Downloads · 30 days
9
23% of all-time downloads
Xiangchen1233/dapo-qwen2.5-vl-3b-virl39k-step202
dapo-qwen2.5-vl-3b-virl39k-step202 is a image-text-to-text model from Xiangchen1233. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as other.
Vanilla DAPO RL baseline (no advantage reshaping of any kind), retrained for baseline comparisons.
Downloads · 30 days
9
23% of all-time downloads
All-time downloads
40
Public
Parameters
4.1B
8.1 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors8.1 GB · 100%
From the Hugging Face model README
Vanilla DAPO RL baseline (no advantage reshaping of any kind), retrained for baseline comparisons.
adv_estimator=dapo (group-relative advantage, no token-level weighting)<think> ... </think> then the final answer in \boxed{}\boxed{} exact match (mathruler grader)global_step_202/actor (scripts/model_merger.py)