Xiaohui Chen comments

Repositories
Issues
Comments

Results 13 comments of


                                            Xiaohui Chen

Reproducing video mme with qwen3-vl-instruct

@ycsun1972 only 2 responses do not finish. Do you mind sharing your script and env?

Reproducing video mme with qwen3-vl-instruct

@lihengtao @ycsun1972 Hey I was able to reproduce the result for thinking model! using transformer backend @ycsun1972 A side question - what's the value of max_frames parameter will be used...

关于V3.5 8B模型的训练时间 the training time of of V3_5 8B model

我把pack ds 关掉了就好了