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 +1,69 @@
+---
+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_muon_lr = {
+        "language": _fallback(getattr(training_args, "language_muon_lr", 2e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 5e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 1e-4), merger_lr),
+    }
+
+    family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), 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 +1,69 @@
+---
+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_muon_lr = {
+        "language": _fallback(getattr(training_args, "language_muon_lr", 2e-5), language_lr),
+        "vision": _fallback(getattr(training_args, "vision_muon_lr", 5e-5), vision_lr),
+        "merger": _fallback(getattr(training_args, "merger_muon_lr", 1e-4), merger_lr),
+    }
+
+    family_to_adamw_lr = {
+        "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), 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": 7374.398147293401,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.8034457557749345,
+    "train_runtime": 9234.9215,
+    "train_samples_per_second": 6.523,
+    "train_steps_per_second": 0.109
+}

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:d6e584525bf707f04c4cf79c209a9385fbfd8dbeca0fca52afe26afae9c24453
+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": 7374.398147293401,
+    "epoch": 4.0,
+    "total_flos": 3.519329211045118e+18,
+    "train_loss": 0.8034457557749345,
+    "train_runtime": 9234.9215,
+    "train_samples_per_second": 6.523,
+    "train_steps_per_second": 0.109
+}

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": 0.8768869042396545,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 2.0941407680511475,
+      "step": 2
+    },
+    {
+      "epoch": 0.01593625498007968,
+      "grad_norm": 0.6598695516586304,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 2.010636329650879,
+      "step": 4
+    },
+    {
+      "epoch": 0.02390438247011952,
+      "grad_norm": 4.373621463775635,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 2.1800665855407715,
+      "step": 6
+    },
+    {
+      "epoch": 0.03187250996015936,
+      "grad_norm": 0.5181940197944641,
+      "learning_rate": 1.3725490196078434e-06,
+      "loss": 1.8901777267456055,
+      "step": 8
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 1.6566247940063477,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 1.4661237001419067,
+      "step": 10
+    },
+    {
+      "epoch": 0.04780876494023904,
+      "grad_norm": 0.5453764796257019,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 1.6929049491882324,
+      "step": 12
+    },
+    {
+      "epoch": 0.055776892430278883,
+      "grad_norm": 1.0317989587783813,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 2.2657735347747803,
+      "step": 14
+    },
+    {
+      "epoch": 0.06374501992031872,
+      "grad_norm": 2.3971784114837646,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 1.8927618265151978,
+      "step": 16
+    },
+    {
+      "epoch": 0.07171314741035857,
+      "grad_norm": 1.1476653814315796,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.6072006225585938,
+      "step": 18
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.49550750851631165,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 1.5377804040908813,
+      "step": 20
+    },
+    {
+      "epoch": 0.08764940239043825,
+      "grad_norm": 0.46495792269706726,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.1792941093444824,
+      "step": 22
+    },
+    {
+      "epoch": 0.09561752988047809,
+      "grad_norm": 0.9762714505195618,
+      "learning_rate": 4.509803921568628e-06,
+      "loss": 1.3525527715682983,
+      "step": 24
+    },
+    {
+      "epoch": 0.10358565737051793,
+      "grad_norm": 0.3185323476791382,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 1.4434500932693481,
+      "step": 26
+    },
+    {
+      "epoch": 0.11155378486055777,
+      "grad_norm": 0.4879108667373657,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 1.6278623342514038,
+      "step": 28
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 2.3721625804901123,
+      "learning_rate": 5.686274509803922e-06,
+      "loss": 1.3184047937393188,
+      "step": 30
+    },
+    {
+      "epoch": 0.12749003984063745,
+      "grad_norm": 0.639505922794342,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 1.1689376831054688,
+      "step": 32
+    },
+    {
+      "epoch": 0.13545816733067728,
+      "grad_norm": 0.5136529803276062,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 1.2432702779769897,
+      "step": 34
+    },
+    {
+      "epoch": 0.14342629482071714,
+      "grad_norm": 0.3695473074913025,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 1.406494140625,
+      "step": 36
+    },
+    {
+      "epoch": 0.15139442231075698,
+      "grad_norm": 3.0504915714263916,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 1.1878796815872192,
+      "step": 38
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 0.5996785163879395,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 1.3199090957641602,
+      "step": 40
+    },
+    {
+      "epoch": 0.16733067729083664,
+      "grad_norm": 0.8212912678718567,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.794780969619751,
+      "step": 42
+    },
+    {
+      "epoch": 0.1752988047808765,
+      "grad_norm": 0.47704246640205383,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.9839233756065369,
+      "step": 44
+    },
+    {
+      "epoch": 0.18326693227091634,
+      "grad_norm": 1.5285656452178955,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.9492515921592712,
+      "step": 46
+    },
+    {
+      "epoch": 0.19123505976095617,
+      "grad_norm": 0.5984688997268677,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 1.1842522621154785,
+      "step": 48
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.3931889533996582,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 1.3357834815979004,
+      "step": 50
+    },
+    {
+      "epoch": 0.20717131474103587,
+      "grad_norm": 0.5188454389572144,
+      "learning_rate": 1e-05,
+      "loss": 1.2938684225082397,
+      "step": 52
+    },
+    {
+      "epoch": 0.2151394422310757,
+      "grad_norm": 1.6206631660461426,
+      "learning_rate": 9.999902196391087e-06,
+      "loss": 1.034663200378418,
+      "step": 54
+    },
+    {
+      "epoch": 0.22310756972111553,
+      "grad_norm": 0.8077843189239502,
+      "learning_rate": 9.999608789815699e-06,
+      "loss": 1.5252913236618042,
+      "step": 56
+    },
+    {
+      "epoch": 0.23107569721115537,
+      "grad_norm": 0.45542818307876587,
+      "learning_rate": 9.999119793027713e-06,
+      "loss": 1.3403213024139404,
+      "step": 58
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.35984286665916443,
+      "learning_rate": 9.998435227282975e-06,
+      "loss": 0.9048821926116943,
+      "step": 60
+    },
+    {
+      "epoch": 0.24701195219123506,
+      "grad_norm": 0.4619295001029968,
+      "learning_rate": 9.997555122338372e-06,
+      "loss": 1.3436535596847534,
+      "step": 62
+    },
+    {
+      "epoch": 0.2549800796812749,
+      "grad_norm": 2.309458017349243,
+      "learning_rate": 9.996479516450544e-06,
+      "loss": 1.1000707149505615,
+      "step": 64
+    },
+    {
+      "epoch": 0.26294820717131473,
+      "grad_norm": 0.3844827115535736,
+      "learning_rate": 9.99520845637422e-06,
+      "loss": 1.2722667455673218,
+      "step": 66
+    },
+    {
+      "epoch": 0.27091633466135456,
+      "grad_norm": 1.0867564678192139,
+      "learning_rate": 9.993741997360185e-06,
+      "loss": 0.8392297029495239,
+      "step": 68
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.5023817420005798,
+      "learning_rate": 9.992080203152872e-06,
+      "loss": 1.5487357378005981,
+      "step": 70
+    },
+    {
+      "epoch": 0.2868525896414343,
+      "grad_norm": 0.8340875506401062,
+      "learning_rate": 9.990223145987608e-06,
+      "loss": 0.8259425163269043,
+      "step": 72
+    },
+    {
+      "epoch": 0.2948207171314741,
+      "grad_norm": 0.9446395635604858,
+      "learning_rate": 9.98817090658745e-06,
+      "loss": 1.1226840019226074,
+      "step": 74
+    },
+    {
+      "epoch": 0.30278884462151395,
+      "grad_norm": 0.4371112585067749,
+      "learning_rate": 9.985923574159696e-06,
+      "loss": 1.4519315958023071,
+      "step": 76
+    },
+    {
+      "epoch": 0.3107569721115538,
+      "grad_norm": 2.011549472808838,
+      "learning_rate": 9.983481246392003e-06,
+      "loss": 0.9990478754043579,
+      "step": 78
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 0.6929413676261902,
+      "learning_rate": 9.980844029448135e-06,
+      "loss": 0.8473486304283142,
+      "step": 80
+    },
+    {
+      "epoch": 0.32669322709163345,
+      "grad_norm": 0.45718321204185486,
+      "learning_rate": 9.97801203796335e-06,
+      "loss": 1.3274242877960205,
+      "step": 82
+    },
+    {
+      "epoch": 0.3346613545816733,
+      "grad_norm": 0.5014246106147766,
+      "learning_rate": 9.974985395039423e-06,
+      "loss": 0.8854591846466064,
+      "step": 84
+    },
+    {
+      "epoch": 0.3426294820717131,
+      "grad_norm": 0.43661314249038696,
+      "learning_rate": 9.971764232239287e-06,
+      "loss": 1.3072928190231323,
+      "step": 86
+    },
+    {
+      "epoch": 0.350597609561753,
+      "grad_norm": 0.9643505811691284,
+      "learning_rate": 9.968348689581315e-06,
+      "loss": 0.9253592491149902,
+      "step": 88
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.3767000734806061,
+      "learning_rate": 9.964738915533242e-06,
+      "loss": 1.3180688619613647,
+      "step": 90
+    },
+    {
+      "epoch": 0.3665338645418327,
+      "grad_norm": 0.7286614179611206,
+      "learning_rate": 9.960935067005706e-06,
+      "loss": 1.1165339946746826,
+      "step": 92
+    },
+    {
+      "epoch": 0.3745019920318725,
+      "grad_norm": 0.3594738841056824,
+      "learning_rate": 9.956937309345418e-06,
+      "loss": 1.2314833402633667,
+      "step": 94
+    },
+    {
+      "epoch": 0.38247011952191234,
+      "grad_norm": 1.4631410837173462,
+      "learning_rate": 9.952745816327998e-06,
+      "loss": 1.083662986755371,
+      "step": 96
+    },
+    {
+      "epoch": 0.3904382470119522,
+      "grad_norm": 1.0809067487716675,
+      "learning_rate": 9.94836077015039e-06,
+      "loss": 1.2512863874435425,
+      "step": 98
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.4656352698802948,
+      "learning_rate": 9.943782361422975e-06,
+      "loss": 0.8630403280258179,
+      "step": 100
+    },
+    {
+      "epoch": 0.4063745019920319,
+      "grad_norm": 0.5072252154350281,
+      "learning_rate": 9.939010789161259e-06,
+      "loss": 1.2671551704406738,
+      "step": 102
+    },
+    {
+      "epoch": 0.41434262948207173,
+      "grad_norm": 0.3878318965435028,
+      "learning_rate": 9.934046260777239e-06,
+      "loss": 0.7883280515670776,
+      "step": 104
+    },
+    {
+      "epoch": 0.42231075697211157,
+      "grad_norm": 0.754079282283783,
+      "learning_rate": 9.928888992070373e-06,
+      "loss": 0.7010841369628906,
+      "step": 106
+    },
+    {
+      "epoch": 0.4302788844621514,
+      "grad_norm": 0.9006465077400208,
+      "learning_rate": 9.923539207218219e-06,
+      "loss": 0.9954584836959839,
+      "step": 108
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.474048376083374,
+      "learning_rate": 9.91799713876667e-06,
+      "loss": 1.2795435190200806,
+      "step": 110
+    },
+    {
+      "epoch": 0.44621513944223107,
+      "grad_norm": 0.47232747077941895,
+      "learning_rate": 9.912263027619856e-06,
+      "loss": 0.8258985280990601,
+      "step": 112
+    },
+    {
+      "epoch": 0.4541832669322709,
+      "grad_norm": 0.5098729729652405,
+      "learning_rate": 9.906337123029675e-06,
+      "loss": 1.1291917562484741,
+      "step": 114
+    },
+    {
+      "epoch": 0.46215139442231074,
+      "grad_norm": 0.5174960494041443,
+      "learning_rate": 9.90021968258495e-06,
+      "loss": 1.28175687789917,
+      "step": 116
+    },
+    {
+      "epoch": 0.4701195219123506,
+      "grad_norm": 0.3036414384841919,
+      "learning_rate": 9.893910972200238e-06,
+      "loss": 1.344761610031128,
+      "step": 118
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.46434810757637024,
+      "learning_rate": 9.887411266104268e-06,
+      "loss": 1.2242240905761719,
+      "step": 120
+    },
+    {
+      "epoch": 0.4860557768924303,
+      "grad_norm": 0.6941715478897095,
+      "learning_rate": 9.880720846828026e-06,
+      "loss": 0.9668956995010376,
+      "step": 122
+    },
+    {
+      "epoch": 0.4940239043824701,
+      "grad_norm": 0.7519447803497314,
+      "learning_rate": 9.873840005192466e-06,
+      "loss": 1.028874158859253,
+      "step": 124
+    },
+    {
+      "epoch": 0.50199203187251,
+      "grad_norm": 1.2101236581802368,
+      "learning_rate": 9.866769040295875e-06,
+      "loss": 0.9155845642089844,
+      "step": 126
+    },
+    {
+      "epoch": 0.5099601593625498,
+      "grad_norm": 0.3666621148586273,
+      "learning_rate": 9.859508259500869e-06,
+      "loss": 1.265921711921692,
+      "step": 128
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.3077186048030853,
+      "learning_rate": 9.852057978421033e-06,
+      "loss": 0.9352775812149048,
+      "step": 130
+    },
+    {
+      "epoch": 0.5258964143426295,
+      "grad_norm": 0.4671645164489746,
+      "learning_rate": 9.844418520907198e-06,
+      "loss": 0.7999334931373596,
+      "step": 132
+    },
+    {
+      "epoch": 0.5338645418326693,
+      "grad_norm": 0.3349776268005371,
+      "learning_rate": 9.836590219033375e-06,
+      "loss": 0.8986390829086304,
+      "step": 134
+    },
+    {
+      "epoch": 0.5418326693227091,
+      "grad_norm": 0.6630865931510925,
+      "learning_rate": 9.828573413082306e-06,
+      "loss": 1.281280279159546,
+      "step": 136
+    },
+    {
+      "epoch": 0.549800796812749,
+      "grad_norm": 0.33890610933303833,
+      "learning_rate": 9.820368451530681e-06,
+      "loss": 0.9360127449035645,
+      "step": 138
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 1.1324018239974976,
+      "learning_rate": 9.811975691033992e-06,
+      "loss": 1.0854263305664062,
+      "step": 140
+    },
+    {
+      "epoch": 0.5657370517928287,
+      "grad_norm": 0.45373809337615967,
+      "learning_rate": 9.80339549641102e-06,
+      "loss": 1.0081974267959595,
+      "step": 142
+    },
+    {
+      "epoch": 0.5737051792828686,
+      "grad_norm": 0.4071025848388672,
+      "learning_rate": 9.794628240627992e-06,
+      "loss": 1.2175102233886719,
+      "step": 144
+    },
+    {
+      "epoch": 0.5816733067729084,
+      "grad_norm": 0.5562829971313477,
+      "learning_rate": 9.785674304782349e-06,
+      "loss": 1.2132571935653687,
+      "step": 146
+    },
+    {
+      "epoch": 0.5896414342629482,
+      "grad_norm": 1.8553446531295776,
+      "learning_rate": 9.776534078086202e-06,
+      "loss": 0.977083683013916,
+      "step": 148
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 2.4018805027008057,
+      "learning_rate": 9.767207957849396e-06,
+      "loss": 0.9409672021865845,
+      "step": 150
+    },
+    {
+      "epoch": 0.6055776892430279,
+      "grad_norm": 1.0131343603134155,
+      "learning_rate": 9.757696349462252e-06,
+      "loss": 0.9264012575149536,
+      "step": 152
+    },
+    {
+      "epoch": 0.6135458167330677,
+      "grad_norm": 0.7705997228622437,
+      "learning_rate": 9.747999666377934e-06,
+      "loss": 1.1069132089614868,
+      "step": 154
+    },
+    {
+      "epoch": 0.6215139442231076,
+      "grad_norm": 1.9844759702682495,
+      "learning_rate": 9.738118330094491e-06,
+      "loss": 1.0613375902175903,
+      "step": 156
+    },
+    {
+      "epoch": 0.6294820717131474,
+      "grad_norm": 1.245699405670166,
+      "learning_rate": 9.728052770136518e-06,
+      "loss": 1.1516886949539185,
+      "step": 158
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.4531320631504059,
+      "learning_rate": 9.717803424036501e-06,
+      "loss": 1.0168956518173218,
+      "step": 160
+    },
+    {
+      "epoch": 0.6454183266932271,
+      "grad_norm": 0.3907632529735565,
+      "learning_rate": 9.707370737315793e-06,
+      "loss": 0.7657440900802612,
+      "step": 162
+    },
+    {
+      "epoch": 0.6533864541832669,
+      "grad_norm": 0.27871450781822205,
+      "learning_rate": 9.69675516346524e-06,
+      "loss": 1.1317344903945923,
+      "step": 164
+    },
+    {
+      "epoch": 0.6613545816733067,
+      "grad_norm": 2.384199857711792,
+      "learning_rate": 9.685957163925479e-06,
+      "loss": 0.8119434118270874,
+      "step": 166
+    },
+    {
+      "epoch": 0.6693227091633466,
+      "grad_norm": 0.8371350765228271,
+      "learning_rate": 9.67497720806688e-06,
+      "loss": 1.2329051494598389,
+      "step": 168
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 1.5197649002075195,
+      "learning_rate": 9.663815773169131e-06,
+      "loss": 0.9320511221885681,
+      "step": 170
+    },
+    {
+      "epoch": 0.6852589641434262,
+      "grad_norm": 0.8917195200920105,
+      "learning_rate": 9.652473344400507e-06,
+      "loss": 0.9010263681411743,
+      "step": 172
+    },
+    {
+      "epoch": 0.6932270916334662,
+      "grad_norm": 0.2711629271507263,
+      "learning_rate": 9.640950414796772e-06,
+      "loss": 1.225519061088562,
+      "step": 174
+    },
+    {
+      "epoch": 0.701195219123506,
+      "grad_norm": 1.2895762920379639,
+      "learning_rate": 9.629247485239747e-06,
+      "loss": 0.7548736929893494,
+      "step": 176
+    },
+    {
+      "epoch": 0.7091633466135459,
+      "grad_norm": 0.9421162009239197,
+      "learning_rate": 9.617365064435544e-06,
+      "loss": 0.9123273491859436,
+      "step": 178
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 1.0172063112258911,
+      "learning_rate": 9.605303668892443e-06,
+      "loss": 1.0832524299621582,
+      "step": 180
+    },
+    {
+      "epoch": 0.7250996015936255,
+      "grad_norm": 0.17812766134738922,
+      "learning_rate": 9.593063822898451e-06,
+      "loss": 0.503821849822998,
+      "step": 182
+    },
+    {
+      "epoch": 0.7330677290836654,
+      "grad_norm": 0.3833879828453064,
+      "learning_rate": 9.580646058498505e-06,
+      "loss": 1.3093364238739014,
+      "step": 184
+    },
+    {
+      "epoch": 0.7410358565737052,
+      "grad_norm": 0.30875980854034424,
+      "learning_rate": 9.56805091547135e-06,
+      "loss": 1.2706356048583984,
+      "step": 186
+    },
+    {
+      "epoch": 0.749003984063745,
+      "grad_norm": 0.3245117664337158,
+      "learning_rate": 9.55527894130607e-06,
+      "loss": 1.1097581386566162,
+      "step": 188
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.6155065298080444,
+      "learning_rate": 9.542330691178296e-06,
+      "loss": 0.8813226819038391,
+      "step": 190
+    },
+    {
+      "epoch": 0.7649402390438247,
+      "grad_norm": 0.7238669395446777,
+      "learning_rate": 9.529206727926071e-06,
+      "loss": 1.2123708724975586,
+      "step": 192
+    },
+    {
+      "epoch": 0.7729083665338645,
+      "grad_norm": 0.37201404571533203,
+      "learning_rate": 9.515907622025376e-06,
+      "loss": 0.8893398642539978,
+      "step": 194
+    },
+    {
+      "epoch": 0.7808764940239044,
+      "grad_norm": 0.7957295179367065,
+      "learning_rate": 9.502433951565347e-06,
+      "loss": 0.618162989616394,
+      "step": 196
+    },
+    {
+      "epoch": 0.7888446215139442,
+      "grad_norm": 1.219519019126892,
+      "learning_rate": 9.48878630222314e-06,
+      "loss": 0.7700294852256775,
+      "step": 198
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 0.873467743396759,
+      "learning_rate": 9.474965267238469e-06,
+      "loss": 1.051179051399231,
+      "step": 200
+    },
+    {
+      "epoch": 0.8047808764940239,
+      "grad_norm": 0.17877569794654846,
+      "learning_rate": 9.460971447387824e-06,
+      "loss": 0.9797074198722839,
+      "step": 202
+    },
+    {
+      "epoch": 0.8127490039840638,
+      "grad_norm": 0.3721548616886139,
+      "learning_rate": 9.446805450958353e-06,
+      "loss": 1.5231633186340332,
+      "step": 204
+    },
+    {
+      "epoch": 0.8207171314741036,
+      "grad_norm": 0.38004380464553833,
+      "learning_rate": 9.43246789372142e-06,
+      "loss": 1.225372552871704,
+      "step": 206
+    },
+    {
+      "epoch": 0.8286852589641435,
+      "grad_norm": 0.2693279981613159,
+      "learning_rate": 9.417959398905846e-06,
+      "loss": 0.586847186088562,
+      "step": 208
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.6006880402565002,
+      "learning_rate": 9.40328059717081e-06,
+      "loss": 1.2704275846481323,
+      "step": 210
+    },
+    {
+      "epoch": 0.8446215139442231,
+      "grad_norm": 0.35779136419296265,
+      "learning_rate": 9.388432126578438e-06,
+      "loss": 1.2139121294021606,
+      "step": 212
+    },
+    {
+      "epoch": 0.852589641434263,
+      "grad_norm": 0.4109326899051666,
+      "learning_rate": 9.373414632566066e-06,
+      "loss": 0.9354465007781982,
+      "step": 214
+    },
+    {
+      "epoch": 0.8605577689243028,
+      "grad_norm": 0.7465597987174988,
+      "learning_rate": 9.35822876791819e-06,
+      "loss": 1.4245351552963257,
+      "step": 216
+    },
+    {
+      "epoch": 0.8685258964143426,
+      "grad_norm": 0.9482311606407166,
+      "learning_rate": 9.342875192738083e-06,
+      "loss": 0.9090552926063538,
+      "step": 218
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.7371097803115845,
+      "learning_rate": 9.327354574419109e-06,
+      "loss": 1.2302675247192383,
+      "step": 220
+    },
+    {
+      "epoch": 0.8844621513944223,
+      "grad_norm": 0.47654393315315247,
+      "learning_rate": 9.311667587615701e-06,
+      "loss": 1.476176142692566,
+      "step": 222
+    },
+    {
+      "epoch": 0.8924302788844621,
+      "grad_norm": 0.4779684841632843,
+      "learning_rate": 9.29581491421405e-06,
+      "loss": 0.7193994522094727,
+      "step": 224
+    },
+    {
+      "epoch": 0.900398406374502,
+      "grad_norm": 0.3918261229991913,
+      "learning_rate": 9.279797243302453e-06,
+      "loss": 0.9889163970947266,
+      "step": 226
+    },
+    {
+      "epoch": 0.9083665338645418,
+      "grad_norm": 0.8190336227416992,
+      "learning_rate": 9.263615271141362e-06,
+      "loss": 1.0910744667053223,
+      "step": 228
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.5420666933059692,
+      "learning_rate": 9.247269701133125e-06,
+      "loss": 1.1785545349121094,
+      "step": 230
+    },
+    {
+      "epoch": 0.9243027888446215,
+      "grad_norm": 0.4055289328098297,
+      "learning_rate": 9.230761243791401e-06,
+      "loss": 1.3215173482894897,
+      "step": 232
+    },
+    {
+      "epoch": 0.9322709163346613,
+      "grad_norm": 0.34705033898353577,
+      "learning_rate": 9.214090616710282e-06,
+      "loss": 0.8221312761306763,
+      "step": 234
+    },
+    {
+      "epoch": 0.9402390438247012,
+      "grad_norm": 0.36026525497436523,
+      "learning_rate": 9.197258544533098e-06,
+      "loss": 0.82416170835495,
+      "step": 236
+    },
+    {
+      "epoch": 0.9482071713147411,
+      "grad_norm": 0.44059935212135315,
+      "learning_rate": 9.180265758920916e-06,
+      "loss": 0.5275893211364746,
+      "step": 238
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 2.5635297298431396,
+      "learning_rate": 9.163112998520741e-06,
+      "loss": 0.8700987696647644,
+      "step": 240
+    },
+    {
+      "epoch": 0.9641434262948207,
+      "grad_norm": 0.8074333071708679,
+      "learning_rate": 9.145801008933403e-06,
+      "loss": 1.1190941333770752,
+      "step": 242
+    },
+    {
+      "epoch": 0.9721115537848606,
+      "grad_norm": 0.2998296022415161,
+      "learning_rate": 9.128330542681154e-06,
+      "loss": 1.2127717733383179,
+      "step": 244
+    },
+    {
+      "epoch": 0.9800796812749004,
+      "grad_norm": 0.41620495915412903,
+      "learning_rate": 9.110702359174944e-06,
+      "loss": 0.9575527906417847,
+      "step": 246
+    },
+    {
+      "epoch": 0.9880478087649402,
+      "grad_norm": 0.8565504550933838,
+      "learning_rate": 9.092917224681427e-06,
+      "loss": 0.9943863153457642,
+      "step": 248
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.8093377947807312,
+      "learning_rate": 9.074975912289641e-06,
+      "loss": 0.9918873310089111,
+      "step": 250
+    },
+    {
+      "epoch": 1.00398406374502,
+      "grad_norm": 0.8429654240608215,
+      "learning_rate": 9.056879201877412e-06,
+      "loss": 0.920735239982605,
+      "step": 252
+    },
+    {
+      "epoch": 1.0119521912350598,
+      "grad_norm": 0.290812224149704,
+      "learning_rate": 9.038627880077441e-06,
+      "loss": 0.7750673890113831,
+      "step": 254
+    },
+    {
+      "epoch": 1.0199203187250996,
+      "grad_norm": 0.423089861869812,
+      "learning_rate": 9.020222740243127e-06,
+      "loss": 1.1213276386260986,
+      "step": 256
+    },
+    {
+      "epoch": 1.0278884462151394,
+      "grad_norm": 0.5509600639343262,
+      "learning_rate": 9.001664582414066e-06,
+      "loss": 1.0930206775665283,
+      "step": 258
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.45708414912223816,
+      "learning_rate": 8.982954213281291e-06,
+      "loss": 1.0447009801864624,
+      "step": 260
+    },
+    {
+      "epoch": 1.043824701195219,
+      "grad_norm": 0.8372029662132263,
+      "learning_rate": 8.964092446152183e-06,
+      "loss": 0.9540250897407532,
+      "step": 262
+    },
+    {
+      "epoch": 1.051792828685259,
+      "grad_norm": 0.6896669864654541,
+      "learning_rate": 8.945080100915145e-06,
+      "loss": 1.0480436086654663,
+      "step": 264
+    },
+    {
+      "epoch": 1.0597609561752988,
+      "grad_norm": 1.3677271604537964,
+      "learning_rate": 8.925918004003941e-06,
+      "loss": 0.6267285943031311,
+      "step": 266
+    },
+    {
+      "epoch": 1.0677290836653386,
+      "grad_norm": 0.2271985411643982,
+      "learning_rate": 8.90660698836179e-06,
+      "loss": 0.8855297565460205,
+      "step": 268
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.27979105710983276,
+      "learning_rate": 8.887147893405145e-06,
+      "loss": 0.5436021089553833,
+      "step": 270
+    },
+    {
+      "epoch": 1.0836653386454183,
+      "grad_norm": 0.5200917720794678,
+      "learning_rate": 8.867541564987212e-06,
+      "loss": 1.212299108505249,
+      "step": 272
+    },
+    {
+      "epoch": 1.091633466135458,
+      "grad_norm": 1.44416081905365,
+      "learning_rate": 8.84778885536118e-06,
+      "loss": 0.5779670476913452,
+      "step": 274
+    },
+    {
+      "epoch": 1.099601593625498,
+      "grad_norm": 0.36050084233283997,
+      "learning_rate": 8.827890623143172e-06,
+      "loss": 0.8575031757354736,
+      "step": 276
+    },
+    {
+      "epoch": 1.1075697211155378,
+      "grad_norm": 0.268659770488739,
+      "learning_rate": 8.80784773327494e-06,
+      "loss": 0.757164478302002,
+      "step": 278
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.5362316370010376,
+      "learning_rate": 8.787661056986236e-06,
+      "loss": 0.9596293568611145,
+      "step": 280
+    },
+    {
+      "epoch": 1.1235059760956174,
+      "grad_norm": 0.6429809927940369,
+      "learning_rate": 8.767331471756975e-06,
+      "loss": 1.1100592613220215,
+      "step": 282
+    },
+    {
+      "epoch": 1.1314741035856573,
+      "grad_norm": 0.4168923497200012,
+      "learning_rate": 8.746859861279068e-06,
+      "loss": 1.141865849494934,
+      "step": 284
+    },
+    {
+      "epoch": 1.139442231075697,
+      "grad_norm": 0.1882515251636505,
+      "learning_rate": 8.726247115418017e-06,
+      "loss": 0.2671031057834625,
+      "step": 286
+    },
+    {
+      "epoch": 1.1474103585657371,
+      "grad_norm": 0.16827291250228882,
+      "learning_rate": 8.70549413017424e-06,
+      "loss": 0.6934893727302551,
+      "step": 288
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 0.7366210222244263,
+      "learning_rate": 8.684601807644114e-06,
+      "loss": 0.8977146744728088,
+      "step": 290
+    },
+    {
+      "epoch": 1.1633466135458168,
+      "grad_norm": 0.5833678841590881,
+      "learning_rate": 8.663571055980769e-06,
+      "loss": 0.4596826136112213,
+      "step": 292
+    },
+    {
+      "epoch": 1.1713147410358566,
+      "grad_norm": 0.27396321296691895,
+      "learning_rate": 8.642402789354609e-06,
+      "loss": 1.1552473306655884,
+      "step": 294
+    },
+    {
+      "epoch": 1.1792828685258965,
+      "grad_norm": 0.3825971782207489,
+      "learning_rate": 8.62109792791358e-06,
+      "loss": 1.2096163034439087,
+      "step": 296
+    },
+    {
+      "epoch": 1.1872509960159363,
+      "grad_norm": 0.2667533755302429,
+      "learning_rate": 8.599657397743163e-06,
+      "loss": 1.272385835647583,
+      "step": 298
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 0.8556350469589233,
+      "learning_rate": 8.578082130826123e-06,
+      "loss": 0.6062456965446472,
+      "step": 300
+    },
+    {
+      "epoch": 1.203187250996016,
+      "grad_norm": 0.4159654378890991,
+      "learning_rate": 8.556373065002006e-06,
+      "loss": 0.875177264213562,
+      "step": 302
+    },
+    {
+      "epoch": 1.2111553784860558,
+      "grad_norm": 0.5554153919219971,
+      "learning_rate": 8.534531143926359e-06,
+      "loss": 0.5195883512496948,
+      "step": 304
+    },
+    {
+      "epoch": 1.2191235059760956,
+      "grad_norm": 0.4305969178676605,
+      "learning_rate": 8.512557317029715e-06,
+      "loss": 1.1448959112167358,
+      "step": 306
+    },
+    {
+      "epoch": 1.2270916334661355,
+      "grad_norm": 1.7159167528152466,
+      "learning_rate": 8.49045253947633e-06,
+      "loss": 0.9804019331932068,
+      "step": 308
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.9805147647857666,
+      "learning_rate": 8.46821777212266e-06,
+      "loss": 1.136406660079956,
+      "step": 310
+    },
+    {
+      "epoch": 1.2430278884462151,
+      "grad_norm": 0.4811799228191376,
+      "learning_rate": 8.445853981475587e-06,
+      "loss": 1.0157673358917236,
+      "step": 312
+    },
+    {
+      "epoch": 1.250996015936255,
+      "grad_norm": 2.8888847827911377,
+      "learning_rate": 8.423362139650414e-06,
+      "loss": 0.8499234318733215,
+      "step": 314
+    },
+    {
+      "epoch": 1.2589641434262948,
+      "grad_norm": 0.4052259624004364,
+      "learning_rate": 8.40074322432861e-06,
+      "loss": 0.7112973928451538,
+      "step": 316
+    },
+    {
+      "epoch": 1.2669322709163346,
+      "grad_norm": 0.5343075394630432,
+      "learning_rate": 8.377998218715311e-06,
+      "loss": 0.6302114725112915,
+      "step": 318
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 1.951466679573059,
+      "learning_rate": 8.355128111496569e-06,
+      "loss": 0.4855978786945343,
+      "step": 320
+    },
+    {
+      "epoch": 1.2828685258964143,
+      "grad_norm": 0.35669681429862976,
+      "learning_rate": 8.332133896796399e-06,
+      "loss": 0.9443033933639526,
+      "step": 322
+    },
+    {
+      "epoch": 1.2908366533864541,
+      "grad_norm": 1.1142667531967163,
+      "learning_rate": 8.309016574133545e-06,
+      "loss": 0.9038307070732117,
+      "step": 324
+    },
+    {
+      "epoch": 1.298804780876494,
+      "grad_norm": 0.6319164037704468,
+      "learning_rate": 8.28577714837805e-06,
+      "loss": 1.2147735357284546,
+      "step": 326
+    },
+    {
+      "epoch": 1.3067729083665338,
+      "grad_norm": 0.3126705586910248,
+      "learning_rate": 8.262416629707555e-06,
+      "loss": 0.6492886543273926,
+      "step": 328
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.24964411556720734,
+      "learning_rate": 8.238936033563415e-06,
+      "loss": 0.819593071937561,
+      "step": 330
+    },
+    {
+      "epoch": 1.3227091633466135,
+      "grad_norm": 0.595253586769104,
+      "learning_rate": 8.215336380606533e-06,
+      "loss": 0.9014972448348999,
+      "step": 332
+    },
+    {
+      "epoch": 1.3306772908366533,
+      "grad_norm": 0.5459438562393188,
+      "learning_rate": 8.191618696673012e-06,
+      "loss": 0.592470645904541,
+      "step": 334
+    },
+    {
+      "epoch": 1.3386454183266931,
+      "grad_norm": 0.7769253849983215,
+      "learning_rate": 8.167784012729555e-06,
+      "loss": 0.8355544805526733,
+      "step": 336
+    },
+    {
+      "epoch": 1.3466135458167332,
+      "grad_norm": 1.0846341848373413,
+      "learning_rate": 8.143833364828654e-06,
+      "loss": 1.0890873670578003,
+      "step": 338
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 0.25273755192756653,
+      "learning_rate": 8.119767794063553e-06,
+      "loss": 1.092644453048706,
+      "step": 340
+    },
+    {
+      "epoch": 1.3625498007968129,
+      "grad_norm": 0.5920921564102173,
+      "learning_rate": 8.095588346523e-06,
+      "loss": 1.1467219591140747,
+      "step": 342
+    },
+    {
+      "epoch": 1.3705179282868527,
+      "grad_norm": 0.4006193280220032,
+      "learning_rate": 8.071296073245759e-06,
+      "loss": 1.1027326583862305,
+      "step": 344
+    },
+    {
+      "epoch": 1.3784860557768925,
+      "grad_norm": 1.362949013710022,
+      "learning_rate": 8.046892030174938e-06,
+      "loss": 0.6859652400016785,
+      "step": 346
+    },
+    {
+      "epoch": 1.3864541832669324,
+      "grad_norm": 0.2718392610549927,
+      "learning_rate": 8.022377278112089e-06,
+      "loss": 1.135136604309082,
+      "step": 348
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.21031154692173004,
+      "learning_rate": 7.997752882671088e-06,
+      "loss": 0.6499249935150146,
+      "step": 350
+    },
+    {
+      "epoch": 1.402390438247012,
+      "grad_norm": 0.41847845911979675,
+      "learning_rate": 7.97301991423182e-06,
+      "loss": 1.1461623907089233,
+      "step": 352
+    },
+    {
+      "epoch": 1.4103585657370519,
+      "grad_norm": 0.5132230520248413,
+      "learning_rate": 7.948179447893652e-06,
+      "loss": 1.0844260454177856,
+      "step": 354
+    },
+    {
+      "epoch": 1.4183266932270917,
+      "grad_norm": 0.509863555431366,
+      "learning_rate": 7.923232563428693e-06,
+      "loss": 0.8674856424331665,
+      "step": 356
+    },
+    {
+      "epoch": 1.4262948207171315,
+      "grad_norm": 0.3798319399356842,
+      "learning_rate": 7.898180345234875e-06,
+      "loss": 1.1289222240447998,
+      "step": 358
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.3714427649974823,
+      "learning_rate": 7.873023882288793e-06,
+      "loss": 0.5660131573677063,
+      "step": 360
+    },
+    {
+      "epoch": 1.4422310756972112,
+      "grad_norm": 0.33141815662384033,
+      "learning_rate": 7.847764268098387e-06,
+      "loss": 0.6731166839599609,
+      "step": 362
+    },
+    {
+      "epoch": 1.450199203187251,
+      "grad_norm": 0.4289546012878418,
+      "learning_rate": 7.822402600655405e-06,
+      "loss": 1.027877688407898,
+      "step": 364
+    },
+    {
+      "epoch": 1.4581673306772909,
+      "grad_norm": 0.26489177346229553,
+      "learning_rate": 7.79693998238767e-06,
+      "loss": 1.0840867757797241,
+      "step": 366
+    },
+    {
+      "epoch": 1.4661354581673307,
+      "grad_norm": 0.4703805446624756,
+      "learning_rate": 7.771377520111163e-06,
+      "loss": 0.595399796962738,
+      "step": 368
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 0.7301031947135925,
+      "learning_rate": 7.745716324981912e-06,
+      "loss": 0.6577028036117554,
+      "step": 370
+    },
+    {
+      "epoch": 1.4820717131474104,
+      "grad_norm": 0.8861334919929504,
+      "learning_rate": 7.71995751244769e-06,
+      "loss": 1.1977068185806274,
+      "step": 372
+    },
+    {
+      "epoch": 1.4900398406374502,
+      "grad_norm": 0.7073877453804016,
+      "learning_rate": 7.694102202199536e-06,
+      "loss": 1.0823616981506348,
+      "step": 374
+    },
+    {
+      "epoch": 1.49800796812749,
+      "grad_norm": 0.25804051756858826,
+      "learning_rate": 7.66815151812307e-06,
+      "loss": 1.092564344406128,
+      "step": 376
+    },
+    {
+      "epoch": 1.5059760956175299,
+      "grad_norm": 2.229679584503174,
+      "learning_rate": 7.642106588249651e-06,
+      "loss": 0.8579627275466919,
+      "step": 378
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.32550248503685,
+      "learning_rate": 7.6159685447073376e-06,
+      "loss": 1.1093991994857788,
+      "step": 380
+    },
+    {
+      "epoch": 1.5219123505976095,
+      "grad_norm": 0.2162555605173111,
+      "learning_rate": 7.589738523671685e-06,
+      "loss": 0.6956920027732849,
+      "step": 382
+    },
+    {
+      "epoch": 1.5298804780876494,
+      "grad_norm": 0.23283204436302185,
+      "learning_rate": 7.563417665316338e-06,
+      "loss": 0.7008070945739746,
+      "step": 384
+    },
+    {
+      "epoch": 1.5378486055776892,
+      "grad_norm": 1.237845540046692,
+      "learning_rate": 7.5370071137634984e-06,
+      "loss": 0.9901597499847412,
+      "step": 386
+    },
+    {
+      "epoch": 1.545816733067729,
+      "grad_norm": 1.176262617111206,
+      "learning_rate": 7.510508017034165e-06,
+      "loss": 1.1066272258758545,
+      "step": 388
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 3.3985369205474854,
+      "learning_rate": 7.483921526998247e-06,
+      "loss": 0.5478018522262573,
+      "step": 390
+    },
+    {
+      "epoch": 1.5617529880478087,
+      "grad_norm": 0.3505213260650635,
+      "learning_rate": 7.4572487993244865e-06,
+      "loss": 0.6140211224555969,
+      "step": 392
+    },
+    {
+      "epoch": 1.5697211155378485,
+      "grad_norm": 0.6949409246444702,
+      "learning_rate": 7.430490993430232e-06,
+      "loss": 1.0416336059570312,
+      "step": 394
+    },
+    {
+      "epoch": 1.5776892430278884,
+      "grad_norm": 0.7350894808769226,
+      "learning_rate": 7.403649272431027e-06,
+      "loss": 0.9971483945846558,
+      "step": 396
+    },
+    {
+      "epoch": 1.5856573705179282,
+      "grad_norm": 0.2538151144981384,
+      "learning_rate": 7.376724803090069e-06,
+      "loss": 1.1101104021072388,
+      "step": 398
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.6578662991523743,
+      "learning_rate": 7.349718755767473e-06,
+      "loss": 0.5809187889099121,
+      "step": 400
+    },
+    {
+      "epoch": 1.6015936254980079,
+      "grad_norm": 0.5323628187179565,
+      "learning_rate": 7.322632304369416e-06,
+      "loss": 1.0850474834442139,
+      "step": 402
+    },
+    {
+      "epoch": 1.6095617529880477,
+      "grad_norm": 0.375665545463562,
+      "learning_rate": 7.2954666262970965e-06,
+      "loss": 1.2244027853012085,
+      "step": 404
+    },
+    {
+      "epoch": 1.6175298804780875,
+      "grad_norm": 0.5385187268257141,
+      "learning_rate": 7.268222902395562e-06,
+      "loss": 0.7118199467658997,
+      "step": 406
+    },
+    {
+      "epoch": 1.6254980079681274,
+      "grad_norm": 0.35690099000930786,
+      "learning_rate": 7.240902316902373e-06,
+      "loss": 0.5543506145477295,
+      "step": 408
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.3967941999435425,
+      "learning_rate": 7.213506057396134e-06,
+      "loss": 0.7297026515007019,
+      "step": 410
+    },
+    {
+      "epoch": 1.641434262948207,
+      "grad_norm": 0.28987917304039,
+      "learning_rate": 7.186035314744872e-06,
+      "loss": 0.48762401938438416,
+      "step": 412
+    },
+    {
+      "epoch": 1.6494023904382469,
+      "grad_norm": 0.21448096632957458,
+      "learning_rate": 7.1584912830542575e-06,
+      "loss": 0.785721480846405,
+      "step": 414
+    },
+    {
+      "epoch": 1.6573705179282867,
+      "grad_norm": 0.5417384505271912,
+      "learning_rate": 7.130875159615716e-06,
+      "loss": 0.7255829572677612,
+      "step": 416
+    },
+    {
+      "epoch": 1.6653386454183265,
+      "grad_norm": 0.4792822003364563,
+      "learning_rate": 7.103188144854376e-06,
+      "loss": 0.9736795425415039,
+      "step": 418
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.3885486125946045,
+      "learning_rate": 7.075431442276888e-06,
+      "loss": 1.1177918910980225,
+      "step": 420
+    },
+    {
+      "epoch": 1.6812749003984062,
+      "grad_norm": 0.3272826373577118,
+      "learning_rate": 7.047606258419108e-06,
+      "loss": 0.9807514548301697,
+      "step": 422
+    },
+    {
+      "epoch": 1.6892430278884463,
+      "grad_norm": 0.5734170079231262,
+      "learning_rate": 7.019713802793663e-06,
+      "loss": 0.5618575215339661,
+      "step": 424
+    },
+    {
+      "epoch": 1.697211155378486,
+      "grad_norm": 0.26961031556129456,
+      "learning_rate": 6.99175528783736e-06,
+      "loss": 1.1945838928222656,
+      "step": 426
+    },
+    {
+      "epoch": 1.705179282868526,
+      "grad_norm": 0.38434457778930664,
+      "learning_rate": 6.963731928858495e-06,
+      "loss": 1.0953164100646973,
+      "step": 428
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.3413749933242798,
+      "learning_rate": 6.935644943984017e-06,
+      "loss": 0.6916750073432922,
+      "step": 430
+    },
+    {
+      "epoch": 1.7211155378486056,
+      "grad_norm": 1.1177042722702026,
+      "learning_rate": 6.9074955541065865e-06,
+      "loss": 0.85158771276474,
+      "step": 432
+    },
+    {
+      "epoch": 1.7290836653386454,
+      "grad_norm": 0.4541873037815094,
+      "learning_rate": 6.8792849828315054e-06,
+      "loss": 0.6357737183570862,
+      "step": 434
+    },
+    {
+      "epoch": 1.7370517928286853,
+      "grad_norm": 0.4778435230255127,
+      "learning_rate": 6.851014456423515e-06,
+      "loss": 0.3721710443496704,
+      "step": 436
+    },
+    {
+      "epoch": 1.745019920318725,
+      "grad_norm": 0.33677831292152405,
+      "learning_rate": 6.822685203753508e-06,
+      "loss": 0.7362313270568848,
+      "step": 438
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 1.3865915536880493,
+      "learning_rate": 6.794298456245111e-06,
+      "loss": 0.6948807835578918,
+      "step": 440
+    },
+    {
+      "epoch": 1.7609561752988048,
+      "grad_norm": 1.5919826030731201,
+      "learning_rate": 6.765855447821144e-06,
+      "loss": 1.3923749923706055,
+      "step": 442
+    },
+    {
+      "epoch": 1.7689243027888446,
+      "grad_norm": 0.36454305052757263,
+      "learning_rate": 6.737357414849993e-06,
+      "loss": 0.7237080931663513,
+      "step": 444
+    },
+    {
+      "epoch": 1.7768924302788844,
+      "grad_norm": 0.26764219999313354,
+      "learning_rate": 6.70880559609187e-06,
+      "loss": 1.0288015604019165,
+      "step": 446
+    },
+    {
+      "epoch": 1.7848605577689243,
+      "grad_norm": 0.38170892000198364,
+      "learning_rate": 6.680201232644958e-06,
+      "loss": 1.068983554840088,
+      "step": 448
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.29805758595466614,
+      "learning_rate": 6.651545567891471e-06,
+      "loss": 0.8502353429794312,
+      "step": 450
+    },
+    {
+      "epoch": 1.800796812749004,
+      "grad_norm": 0.2624618709087372,
+      "learning_rate": 6.622839847443596e-06,
+      "loss": 0.4640595018863678,
+      "step": 452
+    },
+    {
+      "epoch": 1.8087649402390438,
+      "grad_norm": 0.43320029973983765,
+      "learning_rate": 6.59408531908936e-06,
+      "loss": 0.902481198310852,
+      "step": 454
+    },
+    {
+      "epoch": 1.8167330677290838,
+      "grad_norm": 0.6760742664337158,
+      "learning_rate": 6.565283232738386e-06,
+      "loss": 0.6522033214569092,
+      "step": 456
+    },
+    {
+      "epoch": 1.8247011952191237,
+      "grad_norm": 0.45959576964378357,
+      "learning_rate": 6.536434840367555e-06,
+      "loss": 1.0185953378677368,
+      "step": 458
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.5338797569274902,
+      "learning_rate": 6.507541395966591e-06,
+      "loss": 0.7339425086975098,
+      "step": 460
+    },
+    {
+      "epoch": 1.8406374501992033,
+      "grad_norm": 0.4542410671710968,
+      "learning_rate": 6.47860415548356e-06,
+      "loss": 0.7993677258491516,
+      "step": 462
+    },
+    {
+      "epoch": 1.8486055776892432,
+      "grad_norm": 0.34993016719818115,
+      "learning_rate": 6.449624376770259e-06,
+      "loss": 0.12993799149990082,
+      "step": 464
+    },
+    {
+      "epoch": 1.856573705179283,
+      "grad_norm": 0.27006521821022034,
+      "learning_rate": 6.420603319527554e-06,
+      "loss": 1.1290206909179688,
+      "step": 466
+    },
+    {
+      "epoch": 1.8645418326693228,
+      "grad_norm": 0.42941004037857056,
+      "learning_rate": 6.3915422452506146e-06,
+      "loss": 0.5305753946304321,
+      "step": 468
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 0.5610271096229553,
+      "learning_rate": 6.362442417174082e-06,
+      "loss": 1.1121795177459717,
+      "step": 470
+    },
+    {
+      "epoch": 1.8804780876494025,
+      "grad_norm": 0.8278189301490784,
+      "learning_rate": 6.33330510021716e-06,
+      "loss": 0.9842663407325745,
+      "step": 472
+    },
+    {
+      "epoch": 1.8884462151394423,
+      "grad_norm": 0.5474660992622375,
+      "learning_rate": 6.304131560928622e-06,
+      "loss": 1.0636874437332153,
+      "step": 474
+    },
+    {
+      "epoch": 1.8964143426294822,
+      "grad_norm": 0.4168717861175537,
+      "learning_rate": 6.274923067431775e-06,
+      "loss": 0.950800359249115,
+      "step": 476
+    },
+    {
+      "epoch": 1.904382470119522,
+      "grad_norm": 0.6326125264167786,
+      "learning_rate": 6.2456808893693155e-06,
+      "loss": 0.8358727097511292,
+      "step": 478
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 1.0347273349761963,
+      "learning_rate": 6.216406297848155e-06,
+      "loss": 0.5756604075431824,
+      "step": 480
+    },
+    {
+      "epoch": 1.9203187250996017,
+      "grad_norm": 1.9544227123260498,
+      "learning_rate": 6.187100565384162e-06,
+      "loss": 0.7567629814147949,
+      "step": 482
+    },
+    {
+      "epoch": 1.9282868525896415,
+      "grad_norm": 0.2865214943885803,
+      "learning_rate": 6.157764965846841e-06,
+      "loss": 1.0821741819381714,
+      "step": 484
+    },
+    {
+      "epoch": 1.9362549800796813,
+      "grad_norm": 0.36544010043144226,
+      "learning_rate": 6.128400774403974e-06,
+      "loss": 1.0638864040374756,
+      "step": 486
+    },
+    {
+      "epoch": 1.9442231075697212,
+      "grad_norm": 0.3533446490764618,
+      "learning_rate": 6.0990092674661834e-06,
+      "loss": 1.0678397417068481,
+      "step": 488
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.228616401553154,
+      "learning_rate": 6.069591722631446e-06,
+      "loss": 1.0764317512512207,
+      "step": 490
+    },
+    {
+      "epoch": 1.9601593625498008,
+      "grad_norm": 0.29699572920799255,
+      "learning_rate": 6.040149418629562e-06,
+      "loss": 0.7453048825263977,
+      "step": 492
+    },
+    {
+      "epoch": 1.9681274900398407,
+      "grad_norm": 1.6544256210327148,
+      "learning_rate": 6.010683635266569e-06,
+      "loss": 0.7180066108703613,
+      "step": 494
+    },
+    {
+      "epoch": 1.9760956175298805,
+      "grad_norm": 0.40201443433761597,
+      "learning_rate": 5.981195653369113e-06,
+      "loss": 1.0393097400665283,
+      "step": 496
+    },
+    {
+      "epoch": 1.9840637450199203,
+      "grad_norm": 0.8247522115707397,
+      "learning_rate": 5.95168675472877e-06,
+      "loss": 0.8204153180122375,
+      "step": 498
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.26851147413253784,
+      "learning_rate": 5.922158222046334e-06,
+      "loss": 1.1470904350280762,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6024758219718933,
+      "learning_rate": 5.892611338876053e-06,
+      "loss": 0.5995702147483826,
+      "step": 502
+    },
+    {
+      "epoch": 2.00796812749004,
+      "grad_norm": 0.24163378775119781,
+      "learning_rate": 5.8630473895698426e-06,
+      "loss": 0.5664758682250977,
+      "step": 504
+    },
+    {
+      "epoch": 2.0159362549800797,
+      "grad_norm": 0.5623154640197754,
+      "learning_rate": 5.833467659221452e-06,
+      "loss": 0.4919930398464203,
+      "step": 506
+    },
+    {
+      "epoch": 2.0239043824701195,
+      "grad_norm": 0.06899180263280869,
+      "learning_rate": 5.803873433610604e-06,
+      "loss": 0.5043032169342041,
+      "step": 508
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.2758553624153137,
+      "learning_rate": 5.774265999147108e-06,
+      "loss": 0.8106697201728821,
+      "step": 510
+    },
+    {
+      "epoch": 2.039840637450199,
+      "grad_norm": 0.6826291680335999,
+      "learning_rate": 5.744646642814935e-06,
+      "loss": 0.6720895767211914,
+      "step": 512
+    },
+    {
+      "epoch": 2.047808764940239,
+      "grad_norm": 0.4133363664150238,
+      "learning_rate": 5.715016652116284e-06,
+      "loss": 0.2957037091255188,
+      "step": 514
+    },
+    {
+      "epoch": 2.055776892430279,
+      "grad_norm": 0.7343904972076416,
+      "learning_rate": 5.68537731501561e-06,
+      "loss": 0.8447034358978271,
+      "step": 516
+    },
+    {
+      "epoch": 2.0637450199203187,
+      "grad_norm": 0.3274170160293579,
+      "learning_rate": 5.655729919883639e-06,
+      "loss": 0.5230758786201477,
+      "step": 518
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.5749053359031677,
+      "learning_rate": 5.626075755441366e-06,
+      "loss": 0.7700369954109192,
+      "step": 520
+    },
+    {
+      "epoch": 2.0796812749003983,
+      "grad_norm": 0.3032110929489136,
+      "learning_rate": 5.596416110704034e-06,
+      "loss": 0.8992440700531006,
+      "step": 522
+    },
+    {
+      "epoch": 2.087649402390438,
+      "grad_norm": 0.6529467105865479,
+      "learning_rate": 5.566752274925109e-06,
+      "loss": 0.8121512532234192,
+      "step": 524
+    },
+    {
+      "epoch": 2.095617529880478,
+      "grad_norm": 0.41848450899124146,
+      "learning_rate": 5.5370855375402345e-06,
+      "loss": 0.4325498342514038,
+      "step": 526
+    },
+    {
+      "epoch": 2.103585657370518,
+      "grad_norm": 0.29763972759246826,
+      "learning_rate": 5.507417188111176e-06,
+      "loss": 0.7011191248893738,
+      "step": 528
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.14057326316833496,
+      "learning_rate": 5.477748516269779e-06,
+      "loss": 0.5567289590835571,
+      "step": 530
+    },
+    {
+      "epoch": 2.1195219123505975,
+      "grad_norm": 0.26458707451820374,
+      "learning_rate": 5.448080811661897e-06,
+      "loss": 0.1889190673828125,
+      "step": 532
+    },
+    {
+      "epoch": 2.1274900398406373,
+      "grad_norm": 0.324570894241333,
+      "learning_rate": 5.418415363891346e-06,
+      "loss": 0.690158486366272,
+      "step": 534
+    },
+    {
+      "epoch": 2.135458167330677,
+      "grad_norm": 0.7778273224830627,
+      "learning_rate": 5.3887534624638335e-06,
+      "loss": 0.6308239102363586,
+      "step": 536
+    },
+    {
+      "epoch": 2.143426294820717,
+      "grad_norm": 0.6098942756652832,
+      "learning_rate": 5.359096396730924e-06,
+      "loss": 0.648769199848175,
+      "step": 538
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.3307505249977112,
+      "learning_rate": 5.329445455833975e-06,
+      "loss": 0.8563162088394165,
+      "step": 540
+    },
+    {
+      "epoch": 2.1593625498007967,
+      "grad_norm": 0.3674680292606354,
+      "learning_rate": 5.299801928648105e-06,
+      "loss": 0.8154585957527161,
+      "step": 542
+    },
+    {
+      "epoch": 2.1673306772908365,
+      "grad_norm": 0.37481123208999634,
+      "learning_rate": 5.270167103726181e-06,
+      "loss": 0.6335033178329468,
+      "step": 544
+    },
+    {
+      "epoch": 2.1752988047808763,
+      "grad_norm": 0.7283374071121216,
+      "learning_rate": 5.240542269242794e-06,
+      "loss": 0.613978922367096,
+      "step": 546
+    },
+    {
+      "epoch": 2.183266932270916,
+      "grad_norm": 0.7753714919090271,
+      "learning_rate": 5.210928712938262e-06,
+      "loss": 0.8765636086463928,
+      "step": 548
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.584176778793335,
+      "learning_rate": 5.181327722062667e-06,
+      "loss": 0.9395719766616821,
+      "step": 550
+    },
+    {
+      "epoch": 2.199203187250996,
+      "grad_norm": 0.34399765729904175,
+      "learning_rate": 5.151740583319892e-06,
+      "loss": 0.5962715148925781,
+      "step": 552
+    },
+    {
+      "epoch": 2.2071713147410357,
+      "grad_norm": 0.5032690763473511,
+      "learning_rate": 5.122168582811689e-06,
+      "loss": 0.44659191370010376,
+      "step": 554
+    },
+    {
+      "epoch": 2.2151394422310755,
+      "grad_norm": 0.32373669743537903,
+      "learning_rate": 5.092613005981781e-06,
+      "loss": 0.6227583885192871,
+      "step": 556
+    },
+    {
+      "epoch": 2.2231075697211153,
+      "grad_norm": 0.891056478023529,
+      "learning_rate": 5.063075137559979e-06,
+      "loss": 0.8263699412345886,
+      "step": 558
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.2750711441040039,
+      "learning_rate": 5.033556261506343e-06,
+      "loss": 0.8477010726928711,
+      "step": 560
+    },
+    {
+      "epoch": 2.239043824701195,
+      "grad_norm": 0.416301429271698,
+      "learning_rate": 5.0040576609553645e-06,
+      "loss": 0.9591284990310669,
+      "step": 562
+    },
+    {
+      "epoch": 2.247011952191235,
+      "grad_norm": 0.6901566386222839,
+      "learning_rate": 4.974580618160194e-06,
+      "loss": 0.7023578882217407,
+      "step": 564
+    },
+    {
+      "epoch": 2.2549800796812747,
+      "grad_norm": 0.5118059515953064,
+      "learning_rate": 4.945126414436908e-06,
+      "loss": 0.5750749111175537,
+      "step": 566
+    },
+    {
+      "epoch": 2.2629482071713145,
+      "grad_norm": 0.21457579731941223,
+      "learning_rate": 4.915696330108804e-06,
+      "loss": 0.7005121111869812,
+      "step": 568
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 1.0557022094726562,
+      "learning_rate": 4.886291644450751e-06,
+      "loss": 0.9137470126152039,
+      "step": 570
+    },
+    {
+      "epoch": 2.278884462151394,
+      "grad_norm": 0.3720467686653137,
+      "learning_rate": 4.8569136356335845e-06,
+      "loss": 0.5933997631072998,
+      "step": 572
+    },
+    {
+      "epoch": 2.2868525896414345,
+      "grad_norm": 0.312335342168808,
+      "learning_rate": 4.827563580668544e-06,
+      "loss": 0.6618290543556213,
+      "step": 574
+    },
+    {
+      "epoch": 2.2948207171314743,
+      "grad_norm": 0.17091913521289825,
+      "learning_rate": 4.798242755351755e-06,
+      "loss": 0.22380410134792328,
+      "step": 576
+    },
+    {
+      "epoch": 2.302788844621514,
+      "grad_norm": 0.47576746344566345,
+      "learning_rate": 4.768952434208795e-06,
+      "loss": 0.6028680801391602,
+      "step": 578
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.3253422975540161,
+      "learning_rate": 4.7396938904392656e-06,
+      "loss": 0.5843865871429443,
+      "step": 580
+    },
+    {
+      "epoch": 2.318725099601594,
+      "grad_norm": 0.4583335220813751,
+      "learning_rate": 4.71046839586147e-06,
+      "loss": 0.5579714179039001,
+      "step": 582
+    },
+    {
+      "epoch": 2.3266932270916336,
+      "grad_norm": 0.40016406774520874,
+      "learning_rate": 4.681277220857108e-06,
+      "loss": 0.7019505500793457,
+      "step": 584
+    },
+    {
+      "epoch": 2.3346613545816735,
+      "grad_norm": 0.4680396616458893,
+      "learning_rate": 4.6521216343160824e-06,
+      "loss": 0.5866464376449585,
+      "step": 586
+    },
+    {
+      "epoch": 2.3426294820717133,
+      "grad_norm": 0.8911101222038269,
+      "learning_rate": 4.623002903581316e-06,
+      "loss": 0.7820551991462708,
+      "step": 588
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.30490267276763916,
+      "learning_rate": 4.593922294393677e-06,
+      "loss": 0.9146092534065247,
+      "step": 590
+    },
+    {
+      "epoch": 2.358565737051793,
+      "grad_norm": 0.3790825605392456,
+      "learning_rate": 4.564881070836955e-06,
+      "loss": 0.9440783858299255,
+      "step": 592
+    },
+    {
+      "epoch": 2.366533864541833,
+      "grad_norm": 0.3498994708061218,
+      "learning_rate": 4.535880495282916e-06,
+      "loss": 0.613997757434845,
+      "step": 594
+    },
+    {
+      "epoch": 2.3745019920318726,
+      "grad_norm": 0.25710582733154297,
+      "learning_rate": 4.506921828336427e-06,
+      "loss": 0.857670247554779,
+      "step": 596
+    },
+    {
+      "epoch": 2.3824701195219125,
+      "grad_norm": 0.28142789006233215,
+      "learning_rate": 4.478006328780657e-06,
+      "loss": 0.644156277179718,
+      "step": 598
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 1.0823054313659668,
+      "learning_rate": 4.44913525352237e-06,
+      "loss": 0.7619560360908508,
+      "step": 600
+    },
+    {
+      "epoch": 2.398406374501992,
+      "grad_norm": 0.2386792004108429,
+      "learning_rate": 4.420309857537277e-06,
+      "loss": 0.9222441911697388,
+      "step": 602
+    },
+    {
+      "epoch": 2.406374501992032,
+      "grad_norm": 0.6879512071609497,
+      "learning_rate": 4.39153139381549e-06,
+      "loss": 0.5248029828071594,
+      "step": 604
+    },
+    {
+      "epoch": 2.414342629482072,
+      "grad_norm": 1.2131038904190063,
+      "learning_rate": 4.3628011133070605e-06,
+      "loss": 0.22474931180477142,
+      "step": 606
+    },
+    {
+      "epoch": 2.4223107569721116,
+      "grad_norm": 0.12215147167444229,
+      "learning_rate": 4.334120264867596e-06,
+      "loss": 0.42530250549316406,
+      "step": 608
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.15212346613407135,
+      "learning_rate": 4.305490095203979e-06,
+      "loss": 0.8002850413322449,
+      "step": 610
+    },
+    {
+      "epoch": 2.4382470119521913,
+      "grad_norm": 0.48848792910575867,
+      "learning_rate": 4.27691184882017e-06,
+      "loss": 0.6646314859390259,
+      "step": 612
+    },
+    {
+      "epoch": 2.446215139442231,
+      "grad_norm": 0.4406188130378723,
+      "learning_rate": 4.248386767963121e-06,
+      "loss": 0.5400463342666626,
+      "step": 614
+    },
+    {
+      "epoch": 2.454183266932271,
+      "grad_norm": 0.31371790170669556,
+      "learning_rate": 4.219916092568764e-06,
+      "loss": 0.8021553158760071,
+      "step": 616
+    },
+    {
+      "epoch": 2.462151394422311,
+      "grad_norm": 0.8428882956504822,
+      "learning_rate": 4.191501060208121e-06,
+      "loss": 0.6619210839271545,
+      "step": 618
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.5744040608406067,
+      "learning_rate": 4.163142906033512e-06,
+      "loss": 0.7309094071388245,
+      "step": 620
+    },
+    {
+      "epoch": 2.4780876494023905,
+      "grad_norm": 0.3610066771507263,
+      "learning_rate": 4.134842862724854e-06,
+      "loss": 0.7704513669013977,
+      "step": 622
+    },
+    {
+      "epoch": 2.4860557768924303,
+      "grad_norm": 0.8808260560035706,
+      "learning_rate": 4.106602160436088e-06,
+      "loss": 0.7354094386100769,
+      "step": 624
+    },
+    {
+      "epoch": 2.49402390438247,
+      "grad_norm": 0.48165568709373474,
+      "learning_rate": 4.078422026741706e-06,
+      "loss": 0.5148885846138,
+      "step": 626
+    },
+    {
+      "epoch": 2.50199203187251,
+      "grad_norm": 0.4090976417064667,
+      "learning_rate": 4.050303686583386e-06,
+      "loss": 0.9259688854217529,
+      "step": 628
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 0.39066246151924133,
+      "learning_rate": 4.0222483622167465e-06,
+      "loss": 0.4034539759159088,
+      "step": 630
+    },
+    {
+      "epoch": 2.5179282868525896,
+      "grad_norm": 2.5648977756500244,
+      "learning_rate": 3.9942572731582185e-06,
+      "loss": 0.37900853157043457,
+      "step": 632
+    },
+    {
+      "epoch": 2.5258964143426295,
+      "grad_norm": 0.28255292773246765,
+      "learning_rate": 3.966331636132039e-06,
+      "loss": 0.9238250255584717,
+      "step": 634
+    },
+    {
+      "epoch": 2.5338645418326693,
+      "grad_norm": 0.34794875979423523,
+      "learning_rate": 3.938472665017355e-06,
+      "loss": 0.8631218075752258,
+      "step": 636
+    },
+    {
+      "epoch": 2.541832669322709,
+      "grad_norm": 0.6642836928367615,
+      "learning_rate": 3.91068157079546e-06,
+      "loss": 0.9304293990135193,
+      "step": 638
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.17640572786331177,
+      "learning_rate": 3.882959561497159e-06,
+      "loss": 0.4698907434940338,
+      "step": 640
+    },
+    {
+      "epoch": 2.557768924302789,
+      "grad_norm": 1.0454108715057373,
+      "learning_rate": 3.855307842150259e-06,
+      "loss": 0.9286588430404663,
+      "step": 642
+    },
+    {
+      "epoch": 2.5657370517928286,
+      "grad_norm": 0.46497124433517456,
+      "learning_rate": 3.827727614727177e-06,
+      "loss": 0.8863463997840881,
+      "step": 644
+    },
+    {
+      "epoch": 2.5737051792828685,
+      "grad_norm": 0.32941934466362,
+      "learning_rate": 3.8002200780927e-06,
+      "loss": 0.5353308916091919,
+      "step": 646
+    },
+    {
+      "epoch": 2.5816733067729083,
+      "grad_norm": 0.3228965997695923,
+      "learning_rate": 3.772786427951878e-06,
+      "loss": 0.91378253698349,
+      "step": 648
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.23210027813911438,
+      "learning_rate": 3.7454278567980406e-06,
+      "loss": 0.6854094862937927,
+      "step": 650
+    },
+    {
+      "epoch": 2.597609561752988,
+      "grad_norm": 1.4749881029129028,
+      "learning_rate": 3.7181455538609577e-06,
+      "loss": 0.6934884786605835,
+      "step": 652
+    },
+    {
+      "epoch": 2.605577689243028,
+      "grad_norm": 0.37490564584732056,
+      "learning_rate": 3.690940705055163e-06,
+      "loss": 0.6151140332221985,
+      "step": 654
+    },
+    {
+      "epoch": 2.6135458167330676,
+      "grad_norm": 0.3677591383457184,
+      "learning_rate": 3.6638144929283824e-06,
+      "loss": 0.30213919281959534,
+      "step": 656
+    },
+    {
+      "epoch": 2.6215139442231075,
+      "grad_norm": 0.5509175062179565,
+      "learning_rate": 3.6367680966101483e-06,
+      "loss": 0.4912146031856537,
+      "step": 658
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.3833698630332947,
+      "learning_rate": 3.6098026917605334e-06,
+      "loss": 0.7913602590560913,
+      "step": 660
+    },
+    {
+      "epoch": 2.637450199203187,
+      "grad_norm": 0.5717480778694153,
+      "learning_rate": 3.5829194505190534e-06,
+      "loss": 0.8066682815551758,
+      "step": 662
+    },
+    {
+      "epoch": 2.645418326693227,
+      "grad_norm": 0.4665083885192871,
+      "learning_rate": 3.5561195414537164e-06,
+      "loss": 0.5120651125907898,
+      "step": 664
+    },
+    {
+      "epoch": 2.653386454183267,
+      "grad_norm": 0.49055489897727966,
+      "learning_rate": 3.529404129510221e-06,
+      "loss": 0.8346430659294128,
+      "step": 666
+    },
+    {
+      "epoch": 2.6613545816733066,
+      "grad_norm": 0.13943371176719666,
+      "learning_rate": 3.502774375961321e-06,
+      "loss": 0.5798537731170654,
+      "step": 668
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.8238434195518494,
+      "learning_rate": 3.4762314383563533e-06,
+      "loss": 0.659113347530365,
+      "step": 670
+    },
+    {
+      "epoch": 2.6772908366533863,
+      "grad_norm": 0.1340644508600235,
+      "learning_rate": 3.4497764704709143e-06,
+      "loss": 0.31450575590133667,
+      "step": 672
+    },
+    {
+      "epoch": 2.685258964143426,
+      "grad_norm": 0.26370561122894287,
+      "learning_rate": 3.4234106222567055e-06,
+      "loss": 0.927239716053009,
+      "step": 674
+    },
+    {
+      "epoch": 2.6932270916334664,
+      "grad_norm": 2.1115357875823975,
+      "learning_rate": 3.3971350397915504e-06,
+      "loss": 0.6564483046531677,
+      "step": 676
+    },
+    {
+      "epoch": 2.7011952191235062,
+      "grad_norm": 0.6908167004585266,
+      "learning_rate": 3.370950865229583e-06,
+      "loss": 0.5538385510444641,
+      "step": 678
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.28337469696998596,
+      "learning_rate": 3.344859236751585e-06,
+      "loss": 0.9321938157081604,
+      "step": 680
+    },
+    {
+      "epoch": 2.717131474103586,
+      "grad_norm": 0.3970954120159149,
+      "learning_rate": 3.3188612885155257e-06,
+      "loss": 0.9474110007286072,
+      "step": 682
+    },
+    {
+      "epoch": 2.7250996015936257,
+      "grad_norm": 0.31849491596221924,
+      "learning_rate": 3.2929581506072596e-06,
+      "loss": 0.6477122902870178,
+      "step": 684
+    },
+    {
+      "epoch": 2.7330677290836656,
+      "grad_norm": 0.4279606342315674,
+      "learning_rate": 3.267150948991387e-06,
+      "loss": 0.40889236330986023,
+      "step": 686
+    },
+    {
+      "epoch": 2.7410358565737054,
+      "grad_norm": 0.2701772153377533,
+      "learning_rate": 3.241440805462337e-06,
+      "loss": 0.9288727045059204,
+      "step": 688
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.3496691882610321,
+      "learning_rate": 3.215828837595585e-06,
+      "loss": 0.5499006509780884,
+      "step": 690
+    },
+    {
+      "epoch": 2.756972111553785,
+      "grad_norm": 0.32566186785697937,
+      "learning_rate": 3.190316158699084e-06,
+      "loss": 1.0266129970550537,
+      "step": 692
+    },
+    {
+      "epoch": 2.764940239043825,
+      "grad_norm": 0.5196744203567505,
+      "learning_rate": 3.1649038777648622e-06,
+      "loss": 0.8611117005348206,
+      "step": 694
+    },
+    {
+      "epoch": 2.7729083665338647,
+      "grad_norm": 0.9223906993865967,
+      "learning_rate": 3.1395930994208234e-06,
+      "loss": 0.5095608830451965,
+      "step": 696
+    },
+    {
+      "epoch": 2.7808764940239046,
+      "grad_norm": 0.6104478240013123,
+      "learning_rate": 3.114384923882734e-06,
+      "loss": 0.7835039496421814,
+      "step": 698
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 0.2418747842311859,
+      "learning_rate": 3.089280446906387e-06,
+      "loss": 0.9654550552368164,
+      "step": 700
+    },
+    {
+      "epoch": 2.7968127490039842,
+      "grad_norm": 0.1913435459136963,
+      "learning_rate": 3.0642807597399847e-06,
+      "loss": 0.5457607507705688,
+      "step": 702
+    },
+    {
+      "epoch": 2.804780876494024,
+      "grad_norm": 0.9318335652351379,
+      "learning_rate": 3.0393869490766913e-06,
+      "loss": 1.1307293176651,
+      "step": 704
+    },
+    {
+      "epoch": 2.812749003984064,
+      "grad_norm": 0.4223534166812897,
+      "learning_rate": 3.0146000970074087e-06,
+      "loss": 1.0508476495742798,
+      "step": 706
+    },
+    {
+      "epoch": 2.8207171314741037,
+      "grad_norm": 0.5362971425056458,
+      "learning_rate": 2.9899212809737275e-06,
+      "loss": 1.0543378591537476,
+      "step": 708
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.17238877713680267,
+      "learning_rate": 2.9653515737211037e-06,
+      "loss": 0.49260807037353516,
+      "step": 710
+    },
+    {
+      "epoch": 2.8366533864541834,
+      "grad_norm": 0.2600444257259369,
+      "learning_rate": 2.940892043252225e-06,
+      "loss": 0.9753041863441467,
+      "step": 712
+    },
+    {
+      "epoch": 2.8446215139442232,
+      "grad_norm": 0.31612956523895264,
+      "learning_rate": 2.916543752780574e-06,
+      "loss": 0.5686275959014893,
+      "step": 714
+    },
+    {
+      "epoch": 2.852589641434263,
+      "grad_norm": 0.4404768943786621,
+      "learning_rate": 2.892307760684234e-06,
+      "loss": 0.9237236976623535,
+      "step": 716
+    },
+    {
+      "epoch": 2.860557768924303,
+      "grad_norm": 1.2829992771148682,
+      "learning_rate": 2.8681851204598675e-06,
+      "loss": 0.6019064784049988,
+      "step": 718
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.2963235080242157,
+      "learning_rate": 2.844176880676931e-06,
+      "loss": 0.542336106300354,
+      "step": 720
+    },
+    {
+      "epoch": 2.8764940239043826,
+      "grad_norm": 0.4852684438228607,
+      "learning_rate": 2.8202840849320867e-06,
+      "loss": 0.5955496430397034,
+      "step": 722
+    },
+    {
+      "epoch": 2.8844621513944224,
+      "grad_norm": 0.23156464099884033,
+      "learning_rate": 2.7965077718038437e-06,
+      "loss": 0.5873034596443176,
+      "step": 724
+    },
+    {
+      "epoch": 2.8924302788844622,
+      "grad_norm": 0.33354711532592773,
+      "learning_rate": 2.7728489748074193e-06,
+      "loss": 0.8840862512588501,
+      "step": 726
+    },
+    {
+      "epoch": 2.900398406374502,
+      "grad_norm": 0.32931795716285706,
+      "learning_rate": 2.7493087223498015e-06,
+      "loss": 0.5226615071296692,
+      "step": 728
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.2865590751171112,
+      "learning_rate": 2.725888037685057e-06,
+      "loss": 0.877202033996582,
+      "step": 730
+    },
+    {
+      "epoch": 2.9163346613545817,
+      "grad_norm": 0.24265752732753754,
+      "learning_rate": 2.702587938869842e-06,
+      "loss": 0.35362404584884644,
+      "step": 732
+    },
+    {
+      "epoch": 2.9243027888446216,
+      "grad_norm": 0.37618952989578247,
+      "learning_rate": 2.6794094387191617e-06,
+      "loss": 0.9530909657478333,
+      "step": 734
+    },
+    {
+      "epoch": 2.9322709163346614,
+      "grad_norm": 0.8494628667831421,
+      "learning_rate": 2.6563535447623302e-06,
+      "loss": 0.48720675706863403,
+      "step": 736
+    },
+    {
+      "epoch": 2.9402390438247012,
+      "grad_norm": 0.6560754776000977,
+      "learning_rate": 2.6334212591991847e-06,
+      "loss": 0.7162171602249146,
+      "step": 738
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.3006666600704193,
+      "learning_rate": 2.6106135788565248e-06,
+      "loss": 1.0109399557113647,
+      "step": 740
+    },
+    {
+      "epoch": 2.956175298804781,
+      "grad_norm": 0.8110612630844116,
+      "learning_rate": 2.5879314951447688e-06,
+      "loss": 0.3084184229373932,
+      "step": 742
+    },
+    {
+      "epoch": 2.9641434262948207,
+      "grad_norm": 0.2997901737689972,
+      "learning_rate": 2.5653759940148692e-06,
+      "loss": 0.8004326224327087,
+      "step": 744
+    },
+    {
+      "epoch": 2.9721115537848606,
+      "grad_norm": 0.3199838399887085,
+      "learning_rate": 2.542948055915454e-06,
+      "loss": 0.7998283505439758,
+      "step": 746
+    },
+    {
+      "epoch": 2.9800796812749004,
+      "grad_norm": 0.2956474721431732,
+      "learning_rate": 2.5206486557502087e-06,
+      "loss": 0.45612776279449463,
+      "step": 748
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.23385298252105713,
+      "learning_rate": 2.498478762835492e-06,
+      "loss": 0.13049250841140747,
+      "step": 750
+    },
+    {
+      "epoch": 2.99601593625498,
+      "grad_norm": 0.29183921217918396,
+      "learning_rate": 2.476439340858208e-06,
+      "loss": 0.9207424521446228,
+      "step": 752
+    },
+    {
+      "epoch": 3.00398406374502,
+      "grad_norm": 0.3017852008342743,
+      "learning_rate": 2.4545313478339178e-06,
+      "loss": 0.49790331721305847,
+      "step": 754
+    },
+    {
+      "epoch": 3.0119521912350598,
+      "grad_norm": 1.0036803483963013,
+      "learning_rate": 2.4327557360651895e-06,
+      "loss": 0.4992743730545044,
+      "step": 756
+    },
+    {
+      "epoch": 3.0199203187250996,
+      "grad_norm": 0.14302383363246918,
+      "learning_rate": 2.411113452100211e-06,
+      "loss": 0.3142014443874359,
+      "step": 758
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.8177183270454407,
+      "learning_rate": 2.3896054366916365e-06,
+      "loss": 0.5167007446289062,
+      "step": 760
+    },
+    {
+      "epoch": 3.0358565737051793,
+      "grad_norm": 0.6338479518890381,
+      "learning_rate": 2.3682326247557e-06,
+      "loss": 0.38520339131355286,
+      "step": 762
+    },
+    {
+      "epoch": 3.043824701195219,
+      "grad_norm": 0.4269687831401825,
+      "learning_rate": 2.346995945331576e-06,
+      "loss": 0.78280109167099,
+      "step": 764
+    },
+    {
+      "epoch": 3.051792828685259,
+      "grad_norm": 0.30745407938957214,
+      "learning_rate": 2.325896321540993e-06,
+      "loss": 0.1655711829662323,
+      "step": 766
+    },
+    {
+      "epoch": 3.0597609561752988,
+      "grad_norm": 0.34582817554473877,
+      "learning_rate": 2.3049346705481086e-06,
+      "loss": 0.47058436274528503,
+      "step": 768
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.4358985722064972,
+      "learning_rate": 2.284111903519641e-06,
+      "loss": 0.6077016592025757,
+      "step": 770
+    },
+    {
+      "epoch": 3.0756972111553784,
+      "grad_norm": 0.5212319493293762,
+      "learning_rate": 2.26342892558526e-06,
+      "loss": 0.7947466969490051,
+      "step": 772
+    },
+    {
+      "epoch": 3.0836653386454183,
+      "grad_norm": 0.6922328472137451,
+      "learning_rate": 2.2428866357982515e-06,
+      "loss": 0.7335124611854553,
+      "step": 774
+    },
+    {
+      "epoch": 3.091633466135458,
+      "grad_norm": 0.044602397829294205,
+      "learning_rate": 2.222485927096424e-06,
+      "loss": 0.43842315673828125,
+      "step": 776
+    },
+    {
+      "epoch": 3.099601593625498,
+      "grad_norm": 0.5351706743240356,
+      "learning_rate": 2.2022276862633075e-06,
+      "loss": 0.4657999277114868,
+      "step": 778
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.2748660743236542,
+      "learning_rate": 2.1821127938895938e-06,
+      "loss": 0.7745958566665649,
+      "step": 780
+    },
+    {
+      "epoch": 3.1155378486055776,
+      "grad_norm": 0.11634573340415955,
+      "learning_rate": 2.162142124334871e-06,
+      "loss": 0.20473244786262512,
+      "step": 782
+    },
+    {
+      "epoch": 3.1235059760956174,
+      "grad_norm": 0.3060646951198578,
+      "learning_rate": 2.1423165456896068e-06,
+      "loss": 0.846551775932312,
+      "step": 784
+    },
+    {
+      "epoch": 3.1314741035856573,
+      "grad_norm": 0.28841432929039,
+      "learning_rate": 2.1226369197374176e-06,
+      "loss": 0.6212831139564514,
+      "step": 786
+    },
+    {
+      "epoch": 3.139442231075697,
+      "grad_norm": 0.8471628427505493,
+      "learning_rate": 2.103104101917615e-06,
+      "loss": 0.4564535319805145,
+      "step": 788
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.1288202852010727,
+      "learning_rate": 2.0837189412880044e-06,
+      "loss": 0.4034203588962555,
+      "step": 790
+    },
+    {
+      "epoch": 3.1553784860557768,
+      "grad_norm": 0.8081262111663818,
+      "learning_rate": 2.0644822804879964e-06,
+      "loss": 0.560878574848175,
+      "step": 792
+    },
+    {
+      "epoch": 3.1633466135458166,
+      "grad_norm": 0.06400708854198456,
+      "learning_rate": 2.045394955701969e-06,
+      "loss": 0.15414239466190338,
+      "step": 794
+    },
+    {
+      "epoch": 3.1713147410358564,
+      "grad_norm": 0.5684446096420288,
+      "learning_rate": 2.0264577966229222e-06,
+      "loss": 0.48147594928741455,
+      "step": 796
+    },
+    {
+      "epoch": 3.1792828685258963,
+      "grad_norm": 0.4146941304206848,
+      "learning_rate": 2.007671626416412e-06,
+      "loss": 0.4810021221637726,
+      "step": 798
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.42831286787986755,
+      "learning_rate": 1.9890372616847663e-06,
+      "loss": 0.6828551292419434,
+      "step": 800
+    },
+    {
+      "epoch": 3.195219123505976,
+      "grad_norm": 0.22842170298099518,
+      "learning_rate": 1.9705555124315965e-06,
+      "loss": 0.562218964099884,
+      "step": 802
+    },
+    {
+      "epoch": 3.2031872509960158,
+      "grad_norm": 0.5850492715835571,
+      "learning_rate": 1.95222718202658e-06,
+      "loss": 0.49006983637809753,
+      "step": 804
+    },
+    {
+      "epoch": 3.2111553784860556,
+      "grad_norm": 1.846373200416565,
+      "learning_rate": 1.934053067170545e-06,
+      "loss": 0.7168378233909607,
+      "step": 806
+    },
+    {
+      "epoch": 3.2191235059760954,
+      "grad_norm": 1.1755845546722412,
+      "learning_rate": 1.9160339578608305e-06,
+      "loss": 0.11919846385717392,
+      "step": 808
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 0.5742408633232117,
+      "learning_rate": 1.8981706373569602e-06,
+      "loss": 0.5427244901657104,
+      "step": 810
+    },
+    {
+      "epoch": 3.235059760956175,
+      "grad_norm": 0.3801616132259369,
+      "learning_rate": 1.8804638821465821e-06,
+      "loss": 0.3534199297428131,
+      "step": 812
+    },
+    {
+      "epoch": 3.243027888446215,
+      "grad_norm": 0.2760299742221832,
+      "learning_rate": 1.8629144619117234e-06,
+      "loss": 0.6733813285827637,
+      "step": 814
+    },
+    {
+      "epoch": 3.2509960159362548,
+      "grad_norm": 1.4330555200576782,
+      "learning_rate": 1.8455231394953348e-06,
+      "loss": 0.1837894469499588,
+      "step": 816
+    },
+    {
+      "epoch": 3.2589641434262946,
+      "grad_norm": 0.3000759482383728,
+      "learning_rate": 1.8282906708681238e-06,
+      "loss": 0.773911714553833,
+      "step": 818
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 7.039217472076416,
+      "learning_rate": 1.811217805095699e-06,
+      "loss": 0.3553723394870758,
+      "step": 820
+    },
+    {
+      "epoch": 3.2749003984063743,
+      "grad_norm": 1.9292426109313965,
+      "learning_rate": 1.7943052843060133e-06,
+      "loss": 0.48876044154167175,
+      "step": 822
+    },
+    {
+      "epoch": 3.2828685258964145,
+      "grad_norm": 1.1212517023086548,
+      "learning_rate": 1.7775538436570937e-06,
+      "loss": 0.4679087996482849,
+      "step": 824
+    },
+    {
+      "epoch": 3.2908366533864544,
+      "grad_norm": 0.31176719069480896,
+      "learning_rate": 1.7609642113050996e-06,
+      "loss": 0.7786960005760193,
+      "step": 826
+    },
+    {
+      "epoch": 3.298804780876494,
+      "grad_norm": 0.6527950763702393,
+      "learning_rate": 1.7445371083726567e-06,
+      "loss": 0.8031498193740845,
+      "step": 828
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.5439575910568237,
+      "learning_rate": 1.7282732489175224e-06,
+      "loss": 0.6720057129859924,
+      "step": 830
+    },
+    {
+      "epoch": 3.314741035856574,
+      "grad_norm": 0.9103624224662781,
+      "learning_rate": 1.7121733399015406e-06,
+      "loss": 0.3532686233520508,
+      "step": 832
+    },
+    {
+      "epoch": 3.3227091633466137,
+      "grad_norm": 0.3843187987804413,
+      "learning_rate": 1.6962380811599138e-06,
+      "loss": 0.25262296199798584,
+      "step": 834
+    },
+    {
+      "epoch": 3.3306772908366535,
+      "grad_norm": 1.6557152271270752,
+      "learning_rate": 1.6804681653707816e-06,
+      "loss": 0.43335628509521484,
+      "step": 836
+    },
+    {
+      "epoch": 3.3386454183266934,
+      "grad_norm": 0.35895177721977234,
+      "learning_rate": 1.6648642780251096e-06,
+      "loss": 0.6557596921920776,
+      "step": 838
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 1.0864629745483398,
+      "learning_rate": 1.6494270973968974e-06,
+      "loss": 0.1416988968849182,
+      "step": 840
+    },
+    {
+      "epoch": 3.354581673306773,
+      "grad_norm": 0.49732112884521484,
+      "learning_rate": 1.6341572945136902e-06,
+      "loss": 0.48536258935928345,
+      "step": 842
+    },
+    {
+      "epoch": 3.362549800796813,
+      "grad_norm": 0.5417231321334839,
+      "learning_rate": 1.6190555331274122e-06,
+      "loss": 0.5153325796127319,
+      "step": 844
+    },
+    {
+      "epoch": 3.3705179282868527,
+      "grad_norm": 0.3946952521800995,
+      "learning_rate": 1.6041224696855145e-06,
+      "loss": 0.831354558467865,
+      "step": 846
+    },
+    {
+      "epoch": 3.3784860557768925,
+      "grad_norm": 0.33575886487960815,
+      "learning_rate": 1.589358753302439e-06,
+      "loss": 0.7561607956886292,
+      "step": 848
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 0.9700213670730591,
+      "learning_rate": 1.5747650257314068e-06,
+      "loss": 0.7340743541717529,
+      "step": 850
+    },
+    {
+      "epoch": 3.394422310756972,
+      "grad_norm": 0.6107082962989807,
+      "learning_rate": 1.5603419213365146e-06,
+      "loss": 0.7586410641670227,
+      "step": 852
+    },
+    {
+      "epoch": 3.402390438247012,
+      "grad_norm": 0.5999823808670044,
+      "learning_rate": 1.5460900670651705e-06,
+      "loss": 0.44386446475982666,
+      "step": 854
+    },
+    {
+      "epoch": 3.410358565737052,
+      "grad_norm": 0.37810757756233215,
+      "learning_rate": 1.5320100824208306e-06,
+      "loss": 0.7061421871185303,
+      "step": 856
+    },
+    {
+      "epoch": 3.4183266932270917,
+      "grad_norm": 0.3629280626773834,
+      "learning_rate": 1.51810257943608e-06,
+      "loss": 0.8219012022018433,
+      "step": 858
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 1.0155540704727173,
+      "learning_rate": 1.5043681626460196e-06,
+      "loss": 0.8624923825263977,
+      "step": 860
+    },
+    {
+      "epoch": 3.4342629482071714,
+      "grad_norm": 0.9448961615562439,
+      "learning_rate": 1.4908074290619972e-06,
+      "loss": 0.5248640775680542,
+      "step": 862
+    },
+    {
+      "epoch": 3.442231075697211,
+      "grad_norm": 0.7457435131072998,
+      "learning_rate": 1.477420968145652e-06,
+      "loss": 0.4058370292186737,
+      "step": 864
+    },
+    {
+      "epoch": 3.450199203187251,
+      "grad_norm": 0.4264843165874481,
+      "learning_rate": 1.4642093617832853e-06,
+      "loss": 0.650264322757721,
+      "step": 866
+    },
+    {
+      "epoch": 3.458167330677291,
+      "grad_norm": 0.3789123594760895,
+      "learning_rate": 1.4511731842605805e-06,
+      "loss": 0.5618629455566406,
+      "step": 868
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.30681291222572327,
+      "learning_rate": 1.4383130022376298e-06,
+      "loss": 0.438922256231308,
+      "step": 870
+    },
+    {
+      "epoch": 3.4741035856573705,
+      "grad_norm": 0.1074134036898613,
+      "learning_rate": 1.4256293747243065e-06,
+      "loss": 0.46597009897232056,
+      "step": 872
+    },
+    {
+      "epoch": 3.4820717131474104,
+      "grad_norm": 0.16204528510570526,
+      "learning_rate": 1.4131228530559637e-06,
+      "loss": 0.5063215494155884,
+      "step": 874
+    },
+    {
+      "epoch": 3.49003984063745,
+      "grad_norm": 0.2670261263847351,
+      "learning_rate": 1.400793980869468e-06,
+      "loss": 0.40464305877685547,
+      "step": 876
+    },
+    {
+      "epoch": 3.49800796812749,
+      "grad_norm": 0.31625449657440186,
+      "learning_rate": 1.3886432940795744e-06,
+      "loss": 0.4505210816860199,
+      "step": 878
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.12725640833377838,
+      "learning_rate": 1.3766713208556232e-06,
+      "loss": 0.30271726846694946,
+      "step": 880
+    },
+    {
+      "epoch": 3.5139442231075697,
+      "grad_norm": 0.3201422095298767,
+      "learning_rate": 1.3648785815985876e-06,
+      "loss": 0.385304719209671,
+      "step": 882
+    },
+    {
+      "epoch": 3.5219123505976095,
+      "grad_norm": 0.4787532091140747,
+      "learning_rate": 1.353265588918448e-06,
+      "loss": 0.1267271190881729,
+      "step": 884
+    },
+    {
+      "epoch": 3.5298804780876494,
+      "grad_norm": 0.5273363590240479,
+      "learning_rate": 1.3418328476119136e-06,
+      "loss": 0.6429634690284729,
+      "step": 886
+    },
+    {
+      "epoch": 3.537848605577689,
+      "grad_norm": 0.49855756759643555,
+      "learning_rate": 1.330580854640477e-06,
+      "loss": 0.7288384437561035,
+      "step": 888
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.3828088939189911,
+      "learning_rate": 1.3195100991088142e-06,
+      "loss": 0.4124334454536438,
+      "step": 890
+    },
+    {
+      "epoch": 3.553784860557769,
+      "grad_norm": 0.765853226184845,
+      "learning_rate": 1.308621062243524e-06,
+      "loss": 0.5804932117462158,
+      "step": 892
+    },
+    {
+      "epoch": 3.5617529880478087,
+      "grad_norm": 0.33631619811058044,
+      "learning_rate": 1.2979142173722054e-06,
+      "loss": 0.8949434161186218,
+      "step": 894
+    },
+    {
+      "epoch": 3.5697211155378485,
+      "grad_norm": 0.21261455118656158,
+      "learning_rate": 1.2873900299028896e-06,
+      "loss": 0.19207896292209625,
+      "step": 896
+    },
+    {
+      "epoch": 3.5776892430278884,
+      "grad_norm": 0.32976871728897095,
+      "learning_rate": 1.2770489573038073e-06,
+      "loss": 0.49919167160987854,
+      "step": 898
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.26310229301452637,
+      "learning_rate": 1.2668914490834983e-06,
+      "loss": 0.16562813520431519,
+      "step": 900
+    },
+    {
+      "epoch": 3.593625498007968,
+      "grad_norm": 0.49629366397857666,
+      "learning_rate": 1.2569179467712805e-06,
+      "loss": 0.09022018313407898,
+      "step": 902
+    },
+    {
+      "epoch": 3.601593625498008,
+      "grad_norm": 0.41529035568237305,
+      "learning_rate": 1.2471288838980505e-06,
+      "loss": 0.5692539811134338,
+      "step": 904
+    },
+    {
+      "epoch": 3.6095617529880477,
+      "grad_norm": 0.4117538630962372,
+      "learning_rate": 1.237524685977444e-06,
+      "loss": 0.4580269157886505,
+      "step": 906
+    },
+    {
+      "epoch": 3.6175298804780875,
+      "grad_norm": 0.5062305331230164,
+      "learning_rate": 1.2281057704873331e-06,
+      "loss": 0.7049476504325867,
+      "step": 908
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 0.4529809355735779,
+      "learning_rate": 1.2188725468516876e-06,
+      "loss": 0.6734727025032043,
+      "step": 910
+    },
+    {
+      "epoch": 3.633466135458167,
+      "grad_norm": 0.11122507601976395,
+      "learning_rate": 1.2098254164227706e-06,
+      "loss": 0.043384622782468796,
+      "step": 912
+    },
+    {
+      "epoch": 3.641434262948207,
+      "grad_norm": 0.39151912927627563,
+      "learning_rate": 1.200964772463695e-06,
+      "loss": 0.6741068363189697,
+      "step": 914
+    },
+    {
+      "epoch": 3.649402390438247,
+      "grad_norm": 1.576507568359375,
+      "learning_rate": 1.1922910001313312e-06,
+      "loss": 0.6726958751678467,
+      "step": 916
+    },
+    {
+      "epoch": 3.6573705179282867,
+      "grad_norm": 0.8982887268066406,
+      "learning_rate": 1.1838044764595617e-06,
+      "loss": 0.12227790057659149,
+      "step": 918
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 0.39536726474761963,
+      "learning_rate": 1.1755055703428958e-06,
+      "loss": 0.7642967104911804,
+      "step": 920
+    },
+    {
+      "epoch": 3.6733067729083664,
+      "grad_norm": 0.6641975045204163,
+      "learning_rate": 1.1673946425204288e-06,
+      "loss": 0.08700887113809586,
+      "step": 922
+    },
+    {
+      "epoch": 3.681274900398406,
+      "grad_norm": 0.7231969237327576,
+      "learning_rate": 1.1594720455601664e-06,
+      "loss": 0.6467123031616211,
+      "step": 924
+    },
+    {
+      "epoch": 3.6892430278884465,
+      "grad_norm": 0.3027903735637665,
+      "learning_rate": 1.1517381238436986e-06,
+      "loss": 0.20401689410209656,
+      "step": 926
+    },
+    {
+      "epoch": 3.6972111553784863,
+      "grad_norm": 0.34666118025779724,
+      "learning_rate": 1.1441932135512262e-06,
+      "loss": 0.4319857358932495,
+      "step": 928
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.14203816652297974,
+      "learning_rate": 1.1368376426469526e-06,
+      "loss": 0.6425674557685852,
+      "step": 930
+    },
+    {
+      "epoch": 3.713147410358566,
+      "grad_norm": 0.2937178313732147,
+      "learning_rate": 1.1296717308648237e-06,
+      "loss": 0.7667829394340515,
+      "step": 932
+    },
+    {
+      "epoch": 3.721115537848606,
+      "grad_norm": 0.5775947570800781,
+      "learning_rate": 1.1226957896946332e-06,
+      "loss": 0.37052983045578003,
+      "step": 934
+    },
+    {
+      "epoch": 3.7290836653386457,
+      "grad_norm": 0.47730228304862976,
+      "learning_rate": 1.115910122368479e-06,
+      "loss": 0.6962873935699463,
+      "step": 936
+    },
+    {
+      "epoch": 3.7370517928286855,
+      "grad_norm": 0.9174061417579651,
+      "learning_rate": 1.1093150238475847e-06,
+      "loss": 0.3150583803653717,
+      "step": 938
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.33844903111457825,
+      "learning_rate": 1.1029107808094789e-06,
+      "loss": 0.6417381167411804,
+      "step": 940
+    },
+    {
+      "epoch": 3.752988047808765,
+      "grad_norm": 1.444950819015503,
+      "learning_rate": 1.0966976716355294e-06,
+      "loss": 0.4399506151676178,
+      "step": 942
+    },
+    {
+      "epoch": 3.760956175298805,
+      "grad_norm": 0.15434329211711884,
+      "learning_rate": 1.0906759663988475e-06,
+      "loss": 0.42593681812286377,
+      "step": 944
+    },
+    {
+      "epoch": 3.768924302788845,
+      "grad_norm": 0.39969784021377563,
+      "learning_rate": 1.084845926852547e-06,
+      "loss": 0.8348492383956909,
+      "step": 946
+    },
+    {
+      "epoch": 3.7768924302788847,
+      "grad_norm": 0.4904302954673767,
+      "learning_rate": 1.0792078064183652e-06,
+      "loss": 0.44714832305908203,
+      "step": 948
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.2667109966278076,
+      "learning_rate": 1.0737618501756455e-06,
+      "loss": 0.7914283871650696,
+      "step": 950
+    },
+    {
+      "epoch": 3.7928286852589643,
+      "grad_norm": 0.30467745661735535,
+      "learning_rate": 1.0685082948506875e-06,
+      "loss": 0.77543705701828,
+      "step": 952
+    },
+    {
+      "epoch": 3.800796812749004,
+      "grad_norm": 0.3196500837802887,
+      "learning_rate": 1.063447368806457e-06,
+      "loss": 0.41149428486824036,
+      "step": 954
+    },
+    {
+      "epoch": 3.808764940239044,
+      "grad_norm": 0.4178663194179535,
+      "learning_rate": 1.0585792920326565e-06,
+      "loss": 0.5832157135009766,
+      "step": 956
+    },
+    {
+      "epoch": 3.816733067729084,
+      "grad_norm": 0.8809601664543152,
+      "learning_rate": 1.0539042761361646e-06,
+      "loss": 0.4472152590751648,
+      "step": 958
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 0.7658699750900269,
+      "learning_rate": 1.0494225243318373e-06,
+      "loss": 0.4110378623008728,
+      "step": 960
+    },
+    {
+      "epoch": 3.8326693227091635,
+      "grad_norm": 0.33869850635528564,
+      "learning_rate": 1.0451342314336755e-06,
+      "loss": 0.4653624892234802,
+      "step": 962
+    },
+    {
+      "epoch": 3.8406374501992033,
+      "grad_norm": 0.6041797995567322,
+      "learning_rate": 1.0410395838463547e-06,
+      "loss": 0.8142421245574951,
+      "step": 964
+    },
+    {
+      "epoch": 3.848605577689243,
+      "grad_norm": 0.527969479560852,
+      "learning_rate": 1.0371387595571248e-06,
+      "loss": 0.26374515891075134,
+      "step": 966
+    },
+    {
+      "epoch": 3.856573705179283,
+      "grad_norm": 0.545873761177063,
+      "learning_rate": 1.0334319281280715e-06,
+      "loss": 0.4810965061187744,
+      "step": 968
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 0.557353675365448,
+      "learning_rate": 1.0299192506887467e-06,
+      "loss": 0.5515950918197632,
+      "step": 970
+    },
+    {
+      "epoch": 3.8725099601593627,
+      "grad_norm": 0.07388067245483398,
+      "learning_rate": 1.0266008799291635e-06,
+      "loss": 0.4826350510120392,
+      "step": 972
+    },
+    {
+      "epoch": 3.8804780876494025,
+      "grad_norm": 0.09544648975133896,
+      "learning_rate": 1.0234769600931612e-06,
+      "loss": 0.35495859384536743,
+      "step": 974
+    },
+    {
+      "epoch": 3.8884462151394423,
+      "grad_norm": 0.3048069477081299,
+      "learning_rate": 1.0205476269721312e-06,
+      "loss": 0.7828022241592407,
+      "step": 976
+    },
+    {
+      "epoch": 3.896414342629482,
+      "grad_norm": 1.1493760347366333,
+      "learning_rate": 1.0178130078991196e-06,
+      "loss": 0.4746944010257721,
+      "step": 978
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.12432657927274704,
+      "learning_rate": 1.0152732217432875e-06,
+      "loss": 0.31902119517326355,
+      "step": 980
+    },
+    {
+      "epoch": 3.912350597609562,
+      "grad_norm": 0.45761051774024963,
+      "learning_rate": 1.0129283789047473e-06,
+      "loss": 0.32653486728668213,
+      "step": 982
+    },
+    {
+      "epoch": 3.9203187250996017,
+      "grad_norm": 0.38355180621147156,
+      "learning_rate": 1.0107785813097621e-06,
+      "loss": 0.7637002468109131,
+      "step": 984
+    },
+    {
+      "epoch": 3.9282868525896415,
+      "grad_norm": 0.3457726538181305,
+      "learning_rate": 1.0088239224063153e-06,
+      "loss": 0.1351126879453659,
+      "step": 986
+    },
+    {
+      "epoch": 3.9362549800796813,
+      "grad_norm": 1.2162338495254517,
+      "learning_rate": 1.0070644871600495e-06,
+      "loss": 0.5420807600021362,
+      "step": 988
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.10997278243303299,
+      "learning_rate": 1.0055003520505714e-06,
+      "loss": 0.43679797649383545,
+      "step": 990
+    },
+    {
+      "epoch": 3.952191235059761,
+      "grad_norm": 0.4889715909957886,
+      "learning_rate": 1.00413158506813e-06,
+      "loss": 0.7833524942398071,
+      "step": 992
+    },
+    {
+      "epoch": 3.960159362549801,
+      "grad_norm": 1.1590372323989868,
+      "learning_rate": 1.0029582457106588e-06,
+      "loss": 0.2577362358570099,
+      "step": 994
+    },
+    {
+      "epoch": 3.9681274900398407,
+      "grad_norm": 0.3678681254386902,
+      "learning_rate": 1.00198038498119e-06,
+      "loss": 0.6670050621032715,
+      "step": 996
+    },
+    {
+      "epoch": 3.9760956175298805,
+      "grad_norm": 0.3324994444847107,
+      "learning_rate": 1.0011980453856391e-06,
+      "loss": 0.4599224030971527,
+      "step": 998
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 0.38484975695610046,
+      "learning_rate": 1.0006112609309548e-06,
+      "loss": 0.32395753264427185,
+      "step": 1000
+    },
+    {
+      "epoch": 3.99203187250996,
+      "grad_norm": 0.3691813349723816,
+      "learning_rate": 1.0002200571236429e-06,
+      "loss": 0.7829529047012329,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.13473699986934662,
+      "learning_rate": 1.0000244509686564e-06,
+      "loss": 0.19491368532180786,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0,
+      "step": 1004,
+      "total_flos": 3.519329211045118e+18,
+      "train_loss": 0.8034457557749345,
+      "train_runtime": 9234.9215,
+      "train_samples_per_second": 6.523,
+      "train_steps_per_second": 0.109
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1004,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.519329211045118e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

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