grail_v2op_nograph_b64_sft_e3_ck640

.gitattributes

@@ -91,3 +91,4 @@ oracle_reason_s2_b64_sft_e3_ck300/tokenizer.json filter=lfs diff=lfs merge=lfs -
 grail_v2op_trajq_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 grail_v2op_fcursp_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 grail_v2op_trajqsp_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+grail_v2op_nograph_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text

grail_v2op_nograph_b64_sft_e3_ck640/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 1.1248214481925063,
+    "eval_loss": 0.034882038831710815,
+    "eval_runtime": 138.4055,
+    "eval_samples_per_second": 13.843,
+    "eval_steps_per_second": 3.461,
+    "total_flos": 275443268714496.0,
+    "train_loss": 0.09438723342027515,
+    "train_runtime": 7438.5282,
+    "train_samples_per_second": 5.506,
+    "train_steps_per_second": 0.086
+}

grail_v2op_nograph_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_nograph_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_nograph_b64_sft_e3_ck640/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 1.1248214481925063,
+    "eval_loss": 0.034882038831710815,
+    "eval_runtime": 138.4055,
+    "eval_samples_per_second": 13.843,
+    "eval_steps_per_second": 3.461
+}

grail_v2op_nograph_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_nograph_b64_sft_e3_ck640/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87e95c3ca2460b4c5ace18eed4ee7a76dcd42d5514de2ef9851a84d733fce795
+size 17534340584

grail_v2op_nograph_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_nograph_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_nograph_sft_e3, exps/sft/grail_v2op_nograph_sft_e3, /scratch/kwc5354/viewagent/viewagent_private/exps/sft_runs/grail_v2op_nograph_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_nograph_sft_e3
+dataset_dir: exps/sft/grail_v2op_nograph_sft_e3
+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_nograph_b64_sft_e3
+logging_steps: 10
+save_strategy: steps
+save_steps: 640
+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: 42
+val_size: 0.05
+eval_strategy: steps
+eval_steps: 640
+per_device_eval_batch_size: 1
+run_name: grail_v2op_nograph_b64_sft_e3

grail_v2op_nograph_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_nograph_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_nograph_b64_sft_e3_ck640/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.1248214481925063,
+    "total_flos": 275443268714496.0,
+    "train_loss": 0.09438723342027515,
+    "train_runtime": 7438.5282,
+    "train_samples_per_second": 5.506,
+    "train_steps_per_second": 0.086
+}

grail_v2op_nograph_b64_sft_e3_ck640/trainer_state.json

@@ -0,0 +1,499 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.1248214481925063,
+  "eval_steps": 640,
+  "global_step": 640,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.017580485660916384,
+      "grad_norm": 6.978233762409854,
+      "learning_rate": 4.5e-06,
+      "loss": 1.2015214920043946,
+      "step": 10
+    },
+    {
+      "epoch": 0.03516097132183277,
+      "grad_norm": 1.452008299291918,
+      "learning_rate": 9.5e-06,
+      "loss": 0.37156412601470945,
+      "step": 20
+    },
+    {
+      "epoch": 0.052741456982749145,
+      "grad_norm": 1.0074007694166378,
+      "learning_rate": 9.994801642711945e-06,
+      "loss": 0.2173619031906128,
+      "step": 30
+    },
+    {
+      "epoch": 0.07032194264366554,
+      "grad_norm": 0.512453753413536,
+      "learning_rate": 9.976845889479778e-06,
+      "loss": 0.1878380298614502,
+      "step": 40
+    },
+    {
+      "epoch": 0.08790242830458192,
+      "grad_norm": 0.5850242283091878,
+      "learning_rate": 9.946114644227765e-06,
+      "loss": 0.1688234329223633,
+      "step": 50
+    },
+    {
+      "epoch": 0.10548291396549829,
+      "grad_norm": 0.45799474175920296,
+      "learning_rate": 9.90268679362172e-06,
+      "loss": 0.15665593147277831,
+      "step": 60
+    },
+    {
+      "epoch": 0.12306339962641467,
+      "grad_norm": 0.6455604522066029,
+      "learning_rate": 9.846673816331297e-06,
+      "loss": 0.15412521362304688,
+      "step": 70
+    },
+    {
+      "epoch": 0.14064388528733107,
+      "grad_norm": 0.521071532964208,
+      "learning_rate": 9.778219496865864e-06,
+      "loss": 0.13551499843597412,
+      "step": 80
+    },
+    {
+      "epoch": 0.15822437094824746,
+      "grad_norm": 0.6422186689182229,
+      "learning_rate": 9.6974995564817e-06,
+      "loss": 0.1396848440170288,
+      "step": 90
+    },
+    {
+      "epoch": 0.17580485660916384,
+      "grad_norm": 0.7046165571086215,
+      "learning_rate": 9.604721202108084e-06,
+      "loss": 0.12936232089996338,
+      "step": 100
+    },
+    {
+      "epoch": 0.19338534227008022,
+      "grad_norm": 0.6939007108579025,
+      "learning_rate": 9.500122594450056e-06,
+      "loss": 0.1288854718208313,
+      "step": 110
+    },
+    {
+      "epoch": 0.21096582793099658,
+      "grad_norm": 0.6443851287186817,
+      "learning_rate": 9.383972236633308e-06,
+      "loss": 0.11608177423477173,
+      "step": 120
+    },
+    {
+      "epoch": 0.22854631359191296,
+      "grad_norm": 0.7001995885880756,
+      "learning_rate": 9.256568284960513e-06,
+      "loss": 0.11698464155197144,
+      "step": 130
+    },
+    {
+      "epoch": 0.24612679925282935,
+      "grad_norm": 0.7088856965072459,
+      "learning_rate": 9.118237783548376e-06,
+      "loss": 0.105198335647583,
+      "step": 140
+    },
+    {
+      "epoch": 0.26370728491374573,
+      "grad_norm": 0.5904637827473859,
+      "learning_rate": 8.969335824810075e-06,
+      "loss": 0.10124807357788086,
+      "step": 150
+    },
+    {
+      "epoch": 0.28128777057466214,
+      "grad_norm": 0.6781576679735906,
+      "learning_rate": 8.810244637938117e-06,
+      "loss": 0.0992238998413086,
+      "step": 160
+    },
+    {
+      "epoch": 0.2988682562355785,
+      "grad_norm": 0.6377888732477195,
+      "learning_rate": 8.641372607727503e-06,
+      "loss": 0.0938180148601532,
+      "step": 170
+    },
+    {
+      "epoch": 0.3164487418964949,
+      "grad_norm": 0.7791216188215783,
+      "learning_rate": 8.463153226257805e-06,
+      "loss": 0.08974037766456604,
+      "step": 180
+    },
+    {
+      "epoch": 0.33402922755741127,
+      "grad_norm": 0.6185809106273865,
+      "learning_rate": 8.276043980125213e-06,
+      "loss": 0.09216833114624023,
+      "step": 190
+    },
+    {
+      "epoch": 0.3516097132183277,
+      "grad_norm": 0.8397507075312262,
+      "learning_rate": 8.080525176081006e-06,
+      "loss": 0.08774979710578919,
+      "step": 200
+    },
+    {
+      "epoch": 0.36919019887924404,
+      "grad_norm": 0.8387328081513985,
+      "learning_rate": 7.877098708091017e-06,
+      "loss": 0.08532504439353943,
+      "step": 210
+    },
+    {
+      "epoch": 0.38677068454016045,
+      "grad_norm": 0.7202999893173491,
+      "learning_rate": 7.666286768981012e-06,
+      "loss": 0.0794275164604187,
+      "step": 220
+    },
+    {
+      "epoch": 0.4043511702010768,
+      "grad_norm": 0.6546360842943003,
+      "learning_rate": 7.448630509975199e-06,
+      "loss": 0.08199517726898194,
+      "step": 230
+    },
+    {
+      "epoch": 0.42193165586199316,
+      "grad_norm": 0.8132503103723937,
+      "learning_rate": 7.224688651568797e-06,
+      "loss": 0.07835033535957336,
+      "step": 240
+    },
+    {
+      "epoch": 0.43951214152290957,
+      "grad_norm": 0.8611147598765134,
+      "learning_rate": 6.995036049300539e-06,
+      "loss": 0.07843175530433655,
+      "step": 250
+    },
+    {
+      "epoch": 0.45709262718382593,
+      "grad_norm": 0.7081987364775277,
+      "learning_rate": 6.7602622181067425e-06,
+      "loss": 0.0728411078453064,
+      "step": 260
+    },
+    {
+      "epoch": 0.47467311284474234,
+      "grad_norm": 0.6968679049165124,
+      "learning_rate": 6.520969819044933e-06,
+      "loss": 0.07278348207473755,
+      "step": 270
+    },
+    {
+      "epoch": 0.4922535985056587,
+      "grad_norm": 0.7064172627572398,
+      "learning_rate": 6.277773112271545e-06,
+      "loss": 0.07027890682220458,
+      "step": 280
+    },
+    {
+      "epoch": 0.509834084166575,
+      "grad_norm": 0.8323118223258629,
+      "learning_rate": 6.031296380244915e-06,
+      "loss": 0.06504650115966797,
+      "step": 290
+    },
+    {
+      "epoch": 0.5274145698274915,
+      "grad_norm": 0.8809017205402678,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.05899275541305542,
+      "step": 300
+    },
+    {
+      "epoch": 0.5449950554884079,
+      "grad_norm": 0.7995140885494462,
+      "learning_rate": 5.531040445016604e-06,
+      "loss": 0.06032218933105469,
+      "step": 310
+    },
+    {
+      "epoch": 0.5625755411493243,
+      "grad_norm": 0.6957256154443505,
+      "learning_rate": 5.278545391625931e-06,
+      "loss": 0.06509553194046021,
+      "step": 320
+    },
+    {
+      "epoch": 0.5801560268102406,
+      "grad_norm": 0.7857832835437646,
+      "learning_rate": 5.025335316209883e-06,
+      "loss": 0.060322153568267825,
+      "step": 330
+    },
+    {
+      "epoch": 0.597736512471157,
+      "grad_norm": 0.7661010570496796,
+      "learning_rate": 4.7720602054004865e-06,
+      "loss": 0.056724411249160764,
+      "step": 340
+    },
+    {
+      "epoch": 0.6153169981320734,
+      "grad_norm": 0.7309198922473006,
+      "learning_rate": 4.51937021277469e-06,
+      "loss": 0.05603755712509155,
+      "step": 350
+    },
+    {
+      "epoch": 0.6328974837929898,
+      "grad_norm": 0.8563881732369832,
+      "learning_rate": 4.267913989919457e-06,
+      "loss": 0.056651508808135985,
+      "step": 360
+    },
+    {
+      "epoch": 0.6504779694539061,
+      "grad_norm": 0.787858854871785,
+      "learning_rate": 4.018337021352447e-06,
+      "loss": 0.053351521492004395,
+      "step": 370
+    },
+    {
+      "epoch": 0.6680584551148225,
+      "grad_norm": 0.6965425602080381,
+      "learning_rate": 3.7712799675725105e-06,
+      "loss": 0.052681684494018555,
+      "step": 380
+    },
+    {
+      "epoch": 0.685638940775739,
+      "grad_norm": 0.7065265810731874,
+      "learning_rate": 3.5273770204933814e-06,
+      "loss": 0.0459515392780304,
+      "step": 390
+    },
+    {
+      "epoch": 0.7032194264366554,
+      "grad_norm": 0.7216400624525897,
+      "learning_rate": 3.2872542754821334e-06,
+      "loss": 0.04877682626247406,
+      "step": 400
+    },
+    {
+      "epoch": 0.7207999120975717,
+      "grad_norm": 0.7320769417808594,
+      "learning_rate": 3.051528124181342e-06,
+      "loss": 0.04883717894554138,
+      "step": 410
+    },
+    {
+      "epoch": 0.7383803977584881,
+      "grad_norm": 0.9500850937549915,
+      "learning_rate": 2.8208036722405895e-06,
+      "loss": 0.04788933396339416,
+      "step": 420
+    },
+    {
+      "epoch": 0.7559608834194045,
+      "grad_norm": 0.5739271908689518,
+      "learning_rate": 2.595673186018929e-06,
+      "loss": 0.043954139947891234,
+      "step": 430
+    },
+    {
+      "epoch": 0.7735413690803209,
+      "grad_norm": 0.7055558101672434,
+      "learning_rate": 2.376714572245651e-06,
+      "loss": 0.044578784704208375,
+      "step": 440
+    },
+    {
+      "epoch": 0.7911218547412372,
+      "grad_norm": 0.8266895072668017,
+      "learning_rate": 2.1644898945419814e-06,
+      "loss": 0.04426226317882538,
+      "step": 450
+    },
+    {
+      "epoch": 0.8087023404021536,
+      "grad_norm": 0.6866244573309381,
+      "learning_rate": 1.959543930611835e-06,
+      "loss": 0.03959735631942749,
+      "step": 460
+    },
+    {
+      "epoch": 0.82628282606307,
+      "grad_norm": 0.8093480383071349,
+      "learning_rate": 1.7624027738052669e-06,
+      "loss": 0.043867182731628415,
+      "step": 470
+    },
+    {
+      "epoch": 0.8438633117239863,
+      "grad_norm": 0.6742931947463796,
+      "learning_rate": 1.5735724826443788e-06,
+      "loss": 0.037909334897994994,
+      "step": 480
+    },
+    {
+      "epoch": 0.8614437973849027,
+      "grad_norm": 0.6220463609103414,
+      "learning_rate": 1.3935377817783585e-06,
+      "loss": 0.03895248174667358,
+      "step": 490
+    },
+    {
+      "epoch": 0.8790242830458191,
+      "grad_norm": 0.688789373678053,
+      "learning_rate": 1.2227608177022522e-06,
+      "loss": 0.03853694200515747,
+      "step": 500
+    },
+    {
+      "epoch": 0.8966047687067356,
+      "grad_norm": 0.9031274307375294,
+      "learning_rate": 1.0616799724335387e-06,
+      "loss": 0.040706279873847964,
+      "step": 510
+    },
+    {
+      "epoch": 0.9141852543676519,
+      "grad_norm": 0.7263601646000968,
+      "learning_rate": 9.107087381917878e-07,
+      "loss": 0.035372170805931094,
+      "step": 520
+    },
+    {
+      "epoch": 0.9317657400285683,
+      "grad_norm": 0.8248824774330984,
+      "learning_rate": 7.702346559700502e-07,
+      "loss": 0.03963362574577332,
+      "step": 530
+    },
+    {
+      "epoch": 0.9493462256894847,
+      "grad_norm": 0.8040833738066231,
+      "learning_rate": 6.406183207226973e-07,
+      "loss": 0.037721645832061765,
+      "step": 540
+    },
+    {
+      "epoch": 0.9669267113504011,
+      "grad_norm": 0.8021089234750791,
+      "learning_rate": 5.22192455723336e-07,
+      "loss": 0.03438863754272461,
+      "step": 550
+    },
+    {
+      "epoch": 0.9845071970113174,
+      "grad_norm": 0.7590710936875649,
+      "learning_rate": 4.152610584689443e-07,
+      "loss": 0.036246579885482785,
+      "step": 560
+    },
+    {
+      "epoch": 1.0017580485660917,
+      "grad_norm": 0.5605840565452501,
+      "learning_rate": 3.200986203226536e-07,
+      "loss": 0.03206341564655304,
+      "step": 570
+    },
+    {
+      "epoch": 1.019338534227008,
+      "grad_norm": 0.5603038221613628,
+      "learning_rate": 2.3694942189835423e-07,
+      "loss": 0.023385101556777955,
+      "step": 580
+    },
+    {
+      "epoch": 1.0369190198879243,
+      "grad_norm": 0.6100471510261499,
+      "learning_rate": 1.6602690599586747e-07,
+      "loss": 0.021044978499412538,
+      "step": 590
+    },
+    {
+      "epoch": 1.0544995055488409,
+      "grad_norm": 0.8305003651518713,
+      "learning_rate": 1.0751312969633665e-07,
+      "loss": 0.02459432780742645,
+      "step": 600
+    },
+    {
+      "epoch": 1.0720799912097572,
+      "grad_norm": 0.5740624168027063,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.024296101927757264,
+      "step": 610
+    },
+    {
+      "epoch": 1.0896604768706735,
+      "grad_norm": 0.5980115849281283,
+      "learning_rate": 2.828037337617273e-08,
+      "loss": 0.021341198682785036,
+      "step": 620
+    },
+    {
+      "epoch": 1.10724096253159,
+      "grad_norm": 0.6903703195880581,
+      "learning_rate": 7.764782704649532e-09,
+      "loss": 0.02394503653049469,
+      "step": 630
+    },
+    {
+      "epoch": 1.1248214481925063,
+      "grad_norm": 0.7433282902970586,
+      "learning_rate": 6.418823675857777e-11,
+      "loss": 0.02472032457590103,
+      "step": 640
+    },
+    {
+      "epoch": 1.1248214481925063,
+      "eval_loss": 0.034882038831710815,
+      "eval_runtime": 138.5036,
+      "eval_samples_per_second": 13.834,
+      "eval_steps_per_second": 3.458,
+      "step": 640
+    },
+    {
+      "epoch": 1.1248214481925063,
+      "step": 640,
+      "total_flos": 275443268714496.0,
+      "train_loss": 0.09438723342027515,
+      "train_runtime": 7438.5282,
+      "train_samples_per_second": 5.506,
+      "train_steps_per_second": 0.086
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 640,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 640,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 275443268714496.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}