Search results

LVLM Lora相关内容有计划更新吗

Status: Open.

#420 In alibaba/ROLL;

· luyouqi233 opened on Apr 9, 2026

mcore-adapter从VLM的hfconfig中获取参数异常

Status: Open.

#419 In alibaba/ROLL;

· YisuZhou opened on Apr 9, 2026

ValueError: There is no module or parameter named 'base_model' in Qwen2_5_VLForConditionalGeneration

Status: Open.

#418 In alibaba/ROLL;

· luyouqi233 opened on Apr 9, 2026

[Bug] Agentic Training Failed with "rock: command not found"

Status: Open.

#412 In alibaba/ROLL;

· KunWuLuan opened on Mar 30, 2026

Error: Qwen3.5-35B-A3B lora sft with mcore-adapter, error occurs when saving a ckpt.

Status: Open.

#411 In alibaba/ROLL;

· Unofish opened on Mar 30, 2026

Feature Request: Trajectory-level formulation mode for AgentNativeStepEnvManager

Status: Open.

#409 In alibaba/ROLL;

· shamanez opened on Mar 28, 2026

LR scheduler exhausts early in agentic training with AgentNativeStepEnvManager

Status: Open.

#407 In alibaba/ROLL;

· shamanez opened on Mar 27, 2026

[Feature Request] Add official Qwen3.5 Sokoban agentic example

Status: Open.

#406 In alibaba/ROLL;

· yingyibiao opened on Mar 27, 2026

Add OpenReward Intergration

Status: Open.

#405 In alibaba/ROLL;

· shamanez opened on Mar 27, 2026

Torch 2.8.0 + vLLM 0.10.2 开启 use_kl_loss 后训练qwen2.5-vl-3B-agentic崩溃，Torch 2.6.0 + vLLM 0.8.4 正常

Status: Open.

#398 In alibaba/ROLL;

· SF-evil opened on Mar 24, 2026

Very slow asynchronous GRPO training on 8×A100

Status: Open.

#394 In alibaba/ROLL;

· 5SSjw opened on Mar 22, 2026

[Type Hints] Add type annotations to utility functions in roll/utils/functionals.py

Status: Open.

#387 In alibaba/ROLL;

· guozhihao-224 opened on Mar 18, 2026