Search results

Qwen3.5 GRPO training fails with ZeRO-3 + gradient checkpointing enabled

Status: Open.

#257 In 2U1/Qwen-VL-Series-Finetune;

· lpenghao opened on Jul 8, 2026

Why doesn't the process_vision_info function require the input of the image_patch_size parameter?

Status: Open.

#256 In 2U1/Qwen-VL-Series-Finetune;

· PlutoQyl opened on Jun 17, 2026

Question: VRAM required for training Qwen3-VL:2B

Status: Open.

#255 In 2U1/Qwen-VL-Series-Finetune;

· shahedmomenzadeh opened on Jun 9, 2026

AttributeError: 'Qwen3VLForSequenceClassification' object has no attribute 'prepare_inputs_for_generation'

Status: Open.

#252 In 2U1/Qwen-VL-Series-Finetune;

· xixixixihu opened on Mar 25, 2026

StopIteration in get_rope_index for Qwen3-VL video training with transformers==5.3.0

Status: Open.

#250 In 2U1/Qwen-VL-Series-Finetune;

· toannguyen1904 opened on Mar 22, 2026

how to enable VLLM in the grpo training of the Qwen 3.5-9B dense model?

Status: Open.

#249 In 2U1/Qwen-VL-Series-Finetune;

· kongyjjj opened on Mar 20, 2026

Show error code -11

Status: Open.

#248 In 2U1/Qwen-VL-Series-Finetune;

· xjohnxjohn opened on Mar 7, 2026

LoRA based multi gpu training issue: Watchdog timeout

Status: Open.

#243 In 2U1/Qwen-VL-Series-Finetune;

· pulkitkumar95 opened on Feb 18, 2026

Fail to run GRPO for Qwen3-VL

Status: Open.

#242 In 2U1/Qwen-VL-Series-Finetune;

· chenyil6 opened on Feb 16, 2026

Qwen2.5-VL configuration problem

Status: Open.

#240 In 2U1/Qwen-VL-Series-Finetune;

· shilongxiang opened on Feb 2, 2026

request for review

Status: Open.

#238 In 2U1/Qwen-VL-Series-Finetune;

· HavenCTO opened on Jan 11, 2026

It seems that video content is necessary?

Status: Open.

#237 In 2U1/Qwen-VL-Series-Finetune;

· nightrain-vampire opened on Jan 7, 2026