Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / hiyouga/LLaMA-Factory issues and pull requests

#5848 - How to continue training LoRA made without llama factory?

Issue - State: open - Opened by Sehyo 11 days ago
Labels: pending

#5847 - Support ferretui model

Issue - State: open - Opened by dushwe 11 days ago
Labels: pending

#5845 - model.generate与llamafactory-cli train do_predict给出的结果不一致

Issue - State: open - Opened by mzc2113391 11 days ago - 1 comment
Labels: pending

#5844 - PISSA 训练完后如何进行推理

Issue - State: open - Opened by user2311717757 11 days ago
Labels: pending

#5841 - [求助] dpo 训练 72b 模型，显存溢出

Issue - State: open - Opened by empty2enrich 12 days ago
Labels: pending

#5840 - vila support

Issue - State: open - Opened by Creazygao 12 days ago
Labels: pending

#5839 - update wechat.jpg

Pull Request - State: closed - Opened by codemayq 12 days ago

#5838 - dpo qwen2-72b oom，9*8 A800 80G需要怎么设置？

Issue - State: open - Opened by BobTsang1995 12 days ago
Labels: pending

#5837 - a little abnormal grad norm value during sft

Issue - State: open - Opened by SinclairCoder 12 days ago - 1 comment
Labels: pending

#5836 - Llama 3 based models not saving chat template

Issue - State: open - Opened by pumetu 12 days ago
Labels: pending

#5835 - Where to select Unsloth in the webUI?

Issue - State: open - Opened by awesomecoolraj 13 days ago
Labels: pending

#5833 - 在使用liger-kernel时报错undefined symbol: cuModuleGetFunction

Issue - State: open - Opened by HypherX 13 days ago
Labels: pending

#5830 - 如何同时使用yaml文件和命令行参数？

Issue - State: open - Opened by koukoulala 14 days ago
Labels: pending

#5829 - 请问现有的mistral框架可以支持最新出的Ministral-8B吗？

Issue - State: open - Opened by koukoulala 14 days ago
Labels: pending

#5828 - The loss of sharegpt format

Issue - State: open - Opened by ZijunSong 14 days ago
Labels: pending

#5826 - Downloading from modelscope failed when running example qwen demo

Issue - State: closed - Opened by flishwang 14 days ago - 5 comments
Labels: solved

#5825 - Downloading from modelscope failed when running example qwen demo

Issue - State: closed - Opened by flishwang 14 days ago - 1 comment
Labels: duplicate

#5825 - Downloading from modelscope failed when running example qwen demo

Issue - State: closed - Opened by flishwang 14 days ago - 1 comment
Labels: duplicate

#5824 - 执行微调训练时，一直停在0%不动

Issue - State: open - Opened by czhcc 14 days ago - 4 comments
Labels: pending

#5823 - How can I add a new customized model?

Issue - State: closed - Opened by williamium3000 15 days ago - 3 comments
Labels: solved

#5823 - How can I add a new customized model?

Issue - State: closed - Opened by williamium3000 15 days ago - 3 comments
Labels: solved

#5822 - Qwen2-VL 微调不支持同时输入video和image么

Issue - State: open - Opened by zhang122994917 15 days ago
Labels: pending

#5822 - Qwen2-VL 微调不支持同时输入video和image么

Issue - State: open - Opened by zhang122994917 15 days ago
Labels: pending

#5821 - How are you using/loading the tuned models outside LLaMa-Factory?

Issue - State: closed - Opened by 240db 15 days ago - 3 comments
Labels: solved

#5820 - What is the correct meaning of the cutoff_len parameter?

Issue - State: closed - Opened by baiyin 15 days ago
Labels: invalid

#5820 - What is the correct meaning of the cutoff_len parameter?

Issue - State: closed - Opened by baiyin 15 days ago
Labels: invalid

#5819 - Add trust_remote_code Parameter and Set Default to False

Pull Request - State: open - Opened by yafshar 15 days ago
Labels: pending

#5816 - Sample dataset added in dataset_info.json

Pull Request - State: open - Opened by NoumanAhmad448 15 days ago

#5815 - Need Help About Long Context

Issue - State: open - Opened by no-execution 15 days ago - 3 comments
Labels: pending

#5814 - ValueError: Some keys are not used by the HfArgumentParser: ['save_dir']

Issue - State: closed - Opened by HelloWorld506 15 days ago - 2 comments
Labels: solved

#5813 - 请问是否支持对数据提前tokenize，启动后直接读取token id进行训练？

Issue - State: closed - Opened by Mr-lonely0 15 days ago - 1 comment
Labels: solved

#5813 - 请问是否支持对数据提前tokenize，启动后直接读取token id进行训练？

Issue - State: closed - Opened by Mr-lonely0 15 days ago - 1 comment
Labels: solved

#5812 - LLaVA_dpo跑不了

Issue - State: open - Opened by zsworld6 15 days ago - 4 comments
Labels: pending

#5811 - ppo有计划使用trl的ppotrainer_v2吗

Issue - State: open - Opened by kechunFIVE 15 days ago
Labels: pending

#5810 - 询问dataset的colums用法。

Issue - State: closed - Opened by yy7798541 15 days ago - 1 comment
Labels: solved

#5809 - Qwen2.5-32B-Instruct-AWQ微调完全胡言乱语

Issue - State: closed - Opened by syusama 15 days ago - 1 comment
Labels: solved

#5808 - 如何关闭验证集？

Issue - State: open - Opened by GasolSun36 15 days ago
Labels: pending

#5808 - 如何关闭验证集？

Issue - State: open - Opened by GasolSun36 15 days ago
Labels: pending

#5807 - size mismatch for base_model.model.model...

Issue - State: open - Opened by ZijunSong 16 days ago
Labels: pending

#5807 - size mismatch for base_model.model.model...

Issue - State: open - Opened by ZijunSong 16 days ago
Labels: pending

#5806 - Installing unsloth

Issue - State: open - Opened by NathanaelTamirat 16 days ago
Labels: pending

#5806 - Installing unsloth

Issue - State: open - Opened by NathanaelTamirat 16 days ago
Labels: pending

#5805 - Videollama2集成

Issue - State: open - Opened by Evanhimself 16 days ago
Labels: pending

#5805 - Videollama2集成

Issue - State: open - Opened by Evanhimself 16 days ago
Labels: pending

#5803 - 使用webui做evaluate时模型输出出现乱码

Issue - State: open - Opened by Kawai1Ace 16 days ago
Labels: pending

#5803 - 使用webui做evaluate时模型输出出现乱码

Issue - State: open - Opened by Kawai1Ace 16 days ago
Labels: pending

#5802 - Gemma 2 forward pass broken

Issue - State: closed - Opened by Sehyo 16 days ago - 2 comments
Labels: solved

#5802 - Gemma 2 forward pass broken

Issue - State: closed - Opened by Sehyo 16 days ago - 2 comments
Labels: solved

#5800 - Support for the model : ibm-granite/granite-3.0-8b-instruct

Issue - State: open - Opened by ArchchanaKugathasan 16 days ago
Labels: pending

#5800 - Support for the model : ibm-granite/granite-3.0-8b-instruct

Issue - State: open - Opened by ArchchanaKugathasan 16 days ago
Labels: pending

#5799 - Update README.md

Pull Request - State: closed - Opened by NoumanAhmad448 16 days ago - 1 comment
Labels: wontfix

#5799 - Update README.md

Pull Request - State: closed - Opened by NoumanAhmad448 16 days ago - 1 comment
Labels: wontfix

#5798 - 4bit-QLora + Qwen2 72b + 16k cutoff_len

Issue - State: open - Opened by lmc8133 16 days ago - 4 comments
Labels: pending

#5797 - trust_remote_code=True is required when training from scratch

Issue - State: closed - Opened by cunliangkong 16 days ago
Labels: solved

#5797 - trust_remote_code=True is required when training from scratch

Issue - State: closed - Opened by cunliangkong 16 days ago
Labels: solved

#5796 - 请问现在支持 Llama-3.2-11B-Vision-Instruct 吗？

Issue - State: open - Opened by HMacro 16 days ago - 1 comment
Labels: pending

#5796 - 请问现在支持 Llama-3.2-11B-Vision-Instruct 吗？

Issue - State: open - Opened by HMacro 16 days ago - 1 comment
Labels: pending

#5786 - GLM4-9b-chat LoRA微调报错

Issue - State: closed - Opened by 2500035435 17 days ago - 6 comments
Labels: invalid

#5784 - When using the Liger kernel, get an error: 'tensor' object has no attribute 'cast'.

Issue - State: open - Opened by Tendo33 17 days ago - 3 comments
Labels: pending

#5781 - fix getattr bug

Pull Request - State: closed - Opened by dongrixinyu 17 days ago - 2 comments
Labels: wontfix

#5781 - fix getattr bug

Pull Request - State: closed - Opened by dongrixinyu 17 days ago - 2 comments
Labels: wontfix

#5770 - llama_pro二次预训练后的模型微调eval_loss为nan

Issue - State: open - Opened by tammypi 18 days ago - 2 comments
Labels: pending

#5770 - llama_pro二次预训练后的模型微调eval_loss为nan

Issue - State: open - Opened by tammypi 18 days ago - 2 comments
Labels: pending

#5768 - <video>查找判断问题BUG

Issue - State: closed - Opened by cqray1990 18 days ago - 5 comments
Labels: solved

#5766 - 运行代码后，数据处理一直为0%

Issue - State: open - Opened by caoyaru123 18 days ago - 5 comments
Labels: pending

#5763 - 华为NPU适配，依赖冲突。

Issue - State: open - Opened by yangyang6666 18 days ago - 8 comments
Labels: pending, npu

#5763 - 华为NPU适配，依赖冲突。

Issue - State: open - Opened by yangyang6666 18 days ago - 8 comments
Labels: pending, npu

#5746 - Add llava med dataset

Pull Request - State: open - Opened by snova-supasani 21 days ago

#5745 - 如何加载本地的.parquet数据训练，没有看到example？

Issue - State: open - Opened by cqray1990 21 days ago
Labels: pending

#5744 - How to deploy a completion api instead of a chat completion api

Issue - State: closed - Opened by thinkwee 21 days ago - 2 comments
Labels: pending

#5743 - dpo训练system prompt问题

Issue - State: closed - Opened by ccp123456789 21 days ago - 1 comment
Labels: solved

#5742 - When will support for allenai/Molmo be added?

Issue - State: open - Opened by HenryHe0123 22 days ago
Labels: pending

#5741 - 请问什么时候支持P-Tuning V2 呢？

Issue - State: closed - Opened by Timmy-love-you 22 days ago
Labels: wontfix

#5740 - Cannot manually assign eval dataset during sft training

Issue - State: closed - Opened by rocke2020 22 days ago - 1 comment
Labels: solved

#5739 - 用llama_pro预训练得到的模型文件

Issue - State: closed - Opened by tammypi 22 days ago - 1 comment
Labels: solved

#5738 - Does this version support running with terminal instead of yaml config?

Issue - State: closed - Opened by zhaoxu98 22 days ago - 2 comments
Labels: solved

#5736 - 界面的eval验证过程最后卡住，很慢

Issue - State: open - Opened by zzk2021 22 days ago - 1 comment
Labels: pending

#5734 - llama3_lora_sft训练在极短时间内结束

Issue - State: closed - Opened by JusticeJason 22 days ago - 1 comment
Labels: solved

#5732 - Memory Usage and Input Length

Issue - State: closed - Opened by yhy-2000 22 days ago
Labels: invalid

#5731 - STF后发现模型的基础能力丢失

Issue - State: closed - Opened by babybboy 22 days ago
Labels: wontfix

#5730 - 请问这样配置是不是就可以利用llama_pro进行增量预训练了？

Issue - State: closed - Opened by tammypi 23 days ago - 5 comments
Labels: solved

#5729 - 使用wandb之后并没有记录val_loss的图像

Issue - State: closed - Opened by wodelt 23 days ago
Labels: invalid

#5728 - webui不显示新增token

Issue - State: closed - Opened by Cheung-Z 23 days ago - 2 comments
Labels: solved

#5727 - 【HELP】Unable to open the web UI interface deployed on the server.

Issue - State: open - Opened by NGCcolor 23 days ago - 3 comments
Labels: pending

#5726 - 如何在lora训练合并后 新增数据训练？

Issue - State: closed - Opened by Kenwwww 23 days ago - 1 comment
Labels: solved