grail_v2op_fcurlong_b64_sft_e3_ck640

.gitattributes

@@ -81,3 +81,4 @@ grail_v2op_fcur_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -t
 grail_v5_fcur_op48k_b64_sft_e3_ck800/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 grail_v2op_fcur_long_b64_sft_e3_ck1280/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 grail_v2op_fcur_sr_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+grail_v2op_fcurlong_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text

grail_v2op_fcurlong_b64_sft_e3_ck640/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 0.9576358365285701,
+    "eval_loss": 0.03894830867648125,
+    "eval_runtime": 163.4194,
+    "eval_samples_per_second": 13.78,
+    "eval_steps_per_second": 3.445,
+    "total_flos": 283306748805120.0,
+    "train_loss": 0.10237010675482452,
+    "train_runtime": 7722.2236,
+    "train_samples_per_second": 5.304,
+    "train_steps_per_second": 0.083
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/chat_template.jinja

@@ -0,0 +1,120 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {%- if messages[0].content is string %}
+            {{- messages[0].content }}
+        {%- else %}
+            {%- for content in messages[0].content %}
+                {%- if 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' }}
+        {%- if messages[0].content is string %}
+            {{- messages[0].content }}
+        {%- else %}
+            {%- for content in messages[0].content %}
+                {%- if 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- for message in messages %}
+    {%- if message.role == "user" %}
+        {{- '<|im_start|>' + message.role + '\n' }}
+        {%- if message.content is string %}
+            {{- message.content }}
+        {%- else %}
+            {%- for content in message.content %}
+                {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                    {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
+                    <|vision_start|><|image_pad|><|vision_end|>
+                {%- elif content.type == 'video' or 'video' in content %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                    {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
+                    <|vision_start|><|video_pad|><|vision_end|>
+                {%- elif 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role + '\n' }}
+        {%- if message.content is string %}
+            {{- message.content }}
+        {%- else %}
+            {%- for content_item in message.content %}
+                {%- if 'text' in content_item %}
+                    {{- content_item.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and message.content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {%- if message.content is string %}
+            {{- message.content }}
+        {%- else %}
+            {%- for content in message.content %}
+                {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                    {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
+                    <|vision_start|><|image_pad|><|vision_end|>
+                {%- elif content.type == 'video' or 'video' in content %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                    {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
+                    <|vision_start|><|video_pad|><|vision_end|>
+                {%- elif 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

grail_v2op_fcurlong_b64_sft_e3_ck640/config.json

@@ -0,0 +1,70 @@
+{
+  "architectures": [
+    "Qwen3VLForConditionalGeneration"
+  ],
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "hidden_size": 4096,
+  "image_token_id": 151655,
+  "model_type": "qwen3_vl",
+  "pad_token_id": 151643,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 151643,
+    "dtype": "bfloat16",
+    "eos_token_id": 151645,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "initializer_range": 0.02,
+    "intermediate_size": 12288,
+    "max_position_embeddings": 262144,
+    "model_type": "qwen3_vl_text",
+    "num_attention_heads": 32,
+    "num_hidden_layers": 36,
+    "num_key_value_heads": 8,
+    "pad_token_id": null,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        24,
+        20,
+        20
+      ],
+      "rope_theta": 5000000,
+      "rope_type": "default"
+    },
+    "use_cache": false,
+    "vocab_size": 151936
+  },
+  "tie_word_embeddings": false,
+  "transformers_version": "5.8.0",
+  "use_cache": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "deepstack_visual_indexes": [
+      8,
+      16,
+      24
+    ],
+    "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_vl_vision",
+    "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": 151653,
+  "vision_start_token_id": 151652
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 0.9576358365285701,
+    "eval_loss": 0.03894830867648125,
+    "eval_runtime": 163.4194,
+    "eval_samples_per_second": 13.78,
+    "eval_steps_per_second": 3.445
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.0,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.8.0"
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6404f38bb3caacbeb7dd6af6ec84144aee0d9ac5a93fb523a66138f7c1ee3151
+size 17534340584

grail_v2op_fcurlong_b64_sft_e3_ck640/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"
+  }
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/sft.yaml

@@ -0,0 +1,58 @@
+# LLaMA-Factory full-parameter SFT of Qwen3-VL-8B-Instruct on one HER-final dataset.
+# Four L40S (46 GB), one run at a time: DeepSpeed ZeRO-3 with optimizer offload to host
+# RAM (~130 GB per run, so two runs do not fit in 372 GB); the vision tower and projector
+# stay frozen (LLaMA-Factory's default for VLMs), the LLM trains.
+# Filled per run by launch_sft.sh: grail_v2op_fcurlong_sft_e3, exps/sft/grail_v2op_fcurlong_sft_e3_k25515, /scratch/kwc5354/viewagent/viewagent_private/exps/sft_runs/grail_v2op_fcurlong_b64_sft_e3.
+### model
+model_name_or_path: /scratch/kwc5354/viewagent/viewagent_private/exps/models/graph_nemi_sft_e2
+template: qwen3_vl_nothink
+trust_remote_code: true
+image_max_pixels: 65536          # frames are 256x256 already; never upscale
+image_min_pixels: 4096
+
+### method
+stage: sft
+do_train: true
+finetuning_type: full
+freeze_vision_tower: true
+freeze_multi_modal_projector: true
+freeze_language_model: false
+deepspeed: /scratch/kwc5354/viewagent/viewagent_private/LLaMA-Factory/examples/deepspeed/ds_z3_config.json
+
+### dataset
+dataset: grail_v2op_fcurlong_sft_e3
+dataset_dir: exps/sft/grail_v2op_fcurlong_sft_e3_k25515
+media_dir: /scratch/kwc5354/viewagent/viewagent_private
+cutoff_len: 8192
+overwrite_cache: true
+preprocessing_num_workers: 16
+dataloader_num_workers: 4
+
+### output
+output_dir: /scratch/kwc5354/viewagent/viewagent_private/exps/sft_runs/grail_v2op_fcurlong_b64_sft_e3
+logging_steps: 10
+save_strategy: steps
+save_steps: 320
+save_total_limit: 100
+save_only_model: true
+plot_loss: true
+overwrite_output_dir: true
+report_to: wandb
+
+### train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 16
+max_steps: 640
+learning_rate: 1.0e-5
+lr_scheduler_type: cosine
+warmup_ratio: 0.03
+weight_decay: 0.01
+bf16: true
+gradient_checkpointing: true
+ddp_timeout: 180000000
+seed: 0
+val_size: 0.05
+eval_strategy: steps
+eval_steps: 320
+per_device_eval_batch_size: 1
+run_name: grail_v2op_fcurlong_b64_sft_e3

grail_v2op_fcurlong_b64_sft_e3_ck640/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

grail_v2op_fcurlong_b64_sft_e3_ck640/tokenizer_config.json

@@ -0,0 +1,32 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "local_files_only": false,
+  "model_max_length": 262144,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.9576358365285701,
+    "total_flos": 283306748805120.0,
+    "train_loss": 0.10237010675482452,
+    "train_runtime": 7722.2236,
+    "train_samples_per_second": 5.304,
+    "train_steps_per_second": 0.083
+}

grail_v2op_fcurlong_b64_sft_e3_ck640/trainer_state.json

@@ -0,0 +1,507 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9576358365285701,
+  "eval_steps": 320,
+  "global_step": 640,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.014963059945758908,
+      "grad_norm": 6.712064915661895,
+      "learning_rate": 4.5e-06,
+      "loss": 1.1995722770690918,
+      "step": 10
+    },
+    {
+      "epoch": 0.029926119891517815,
+      "grad_norm": 0.9496274265303064,
+      "learning_rate": 9.5e-06,
+      "loss": 0.37840962409973145,
+      "step": 20
+    },
+    {
+      "epoch": 0.04488917983727672,
+      "grad_norm": 0.9432001414342038,
+      "learning_rate": 9.994801642711945e-06,
+      "loss": 0.21686108112335206,
+      "step": 30
+    },
+    {
+      "epoch": 0.05985223978303563,
+      "grad_norm": 0.6710818393370306,
+      "learning_rate": 9.976845889479778e-06,
+      "loss": 0.1912616491317749,
+      "step": 40
+    },
+    {
+      "epoch": 0.07481529972879454,
+      "grad_norm": 0.6223348357038324,
+      "learning_rate": 9.946114644227765e-06,
+      "loss": 0.17536349296569825,
+      "step": 50
+    },
+    {
+      "epoch": 0.08977835967455344,
+      "grad_norm": 0.611512940619968,
+      "learning_rate": 9.90268679362172e-06,
+      "loss": 0.16727099418640137,
+      "step": 60
+    },
+    {
+      "epoch": 0.10474141962031236,
+      "grad_norm": 0.4862819097650955,
+      "learning_rate": 9.846673816331297e-06,
+      "loss": 0.15500309467315673,
+      "step": 70
+    },
+    {
+      "epoch": 0.11970447956607126,
+      "grad_norm": 0.624966172476472,
+      "learning_rate": 9.778219496865864e-06,
+      "loss": 0.1510760545730591,
+      "step": 80
+    },
+    {
+      "epoch": 0.13466753951183016,
+      "grad_norm": 0.6902242186854266,
+      "learning_rate": 9.6974995564817e-06,
+      "loss": 0.14040124416351318,
+      "step": 90
+    },
+    {
+      "epoch": 0.14963059945758908,
+      "grad_norm": 0.7788203246126214,
+      "learning_rate": 9.604721202108084e-06,
+      "loss": 0.1432424783706665,
+      "step": 100
+    },
+    {
+      "epoch": 0.164593659403348,
+      "grad_norm": 0.5852420867540118,
+      "learning_rate": 9.500122594450056e-06,
+      "loss": 0.13022279739379883,
+      "step": 110
+    },
+    {
+      "epoch": 0.17955671934910689,
+      "grad_norm": 0.6148172005879359,
+      "learning_rate": 9.383972236633308e-06,
+      "loss": 0.12292717695236206,
+      "step": 120
+    },
+    {
+      "epoch": 0.1945197792948658,
+      "grad_norm": 0.6070878255523858,
+      "learning_rate": 9.256568284960513e-06,
+      "loss": 0.12246812582015991,
+      "step": 130
+    },
+    {
+      "epoch": 0.20948283924062472,
+      "grad_norm": 0.6422608121331613,
+      "learning_rate": 9.118237783548376e-06,
+      "loss": 0.11907356977462769,
+      "step": 140
+    },
+    {
+      "epoch": 0.2244458991863836,
+      "grad_norm": 0.5163340697258343,
+      "learning_rate": 8.969335824810075e-06,
+      "loss": 0.11318811178207397,
+      "step": 150
+    },
+    {
+      "epoch": 0.23940895913214252,
+      "grad_norm": 0.6177734589885724,
+      "learning_rate": 8.810244637938117e-06,
+      "loss": 0.10748043060302734,
+      "step": 160
+    },
+    {
+      "epoch": 0.2543720190779014,
+      "grad_norm": 0.5944876414891593,
+      "learning_rate": 8.641372607727503e-06,
+      "loss": 0.10588945150375366,
+      "step": 170
+    },
+    {
+      "epoch": 0.26933507902366033,
+      "grad_norm": 0.7755321978339716,
+      "learning_rate": 8.463153226257805e-06,
+      "loss": 0.10246493816375732,
+      "step": 180
+    },
+    {
+      "epoch": 0.28429813896941925,
+      "grad_norm": 0.7232569376663098,
+      "learning_rate": 8.276043980125213e-06,
+      "loss": 0.09700657725334168,
+      "step": 190
+    },
+    {
+      "epoch": 0.29926119891517816,
+      "grad_norm": 0.7837189952328507,
+      "learning_rate": 8.080525176081006e-06,
+      "loss": 0.09719915390014648,
+      "step": 200
+    },
+    {
+      "epoch": 0.3142242588609371,
+      "grad_norm": 0.5525191148656788,
+      "learning_rate": 7.877098708091017e-06,
+      "loss": 0.09253854751586914,
+      "step": 210
+    },
+    {
+      "epoch": 0.329187318806696,
+      "grad_norm": 0.7908831498715817,
+      "learning_rate": 7.666286768981012e-06,
+      "loss": 0.09041429758071899,
+      "step": 220
+    },
+    {
+      "epoch": 0.34415037875245486,
+      "grad_norm": 0.7624706231808076,
+      "learning_rate": 7.448630509975199e-06,
+      "loss": 0.08354318141937256,
+      "step": 230
+    },
+    {
+      "epoch": 0.35911343869821377,
+      "grad_norm": 0.7465025086242588,
+      "learning_rate": 7.224688651568797e-06,
+      "loss": 0.08402214050292969,
+      "step": 240
+    },
+    {
+      "epoch": 0.3740764986439727,
+      "grad_norm": 0.660348987342,
+      "learning_rate": 6.995036049300539e-06,
+      "loss": 0.08162661790847778,
+      "step": 250
+    },
+    {
+      "epoch": 0.3890395585897316,
+      "grad_norm": 0.8734601802580341,
+      "learning_rate": 6.7602622181067425e-06,
+      "loss": 0.08035560846328735,
+      "step": 260
+    },
+    {
+      "epoch": 0.4040026185354905,
+      "grad_norm": 0.6868416501195594,
+      "learning_rate": 6.520969819044933e-06,
+      "loss": 0.0774998426437378,
+      "step": 270
+    },
+    {
+      "epoch": 0.41896567848124944,
+      "grad_norm": 0.8316693984204528,
+      "learning_rate": 6.277773112271545e-06,
+      "loss": 0.07430893778800965,
+      "step": 280
+    },
+    {
+      "epoch": 0.4339287384270083,
+      "grad_norm": 0.7439269465248411,
+      "learning_rate": 6.031296380244915e-06,
+      "loss": 0.07956287264823914,
+      "step": 290
+    },
+    {
+      "epoch": 0.4488917983727672,
+      "grad_norm": 0.7110813815315041,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.07669154405593873,
+      "step": 300
+    },
+    {
+      "epoch": 0.46385485831852613,
+      "grad_norm": 0.7160143070153896,
+      "learning_rate": 5.531040445016604e-06,
+      "loss": 0.07181846499443054,
+      "step": 310
+    },
+    {
+      "epoch": 0.47881791826428505,
+      "grad_norm": 0.773863838790316,
+      "learning_rate": 5.278545391625931e-06,
+      "loss": 0.06745543479919433,
+      "step": 320
+    },
+    {
+      "epoch": 0.47881791826428505,
+      "eval_loss": 0.0677054300904274,
+      "eval_runtime": 164.4593,
+      "eval_samples_per_second": 13.693,
+      "eval_steps_per_second": 3.423,
+      "step": 320
+    },
+    {
+      "epoch": 0.49378097821004396,
+      "grad_norm": 0.7131694654525507,
+      "learning_rate": 5.025335316209883e-06,
+      "loss": 0.06523659229278564,
+      "step": 330
+    },
+    {
+      "epoch": 0.5087440381558028,
+      "grad_norm": 0.7205242094544011,
+      "learning_rate": 4.7720602054004865e-06,
+      "loss": 0.06353209018707276,
+      "step": 340
+    },
+    {
+      "epoch": 0.5237070981015618,
+      "grad_norm": 0.7134129544026636,
+      "learning_rate": 4.51937021277469e-06,
+      "loss": 0.06176872849464417,
+      "step": 350
+    },
+    {
+      "epoch": 0.5386701580473207,
+      "grad_norm": 0.7297557198575642,
+      "learning_rate": 4.267913989919457e-06,
+      "loss": 0.05794233679771423,
+      "step": 360
+    },
+    {
+      "epoch": 0.5536332179930796,
+      "grad_norm": 1.175566336234018,
+      "learning_rate": 4.018337021352447e-06,
+      "loss": 0.05821115970611572,
+      "step": 370
+    },
+    {
+      "epoch": 0.5685962779388385,
+      "grad_norm": 0.6877611316331875,
+      "learning_rate": 3.7712799675725105e-06,
+      "loss": 0.05985587239265442,
+      "step": 380
+    },
+    {
+      "epoch": 0.5835593378845974,
+      "grad_norm": 0.7409321414744433,
+      "learning_rate": 3.5273770204933814e-06,
+      "loss": 0.05761619806289673,
+      "step": 390
+    },
+    {
+      "epoch": 0.5985223978303563,
+      "grad_norm": 0.6154664279949535,
+      "learning_rate": 3.2872542754821334e-06,
+      "loss": 0.05558090209960938,
+      "step": 400
+    },
+    {
+      "epoch": 0.6134854577761152,
+      "grad_norm": 0.6920333862613972,
+      "learning_rate": 3.051528124181342e-06,
+      "loss": 0.05565959811210632,
+      "step": 410
+    },
+    {
+      "epoch": 0.6284485177218742,
+      "grad_norm": 0.9792034808913532,
+      "learning_rate": 2.8208036722405895e-06,
+      "loss": 0.05608475208282471,
+      "step": 420
+    },
+    {
+      "epoch": 0.643411577667633,
+      "grad_norm": 0.759793896388316,
+      "learning_rate": 2.595673186018929e-06,
+      "loss": 0.051554566621780394,
+      "step": 430
+    },
+    {
+      "epoch": 0.658374637613392,
+      "grad_norm": 0.8964310627102796,
+      "learning_rate": 2.376714572245651e-06,
+      "loss": 0.052475863695144655,
+      "step": 440
+    },
+    {
+      "epoch": 0.6733376975591508,
+      "grad_norm": 0.9235822737680275,
+      "learning_rate": 2.1644898945419814e-06,
+      "loss": 0.050426656007766725,
+      "step": 450
+    },
+    {
+      "epoch": 0.6883007575049097,
+      "grad_norm": 0.794468091178141,
+      "learning_rate": 1.959543930611835e-06,
+      "loss": 0.04778494238853455,
+      "step": 460
+    },
+    {
+      "epoch": 0.7032638174506687,
+      "grad_norm": 0.7759291760854061,
+      "learning_rate": 1.7624027738052669e-06,
+      "loss": 0.05018991231918335,
+      "step": 470
+    },
+    {
+      "epoch": 0.7182268773964275,
+      "grad_norm": 0.7833656042097829,
+      "learning_rate": 1.5735724826443788e-06,
+      "loss": 0.04593663811683655,
+      "step": 480
+    },
+    {
+      "epoch": 0.7331899373421865,
+      "grad_norm": 0.717694534494841,
+      "learning_rate": 1.3935377817783585e-06,
+      "loss": 0.04782665967941284,
+      "step": 490
+    },
+    {
+      "epoch": 0.7481529972879454,
+      "grad_norm": 0.6982897311158559,
+      "learning_rate": 1.2227608177022522e-06,
+      "loss": 0.041572964191436766,
+      "step": 500
+    },
+    {
+      "epoch": 0.7631160572337043,
+      "grad_norm": 0.775286601949971,
+      "learning_rate": 1.0616799724335387e-06,
+      "loss": 0.04551121294498443,
+      "step": 510
+    },
+    {
+      "epoch": 0.7780791171794632,
+      "grad_norm": 0.8765090195242348,
+      "learning_rate": 9.107087381917878e-07,
+      "loss": 0.0452000230550766,
+      "step": 520
+    },
+    {
+      "epoch": 0.7930421771252221,
+      "grad_norm": 0.9450458740458394,
+      "learning_rate": 7.702346559700502e-07,
+      "loss": 0.04239870011806488,
+      "step": 530
+    },
+    {
+      "epoch": 0.808005237070981,
+      "grad_norm": 1.1084902010463304,
+      "learning_rate": 6.406183207226973e-07,
+      "loss": 0.046213507652282715,
+      "step": 540
+    },
+    {
+      "epoch": 0.8229682970167399,
+      "grad_norm": 0.8888364483744429,
+      "learning_rate": 5.22192455723336e-07,
+      "loss": 0.04177500307559967,
+      "step": 550
+    },
+    {
+      "epoch": 0.8379313569624989,
+      "grad_norm": 0.7800564779827981,
+      "learning_rate": 4.152610584689443e-07,
+      "loss": 0.03991095423698425,
+      "step": 560
+    },
+    {
+      "epoch": 0.8528944169082577,
+      "grad_norm": 0.9844258193628123,
+      "learning_rate": 3.200986203226536e-07,
+      "loss": 0.041886287927627566,
+      "step": 570
+    },
+    {
+      "epoch": 0.8678574768540166,
+      "grad_norm": 0.8309489114957327,
+      "learning_rate": 2.3694942189835423e-07,
+      "loss": 0.041270244121551516,
+      "step": 580
+    },
+    {
+      "epoch": 0.8828205367997756,
+      "grad_norm": 0.7443735511453847,
+      "learning_rate": 1.6602690599586747e-07,
+      "loss": 0.03482242822647095,
+      "step": 590
+    },
+    {
+      "epoch": 0.8977835967455344,
+      "grad_norm": 0.6246977724942839,
+      "learning_rate": 1.0751312969633665e-07,
+      "loss": 0.03933063745498657,
+      "step": 600
+    },
+    {
+      "epoch": 0.9127466566912934,
+      "grad_norm": 0.9075497829268706,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.03947060108184815,
+      "step": 610
+    },
+    {
+      "epoch": 0.9277097166370523,
+      "grad_norm": 0.7369260799683012,
+      "learning_rate": 2.828037337617273e-08,
+      "loss": 0.03887377083301544,
+      "step": 620
+    },
+    {
+      "epoch": 0.9426727765828112,
+      "grad_norm": 0.7638096393890356,
+      "learning_rate": 7.764782704649532e-09,
+      "loss": 0.03906640410423279,
+      "step": 630
+    },
+    {
+      "epoch": 0.9576358365285701,
+      "grad_norm": 0.9202837566996431,
+      "learning_rate": 6.418823675857777e-11,
+      "loss": 0.040480810403823855,
+      "step": 640
+    },
+    {
+      "epoch": 0.9576358365285701,
+      "eval_loss": 0.03894830867648125,
+      "eval_runtime": 163.8545,
+      "eval_samples_per_second": 13.744,
+      "eval_steps_per_second": 3.436,
+      "step": 640
+    },
+    {
+      "epoch": 0.9576358365285701,
+      "step": 640,
+      "total_flos": 283306748805120.0,
+      "train_loss": 0.10237010675482452,
+      "train_runtime": 7722.2236,
+      "train_samples_per_second": 5.304,
+      "train_steps_per_second": 0.083
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 640,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 320,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 283306748805120.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}