Training in progress, step 10

README.md

@@ -1,18 +1,17 @@
 ---
 base_model: Qwen/Qwen2.5-VL-3B-Instruct
-datasets: lmms-lab/multimodal-open-r1-8k-verified
 library_name: transformers
 model_name: Qwen2.5-VL-3B-Instruct-Thinking
 tags:
 - generated_from_trainer
-- grpo
 - trl
+- grpo
 licence: license
 ---
 
 # Model Card for Qwen2.5-VL-3B-Instruct-Thinking
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct) on the [lmms-lab/multimodal-open-r1-8k-verified](https://huggingface.co/datasets/lmms-lab/multimodal-open-r1-8k-verified) dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -36,7 +35,7 @@ This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing
 ### Framework versions
 
 - TRL: 0.20.0.dev0
-- Transformers: 4.53.1
+- Transformers: 4.53.2
 - Pytorch: 2.6.0+cu124
 - Datasets: 4.0.0
 - Tokenizers: 0.21.2

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82f160d43a4bf78d2e10880121da85c0ec422fe1293d6c0c15ec29032b845269
+oid sha256:457cb18179dd90f1bcaa5e1f170f9abb26edf4d4e4d40b1911f59ac0f5e35305
 size 7393888

runs/Jul14_16-18-04_61ee1e7cbf16/events.out.tfevents.1752509917.61ee1e7cbf16.250.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c4a4c0e6fa55fbf7bcafed66dfedc97ada2bef9f31ee6812928410b473d815d
+size 10523

tokenizer_config.json

@@ -201,6 +201,7 @@
   "extra_special_tokens": {},
   "model_max_length": 131072,
   "pad_token": "<|endoftext|>",
+  "padding_side": "left",
   "processor_class": "Qwen2_5_VLProcessor",
   "split_special_tokens": false,
   "tokenizer_class": "Qwen2Tokenizer",

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6930c4e035bc08d325997bb6028fa3d659dbdf2537e5b142f78c883821a2ca94
+oid sha256:2b43383700005c240a83cb4d7d630ca15ecd0071a6c6f1a6be238d7b81be4c19
 size 6520