Upload folder using huggingface_hub

.gitattributes

@@ -57,3 +57,36 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+17_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+18_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+19_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+20_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+21_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+22_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+23_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+24_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+25_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+26_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+27_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+28_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+29_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+30_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+31_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+32_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+33_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+34_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+35_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+36_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+37_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+38_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+39_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+40_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+41_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+42_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+43_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+44_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+45_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+46_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+47_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+48_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+49_lora_sft_ttt/tokenizer.json filter=lfs diff=lfs merge=lfs -text

100_lora_sft_ttt/README.md

@@ -0,0 +1,74 @@
+---
+library_name: peft
+license: other
+base_model: /home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: 100_lora_sft_ttt
+  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. -->
+
+# 100_lora_sft_ttt
+
+This model is a fine-tuned version of [/home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/](https://huggingface.co//home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/) on the bongard_ttt dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0001
+
+## 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:
+- learning_rate: 0.0001
+- train_batch_size: 1
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 4
+- total_eval_batch_size: 4
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 2.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 0.0798        | 0.2137 | 100  | 0.0591          |
+| 0.0566        | 0.4274 | 200  | 0.0586          |
+| 0.0444        | 0.6410 | 300  | 0.0120          |
+| 0.0235        | 0.8547 | 400  | 0.0065          |
+| 0.0066        | 1.0684 | 500  | 0.0003          |
+| 0.0021        | 1.2821 | 600  | 0.0002          |
+| 0.0005        | 1.4957 | 700  | 0.0001          |
+| 0.0008        | 1.7094 | 800  | 0.0001          |
+| 0.0021        | 1.9231 | 900  | 0.0001          |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.45.0.dev0
+- Pytorch 2.5.1+cu124
+- Datasets 3.1.0
+- Tokenizers 0.19.1

100_lora_sft_ttt/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": "^(?!.*visual.patch_embed|visual.blocks).*(?:up_proj|v_proj|q_proj|k_proj|gate_proj|down_proj|o_proj).*",
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

100_lora_sft_ttt/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:659eb55884e992e8bad276baeec369e4f6d8cc735a8f8fe95c9d727357295c32
+size 147770888

100_lora_sft_ttt/added_tokens.json

@@ -0,0 +1,16 @@
+{
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

100_lora_sft_ttt/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 7.351845124503598e-05,
+    "eval_runtime": 17.1828,
+    "eval_samples_per_second": 1.106,
+    "eval_steps_per_second": 0.291,
+    "total_flos": 34102773284864.0,
+    "train_loss": 0.02313395333948991,
+    "train_runtime": 4154.1424,
+    "train_samples_per_second": 0.901,
+    "train_steps_per_second": 0.225
+}

100_lora_sft_ttt/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 7.351845124503598e-05,
+    "eval_runtime": 17.1828,
+    "eval_samples_per_second": 1.106,
+    "eval_steps_per_second": 0.291
+}

100_lora_sft_ttt/preprocessor_config.json

@@ -0,0 +1,29 @@
+{
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.48145466,
+    0.4578275,
+    0.40821073
+  ],
+  "image_processor_type": "Qwen2VLImageProcessor",
+  "image_std": [
+    0.26862954,
+    0.26130258,
+    0.27577711
+  ],
+  "max_pixels": 12845056,
+  "merge_size": 2,
+  "min_pixels": 3136,
+  "patch_size": 14,
+  "processor_class": "Qwen2VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "max_pixels": 12845056,
+    "min_pixels": 3136
+  },
+  "temporal_patch_size": 2
+}

100_lora_sft_ttt/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

100_lora_sft_ttt/tokenizer_config.json

@@ -0,0 +1,144 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% 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 %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "processor_class": "Qwen2VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

100_lora_sft_ttt/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "total_flos": 34102773284864.0,
+    "train_loss": 0.02313395333948991,
+    "train_runtime": 4154.1424,
+    "train_samples_per_second": 0.901,
+    "train_steps_per_second": 0.225
+}

100_lora_sft_ttt/trainer_log.jsonl

@@ -0,0 +1,19 @@
+{"current_steps": 100, "total_steps": 936, "loss": 0.0798, "lr": 9.998747147528374e-05, "epoch": 0.21367521367521367, "percentage": 10.68, "elapsed_time": "0:07:24", "remaining_time": "1:01:56"}
+{"current_steps": 100, "total_steps": 936, "eval_loss": 0.05914049223065376, "epoch": 0.21367521367521367, "percentage": 10.68, "elapsed_time": "0:07:46", "remaining_time": "1:05:01"}
+{"current_steps": 200, "total_steps": 936, "loss": 0.0566, "lr": 9.614025209023084e-05, "epoch": 0.42735042735042733, "percentage": 21.37, "elapsed_time": "0:14:47", "remaining_time": "0:54:24"}
+{"current_steps": 200, "total_steps": 936, "eval_loss": 0.058623503893613815, "epoch": 0.42735042735042733, "percentage": 21.37, "elapsed_time": "0:15:04", "remaining_time": "0:55:27"}
+{"current_steps": 300, "total_steps": 936, "loss": 0.0444, "lr": 8.594394348255238e-05, "epoch": 0.6410256410256411, "percentage": 32.05, "elapsed_time": "0:22:06", "remaining_time": "0:46:51"}
+{"current_steps": 300, "total_steps": 936, "eval_loss": 0.011989740654826164, "epoch": 0.6410256410256411, "percentage": 32.05, "elapsed_time": "0:22:23", "remaining_time": "0:47:28"}
+{"current_steps": 400, "total_steps": 936, "loss": 0.0235, "lr": 7.080159982206471e-05, "epoch": 0.8547008547008547, "percentage": 42.74, "elapsed_time": "0:29:29", "remaining_time": "0:39:31"}
+{"current_steps": 400, "total_steps": 936, "eval_loss": 0.006506655365228653, "epoch": 0.8547008547008547, "percentage": 42.74, "elapsed_time": "0:29:46", "remaining_time": "0:39:54"}
+{"current_steps": 500, "total_steps": 936, "loss": 0.0066, "lr": 5.279687012637799e-05, "epoch": 1.0683760683760684, "percentage": 53.42, "elapsed_time": "0:36:52", "remaining_time": "0:32:09"}
+{"current_steps": 500, "total_steps": 936, "eval_loss": 0.00026314370916225016, "epoch": 1.0683760683760684, "percentage": 53.42, "elapsed_time": "0:37:09", "remaining_time": "0:32:24"}
+{"current_steps": 600, "total_steps": 936, "loss": 0.0021, "lr": 3.4407279551696846e-05, "epoch": 1.282051282051282, "percentage": 64.1, "elapsed_time": "0:44:12", "remaining_time": "0:24:45"}
+{"current_steps": 600, "total_steps": 936, "eval_loss": 0.00022907418315298855, "epoch": 1.282051282051282, "percentage": 64.1, "elapsed_time": "0:44:30", "remaining_time": "0:24:55"}
+{"current_steps": 700, "total_steps": 936, "loss": 0.0005, "lr": 1.8163311700448898e-05, "epoch": 1.4957264957264957, "percentage": 74.79, "elapsed_time": "0:51:35", "remaining_time": "0:17:23"}
+{"current_steps": 700, "total_steps": 936, "eval_loss": 9.96691087493673e-05, "epoch": 1.4957264957264957, "percentage": 74.79, "elapsed_time": "0:51:52", "remaining_time": "0:17:29"}
+{"current_steps": 800, "total_steps": 936, "loss": 0.0008, "lr": 6.300203628022272e-06, "epoch": 1.7094017094017095, "percentage": 85.47, "elapsed_time": "0:58:58", "remaining_time": "0:10:01"}
+{"current_steps": 800, "total_steps": 936, "eval_loss": 6.905790360178798e-05, "epoch": 1.7094017094017095, "percentage": 85.47, "elapsed_time": "0:59:15", "remaining_time": "0:10:04"}
+{"current_steps": 900, "total_steps": 936, "loss": 0.0021, "lr": 4.5036799358910697e-07, "epoch": 1.9230769230769231, "percentage": 96.15, "elapsed_time": "1:06:21", "remaining_time": "0:02:39"}
+{"current_steps": 900, "total_steps": 936, "eval_loss": 7.419099711114541e-05, "epoch": 1.9230769230769231, "percentage": 96.15, "elapsed_time": "1:06:38", "remaining_time": "0:02:39"}
+{"current_steps": 936, "total_steps": 936, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "1:09:13", "remaining_time": "0:00:00"}

100_lora_sft_ttt/trainer_state.json

@@ -0,0 +1,177 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 100,
+  "global_step": 936,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.21367521367521367,
+      "grad_norm": 0.13362472032354034,
+      "learning_rate": 9.998747147528374e-05,
+      "loss": 0.0798,
+      "step": 100
+    },
+    {
+      "epoch": 0.21367521367521367,
+      "eval_loss": 0.05914049223065376,
+      "eval_runtime": 22.1821,
+      "eval_samples_per_second": 0.857,
+      "eval_steps_per_second": 0.225,
+      "step": 100
+    },
+    {
+      "epoch": 0.42735042735042733,
+      "grad_norm": 0.5855884517843695,
+      "learning_rate": 9.614025209023084e-05,
+      "loss": 0.0566,
+      "step": 200
+    },
+    {
+      "epoch": 0.42735042735042733,
+      "eval_loss": 0.058623503893613815,
+      "eval_runtime": 17.0992,
+      "eval_samples_per_second": 1.111,
+      "eval_steps_per_second": 0.292,
+      "step": 200
+    },
+    {
+      "epoch": 0.6410256410256411,
+      "grad_norm": 0.2980642878698098,
+      "learning_rate": 8.594394348255238e-05,
+      "loss": 0.0444,
+      "step": 300
+    },
+    {
+      "epoch": 0.6410256410256411,
+      "eval_loss": 0.011989740654826164,
+      "eval_runtime": 17.2205,
+      "eval_samples_per_second": 1.103,
+      "eval_steps_per_second": 0.29,
+      "step": 300
+    },
+    {
+      "epoch": 0.8547008547008547,
+      "grad_norm": 0.15631837239845456,
+      "learning_rate": 7.080159982206471e-05,
+      "loss": 0.0235,
+      "step": 400
+    },
+    {
+      "epoch": 0.8547008547008547,
+      "eval_loss": 0.006506655365228653,
+      "eval_runtime": 17.0799,
+      "eval_samples_per_second": 1.112,
+      "eval_steps_per_second": 0.293,
+      "step": 400
+    },
+    {
+      "epoch": 1.0683760683760684,
+      "grad_norm": 0.0026226211310534132,
+      "learning_rate": 5.279687012637799e-05,
+      "loss": 0.0066,
+      "step": 500
+    },
+    {
+      "epoch": 1.0683760683760684,
+      "eval_loss": 0.00026314370916225016,
+      "eval_runtime": 17.0923,
+      "eval_samples_per_second": 1.112,
+      "eval_steps_per_second": 0.293,
+      "step": 500
+    },
+    {
+      "epoch": 1.282051282051282,
+      "grad_norm": 0.0006464192921040726,
+      "learning_rate": 3.4407279551696846e-05,
+      "loss": 0.0021,
+      "step": 600
+    },
+    {
+      "epoch": 1.282051282051282,
+      "eval_loss": 0.00022907418315298855,
+      "eval_runtime": 17.2201,
+      "eval_samples_per_second": 1.103,
+      "eval_steps_per_second": 0.29,
+      "step": 600
+    },
+    {
+      "epoch": 1.4957264957264957,
+      "grad_norm": 0.3605781591352982,
+      "learning_rate": 1.8163311700448898e-05,
+      "loss": 0.0005,
+      "step": 700
+    },
+    {
+      "epoch": 1.4957264957264957,
+      "eval_loss": 9.96691087493673e-05,
+      "eval_runtime": 17.1121,
+      "eval_samples_per_second": 1.11,
+      "eval_steps_per_second": 0.292,
+      "step": 700
+    },
+    {
+      "epoch": 1.7094017094017095,
+      "grad_norm": 0.0007160112841758006,
+      "learning_rate": 6.300203628022272e-06,
+      "loss": 0.0008,
+      "step": 800
+    },
+    {
+      "epoch": 1.7094017094017095,
+      "eval_loss": 6.905790360178798e-05,
+      "eval_runtime": 17.0966,
+      "eval_samples_per_second": 1.111,
+      "eval_steps_per_second": 0.292,
+      "step": 800
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.0009131187233137881,
+      "learning_rate": 4.5036799358910697e-07,
+      "loss": 0.0021,
+      "step": 900
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "eval_loss": 7.419099711114541e-05,
+      "eval_runtime": 17.1017,
+      "eval_samples_per_second": 1.111,
+      "eval_steps_per_second": 0.292,
+      "step": 900
+    },
+    {
+      "epoch": 2.0,
+      "step": 936,
+      "total_flos": 34102773284864.0,
+      "train_loss": 0.02313395333948991,
+      "train_runtime": 4154.1424,
+      "train_samples_per_second": 0.901,
+      "train_steps_per_second": 0.225
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 936,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 34102773284864.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

100_lora_sft_ttt/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:572a7a3a906125e6efb9c23c8b3040f69eba08bbb847d605fe362b6bc638e0f4
+size 7352

10_lora_sft_ttt/README.md

@@ -0,0 +1,68 @@
+---
+library_name: peft
+license: other
+base_model: /home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: 10_lora_sft_ttt
+  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. -->
+
+# 10_lora_sft_ttt
+
+This model is a fine-tuned version of [/home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/](https://huggingface.co//home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/) on the bongard_ttt dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0004
+
+## 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:
+- learning_rate: 0.0001
+- train_batch_size: 1
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 4
+- total_eval_batch_size: 4
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 2.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 0.0456        | 0.6410 | 100  | 0.0040          |
+| 0.0076        | 1.2821 | 200  | 0.0007          |
+| 0.0003        | 1.9231 | 300  | 0.0004          |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.45.0.dev0
+- Pytorch 2.5.1+cu124
+- Datasets 3.1.0
+- Tokenizers 0.19.1

10_lora_sft_ttt/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": "^(?!.*visual.patch_embed|visual.blocks).*(?:gate_proj|q_proj|down_proj|v_proj|o_proj|up_proj|k_proj).*",
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

10_lora_sft_ttt/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2416ea89271cf77a1c7f4bb4b492f1f33df81470234aae1f2d14cf226ec2d7c
+size 147770888

10_lora_sft_ttt/added_tokens.json

@@ -0,0 +1,16 @@
+{
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

10_lora_sft_ttt/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 0.0003799094120040536,
+    "eval_runtime": 11.0452,
+    "eval_samples_per_second": 0.634,
+    "eval_steps_per_second": 0.181,
+    "total_flos": 11399228555264.0,
+    "train_loss": 0.017151810213899575,
+    "train_runtime": 3811.5722,
+    "train_samples_per_second": 0.327,
+    "train_steps_per_second": 0.082
+}

10_lora_sft_ttt/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 0.0003799094120040536,
+    "eval_runtime": 11.0452,
+    "eval_samples_per_second": 0.634,
+    "eval_steps_per_second": 0.181
+}

10_lora_sft_ttt/preprocessor_config.json

@@ -0,0 +1,29 @@
+{
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.48145466,
+    0.4578275,
+    0.40821073
+  ],
+  "image_processor_type": "Qwen2VLImageProcessor",
+  "image_std": [
+    0.26862954,
+    0.26130258,
+    0.27577711
+  ],
+  "max_pixels": 12845056,
+  "merge_size": 2,
+  "min_pixels": 3136,
+  "patch_size": 14,
+  "processor_class": "Qwen2VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "max_pixels": 12845056,
+    "min_pixels": 3136
+  },
+  "temporal_patch_size": 2
+}

10_lora_sft_ttt/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

10_lora_sft_ttt/tokenizer_config.json

@@ -0,0 +1,144 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% 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 %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "processor_class": "Qwen2VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

10_lora_sft_ttt/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "total_flos": 11399228555264.0,
+    "train_loss": 0.017151810213899575,
+    "train_runtime": 3811.5722,
+    "train_samples_per_second": 0.327,
+    "train_steps_per_second": 0.082
+}

10_lora_sft_ttt/trainer_log.jsonl

@@ -0,0 +1,7 @@
+{"current_steps": 100, "total_steps": 312, "loss": 0.0456, "lr": 8.613974319136958e-05, "epoch": 0.6410256410256411, "percentage": 32.05, "elapsed_time": "0:24:18", "remaining_time": "0:51:31"}
+{"current_steps": 100, "total_steps": 312, "eval_loss": 0.004042092710733414, "epoch": 0.6410256410256411, "percentage": 32.05, "elapsed_time": "0:24:41", "remaining_time": "0:52:20"}
+{"current_steps": 200, "total_steps": 312, "loss": 0.0076, "lr": 3.4549150281252636e-05, "epoch": 1.282051282051282, "percentage": 64.1, "elapsed_time": "0:42:26", "remaining_time": "0:23:45"}
+{"current_steps": 200, "total_steps": 312, "eval_loss": 0.0006544874631799757, "epoch": 1.282051282051282, "percentage": 64.1, "elapsed_time": "0:42:44", "remaining_time": "0:23:56"}
+{"current_steps": 300, "total_steps": 312, "loss": 0.0003, "lr": 4.52511911603265e-07, "epoch": 1.9230769230769231, "percentage": 96.15, "elapsed_time": "1:01:07", "remaining_time": "0:02:26"}
+{"current_steps": 300, "total_steps": 312, "eval_loss": 0.00038461596705019474, "epoch": 1.9230769230769231, "percentage": 96.15, "elapsed_time": "1:01:28", "remaining_time": "0:02:27"}
+{"current_steps": 312, "total_steps": 312, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "1:03:30", "remaining_time": "0:00:00"}

10_lora_sft_ttt/trainer_state.json

@@ -0,0 +1,87 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 100,
+  "global_step": 312,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.6410256410256411,
+      "grad_norm": 0.5753826484194268,
+      "learning_rate": 8.613974319136958e-05,
+      "loss": 0.0456,
+      "step": 100
+    },
+    {
+      "epoch": 0.6410256410256411,
+      "eval_loss": 0.004042092710733414,
+      "eval_runtime": 23.234,
+      "eval_samples_per_second": 0.301,
+      "eval_steps_per_second": 0.086,
+      "step": 100
+    },
+    {
+      "epoch": 1.282051282051282,
+      "grad_norm": 1.7751791918185151,
+      "learning_rate": 3.4549150281252636e-05,
+      "loss": 0.0076,
+      "step": 200
+    },
+    {
+      "epoch": 1.282051282051282,
+      "eval_loss": 0.0006544874631799757,
+      "eval_runtime": 18.0146,
+      "eval_samples_per_second": 0.389,
+      "eval_steps_per_second": 0.111,
+      "step": 200
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.00147995503102493,
+      "learning_rate": 4.52511911603265e-07,
+      "loss": 0.0003,
+      "step": 300
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "eval_loss": 0.00038461596705019474,
+      "eval_runtime": 21.005,
+      "eval_samples_per_second": 0.333,
+      "eval_steps_per_second": 0.095,
+      "step": 300
+    },
+    {
+      "epoch": 2.0,
+      "step": 312,
+      "total_flos": 11399228555264.0,
+      "train_loss": 0.017151810213899575,
+      "train_runtime": 3811.5722,
+      "train_samples_per_second": 0.327,
+      "train_steps_per_second": 0.082
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 312,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 11399228555264.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

10_lora_sft_ttt/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68405e2a28bc2350be4fde1bcd4aa1dabc39c88fb67acf6571c4b35ed572b937
+size 7352

11_lora_sft_ttt/README.md

@@ -0,0 +1,68 @@
+---
+library_name: peft
+license: other
+base_model: /home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: 11_lora_sft_ttt
+  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. -->
+
+# 11_lora_sft_ttt
+
+This model is a fine-tuned version of [/home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/](https://huggingface.co//home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/) on the bongard_ttt dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0001
+
+## 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:
+- learning_rate: 0.0001
+- train_batch_size: 1
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 4
+- total_eval_batch_size: 4
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 2.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 0.0471        | 0.5848 | 100  | 0.0021          |
+| 0.0135        | 1.1696 | 200  | 0.0002          |
+| 0.0011        | 1.7544 | 300  | 0.0001          |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.45.0.dev0
+- Pytorch 2.5.1+cu124
+- Datasets 3.1.0
+- Tokenizers 0.19.1

11_lora_sft_ttt/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/kh844/icml/LLaMA-Factory/models/qwen2_vl_lora_sft_v1/",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 64,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": "^(?!.*visual.patch_embed|visual.blocks).*(?:gate_proj|up_proj|k_proj|q_proj|v_proj|down_proj|o_proj).*",
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

11_lora_sft_ttt/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be58ae24a83ba0af38cfd2a6d062fc2a09b32965cb67db90e479f8a5f6215692
+size 147770888

11_lora_sft_ttt/added_tokens.json

@@ -0,0 +1,16 @@
+{
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

11_lora_sft_ttt/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 9.797404345590621e-05,
+    "eval_runtime": 15.6859,
+    "eval_samples_per_second": 0.446,
+    "eval_steps_per_second": 0.128,
+    "total_flos": 12490745577472.0,
+    "train_loss": 0.018072671783891946,
+    "train_runtime": 3749.7125,
+    "train_samples_per_second": 0.364,
+    "train_steps_per_second": 0.091
+}

11_lora_sft_ttt/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 9.797404345590621e-05,
+    "eval_runtime": 15.6859,
+    "eval_samples_per_second": 0.446,
+    "eval_steps_per_second": 0.128
+}

11_lora_sft_ttt/preprocessor_config.json

@@ -0,0 +1,29 @@
+{
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.48145466,
+    0.4578275,
+    0.40821073
+  ],
+  "image_processor_type": "Qwen2VLImageProcessor",
+  "image_std": [
+    0.26862954,
+    0.26130258,
+    0.27577711
+  ],
+  "max_pixels": 12845056,
+  "merge_size": 2,
+  "min_pixels": 3136,
+  "patch_size": 14,
+  "processor_class": "Qwen2VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "max_pixels": 12845056,
+    "min_pixels": 3136
+  },
+  "temporal_patch_size": 2
+}

11_lora_sft_ttt/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

11_lora_sft_ttt/tokenizer_config.json

@@ -0,0 +1,144 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% 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 %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "processor_class": "Qwen2VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

11_lora_sft_ttt/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "total_flos": 12490745577472.0,
+    "train_loss": 0.018072671783891946,
+    "train_runtime": 3749.7125,
+    "train_samples_per_second": 0.364,
+    "train_steps_per_second": 0.091
+}

11_lora_sft_ttt/trainer_log.jsonl

@@ -0,0 +1,7 @@
+{"current_steps": 100, "total_steps": 342, "loss": 0.0471, "lr": 8.934096234129843e-05, "epoch": 0.5847953216374269, "percentage": 29.24, "elapsed_time": "0:20:21", "remaining_time": "0:49:15"}
+{"current_steps": 100, "total_steps": 342, "eval_loss": 0.002146613784134388, "epoch": 0.5847953216374269, "percentage": 29.24, "elapsed_time": "0:20:49", "remaining_time": "0:50:24"}
+{"current_steps": 200, "total_steps": 342, "loss": 0.0135, "lr": 4.412948159326646e-05, "epoch": 1.1695906432748537, "percentage": 58.48, "elapsed_time": "0:40:20", "remaining_time": "0:28:38"}
+{"current_steps": 200, "total_steps": 342, "eval_loss": 0.00021622402709908783, "epoch": 1.1695906432748537, "percentage": 58.48, "elapsed_time": "0:40:28", "remaining_time": "0:28:44"}
+{"current_steps": 300, "total_steps": 342, "loss": 0.0011, "lr": 4.547428767819578e-06, "epoch": 1.7543859649122808, "percentage": 87.72, "elapsed_time": "0:55:52", "remaining_time": "0:07:49"}
+{"current_steps": 300, "total_steps": 342, "eval_loss": 9.790992044145241e-05, "epoch": 1.7543859649122808, "percentage": 87.72, "elapsed_time": "0:56:09", "remaining_time": "0:07:51"}
+{"current_steps": 342, "total_steps": 342, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "1:02:28", "remaining_time": "0:00:00"}