Rin

hu5enpai's activity

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

upvoted a paper 28 days ago

On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability

Paper • 2608.30320 • Published 29 days ago • 63

New activity in Qwen/Qwen3.8-Flash-Next about 1 month ago

🚀 Try training Qwen3.8-Flash-Next with ms-swift (Day-0 Support)

🚀❤️ 3

#19 opened about 1 month ago by

hu5enpai

New activity in deepseek-ai/DeepSeek-OCR-2 8 months ago

ms-swift has supported inference, deployment, and fine-tuning of the DeepSeek-OCR-2 model.

#5 opened 8 months ago by

hu5enpai

upvoted an article 8 months ago

view article

Article

Benchmark Smarter: Tailor Your Model Evaluation Suite with EvalScope

• 

kelseye

•

Jan 22

• 7

upvoted an article 10 months ago

view article

Article

Qwen-Image-i2L: Training Strategies for Image-to-LoRA Generation

• 

kelseye

•

Dec 16, 2025

• 59

commented a paper 10 months ago

Soft Adaptive Policy Optimization

Paper • 2511.20347 • Published Nov 25, 2025 • 44 •

6

upvoted a paper 10 months ago

Soft Adaptive Policy Optimization

Paper • 2511.20347 • Published Nov 25, 2025 • 44

New activity in PaddlePaddle/PaddleOCR-VL 11 months ago

ms-swift has supported inference, deployment, and fine-tuning of the PaddleOCR-VL model.

🚀🔥 7

4

#42 opened 11 months ago by

hu5enpai

upvoted a paper about 1 year ago

Parallel-R1: Towards Parallel Thinking via Reinforcement Learning

Paper • 2509.07980 • Published Sep 9, 2025 • 106

commented 4 papers about 1 year ago

On-Policy RL Meets Off-Policy Experts: Harmonizing Supervised Fine-Tuning and Reinforcement Learning via Dynamic Weighting

Paper • 2508.11408 • Published Aug 15, 2025 • 8 •

6

On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification

Paper • 2508.05629 • Published Aug 7, 2025 • 191 •

21

DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning

Paper • 2505.14362 • Published May 20, 2025 • 6 •

2

Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning

Paper • 2506.01939 • Published Jun 2, 2025 • 190 •

7

New activity in Qwen/Qwen3-Coder-480B-A35B-Instruct about 1 year ago

👍👍

👍 2

1

#19 opened about 1 year ago by

hu5enpai

New activity in ChenShawn/DeepEyes-Datasets-47k about 1 year ago

Unable to load the dataset

#2 opened about 1 year ago by

hu5enpai

New activity in microsoft/Florence-2-large-ft about 2 years ago

Swift now supports inference, training, and deployment of the Florence models.

3

#14 opened over 2 years ago by

hu5enpai

New activity in microsoft/Florence-2-large over 2 years ago

How to Finetune?

👍👀 18

5

#19 opened over 2 years ago by

vody-am

Fix incorrect bos_token, eos_token, and pad_token ids in config.json

3

#17 opened over 2 years ago by

Xenova

New activity in liuhaotian/LLaVA-Instruct-150K over 2 years ago

Unable to load dataset.

👍 4

2

#10 opened over 2 years ago by

ToletiSri

New activity in OpenGVLab/InternVL-Chat-V1-5 over 2 years ago

Swift now supports inference, training of InternVL-Chat-V1-5

❤️ 1

3

#11 opened over 2 years ago by

hu5enpai