Search results

训练lora之后，用VLLM部署报错：ValueError: Call to add_lora method failed: vLLM only supports modules_to_save being None.

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9545 In hiyouga/LLaMA-Factory;

· Tinystormjojo opened on Nov 27, 2025

Dose vllm serve support Reward model trained by Llama Factory

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9540 In hiyouga/LLaMA-Factory;

· silence0zong-alt opened on Nov 26, 2025

我正在尝试将Intel最新的B60显卡与本项目更好的结合

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9537 In hiyouga/LLaMA-Factory;

· hzdzkjdxyjs opened on Nov 26, 2025

[Feature Request] Support LoRA/PEFT fine-tuning with Megatron (mcore_adapter) backend

enhancementNew feature or requestNew feature or request

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

Status: Open.

#9535 In hiyouga/LLaMA-Factory;

· Fang-Lansheng opened on Nov 25, 2025

CPU RAM Keeps Increasing During Tokenization during SFT with LoRA on large scale video-vqa dataset

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9531 In hiyouga/LLaMA-Factory;

· nimeshagrawal opened on Nov 25, 2025

支持Qwen3-235B模型8卡ktransformers训练配置

enhancementNew feature or requestNew feature or request

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

Status: Open.

#9529 In hiyouga/LLaMA-Factory;

· abc123000111 opened on Nov 24, 2025

在npu上，lora+zero3，训练时长异常

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.

#9528 In hiyouga/LLaMA-Factory;

· ble-ZY opened on Nov 24, 2025

OmniVinci Support | 是否支持OmniVinci的lora finetune？

enhancementNew feature or requestNew feature or request

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

Status: Open.

#9527 In hiyouga/LLaMA-Factory;

· JOY-SWang opened on Nov 24, 2025

Intern-s1 image_processor默认参数未加载，无法使用图片patch

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9526 In hiyouga/LLaMA-Factory;

· liuwei130 opened on Nov 23, 2025

目前是否支持对 DeepSeek-R1-0528-Qwen3-8B 使用ktransformers

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9525 In hiyouga/LLaMA-Factory;

· tebie6 opened on Nov 23, 2025

How to increase rope_theta for long context fine-tuning?

enhancementNew feature or requestNew feature or request

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

Status: Open.

#9524 In hiyouga/LLaMA-Factory;

· harshraj172 opened on Nov 23, 2025

在WSL的Ubuntu中使用docker/docker-cuda/Dockerfile.megatron创建镜像报错

bugSomething isn't workingSomething isn't working

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

Status: Open.

#9523 In hiyouga/LLaMA-Factory;

· Nemo0223 opened on Nov 21, 2025