All issues

Search results

How can VLM with a small number of parameters, like under 1B, improve visual reasoning ability?

Status: Open.

#49 In tanhuajie/Reason-RFT;

· zhouyizhuo opened on Jun 3, 2026

Confirmation over settings on clevr_math_sft and its RL version

Status: Open.

#48 In tanhuajie/Reason-RFT;

· serser opened on Jan 26, 2026

RL阶段冻结相关问题

Status: Open.

#47 In tanhuajie/Reason-RFT;

· alalajili opened on Jan 8, 2026

求助！如何使用Lora实现RFT

Status: Open.

#46 In tanhuajie/Reason-RFT;

· Rhtoric opened on Nov 6, 2025

请求开源SFT训练后的模型～

Status: Open.

#44 In tanhuajie/Reason-RFT;

· yanyc428 opened on Sep 27, 2025