Congmin(Xavier) Qiu comments

Repositories
Issues
Comments

Results 4 comments of


                                            Congmin(Xavier) Qiu

Agentic RL Training example: KeyError: 'solve_equation'

Actually, there is a try catch that avoid training failure, if the tool name not exist, the reward will be 0.0. In this case, the model will be able to...

Mismatch between trajectory and reward_extra_infos_dict when using _log_rollout_data

为啥不一致呢？后面计算 reward_extra_infos_dict 不是按照balance过的batch去计算的吗？ ``` # === 配置 === config = { "trainer": { "balance_batch": True, # ✅ 开启 "rollout_data_dir": "./rollout" # ✅ 开启 } } # === Step 0: 原始...

Tool call prase Error in slang_rollout

Closed my PR because SPMD will be deprecated https://github.com/volcengine/verl/pull/4106

[Learning Verl] what is max_colocate_count?

@wuxibin89 @vermouth1992 can you help review my fix? 🙏