Search results

Support Liger Kernel for Qwen3.5

enhancementNew feature or requestNew feature or request

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10253 In hiyouga/LlamaFactory;

· wyt2000 opened on Mar 6, 2026

[BUG] Ray Demo Unused Keys

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10251 In hiyouga/LlamaFactory;

· SnowCharmQ opened on Mar 5, 2026

TypeError when saving predictions with HuggingFace Dataset

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10250 In hiyouga/LlamaFactory;

· pyxnpyx opened on Mar 5, 2026

华为昇腾使用accelerate调度器微调qwen3-next-80B，但是在配置文件中deepspeed_moe_layer_cls_names不知道应该写什么？

bugSomething isn't workingSomething isn't working

npuThis problem is related to NPU devicesThis problem is related to NPU devices

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10249 In hiyouga/LlamaFactory;

· gongyinfeng0206-design opened on Mar 5, 2026

transformers5.2.0训练moe模型，如果使用torch_npu的moe优化算子，loss会变成nan

bugSomething isn't workingSomething isn't working

npuThis problem is related to NPU devicesThis problem is related to NPU devices

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10248 In hiyouga/LlamaFactory;

· piekey1994 opened on Mar 5, 2026

ChatML template bug in test chat with unsloth

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10247 In hiyouga/LlamaFactory;

· mykeehu opened on Mar 4, 2026

ValueError: could not determine the shape of object type 'VideoMetadata'

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10246 In hiyouga/LlamaFactory;

· markmochi200 opened on Mar 4, 2026

lora微调训练没问题，导出的模型运行后一对话就报错 total bytes of NDArray > 2**32'

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10245 In hiyouga/LlamaFactory;

· Coomfu opened on Mar 4, 2026

运行vllm_infer.py 报错：torch.AcceleratorError: CUDA error: CUDA-capable device(s) is/are busy or unavailable

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10240 In hiyouga/LlamaFactory;

· skfeng36 opened on Mar 3, 2026

Error on running the example NLG evaluation: TypeError: argument 'ids': 'list' object cannot be interpreted as an integer

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10239 In hiyouga/LlamaFactory;

· joshchristo opened on Mar 3, 2026

transformers version promblem when vllm_infer.py run Qwen3-VL model

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10238 In hiyouga/LlamaFactory;

· TouchDeeper opened on Mar 3, 2026

vqa数据lora微调Qwen3.5-27B, 之前整理的vqa数据微调qwen3vl的，怎么微调3.5报错了，怎么改

bugSomething isn't workingSomething isn't working

pendingThis problem is yet to be addressedThis problem is yet to be addressed

Status: Open.

#10235 In hiyouga/LlamaFactory;

· zkailinzhang opened on Mar 2, 2026