Nguyen Quang Trung

ngqtrung's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

updated a collection 5 days ago

Video HopChain

Collection

Multi-hop video questions for RLVR, and the 8B model trained on them with Second-Wave Exploration. • 2 items • Updated 5 days ago

updated a model 5 days ago

ngqtrung/video-hopchain-8b

Video-Text-to-Text • 9B • Updated 5 days ago • 34

updated a dataset 5 days ago

ngqtrung/video-hopchain

Viewer • Updated 5 days ago • 23.6k • 52

published a model 5 days ago

ngqtrung/video-hopchain-8b

Video-Text-to-Text • 9B • Updated 5 days ago • 34

published a dataset 5 days ago

ngqtrung/video-hopchain

Viewer • Updated 5 days ago • 23.6k • 52

liked a Space 9 days ago

Running on CPU Upgrade

Featured

3.3k

The Smol Training Playbook

📚

3.3k

The secrets to building world-class LLMs

upvoted a paper 15 days ago

LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training

Paper • 2509.23661 • Published Sep 28, 2025 • 52

updated a collection 16 days ago

Video RLVR — final training data

Collection

The datasets behind our Qwen3-VL-8B video RLVR runs: the base 24f/100k mixture and the HopChain v5 multi-hop corpora. • 2 items • Updated 15 days ago

updated a dataset 16 days ago

ngqtrung/video-r1-mc-24f100k-combined

Viewer • Updated 16 days ago • 106k • 132

published a dataset 16 days ago

ngqtrung/video-r1-mc-24f100k-combined

Viewer • Updated 16 days ago • 106k • 132

updated a collection 16 days ago

videorl

Collection

11 items • Updated 16 days ago

published a dataset 16 days ago

ngqtrung/hopchain-v5

Viewer • Updated 16 days ago • 23.6k • 99

updated a dataset 16 days ago

ngqtrung/hopchain-v5

Viewer • Updated 16 days ago • 23.6k • 99

upvoted a paper 21 days ago

VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning

Paper • 2608.26105 • Published 24 days ago • 272

liked a model 22 days ago

Qwen/Qwen3.8-Flash-Next

Image-Text-to-Text • 180B • Updated 22 days ago • 724k • • 5.39k

updated a collection 23 days ago

Qwen3-VL-8B video RLVR - GRPO checkpoint ladder

Collection

Every published checkpoint of the 8B video-QA GRPO campaign, ranked by core-3 mean_accuracy (5,645 rows). Untrained base = 0.4426. • 9 items • Updated 23 days ago