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", 1e-6), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), 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", 1e-6), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_adamw_lr", 5e-6), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), 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": 7436.254647328101,
+    "epoch": 4.0,
+    "total_flos": 4.038502240003031e+18,
+    "train_loss": 1.0441685825645686,
+    "train_runtime": 10534.5209,
+    "train_samples_per_second": 5.718,
+    "train_steps_per_second": 0.095
+}

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 sha256:95783b4ce87743d0e226f998ca18589279c2869cd29ed336d238f112d60f5707
+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": 7436.254647328101,
+    "epoch": 4.0,
+    "total_flos": 4.038502240003031e+18,
+    "train_loss": 1.0441685825645686,
+    "train_runtime": 10534.5209,
+    "train_samples_per_second": 5.718,
+    "train_steps_per_second": 0.095
+}

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": 0.8425382375717163,
+      "learning_rate": 1.9607843137254902e-08,
+      "loss": 2.0922725200653076,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.803239107131958,
+      "learning_rate": 5.88235294117647e-08,
+      "loss": 2.010118246078491,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 2.5188472270965576,
+      "learning_rate": 9.80392156862745e-08,
+      "loss": 2.2071139812469482,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.8189099431037903,
+      "learning_rate": 1.3725490196078432e-07,
+      "loss": 1.9252610206604004,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 1.6914823055267334,
+      "learning_rate": 1.764705882352941e-07,
+      "loss": 1.6076843738555908,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 1.1839359998703003,
+      "learning_rate": 2.156862745098039e-07,
+      "loss": 1.8892102241516113,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 2.1418583393096924,
+      "learning_rate": 2.549019607843137e-07,
+      "loss": 2.9728989601135254,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 3.9288241863250732,
+      "learning_rate": 2.941176470588235e-07,
+      "loss": 5.759878635406494,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 1.1443700790405273,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": 2.0372538566589355,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.8676067590713501,
+      "learning_rate": 3.7254901960784315e-07,
+      "loss": 1.881750464439392,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 1.101248860359192,
+      "learning_rate": 4.117647058823529e-07,
+      "loss": 2.997101068496704,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 1.9547584056854248,
+      "learning_rate": 4.5098039215686274e-07,
+      "loss": 1.8748698234558105,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.5487688183784485,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 1.868337869644165,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 1.0396053791046143,
+      "learning_rate": 5.294117647058823e-07,
+      "loss": 1.952110767364502,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 2.0077109336853027,
+      "learning_rate": 5.686274509803921e-07,
+      "loss": 1.9521887302398682,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 1.8067870140075684,
+      "learning_rate": 6.078431372549019e-07,
+      "loss": 1.727762222290039,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 1.690971851348877,
+      "learning_rate": 6.470588235294117e-07,
+      "loss": 1.7112396955490112,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.4531589448451996,
+      "learning_rate": 6.862745098039216e-07,
+      "loss": 1.721122145652771,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 0.8850814700126648,
+      "learning_rate": 7.254901960784313e-07,
+      "loss": 1.4740442037582397,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.3963426649570465,
+      "learning_rate": 7.647058823529411e-07,
+      "loss": 1.6165574789047241,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 2.674665927886963,
+      "learning_rate": 8.03921568627451e-07,
+      "loss": 1.2724214792251587,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.38429734110832214,
+      "learning_rate": 8.431372549019608e-07,
+      "loss": 1.3460066318511963,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 1.1492419242858887,
+      "learning_rate": 8.823529411764705e-07,
+      "loss": 1.118135690689087,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 1.7088249921798706,
+      "learning_rate": 9.215686274509803e-07,
+      "loss": 1.3846871852874756,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.6297861933708191,
+      "learning_rate": 9.607843137254902e-07,
+      "loss": 1.5227962732315063,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.732559084892273,
+      "learning_rate": 1e-06,
+      "loss": 1.4811941385269165,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.7871214151382446,
+      "learning_rate": 9.999902196391085e-07,
+      "loss": 1.3609808683395386,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.7547661662101746,
+      "learning_rate": 9.999608789815699e-07,
+      "loss": 1.6412900686264038,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.6946903467178345,
+      "learning_rate": 9.999119793027712e-07,
+      "loss": 1.5298701524734497,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.4203775227069855,
+      "learning_rate": 9.998435227282974e-07,
+      "loss": 1.0508219003677368,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.7225731611251831,
+      "learning_rate": 9.997555122338371e-07,
+      "loss": 1.5216001272201538,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 1.30720055103302,
+      "learning_rate": 9.996479516450542e-07,
+      "loss": 1.2563985586166382,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.3728600740432739,
+      "learning_rate": 9.99520845637422e-07,
+      "loss": 1.418299674987793,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 1.6880732774734497,
+      "learning_rate": 9.993741997360183e-07,
+      "loss": 1.0281386375427246,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.7022993564605713,
+      "learning_rate": 9.992080203152872e-07,
+      "loss": 1.6319724321365356,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 1.1037302017211914,
+      "learning_rate": 9.990223145987605e-07,
+      "loss": 0.9391559958457947,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 1.4968682527542114,
+      "learning_rate": 9.988170906587447e-07,
+      "loss": 1.2250131368637085,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.5804729461669922,
+      "learning_rate": 9.985923574159694e-07,
+      "loss": 1.5532232522964478,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 3.3081233501434326,
+      "learning_rate": 9.983481246392002e-07,
+      "loss": 1.0145370960235596,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.4309065341949463,
+      "learning_rate": 9.980844029448134e-07,
+      "loss": 0.8697817325592041,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.316090852022171,
+      "learning_rate": 9.978012037963349e-07,
+      "loss": 1.3778233528137207,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.7420907616615295,
+      "learning_rate": 9.974985395039423e-07,
+      "loss": 0.9268777966499329,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.4883190989494324,
+      "learning_rate": 9.971764232239285e-07,
+      "loss": 1.3983604907989502,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 1.7797441482543945,
+      "learning_rate": 9.968348689581315e-07,
+      "loss": 0.9551000595092773,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.6802905797958374,
+      "learning_rate": 9.964738915533242e-07,
+      "loss": 1.4219110012054443,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 1.2425458431243896,
+      "learning_rate": 9.960935067005704e-07,
+      "loss": 1.196105718612671,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.4075991213321686,
+      "learning_rate": 9.956937309345418e-07,
+      "loss": 1.3398497104644775,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 0.8651424646377563,
+      "learning_rate": 9.952745816327997e-07,
+      "loss": 1.1787023544311523,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 0.7215142250061035,
+      "learning_rate": 9.94836077015039e-07,
+      "loss": 1.3582069873809814,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.5184227228164673,
+      "learning_rate": 9.943782361422975e-07,
+      "loss": 0.9418277144432068,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.7048274874687195,
+      "learning_rate": 9.939010789161259e-07,
+      "loss": 1.3647910356521606,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.49403125047683716,
+      "learning_rate": 9.934046260777237e-07,
+      "loss": 0.821587860584259,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 1.0906405448913574,
+      "learning_rate": 9.928888992070372e-07,
+      "loss": 0.779825747013092,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.9795541763305664,
+      "learning_rate": 9.923539207218217e-07,
+      "loss": 1.0549049377441406,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.44692325592041016,
+      "learning_rate": 9.91799713876667e-07,
+      "loss": 1.3694568872451782,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.5556872487068176,
+      "learning_rate": 9.912263027619856e-07,
+      "loss": 0.8702759742736816,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 1.4165143966674805,
+      "learning_rate": 9.906337123029674e-07,
+      "loss": 1.1946947574615479,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 1.068716049194336,
+      "learning_rate": 9.900219682584949e-07,
+      "loss": 1.374321460723877,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.5233946442604065,
+      "learning_rate": 9.893910972200236e-07,
+      "loss": 1.422488808631897,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.4229278266429901,
+      "learning_rate": 9.887411266104268e-07,
+      "loss": 1.3195555210113525,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.38519975543022156,
+      "learning_rate": 9.880720846828027e-07,
+      "loss": 0.9988375306129456,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 1.230621576309204,
+      "learning_rate": 9.873840005192465e-07,
+      "loss": 1.0740467309951782,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 1.4735175371170044,
+      "learning_rate": 9.866769040295874e-07,
+      "loss": 0.9878921508789062,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.6711484789848328,
+      "learning_rate": 9.859508259500868e-07,
+      "loss": 1.3353742361068726,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.4325737953186035,
+      "learning_rate": 9.85205797842103e-07,
+      "loss": 0.9991917014122009,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.45981335639953613,
+      "learning_rate": 9.844418520907198e-07,
+      "loss": 0.8318514227867126,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.6961105465888977,
+      "learning_rate": 9.836590219033375e-07,
+      "loss": 0.9709692001342773,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.7408419847488403,
+      "learning_rate": 9.828573413082303e-07,
+      "loss": 1.315981149673462,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.7233710885047913,
+      "learning_rate": 9.82036845153068e-07,
+      "loss": 0.9771102070808411,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.1119574308395386,
+      "learning_rate": 9.811975691033992e-07,
+      "loss": 1.133639931678772,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.5853333473205566,
+      "learning_rate": 9.80339549641102e-07,
+      "loss": 1.0645649433135986,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.4608076810836792,
+      "learning_rate": 9.79462824062799e-07,
+      "loss": 1.2968108654022217,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.9481902718544006,
+      "learning_rate": 9.785674304782349e-07,
+      "loss": 1.6632933616638184,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 0.6470722556114197,
+      "learning_rate": 9.776534078086203e-07,
+      "loss": 0.9678738117218018,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.8709889054298401,
+      "learning_rate": 9.767207957849396e-07,
+      "loss": 0.9982343316078186,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 1.292151927947998,
+      "learning_rate": 9.757696349462252e-07,
+      "loss": 1.002139925956726,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.9532095789909363,
+      "learning_rate": 9.747999666377932e-07,
+      "loss": 1.1617112159729004,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 2.9578909873962402,
+      "learning_rate": 9.73811833009449e-07,
+      "loss": 0.8545269966125488,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 0.5140585899353027,
+      "learning_rate": 9.728052770136516e-07,
+      "loss": 1.1517466306686401,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.6556440591812134,
+      "learning_rate": 9.7178034240365e-07,
+      "loss": 1.0348566770553589,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.3231322765350342,
+      "learning_rate": 9.707370737315792e-07,
+      "loss": 0.7647264003753662,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.4657403826713562,
+      "learning_rate": 9.696755163465239e-07,
+      "loss": 1.1963428258895874,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 0.5329248309135437,
+      "learning_rate": 9.685957163925477e-07,
+      "loss": 0.8025826215744019,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.3986462354660034,
+      "learning_rate": 9.674977208066878e-07,
+      "loss": 1.2882846593856812,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 1.5090875625610352,
+      "learning_rate": 9.663815773169129e-07,
+      "loss": 0.8809756636619568,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.38049158453941345,
+      "learning_rate": 9.652473344400506e-07,
+      "loss": 0.8575432300567627,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 1.3259919881820679,
+      "learning_rate": 9.64095041479677e-07,
+      "loss": 1.2869794368743896,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 1.4261211156845093,
+      "learning_rate": 9.629247485239747e-07,
+      "loss": 0.7611839771270752,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 0.9687036871910095,
+      "learning_rate": 9.617365064435542e-07,
+      "loss": 0.8994767069816589,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 1.4794148206710815,
+      "learning_rate": 9.605303668892442e-07,
+      "loss": 1.0612611770629883,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.2877807319164276,
+      "learning_rate": 9.593063822898449e-07,
+      "loss": 0.507901668548584,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.4127897322177887,
+      "learning_rate": 9.580646058498502e-07,
+      "loss": 1.3074630498886108,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.48773902654647827,
+      "learning_rate": 9.568050915471348e-07,
+      "loss": 1.2987581491470337,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 1.8735255002975464,
+      "learning_rate": 9.55527894130607e-07,
+      "loss": 1.1417641639709473,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.7664352059364319,
+      "learning_rate": 9.542330691178296e-07,
+      "loss": 0.880364716053009,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 1.3381983041763306,
+      "learning_rate": 9.529206727926069e-07,
+      "loss": 1.2473231554031372,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.6738878488540649,
+      "learning_rate": 9.515907622025374e-07,
+      "loss": 0.9005473256111145,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 0.933052122592926,
+      "learning_rate": 9.502433951565347e-07,
+      "loss": 0.6045718193054199,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 1.354889988899231,
+      "learning_rate": 9.48878630222314e-07,
+      "loss": 0.8135166168212891,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 0.6516144275665283,
+      "learning_rate": 9.474965267238469e-07,
+      "loss": 1.063294529914856,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.23903584480285645,
+      "learning_rate": 9.460971447387823e-07,
+      "loss": 0.9724904298782349,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.6304468512535095,
+      "learning_rate": 9.446805450958352e-07,
+      "loss": 1.56329345703125,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.4726199507713318,
+      "learning_rate": 9.432467893721418e-07,
+      "loss": 1.2850979566574097,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.5835171341896057,
+      "learning_rate": 9.417959398905845e-07,
+      "loss": 0.6242196559906006,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.46454572677612305,
+      "learning_rate": 9.40328059717081e-07,
+      "loss": 1.3238112926483154,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.4172280728816986,
+      "learning_rate": 9.388432126578437e-07,
+      "loss": 1.2836631536483765,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.3328154683113098,
+      "learning_rate": 9.373414632566065e-07,
+      "loss": 0.9749572277069092,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.7151919603347778,
+      "learning_rate": 9.35822876791819e-07,
+      "loss": 1.4677996635437012,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 1.3108896017074585,
+      "learning_rate": 9.342875192738082e-07,
+      "loss": 0.8482100367546082,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.5240399241447449,
+      "learning_rate": 9.327354574419107e-07,
+      "loss": 1.2989330291748047,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.6148662567138672,
+      "learning_rate": 9.311667587615701e-07,
+      "loss": 1.5430316925048828,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 1.194664478302002,
+      "learning_rate": 9.29581491421405e-07,
+      "loss": 0.7006524801254272,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.4551745057106018,
+      "learning_rate": 9.279797243302452e-07,
+      "loss": 1.0084503889083862,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 1.3066445589065552,
+      "learning_rate": 9.26361527114136e-07,
+      "loss": 1.1240051984786987,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 1.429513692855835,
+      "learning_rate": 9.247269701133124e-07,
+      "loss": 1.2385070323944092,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.9680597186088562,
+      "learning_rate": 9.2307612437914e-07,
+      "loss": 1.3867727518081665,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.4233076274394989,
+      "learning_rate": 9.214090616710282e-07,
+      "loss": 0.818595290184021,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.5137848854064941,
+      "learning_rate": 9.197258544533097e-07,
+      "loss": 0.843373715877533,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 0.5644413232803345,
+      "learning_rate": 9.180265758920915e-07,
+      "loss": 0.4330376088619232,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 1.0108284950256348,
+      "learning_rate": 9.16311299852074e-07,
+      "loss": 0.8451502919197083,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.5121577978134155,
+      "learning_rate": 9.145801008933403e-07,
+      "loss": 1.2013404369354248,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.44426608085632324,
+      "learning_rate": 9.128330542681153e-07,
+      "loss": 1.274540901184082,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 1.4995366334915161,
+      "learning_rate": 9.110702359174942e-07,
+      "loss": 0.9922599196434021,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.38720613718032837,
+      "learning_rate": 9.092917224681426e-07,
+      "loss": 1.0016952753067017,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.44578617811203003,
+      "learning_rate": 9.074975912289641e-07,
+      "loss": 1.0048573017120361,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 1.6150213479995728,
+      "learning_rate": 9.056879201877411e-07,
+      "loss": 0.9475053548812866,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.34185007214546204,
+      "learning_rate": 9.03862788007744e-07,
+      "loss": 0.7679318785667419,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.8040241003036499,
+      "learning_rate": 9.020222740243126e-07,
+      "loss": 1.2764217853546143,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.6910593509674072,
+      "learning_rate": 9.001664582414066e-07,
+      "loss": 1.2588551044464111,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.6398411393165588,
+      "learning_rate": 8.98295421328129e-07,
+      "loss": 1.2290822267532349,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.4308067262172699,
+      "learning_rate": 8.964092446152181e-07,
+      "loss": 1.1361215114593506,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.5899179577827454,
+      "learning_rate": 8.945080100915143e-07,
+      "loss": 1.161683440208435,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 1.2997397184371948,
+      "learning_rate": 8.925918004003941e-07,
+      "loss": 0.7155571579933167,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.29507866501808167,
+      "learning_rate": 8.90660698836179e-07,
+      "loss": 1.028916358947754,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 1.501222014427185,
+      "learning_rate": 8.887147893405145e-07,
+      "loss": 0.7137823700904846,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 4.791189670562744,
+      "learning_rate": 8.867541564987211e-07,
+      "loss": 1.3115404844284058,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 0.6978177428245544,
+      "learning_rate": 8.847788855361178e-07,
+      "loss": 0.6232706904411316,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.7418224811553955,
+      "learning_rate": 8.827890623143172e-07,
+      "loss": 1.0319249629974365,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.3864130675792694,
+      "learning_rate": 8.807847733274938e-07,
+      "loss": 0.9418960213661194,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.5017346143722534,
+      "learning_rate": 8.787661056986235e-07,
+      "loss": 1.2005664110183716,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.4901583194732666,
+      "learning_rate": 8.767331471756975e-07,
+      "loss": 1.3002249002456665,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.8549445867538452,
+      "learning_rate": 8.746859861279067e-07,
+      "loss": 1.2881137132644653,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.4619888365268707,
+      "learning_rate": 8.726247115418016e-07,
+      "loss": 0.32264673709869385,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.3754831850528717,
+      "learning_rate": 8.705494130174237e-07,
+      "loss": 0.8273866772651672,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 1.918715000152588,
+      "learning_rate": 8.684601807644112e-07,
+      "loss": 1.0249416828155518,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.557555079460144,
+      "learning_rate": 8.663571055980767e-07,
+      "loss": 0.8922566771507263,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.5217064023017883,
+      "learning_rate": 8.642402789354608e-07,
+      "loss": 1.287339448928833,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.4360116124153137,
+      "learning_rate": 8.62109792791358e-07,
+      "loss": 1.3090788125991821,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.4195604622364044,
+      "learning_rate": 8.599657397743161e-07,
+      "loss": 1.4944058656692505,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.8361397385597229,
+      "learning_rate": 8.578082130826123e-07,
+      "loss": 0.6945460438728333,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.5296726822853088,
+      "learning_rate": 8.556373065002005e-07,
+      "loss": 1.0002961158752441,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 1.0679765939712524,
+      "learning_rate": 8.534531143926357e-07,
+      "loss": 0.7207995057106018,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.4727639853954315,
+      "learning_rate": 8.512557317029714e-07,
+      "loss": 1.2678115367889404,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 1.296967625617981,
+      "learning_rate": 8.49045253947633e-07,
+      "loss": 1.0937469005584717,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.6083080768585205,
+      "learning_rate": 8.46821777212266e-07,
+      "loss": 1.3267154693603516,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.9509052634239197,
+      "learning_rate": 8.445853981475585e-07,
+      "loss": 1.127556324005127,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 0.8842674493789673,
+      "learning_rate": 8.423362139650413e-07,
+      "loss": 0.9595845341682434,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.6711766123771667,
+      "learning_rate": 8.40074322432861e-07,
+      "loss": 0.7757585644721985,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 9.768728256225586,
+      "learning_rate": 8.37799821871531e-07,
+      "loss": 0.6862208247184753,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 2.1323413848876953,
+      "learning_rate": 8.355128111496567e-07,
+      "loss": 0.656395673751831,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.5895070433616638,
+      "learning_rate": 8.332133896796397e-07,
+      "loss": 1.0738224983215332,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 3.6088438034057617,
+      "learning_rate": 8.309016574133543e-07,
+      "loss": 1.0024151802062988,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 0.5182656645774841,
+      "learning_rate": 8.285777148378048e-07,
+      "loss": 1.4279625415802002,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.4689061939716339,
+      "learning_rate": 8.262416629707555e-07,
+      "loss": 0.7772690057754517,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.3927735984325409,
+      "learning_rate": 8.238936033563414e-07,
+      "loss": 1.0451743602752686,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.7997499704360962,
+      "learning_rate": 8.215336380606531e-07,
+      "loss": 1.0469849109649658,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 0.7120677828788757,
+      "learning_rate": 8.19161869667301e-07,
+      "loss": 0.6978618502616882,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.7115157842636108,
+      "learning_rate": 8.167784012729554e-07,
+      "loss": 0.9717881679534912,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 0.45555925369262695,
+      "learning_rate": 8.143833364828654e-07,
+      "loss": 1.2749401330947876,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.4267056882381439,
+      "learning_rate": 8.119767794063552e-07,
+      "loss": 1.2420390844345093,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.3638407289981842,
+      "learning_rate": 8.095588346522998e-07,
+      "loss": 1.2725023031234741,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.45394954085350037,
+      "learning_rate": 8.071296073245757e-07,
+      "loss": 1.2465835809707642,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 1.3502168655395508,
+      "learning_rate": 8.046892030174937e-07,
+      "loss": 0.7819681167602539,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.5712032914161682,
+      "learning_rate": 8.022377278112088e-07,
+      "loss": 1.2594401836395264,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.6428905129432678,
+      "learning_rate": 7.997752882671088e-07,
+      "loss": 0.8081700801849365,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.41046562790870667,
+      "learning_rate": 7.973019914231819e-07,
+      "loss": 1.3009734153747559,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.8816495537757874,
+      "learning_rate": 7.94817944789365e-07,
+      "loss": 1.2258024215698242,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.6734359860420227,
+      "learning_rate": 7.923232563428692e-07,
+      "loss": 1.037970781326294,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.5581151247024536,
+      "learning_rate": 7.898180345234874e-07,
+      "loss": 1.2316817045211792,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.5218549370765686,
+      "learning_rate": 7.873023882288793e-07,
+      "loss": 0.6890713572502136,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.47731778025627136,
+      "learning_rate": 7.847764268098387e-07,
+      "loss": 0.8672411441802979,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.5206546783447266,
+      "learning_rate": 7.822402600655405e-07,
+      "loss": 1.1427444219589233,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.3739563822746277,
+      "learning_rate": 7.796939982387669e-07,
+      "loss": 1.2557984590530396,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.6617816090583801,
+      "learning_rate": 7.771377520111162e-07,
+      "loss": 0.6997984647750854,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.6152245402336121,
+      "learning_rate": 7.74571632498191e-07,
+      "loss": 0.8375219106674194,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.9147626161575317,
+      "learning_rate": 7.71995751244769e-07,
+      "loss": 1.316990613937378,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.5100573897361755,
+      "learning_rate": 7.694102202199536e-07,
+      "loss": 1.272545576095581,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.384264200925827,
+      "learning_rate": 7.668151518123069e-07,
+      "loss": 1.2523423433303833,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 0.8864626288414001,
+      "learning_rate": 7.64210658824965e-07,
+      "loss": 1.0491065979003906,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.49264994263648987,
+      "learning_rate": 7.615968544707337e-07,
+      "loss": 1.2036755084991455,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.39525148272514343,
+      "learning_rate": 7.589738523671684e-07,
+      "loss": 0.934528112411499,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.4015274941921234,
+      "learning_rate": 7.563417665316337e-07,
+      "loss": 0.9341564774513245,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 0.9333391785621643,
+      "learning_rate": 7.537007113763497e-07,
+      "loss": 1.1456751823425293,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 1.0230107307434082,
+      "learning_rate": 7.510508017034164e-07,
+      "loss": 1.2499125003814697,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 1.1917494535446167,
+      "learning_rate": 7.483921526998246e-07,
+      "loss": 0.6517433524131775,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.5881509184837341,
+      "learning_rate": 7.457248799324485e-07,
+      "loss": 0.6578674912452698,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 0.8851764798164368,
+      "learning_rate": 7.430490993430231e-07,
+      "loss": 1.2653934955596924,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.7265904545783997,
+      "learning_rate": 7.403649272431027e-07,
+      "loss": 1.1638611555099487,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.3751894235610962,
+      "learning_rate": 7.376724803090068e-07,
+      "loss": 1.2272953987121582,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.626318097114563,
+      "learning_rate": 7.349718755767473e-07,
+      "loss": 0.6993715167045593,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.36690616607666016,
+      "learning_rate": 7.322632304369415e-07,
+      "loss": 1.2126752138137817,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.5017926096916199,
+      "learning_rate": 7.295466626297095e-07,
+      "loss": 1.373421311378479,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 0.7983066439628601,
+      "learning_rate": 7.268222902395561e-07,
+      "loss": 0.8057880997657776,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 1.127365231513977,
+      "learning_rate": 7.240902316902372e-07,
+      "loss": 0.6592608094215393,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.4174342155456543,
+      "learning_rate": 7.213506057396133e-07,
+      "loss": 0.9217919111251831,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.6697507500648499,
+      "learning_rate": 7.186035314744871e-07,
+      "loss": 0.6113168001174927,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.34020891785621643,
+      "learning_rate": 7.158491283054257e-07,
+      "loss": 0.964951753616333,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.7765717506408691,
+      "learning_rate": 7.130875159615714e-07,
+      "loss": 0.8281651735305786,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.6381047368049622,
+      "learning_rate": 7.103188144854375e-07,
+      "loss": 1.1599037647247314,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.45585206151008606,
+      "learning_rate": 7.075431442276887e-07,
+      "loss": 1.2157034873962402,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.9851412773132324,
+      "learning_rate": 7.047606258419107e-07,
+      "loss": 1.1312099695205688,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 5.436736106872559,
+      "learning_rate": 7.019713802793662e-07,
+      "loss": 0.6995736956596375,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.4824753403663635,
+      "learning_rate": 6.991755287837358e-07,
+      "loss": 1.3122954368591309,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.35564494132995605,
+      "learning_rate": 6.963731928858494e-07,
+      "loss": 1.246601939201355,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.6360551118850708,
+      "learning_rate": 6.935644943984016e-07,
+      "loss": 0.8579744696617126,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 1.4960342645645142,
+      "learning_rate": 6.907495554106586e-07,
+      "loss": 1.1617209911346436,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.6668098568916321,
+      "learning_rate": 6.879284982831505e-07,
+      "loss": 0.7668188214302063,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 0.7893583178520203,
+      "learning_rate": 6.851014456423514e-07,
+      "loss": 0.4187908172607422,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 1.4070234298706055,
+      "learning_rate": 6.822685203753508e-07,
+      "loss": 0.83896404504776,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 2.736166000366211,
+      "learning_rate": 6.79429845624511e-07,
+      "loss": 0.7182890772819519,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 0.5745601654052734,
+      "learning_rate": 6.765855447821143e-07,
+      "loss": 1.679482340812683,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.42770564556121826,
+      "learning_rate": 6.737357414849993e-07,
+      "loss": 0.9792550802230835,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.37946099042892456,
+      "learning_rate": 6.708805596091868e-07,
+      "loss": 1.1162333488464355,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.5678144097328186,
+      "learning_rate": 6.680201232644957e-07,
+      "loss": 1.1947766542434692,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.3277341425418854,
+      "learning_rate": 6.65154556789147e-07,
+      "loss": 0.943204939365387,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.29292938113212585,
+      "learning_rate": 6.622839847443596e-07,
+      "loss": 0.6688568592071533,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 1.0823400020599365,
+      "learning_rate": 6.594085319089359e-07,
+      "loss": 1.0446151494979858,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 2.08892560005188,
+      "learning_rate": 6.565283232738386e-07,
+      "loss": 0.7744835615158081,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.5839812755584717,
+      "learning_rate": 6.536434840367554e-07,
+      "loss": 1.1386644840240479,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.6158760190010071,
+      "learning_rate": 6.507541395966591e-07,
+      "loss": 0.7886677384376526,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.4198272228240967,
+      "learning_rate": 6.47860415548356e-07,
+      "loss": 0.8355951309204102,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.3009353578090668,
+      "learning_rate": 6.449624376770259e-07,
+      "loss": 0.14103345572948456,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.4450838565826416,
+      "learning_rate": 6.420603319527553e-07,
+      "loss": 1.2537599802017212,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 1.1065186262130737,
+      "learning_rate": 6.391542245250615e-07,
+      "loss": 0.621747612953186,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 1.7257264852523804,
+      "learning_rate": 6.362442417174081e-07,
+      "loss": 1.2340410947799683,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.6690614223480225,
+      "learning_rate": 6.333305100217159e-07,
+      "loss": 1.0991343259811401,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 0.5866219401359558,
+      "learning_rate": 6.304131560928621e-07,
+      "loss": 1.3845065832138062,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 1.2586193084716797,
+      "learning_rate": 6.274923067431775e-07,
+      "loss": 1.0518014430999756,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.5666131377220154,
+      "learning_rate": 6.245680889369315e-07,
+      "loss": 0.9778220653533936,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 1.1292035579681396,
+      "learning_rate": 6.216406297848154e-07,
+      "loss": 0.6951173543930054,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 0.7351351976394653,
+      "learning_rate": 6.187100565384161e-07,
+      "loss": 0.7900633811950684,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.7023295164108276,
+      "learning_rate": 6.15776496584684e-07,
+      "loss": 1.2109326124191284,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.5147722363471985,
+      "learning_rate": 6.128400774403973e-07,
+      "loss": 1.3600739240646362,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.4234119653701782,
+      "learning_rate": 6.099009267466183e-07,
+      "loss": 1.2092632055282593,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.3030948340892792,
+      "learning_rate": 6.069591722631445e-07,
+      "loss": 1.2274162769317627,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.43695008754730225,
+      "learning_rate": 6.04014941862956e-07,
+      "loss": 0.9162517786026001,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 2.7067582607269287,
+      "learning_rate": 6.010683635266568e-07,
+      "loss": 0.5965803265571594,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.7394753694534302,
+      "learning_rate": 5.981195653369112e-07,
+      "loss": 1.1902916431427002,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 2.5571165084838867,
+      "learning_rate": 5.95168675472877e-07,
+      "loss": 0.9461820125579834,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.7218238115310669,
+      "learning_rate": 5.922158222046332e-07,
+      "loss": 1.2422934770584106,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.271190881729126,
+      "learning_rate": 5.892611338876052e-07,
+      "loss": 0.7214773297309875,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.528483510017395,
+      "learning_rate": 5.863047389569842e-07,
+      "loss": 0.8567409515380859,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 1.0696465969085693,
+      "learning_rate": 5.833467659221452e-07,
+      "loss": 0.821619987487793,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.16100502014160156,
+      "learning_rate": 5.803873433610603e-07,
+      "loss": 0.652566134929657,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.40720587968826294,
+      "learning_rate": 5.774265999147108e-07,
+      "loss": 1.0592925548553467,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.7444103360176086,
+      "learning_rate": 5.744646642814934e-07,
+      "loss": 0.9694873094558716,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 3.214693307876587,
+      "learning_rate": 5.715016652116284e-07,
+      "loss": 0.31219401955604553,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.4227757453918457,
+      "learning_rate": 5.685377315015609e-07,
+      "loss": 1.039976716041565,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 0.8102324604988098,
+      "learning_rate": 5.655729919883639e-07,
+      "loss": 0.6844033598899841,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.7231845855712891,
+      "learning_rate": 5.626075755441365e-07,
+      "loss": 1.0793242454528809,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.5198511481285095,
+      "learning_rate": 5.596416110704033e-07,
+      "loss": 1.1956276893615723,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.7063606381416321,
+      "learning_rate": 5.566752274925108e-07,
+      "loss": 1.2138276100158691,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.5600472688674927,
+      "learning_rate": 5.537085537540233e-07,
+      "loss": 0.5733051300048828,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.43647804856300354,
+      "learning_rate": 5.507417188111175e-07,
+      "loss": 1.2009754180908203,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.6653239130973816,
+      "learning_rate": 5.477748516269778e-07,
+      "loss": 0.9144095182418823,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.3898330628871918,
+      "learning_rate": 5.448080811661897e-07,
+      "loss": 0.614565908908844,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.6056618690490723,
+      "learning_rate": 5.418415363891345e-07,
+      "loss": 1.0166387557983398,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 1.1628928184509277,
+      "learning_rate": 5.388753462463833e-07,
+      "loss": 0.7345898747444153,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 0.5768036842346191,
+      "learning_rate": 5.359096396730923e-07,
+      "loss": 0.7580212950706482,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.6307709217071533,
+      "learning_rate": 5.329445455833974e-07,
+      "loss": 1.1226731538772583,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.43720167875289917,
+      "learning_rate": 5.299801928648104e-07,
+      "loss": 1.1661862134933472,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.712365984916687,
+      "learning_rate": 5.270167103726181e-07,
+      "loss": 1.2349671125411987,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.5683341026306152,
+      "learning_rate": 5.240542269242793e-07,
+      "loss": 0.8632749319076538,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.5119061470031738,
+      "learning_rate": 5.210928712938261e-07,
+      "loss": 1.187488079071045,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.5245291590690613,
+      "learning_rate": 5.181327722062666e-07,
+      "loss": 1.244426965713501,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.5254369378089905,
+      "learning_rate": 5.151740583319891e-07,
+      "loss": 0.981595516204834,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 0.9996938109397888,
+      "learning_rate": 5.122168582811689e-07,
+      "loss": 0.5546106696128845,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.478613942861557,
+      "learning_rate": 5.092613005981781e-07,
+      "loss": 0.7716253995895386,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 1.2469408512115479,
+      "learning_rate": 5.063075137559979e-07,
+      "loss": 0.9730294346809387,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.4390407204627991,
+      "learning_rate": 5.033556261506342e-07,
+      "loss": 1.1022344827651978,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 1.3295910358428955,
+      "learning_rate": 5.004057660955364e-07,
+      "loss": 1.1952877044677734,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 1.2258836030960083,
+      "learning_rate": 4.974580618160194e-07,
+      "loss": 1.0791444778442383,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.5492210984230042,
+      "learning_rate": 4.945126414436908e-07,
+      "loss": 0.8663148880004883,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.38257676362991333,
+      "learning_rate": 4.915696330108804e-07,
+      "loss": 0.954747200012207,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.3830359876155853,
+      "learning_rate": 4.88629164445075e-07,
+      "loss": 1.1951080560684204,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 0.8451001644134521,
+      "learning_rate": 4.856913635633584e-07,
+      "loss": 0.7459049224853516,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.33643773198127747,
+      "learning_rate": 4.827563580668543e-07,
+      "loss": 0.9163620471954346,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.46159231662750244,
+      "learning_rate": 4.798242755351754e-07,
+      "loss": 0.37290844321250916,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 2.699605703353882,
+      "learning_rate": 4.768952434208794e-07,
+      "loss": 0.7714478969573975,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.5102899074554443,
+      "learning_rate": 4.7396938904392645e-07,
+      "loss": 0.7445698976516724,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 0.8967058658599854,
+      "learning_rate": 4.710468395861469e-07,
+      "loss": 0.8099274635314941,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.4165751338005066,
+      "learning_rate": 4.6812772208571073e-07,
+      "loss": 1.0066813230514526,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.6290020942687988,
+      "learning_rate": 4.6521216343160816e-07,
+      "loss": 0.7436711192131042,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.5180162787437439,
+      "learning_rate": 4.6230029035813156e-07,
+      "loss": 1.0371257066726685,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.6652600169181824,
+      "learning_rate": 4.593922294393676e-07,
+      "loss": 1.180807113647461,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.515034019947052,
+      "learning_rate": 4.5648810708369543e-07,
+      "loss": 1.2112433910369873,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 1.6332746744155884,
+      "learning_rate": 4.5358804952829153e-07,
+      "loss": 1.0091413259506226,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.4357254207134247,
+      "learning_rate": 4.506921828336426e-07,
+      "loss": 1.2268762588500977,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.45544084906578064,
+      "learning_rate": 4.478006328780657e-07,
+      "loss": 0.9058987498283386,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 1.723895788192749,
+      "learning_rate": 4.449135253522369e-07,
+      "loss": 1.1184284687042236,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.42880570888519287,
+      "learning_rate": 4.420309857537277e-07,
+      "loss": 1.1904247999191284,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.25226956605911255,
+      "learning_rate": 4.3915313938154897e-07,
+      "loss": 0.713387668132782,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 0.9065414071083069,
+      "learning_rate": 4.36280111330706e-07,
+      "loss": 0.31334733963012695,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.252196341753006,
+      "learning_rate": 4.3341202648675956e-07,
+      "loss": 0.8130457997322083,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.19010767340660095,
+      "learning_rate": 4.3054900952039783e-07,
+      "loss": 1.1392337083816528,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.4542621076107025,
+      "learning_rate": 4.2769118488201694e-07,
+      "loss": 0.9002693891525269,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.5231760740280151,
+      "learning_rate": 4.2483867679631203e-07,
+      "loss": 0.6867619156837463,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.6279378533363342,
+      "learning_rate": 4.219916092568764e-07,
+      "loss": 1.215380311012268,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.607097327709198,
+      "learning_rate": 4.1915010602081203e-07,
+      "loss": 0.9204408526420593,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.9494768977165222,
+      "learning_rate": 4.163142906033511e-07,
+      "loss": 1.07428777217865,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.8995746374130249,
+      "learning_rate": 4.134842862724853e-07,
+      "loss": 1.0317769050598145,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.5626697540283203,
+      "learning_rate": 4.106602160436088e-07,
+      "loss": 0.9463708400726318,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.5927757024765015,
+      "learning_rate": 4.078422026741706e-07,
+      "loss": 0.9035646319389343,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.5040337443351746,
+      "learning_rate": 4.050303686583386e-07,
+      "loss": 1.1803009510040283,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.49055758118629456,
+      "learning_rate": 4.022248362216746e-07,
+      "loss": 0.7859451770782471,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 1.2180066108703613,
+      "learning_rate": 3.9942572731582183e-07,
+      "loss": 0.6770949959754944,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.4659102261066437,
+      "learning_rate": 3.9663316361320385e-07,
+      "loss": 1.176295280456543,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.576202392578125,
+      "learning_rate": 3.938472665017354e-07,
+      "loss": 1.126934289932251,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.5327000617980957,
+      "learning_rate": 3.91068157079546e-07,
+      "loss": 1.1344680786132812,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.46229591965675354,
+      "learning_rate": 3.882959561497159e-07,
+      "loss": 1.0722061395645142,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 0.8892953991889954,
+      "learning_rate": 3.8553078421502585e-07,
+      "loss": 1.1972440481185913,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.36197274923324585,
+      "learning_rate": 3.8277276147271764e-07,
+      "loss": 1.4118647575378418,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.7987051010131836,
+      "learning_rate": 3.8002200780926995e-07,
+      "loss": 0.6525497436523438,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.5063230991363525,
+      "learning_rate": 3.7727864279518775e-07,
+      "loss": 1.218778371810913,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.5883747935295105,
+      "learning_rate": 3.7454278567980403e-07,
+      "loss": 1.010594129562378,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 0.5669489502906799,
+      "learning_rate": 3.718145553860957e-07,
+      "loss": 1.1629045009613037,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 1.0424724817276,
+      "learning_rate": 3.690940705055162e-07,
+      "loss": 0.7290706038475037,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.4018857479095459,
+      "learning_rate": 3.6638144929283823e-07,
+      "loss": 0.5022093057632446,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.5531585216522217,
+      "learning_rate": 3.636768096610148e-07,
+      "loss": 0.6555119752883911,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.5517063140869141,
+      "learning_rate": 3.609802691760533e-07,
+      "loss": 1.0602622032165527,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 0.9075046181678772,
+      "learning_rate": 3.582919450519053e-07,
+      "loss": 1.372468113899231,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.7803366184234619,
+      "learning_rate": 3.556119541453716e-07,
+      "loss": 0.7533060908317566,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.8778672218322754,
+      "learning_rate": 3.5294041295102207e-07,
+      "loss": 1.1376829147338867,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.26959022879600525,
+      "learning_rate": 3.5027743759613205e-07,
+      "loss": 0.637627363204956,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 1.0191822052001953,
+      "learning_rate": 3.476231438356353e-07,
+      "loss": 0.890547513961792,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.4155731797218323,
+      "learning_rate": 3.449776470470914e-07,
+      "loss": 0.6647993922233582,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.36803483963012695,
+      "learning_rate": 3.423410622256705e-07,
+      "loss": 1.1900908946990967,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 1.4192544221878052,
+      "learning_rate": 3.39713503979155e-07,
+      "loss": 0.9067127704620361,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.6935244798660278,
+      "learning_rate": 3.370950865229583e-07,
+      "loss": 0.7337151765823364,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.35887786746025085,
+      "learning_rate": 3.3448592367515843e-07,
+      "loss": 1.1841096878051758,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 1.2133479118347168,
+      "learning_rate": 3.3188612885155253e-07,
+      "loss": 1.2082436084747314,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.38288289308547974,
+      "learning_rate": 3.292958150607259e-07,
+      "loss": 0.8327260613441467,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 0.6938713788986206,
+      "learning_rate": 3.267150948991387e-07,
+      "loss": 0.6335185766220093,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.385933518409729,
+      "learning_rate": 3.241440805462337e-07,
+      "loss": 1.2009639739990234,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.789777934551239,
+      "learning_rate": 3.2158288375955845e-07,
+      "loss": 0.6830637454986572,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.6893877983093262,
+      "learning_rate": 3.1903161586990834e-07,
+      "loss": 1.2352404594421387,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.8104536533355713,
+      "learning_rate": 3.164903877764862e-07,
+      "loss": 1.032488465309143,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.7075826525688171,
+      "learning_rate": 3.139593099420823e-07,
+      "loss": 0.9938370585441589,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.539611279964447,
+      "learning_rate": 3.1143849238827336e-07,
+      "loss": 1.1270115375518799,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.5532920956611633,
+      "learning_rate": 3.0892804469063864e-07,
+      "loss": 1.190551996231079,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.8707150220870972,
+      "learning_rate": 3.0642807597399845e-07,
+      "loss": 0.8870285749435425,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.6733996868133545,
+      "learning_rate": 3.039386949076691e-07,
+      "loss": 1.4412972927093506,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.2842189073562622,
+      "learning_rate": 3.014600097007408e-07,
+      "loss": 1.4471794366836548,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.7518956661224365,
+      "learning_rate": 2.9899212809737273e-07,
+      "loss": 1.2614086866378784,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.39055633544921875,
+      "learning_rate": 2.965351573721103e-07,
+      "loss": 0.6366642117500305,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.5946579575538635,
+      "learning_rate": 2.9408920432522246e-07,
+      "loss": 1.2139933109283447,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.4142098128795624,
+      "learning_rate": 2.9165437527805734e-07,
+      "loss": 0.7183558940887451,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.3885643184185028,
+      "learning_rate": 2.892307760684234e-07,
+      "loss": 1.190184473991394,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 2.370802879333496,
+      "learning_rate": 2.868185120459867e-07,
+      "loss": 0.7851237654685974,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.4130178987979889,
+      "learning_rate": 2.8441768806769305e-07,
+      "loss": 0.7491707801818848,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.5927749276161194,
+      "learning_rate": 2.8202840849320864e-07,
+      "loss": 1.0182961225509644,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.3625214397907257,
+      "learning_rate": 2.796507771803843e-07,
+      "loss": 0.7264545559883118,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.3499440848827362,
+      "learning_rate": 2.772848974807419e-07,
+      "loss": 1.1779705286026,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.493435263633728,
+      "learning_rate": 2.7493087223498015e-07,
+      "loss": 0.686111330986023,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.4593299925327301,
+      "learning_rate": 2.7258880376850567e-07,
+      "loss": 1.1069648265838623,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.5402621626853943,
+      "learning_rate": 2.702587938869842e-07,
+      "loss": 0.6000016927719116,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.7403383255004883,
+      "learning_rate": 2.679409438719161e-07,
+      "loss": 1.2140743732452393,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 1.8978993892669678,
+      "learning_rate": 2.6563535447623297e-07,
+      "loss": 0.5592060089111328,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 0.8265647292137146,
+      "learning_rate": 2.6334212591991844e-07,
+      "loss": 1.1739059686660767,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.5640074610710144,
+      "learning_rate": 2.6106135788565244e-07,
+      "loss": 1.3693397045135498,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 0.8951840996742249,
+      "learning_rate": 2.5879314951447684e-07,
+      "loss": 0.6989073753356934,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.4623081386089325,
+      "learning_rate": 2.565375994014869e-07,
+      "loss": 1.1248592138290405,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.523663341999054,
+      "learning_rate": 2.5429480559154536e-07,
+      "loss": 1.1039721965789795,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 1.67033052444458,
+      "learning_rate": 2.520648655750208e-07,
+      "loss": 0.7559752464294434,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.550735354423523,
+      "learning_rate": 2.4984787628354915e-07,
+      "loss": 0.3449558913707733,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 1.1384780406951904,
+      "learning_rate": 2.476439340858208e-07,
+      "loss": 1.1599386930465698,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.6758951544761658,
+      "learning_rate": 2.4545313478339177e-07,
+      "loss": 0.6901751160621643,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 0.4316188097000122,
+      "learning_rate": 2.432755736065189e-07,
+      "loss": 0.8225246667861938,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.44040101766586304,
+      "learning_rate": 2.4111134521002105e-07,
+      "loss": 0.5468270778656006,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.6496953964233398,
+      "learning_rate": 2.3896054366916363e-07,
+      "loss": 0.9568942785263062,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 3.118239641189575,
+      "learning_rate": 2.3682326247556998e-07,
+      "loss": 0.6734780073165894,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.45301729440689087,
+      "learning_rate": 2.3469959453315754e-07,
+      "loss": 1.1819524765014648,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.48368385434150696,
+      "learning_rate": 2.3258963215409923e-07,
+      "loss": 0.2418724000453949,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.3687514364719391,
+      "learning_rate": 2.3049346705481084e-07,
+      "loss": 1.0057097673416138,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.5440084934234619,
+      "learning_rate": 2.2841119035196407e-07,
+      "loss": 1.117985725402832,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.4214663505554199,
+      "learning_rate": 2.26342892558526e-07,
+      "loss": 1.182395100593567,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.6138442754745483,
+      "learning_rate": 2.2428866357982513e-07,
+      "loss": 1.0611752271652222,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.29164546728134155,
+      "learning_rate": 2.2224859270964235e-07,
+      "loss": 0.8244748115539551,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.5247227549552917,
+      "learning_rate": 2.202227686263307e-07,
+      "loss": 0.8076885938644409,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.4827449321746826,
+      "learning_rate": 2.1821127938895936e-07,
+      "loss": 1.164618968963623,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.27927151322364807,
+      "learning_rate": 2.162142124334871e-07,
+      "loss": 0.6487075686454773,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 2.0876452922821045,
+      "learning_rate": 2.1423165456896063e-07,
+      "loss": 1.2239673137664795,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.58223557472229,
+      "learning_rate": 2.1226369197374174e-07,
+      "loss": 1.0446572303771973,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.7666153907775879,
+      "learning_rate": 2.103104101917615e-07,
+      "loss": 0.9695687294006348,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.17971999943256378,
+      "learning_rate": 2.083718941288004e-07,
+      "loss": 0.7158756256103516,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.5609447956085205,
+      "learning_rate": 2.0644822804879964e-07,
+      "loss": 1.0817302465438843,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.14870183169841766,
+      "learning_rate": 2.045394955701969e-07,
+      "loss": 0.4572441577911377,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.45732492208480835,
+      "learning_rate": 2.0264577966229222e-07,
+      "loss": 0.9440208673477173,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 0.30812424421310425,
+      "learning_rate": 2.0076716264164116e-07,
+      "loss": 0.6523926854133606,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.5582972168922424,
+      "learning_rate": 1.989037261684766e-07,
+      "loss": 1.073933720588684,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.3473052978515625,
+      "learning_rate": 1.9705555124315965e-07,
+      "loss": 0.9868369102478027,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 1.0986326932907104,
+      "learning_rate": 1.9522271820265802e-07,
+      "loss": 0.9842031598091125,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 0.9676355719566345,
+      "learning_rate": 1.9340530671705447e-07,
+      "loss": 1.4722096920013428,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 1.6234230995178223,
+      "learning_rate": 1.9160339578608303e-07,
+      "loss": 0.7529752254486084,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.4372311234474182,
+      "learning_rate": 1.89817063735696e-07,
+      "loss": 1.0224361419677734,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 0.6926037073135376,
+      "learning_rate": 1.8804638821465818e-07,
+      "loss": 0.5187029838562012,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.34467005729675293,
+      "learning_rate": 1.8629144619117233e-07,
+      "loss": 1.0658881664276123,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 1.1070398092269897,
+      "learning_rate": 1.8455231394953346e-07,
+      "loss": 0.7056020498275757,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.45471855998039246,
+      "learning_rate": 1.8282906708681236e-07,
+      "loss": 1.150197148323059,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 1.0237236022949219,
+      "learning_rate": 1.8112178050956988e-07,
+      "loss": 0.9537737369537354,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 0.6596385836601257,
+      "learning_rate": 1.794305284306013e-07,
+      "loss": 1.0868531465530396,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 0.7938012480735779,
+      "learning_rate": 1.7775538436570937e-07,
+      "loss": 0.7373219132423401,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.34923869371414185,
+      "learning_rate": 1.7609642113050992e-07,
+      "loss": 1.1872001886367798,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.4653020203113556,
+      "learning_rate": 1.7445371083726564e-07,
+      "loss": 1.1614234447479248,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.4868079125881195,
+      "learning_rate": 1.7282732489175224e-07,
+      "loss": 1.1662194728851318,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 0.7877575755119324,
+      "learning_rate": 1.7121733399015402e-07,
+      "loss": 0.91205233335495,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 2.043727397918701,
+      "learning_rate": 1.6962380811599137e-07,
+      "loss": 0.5424389839172363,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.0781350135803223,
+      "learning_rate": 1.6804681653707814e-07,
+      "loss": 0.8418402671813965,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 0.7440342307090759,
+      "learning_rate": 1.6648642780251092e-07,
+      "loss": 1.0298240184783936,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 0.8263541460037231,
+      "learning_rate": 1.649427097396897e-07,
+      "loss": 0.42005059123039246,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.5452672243118286,
+      "learning_rate": 1.63415729451369e-07,
+      "loss": 1.0754554271697998,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.5182120203971863,
+      "learning_rate": 1.619055533127412e-07,
+      "loss": 0.9780777096748352,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.48941749334335327,
+      "learning_rate": 1.6041224696855143e-07,
+      "loss": 1.1990125179290771,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.38062039017677307,
+      "learning_rate": 1.589358753302439e-07,
+      "loss": 1.1686046123504639,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 0.9446031451225281,
+      "learning_rate": 1.5747650257314066e-07,
+      "loss": 1.432015061378479,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 1.1120773553848267,
+      "learning_rate": 1.5603419213365144e-07,
+      "loss": 1.1429520845413208,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.538745105266571,
+      "learning_rate": 1.5460900670651704e-07,
+      "loss": 0.7400649785995483,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.5193626880645752,
+      "learning_rate": 1.5320100824208305e-07,
+      "loss": 1.1519125699996948,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.3940195143222809,
+      "learning_rate": 1.5181025794360798e-07,
+      "loss": 1.3227684497833252,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.5315623879432678,
+      "learning_rate": 1.5043681626460196e-07,
+      "loss": 1.2551010847091675,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 2.863178014755249,
+      "learning_rate": 1.490807429061997e-07,
+      "loss": 1.043013572692871,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 5.395115375518799,
+      "learning_rate": 1.4774209681456517e-07,
+      "loss": 0.9645116329193115,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.8922992944717407,
+      "learning_rate": 1.464209361783285e-07,
+      "loss": 1.0984504222869873,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.5256724953651428,
+      "learning_rate": 1.4511731842605803e-07,
+      "loss": 0.9979899525642395,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 20.09447479248047,
+      "learning_rate": 1.4383130022376296e-07,
+      "loss": 0.6578091382980347,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.2142326384782791,
+      "learning_rate": 1.4256293747243065e-07,
+      "loss": 0.8180892467498779,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.21468740701675415,
+      "learning_rate": 1.4131228530559635e-07,
+      "loss": 0.9822489619255066,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.7827680110931396,
+      "learning_rate": 1.400793980869468e-07,
+      "loss": 0.8261870741844177,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 1.5622820854187012,
+      "learning_rate": 1.3886432940795743e-07,
+      "loss": 0.6720483303070068,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.8509066700935364,
+      "learning_rate": 1.376671320855623e-07,
+      "loss": 0.7396157383918762,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.5573006272315979,
+      "learning_rate": 1.3648785815985874e-07,
+      "loss": 1.1854437589645386,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 3.8408010005950928,
+      "learning_rate": 1.3532655889184476e-07,
+      "loss": 0.5168848633766174,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 2.9168479442596436,
+      "learning_rate": 1.3418328476119135e-07,
+      "loss": 1.034480333328247,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.39456796646118164,
+      "learning_rate": 1.3305808546404768e-07,
+      "loss": 1.1958153247833252,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.4726790487766266,
+      "learning_rate": 1.319510099108814e-07,
+      "loss": 0.62462317943573,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.7686280608177185,
+      "learning_rate": 1.308621062243524e-07,
+      "loss": 1.4123520851135254,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.43162164092063904,
+      "learning_rate": 1.2979142173722052e-07,
+      "loss": 1.2093850374221802,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.6851413249969482,
+      "learning_rate": 1.2873900299028894e-07,
+      "loss": 0.4169965088367462,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.37487801909446716,
+      "learning_rate": 1.2770489573038072e-07,
+      "loss": 0.8792660236358643,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.42972782254219055,
+      "learning_rate": 1.2668914490834982e-07,
+      "loss": 0.5300610661506653,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 5.69271993637085,
+      "learning_rate": 1.2569179467712804e-07,
+      "loss": 0.1809053122997284,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.4998655915260315,
+      "learning_rate": 1.2471288838980504e-07,
+      "loss": 1.043647050857544,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.5562883615493774,
+      "learning_rate": 1.237524685977444e-07,
+      "loss": 0.9470633268356323,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.45538005232810974,
+      "learning_rate": 1.228105770487333e-07,
+      "loss": 1.179053544998169,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.5423485636711121,
+      "learning_rate": 1.2188725468516874e-07,
+      "loss": 1.1503219604492188,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.5560302734375,
+      "learning_rate": 1.2098254164227705e-07,
+      "loss": 0.2553570568561554,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.3783654272556305,
+      "learning_rate": 1.200964772463695e-07,
+      "loss": 1.0606558322906494,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 1.3476688861846924,
+      "learning_rate": 1.192291000131331e-07,
+      "loss": 1.0885281562805176,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 1.2853106260299683,
+      "learning_rate": 1.1838044764595615e-07,
+      "loss": 0.5078513622283936,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.5647287964820862,
+      "learning_rate": 1.1755055703428956e-07,
+      "loss": 1.2124147415161133,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.4864143133163452,
+      "learning_rate": 1.1673946425204286e-07,
+      "loss": 0.43091651797294617,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.6131622791290283,
+      "learning_rate": 1.1594720455601663e-07,
+      "loss": 1.3501334190368652,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 1.0398799180984497,
+      "learning_rate": 1.1517381238436986e-07,
+      "loss": 0.5815224051475525,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.7707909941673279,
+      "learning_rate": 1.144193213551226e-07,
+      "loss": 0.9891685247421265,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.2545696198940277,
+      "learning_rate": 1.1368376426469523e-07,
+      "loss": 1.1158779859542847,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.5492184162139893,
+      "learning_rate": 1.1296717308648236e-07,
+      "loss": 1.1675889492034912,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.9168418049812317,
+      "learning_rate": 1.1226957896946332e-07,
+      "loss": 0.8874309062957764,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.5822036266326904,
+      "learning_rate": 1.1159101223684788e-07,
+      "loss": 1.1191778182983398,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 2.816157579421997,
+      "learning_rate": 1.1093150238475845e-07,
+      "loss": 1.1562682390213013,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.37556105852127075,
+      "learning_rate": 1.1029107808094788e-07,
+      "loss": 1.0539166927337646,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 0.7887294292449951,
+      "learning_rate": 1.0966976716355293e-07,
+      "loss": 1.166836142539978,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.9350002408027649,
+      "learning_rate": 1.0906759663988474e-07,
+      "loss": 0.6714451909065247,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.5544145107269287,
+      "learning_rate": 1.084845926852547e-07,
+      "loss": 1.1760157346725464,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 1.4443753957748413,
+      "learning_rate": 1.079207806418365e-07,
+      "loss": 0.9075829982757568,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.42774176597595215,
+      "learning_rate": 1.0737618501756453e-07,
+      "loss": 1.1582362651824951,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.46223682165145874,
+      "learning_rate": 1.0685082948506874e-07,
+      "loss": 1.188973069190979,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.5352646112442017,
+      "learning_rate": 1.063447368806457e-07,
+      "loss": 0.8787344098091125,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 2.997920513153076,
+      "learning_rate": 1.0585792920326563e-07,
+      "loss": 1.0658434629440308,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 1.961532711982727,
+      "learning_rate": 1.0539042761361646e-07,
+      "loss": 0.8814696073532104,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.9376783967018127,
+      "learning_rate": 1.049422524331837e-07,
+      "loss": 1.0061813592910767,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.6350404620170593,
+      "learning_rate": 1.0451342314336753e-07,
+      "loss": 0.9253600239753723,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 1.0082424879074097,
+      "learning_rate": 1.0410395838463545e-07,
+      "loss": 1.2256076335906982,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.7602862119674683,
+      "learning_rate": 1.0371387595571246e-07,
+      "loss": 0.5702172517776489,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.5603384375572205,
+      "learning_rate": 1.0334319281280713e-07,
+      "loss": 0.7256681323051453,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 1.6261464357376099,
+      "learning_rate": 1.0299192506887466e-07,
+      "loss": 1.083208680152893,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.40230756998062134,
+      "learning_rate": 1.0266008799291634e-07,
+      "loss": 0.8354797959327698,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.18954500555992126,
+      "learning_rate": 1.0234769600931611e-07,
+      "loss": 0.6113368272781372,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.5848817825317383,
+      "learning_rate": 1.0205476269721311e-07,
+      "loss": 1.167410969734192,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 0.8019612431526184,
+      "learning_rate": 1.0178130078991195e-07,
+      "loss": 0.6739932298660278,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.6015135049819946,
+      "learning_rate": 1.0152732217432873e-07,
+      "loss": 0.5854436159133911,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.780681848526001,
+      "learning_rate": 1.0129283789047472e-07,
+      "loss": 0.8006756901741028,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.37933349609375,
+      "learning_rate": 1.0107785813097619e-07,
+      "loss": 1.1144965887069702,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.8300569653511047,
+      "learning_rate": 1.0088239224063151e-07,
+      "loss": 0.2734246253967285,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 1.4782869815826416,
+      "learning_rate": 1.0070644871600493e-07,
+      "loss": 0.7137698531150818,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.3787873387336731,
+      "learning_rate": 1.0055003520505713e-07,
+      "loss": 0.8101165294647217,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.5657850503921509,
+      "learning_rate": 1.0041315850681298e-07,
+      "loss": 1.2873133420944214,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 0.21417276561260223,
+      "learning_rate": 1.0029582457106586e-07,
+      "loss": 0.3951149582862854,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.49882808327674866,
+      "learning_rate": 1.00198038498119e-07,
+      "loss": 1.1357934474945068,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.3907023072242737,
+      "learning_rate": 1.001198045385639e-07,
+      "loss": 0.8105093836784363,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.8356418013572693,
+      "learning_rate": 1.0006112609309548e-07,
+      "loss": 0.6233622431755066,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.475093275308609,
+      "learning_rate": 1.0002200571236428e-07,
+      "loss": 1.172187328338623,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.1982579082250595,
+      "learning_rate": 1.0000244509686563e-07,
+      "loss": 0.5022701025009155,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 4.038502240003031e+18,
+      "train_loss": 1.0441685825645686,
+      "train_runtime": 10534.5209,
+      "train_samples_per_second": 5.718,
+      "train_steps_per_second": 0.095
+    }
+  ],
+  "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": 4.038502240003031e+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 sha256:98ca1a3153d0bb381eded7d815a0d6c846f865dfd739454844594d8ea37ee83d
+size 5649