Search results

Qwen3-VL models don't have <code>max_position_embeddings</code>

bugSomething isn't workingSomething isn't working

Status: Open.

#4807 In volcengine/verl;

· shirley-wu opened on Jan 6, 2026

AttributeError: 'FullyAsyncTrainer' object has no attribute 'train_dataloader'

bugSomething isn't workingSomething isn't working

Status: Open.

#4794 In volcengine/verl;

· Silentssss opened on Jan 5, 2026

[RFC] Suffix Decoding + LSTM Speculative Decoding for Faster Generation

Status: Open.

#4791 In volcengine/verl;

· vx120 opened on Jan 5, 2026

是否支持多轮检索的RL

Status: Open.

#4790 In volcengine/verl;

· ZhangXcho opened on Jan 5, 2026

DAPO's std calculation does not account for continuous rewards.

bugSomething isn't workingSomething isn't working

Status: Open.

#4786 In volcengine/verl;

· TsingZ0 opened on Jan 5, 2026

Multiple Tasks Checkpoint Writing Error

Status: Open.

#4772 In volcengine/verl;

· lzxdjb opened on Jan 3, 2026

求大佬解答，当前SFT支持流式数据读取吗？

Status: Open.

#4765 In volcengine/verl;

· Marco692 opened on Jan 1, 2026

TypeError: Qwen3VLMultimodalRotaryEmbedding.forward() got an unexpected keyword argument 'packed_seq_params'

bugSomething isn't workingSomething isn't working

Status: Open.

#4763 In volcengine/verl;

· enbacoo opened on Jan 1, 2026

[Bug] no merged model under the folder <code>huggingface</code>

bugSomething isn't workingSomething isn't working

Status: Open.

#4761 In volcengine/verl;

· YongchengYAO opened on Jan 1, 2026

Invalid tool call parsing in thinking token for hermes format in tool_parser.py in agent_loop from experiment folder.

Status: Open.

#4757 In volcengine/verl;

· chetandhembre opened on Dec 31, 2025

RuntimeError: aten.embedding.default: got mixed torch.Tensor and DTensor

bugSomething isn't workingSomething isn't working

Status: Open.

#4756 In volcengine/verl;

· AIprogrammer opened on Dec 31, 2025

Qwen3-VL MFU

bugSomething isn't workingSomething isn't working

Status: Open.

#4748 In volcengine/verl;

· tardis-key opened on Dec 31, 2025