481607704964e-06,
+      "loss": 0.18817652761936188,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.32912683486938477,
+      "learning_rate": 1.1524673352740543e-06,
+      "loss": 0.5973628163337708,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.8259820938110352,
+      "learning_rate": 1.1420559517598206e-06,
+      "loss": 0.8266685009002686,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.3163120150566101,
+      "learning_rate": 1.13171446279263e-06,
+      "loss": 0.9833706021308899,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.8297820091247559,
+      "learning_rate": 1.1214433178991258e-06,
+      "loss": 0.8756715059280396,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.25764045119285583,
+      "learning_rate": 1.111242963548212e-06,
+      "loss": 0.5852322578430176,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.4293316602706909,
+      "learning_rate": 1.1011138431316538e-06,
+      "loss": 0.61106276512146,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.2946054935455322,
+      "learning_rate": 1.0910563969447969e-06,
+      "loss": 0.9739710092544556,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.15605485439300537,
+      "learning_rate": 1.0810710621674355e-06,
+      "loss": 0.40483349561691284,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.8007962107658386,
+      "learning_rate": 1.0711582728448034e-06,
+      "loss": 1.0255299806594849,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.26464608311653137,
+      "learning_rate": 1.0613184598687088e-06,
+      "loss": 0.8855907320976257,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 1.0915319919586182,
+      "learning_rate": 1.0515520509588076e-06,
+      "loss": 0.6511921882629395,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.15582917630672455,
+      "learning_rate": 1.0418594706440022e-06,
+      "loss": 0.5405547618865967,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.9305787682533264,
+      "learning_rate": 1.0322411402439982e-06,
+      "loss": 0.7616873979568481,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.08206583559513092,
+      "learning_rate": 1.0226974778509846e-06,
+      "loss": 0.30544817447662354,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.39827874302864075,
+      "learning_rate": 1.0132288983114611e-06,
+      "loss": 0.6539348363876343,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 1.0456311702728271,
+      "learning_rate": 1.003835813208206e-06,
+      "loss": 0.5528361201286316,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 2.366809129714966,
+      "learning_rate": 9.945186308423832e-07,
+      "loss": 0.8501113653182983,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.20413146913051605,
+      "learning_rate": 9.852777562157983e-07,
+      "loss": 0.744661271572113,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 0.971601128578186,
+      "learning_rate": 9.7611359101329e-07,
+      "loss": 0.6784732341766357,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 1.0325769186019897,
+      "learning_rate": 9.670265335852725e-07,
+      "loss": 1.1338965892791748,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 0.8139323592185974,
+      "learning_rate": 9.580169789304152e-07,
+      "loss": 0.3782649040222168,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.3383195698261261,
+      "learning_rate": 9.490853186784801e-07,
+      "loss": 0.7245016694068909,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 0.9236894249916077,
+      "learning_rate": 9.402319410732911e-07,
+      "loss": 0.3917732834815979,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.4874762296676636,
+      "learning_rate": 9.314572309558617e-07,
+      "loss": 0.8519300818443298,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 1.5864741802215576,
+      "learning_rate": 9.227615697476674e-07,
+      "loss": 0.3443653881549835,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.247799351811409,
+      "learning_rate": 9.141453354340619e-07,
+      "loss": 0.9757431149482727,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 1.8981910943984985,
+      "learning_rate": 9.056089025478495e-07,
+      "loss": 0.5854407548904419,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 1.3040696382522583,
+      "learning_rate": 8.971526421530067e-07,
+      "loss": 0.7975965142250061,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 1.1990466117858887,
+      "learning_rate": 8.887769218285469e-07,
+      "loss": 0.5837783217430115,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.353559672832489,
+      "learning_rate": 8.804821056525498e-07,
+      "loss": 0.9902276992797852,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.7336722612380981,
+      "learning_rate": 8.722685541863284e-07,
+      "loss": 0.9774597883224487,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.710015058517456,
+      "learning_rate": 8.641366244587612e-07,
+      "loss": 0.9188675284385681,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 1.9346073865890503,
+      "learning_rate": 8.560866699507703e-07,
+      "loss": 0.5201735496520996,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 1.0097771883010864,
+      "learning_rate": 8.481190405799569e-07,
+      "loss": 0.3331214487552643,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.9304558038711548,
+      "learning_rate": 8.402340826853908e-07,
+      "loss": 0.5550777316093445,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 0.5220862030982971,
+      "learning_rate": 8.324321390125548e-07,
+      "loss": 0.786139190196991,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 2.224722385406494,
+      "learning_rate": 8.247135486984487e-07,
+      "loss": 0.2027917355298996,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.8310621976852417,
+      "learning_rate": 8.170786472568451e-07,
+      "loss": 0.7440991997718811,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 2.902597665786743,
+      "learning_rate": 8.095277665637061e-07,
+      "loss": 0.6892938613891602,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.3163703680038452,
+      "learning_rate": 8.020612348427572e-07,
+      "loss": 1.0154277086257935,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.28178870677948,
+      "learning_rate": 7.946793766512195e-07,
+      "loss": 0.9631707072257996,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 1.025127649307251,
+      "learning_rate": 7.873825128657034e-07,
+      "loss": 1.1313951015472412,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.3478102385997772,
+      "learning_rate": 7.801709606682573e-07,
+      "loss": 0.959987223148346,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 1.1878464221954346,
+      "learning_rate": 7.730450335325852e-07,
+      "loss": 0.572114109992981,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.4103242754936218,
+      "learning_rate": 7.660050412104153e-07,
+      "loss": 0.9260150790214539,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.35635825991630554,
+      "learning_rate": 7.5905128971804e-07,
+      "loss": 1.0890753269195557,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.43806806206703186,
+      "learning_rate": 7.521840813230098e-07,
+      "loss": 1.0328636169433594,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.3665009140968323,
+      "learning_rate": 7.454037145309986e-07,
+      "loss": 0.7346044778823853,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.8226467370986938,
+      "learning_rate": 7.38710484072826e-07,
+      "loss": 0.5700799226760864,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.47432562708854675,
+      "learning_rate": 7.321046808916427e-07,
+      "loss": 0.8751451969146729,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.40358126163482666,
+      "learning_rate": 7.255865921302902e-07,
+      "loss": 0.7393505573272705,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.48215898871421814,
+      "learning_rate": 7.191565011188149e-07,
+      "loss": 0.5352509617805481,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.12182939052581787,
+      "learning_rate": 7.128146873621533e-07,
+      "loss": 0.5865047574043274,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.17820458114147186,
+      "learning_rate": 7.065614265279819e-07,
+      "loss": 0.7133850455284119,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.3160170912742615,
+      "learning_rate": 7.00396990434734e-07,
+      "loss": 0.5455622673034668,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.351772665977478,
+      "learning_rate": 6.943216470397872e-07,
+      "loss": 0.5637160539627075,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.17558270692825317,
+      "learning_rate": 6.883356604278116e-07,
+      "loss": 0.42473936080932617,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.4387773275375366,
+      "learning_rate": 6.824392907992938e-07,
+      "loss": 0.6959361433982849,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 0.7810353636741638,
+      "learning_rate": 6.76632794459224e-07,
+      "loss": 0.28893864154815674,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 0.6736273169517517,
+      "learning_rate": 6.709164238059568e-07,
+      "loss": 0.7922133207321167,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.4203196167945862,
+      "learning_rate": 6.652904273202385e-07,
+      "loss": 0.95040363073349,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.3402877748012543,
+      "learning_rate": 6.597550495544071e-07,
+      "loss": 0.507298469543457,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.7668086886405945,
+      "learning_rate": 6.54310531121762e-07,
+      "loss": 0.9591726064682007,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.38040411472320557,
+      "learning_rate": 6.489571086861027e-07,
+      "loss": 1.0543078184127808,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.1934942752122879,
+      "learning_rate": 6.436950149514448e-07,
+      "loss": 0.248811274766922,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.31154578924179077,
+      "learning_rate": 6.385244786519036e-07,
+      "loss": 0.6420713067054749,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.3321472704410553,
+      "learning_rate": 6.334457245417491e-07,
+      "loss": 0.2751573324203491,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 3.3124876022338867,
+      "learning_rate": 6.284589733856403e-07,
+      "loss": 0.10058709979057312,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.8314961791038513,
+      "learning_rate": 6.235644419490253e-07,
+      "loss": 0.8496547937393188,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 1.726009726524353,
+      "learning_rate": 6.18762342988722e-07,
+      "loss": 0.6637760400772095,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.3086608350276947,
+      "learning_rate": 6.140528852436666e-07,
+      "loss": 0.9243069887161255,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 1.2038524150848389,
+      "learning_rate": 6.094362734258438e-07,
+      "loss": 1.0163464546203613,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.21963441371917725,
+      "learning_rate": 6.049127082113853e-07,
+      "loss": 0.04721667245030403,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.44886183738708496,
+      "learning_rate": 6.004823862318475e-07,
+      "loss": 0.850946843624115,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 2.2290444374084473,
+      "learning_rate": 5.961455000656656e-07,
+      "loss": 0.8006585836410522,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 2.3044376373291016,
+      "learning_rate": 5.919022382297808e-07,
+      "loss": 0.21657533943653107,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.33850589394569397,
+      "learning_rate": 5.877527851714479e-07,
+      "loss": 0.9726195931434631,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.38078171014785767,
+      "learning_rate": 5.836973212602144e-07,
+      "loss": 0.15247878432273865,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.8443194627761841,
+      "learning_rate": 5.797360227800832e-07,
+      "loss": 0.9894143342971802,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.8032966256141663,
+      "learning_rate": 5.758690619218493e-07,
+      "loss": 0.32026636600494385,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.21422258019447327,
+      "learning_rate": 5.720966067756131e-07,
+      "loss": 0.5714766979217529,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.19034920632839203,
+      "learning_rate": 5.684188213234763e-07,
+      "loss": 0.8764244914054871,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.3586026430130005,
+      "learning_rate": 5.648358654324119e-07,
+      "loss": 0.975957453250885,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.6390727758407593,
+      "learning_rate": 5.613478948473166e-07,
+      "loss": 0.5872451066970825,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.33791130781173706,
+      "learning_rate": 5.579550611842395e-07,
+      "loss": 0.8956487774848938,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 0.812251627445221,
+      "learning_rate": 5.546575119237923e-07,
+      "loss": 0.6741803288459778,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.45107942819595337,
+      "learning_rate": 5.514553904047394e-07,
+      "loss": 0.8398787975311279,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 6.392262935638428,
+      "learning_rate": 5.483488358177647e-07,
+      "loss": 0.7718819975852966,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.3152640759944916,
+      "learning_rate": 5.453379831994238e-07,
+      "loss": 0.5291160345077515,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.330775648355484,
+      "learning_rate": 5.424229634262735e-07,
+      "loss": 1.0034441947937012,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.4409928619861603,
+      "learning_rate": 5.396039032091826e-07,
+      "loss": 0.5757849812507629,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.26044970750808716,
+      "learning_rate": 5.368809250878228e-07,
+      "loss": 0.9730173945426941,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.30905112624168396,
+      "learning_rate": 5.342541474253438e-07,
+      "loss": 0.9819098711013794,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.32488954067230225,
+      "learning_rate": 5.317236844032285e-07,
+      "loss": 0.5931157469749451,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 1.3462376594543457,
+      "learning_rate": 5.292896460163282e-07,
+      "loss": 0.7901210188865662,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 1.3947076797485352,
+      "learning_rate": 5.269521380680823e-07,
+      "loss": 0.6023671627044678,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.5845421552658081,
+      "learning_rate": 5.247112621659186e-07,
+      "loss": 0.7232294082641602,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.7334157228469849,
+      "learning_rate": 5.225671157168378e-07,
+      "loss": 0.5757493376731873,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.6218270659446716,
+      "learning_rate": 5.205197919231773e-07,
+      "loss": 0.9987547397613525,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.5024552345275879,
+      "learning_rate": 5.185693797785624e-07,
+      "loss": 0.38397306203842163,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.3402678966522217,
+      "learning_rate": 5.167159640640358e-07,
+      "loss": 0.5714554786682129,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 1.142331600189209,
+      "learning_rate": 5.149596253443733e-07,
+      "loss": 0.7634201049804688,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.5436675548553467,
+      "learning_rate": 5.133004399645818e-07,
+      "loss": 0.5609928369522095,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.21825340390205383,
+      "learning_rate": 5.117384800465806e-07,
+      "loss": 0.44919031858444214,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.269243061542511,
+      "learning_rate": 5.102738134860656e-07,
+      "loss": 0.9820560812950134,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 1.8968443870544434,
+      "learning_rate": 5.089065039495598e-07,
+      "loss": 0.5496000051498413,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.2789445221424103,
+      "learning_rate": 5.076366108716438e-07,
+      "loss": 0.41555333137512207,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.47752612829208374,
+      "learning_rate": 5.064641894523737e-07,
+      "loss": 0.49113360047340393,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.4510054886341095,
+      "learning_rate": 5.053892906548811e-07,
+      "loss": 0.9245374202728271,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.8031210899353027,
+      "learning_rate": 5.044119612031576e-07,
+      "loss": 0.15530680119991302,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 0.6413305401802063,
+      "learning_rate": 5.035322435800248e-07,
+      "loss": 0.6123966574668884,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.22468869388103485,
+      "learning_rate": 5.027501760252857e-07,
+      "loss": 0.5892717242240906,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.6182029843330383,
+      "learning_rate": 5.02065792534065e-07,
+      "loss": 1.0206186771392822,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.27039653062820435,
+      "learning_rate": 5.014791228553294e-07,
+      "loss": 0.24477030336856842,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.5045114755630493,
+      "learning_rate": 5.00990192490595e-07,
+      "loss": 0.9143338799476624,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.23271803557872772,
+      "learning_rate": 5.005990226928196e-07,
+      "loss": 0.5732707381248474,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.9926494359970093,
+      "learning_rate": 5.003056304654774e-07,
+      "loss": 0.4782474935054779,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.42918381094932556,
+      "learning_rate": 5.001100285618214e-07,
+      "loss": 0.9882228970527649,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2043452113866806,
+      "learning_rate": 5.000122254843282e-07,
+      "loss": 0.3125912547111511,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 3.519329211045118e+18,
+      "train_loss": 0.888545433905317,
+      "train_runtime": 9327.0059,
+      "train_samples_per_second": 6.459,
+      "train_steps_per_second": 0.108
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.519329211045118e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2dae75b55731660e84e9bdaf07e0f5add551266c5c06e6965e9cc1e4dcced9a
+size 5649