Discussions

• 1 You must be logged in to vote

  🙏

  使用llamafactory-cli webchat启动时，执行的是哪个脚本呢？求告知

  nonoDamn asked Oct 29, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  有思考过程的数据，训练数据是什么格式呢？有模版吗？

  cy565025164 asked Oct 28, 2025 in Q&A · Unanswered

  

  

  0
• 2 You must be logged in to vote

  🙏

  多模态数据包含history的对话数据集格式

  duplicate This issue or pull request already exists

  ZrW00 asked Aug 2, 2025 in Q&A · Unanswered

  

  

  

  

  6
• 1 You must be logged in to vote

  🙏

  Which type is selected if the argument "bf16" is set false in sft yaml? float16? 在sft训练yaml中把bf16这一参数改成false会默认用什么类型训练？

  tendyfang asked Oct 24, 2025 in Q&A · Unanswered

  

  

  0
• 0 You must be logged in to vote

  🙏

  [help] Weird loss qwen3-vl-8b

  Damon-GSY asked Oct 24, 2025 in Q&A · Unanswered

  

  

  5
• 1 You must be logged in to vote

  🙏

  NotImplementedError("Evaluation will be deprecated in the future.") NotImplementedError: Evaluation will be deprecated in the future.

  China-wuxiaofei asked Oct 23, 2025 in Q&A · Unanswered

  

  

  1
• 2 You must be logged in to vote

  🙏

  qwen2.5-vl 3B inference is OOM, but qwen2-vl 7B does not

  bug Something isn't working pending This problem is yet to be addressed

  missTL asked Mar 3, 2025 in Q&A · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🙏

  冻结部分参数微调

  jianbaba asked Oct 21, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  DPO训练在一百个step内就会loss降到0，acc变成1

  bingo-G asked Oct 17, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  大规模多模态数据集训练

  Lin-Tianwei asked Oct 14, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  💬

  packing and neat_packing

  AriesJin started Oct 11, 2025 in General

  

  

  0
• 1 You must be logged in to vote

  🙏

  怎么添加除了loss之外的指标

  pending This problem is yet to be addressed

  yzd11 asked Dec 25, 2024 in Q&A · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🙏

  微调模型部署问题

  yangzishy asked Oct 9, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  Video max pixels 设置

  kang-1-2-3 asked Oct 7, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  这个支持多模态的ppo训练吗？

  onex7777 asked Sep 29, 2025 in Q&A · Unanswered

  

  

  0
• 2 You must be logged in to vote

  🙏

  Qwen3 Moe 训练效率低

  stainless-steel-rat asked Sep 28, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  华为晟腾现在支持哪些量化方法？

  InvincibleMinions asked Sep 22, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  💬

  请教一下llamafactory微调模型后的推理问题？

  MengHao666 started Sep 21, 2025 in General

  

  

  0
• 1 You must be logged in to vote

  🙏

  请教多模态模型的图片处理参数image_max_pixels问题？

  MengHao666 asked Sep 21, 2025 in Q&A · Unanswered

  

  

  0
• 2 You must be logged in to vote

  🙏

  set cutoff_len and image_max_pixels failed

  AriesJin asked Sep 19, 2025 in Q&A · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🙏

  咨询：多个数据集怎么设置比例

  HongyongZZZ asked Sep 20, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  add_tokens和add_special_tokens训练过程中报错：

  LiziLiziok asked Sep 17, 2025 in Q&A · Unanswered

  

  

  2
• 0 You must be logged in to vote

  🙏

  咨询：是否支持根据train dataset按比例split为eval dataset

  duplicate This issue or pull request already exists

  TracyGuo2001 asked Sep 16, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  Qwen3-30b-a3b lora训练结束后保存最终checkpoint失败

  alex-wll asked Sep 15, 2025 in Q&A · Unanswered

  

  

  0
• 1 You must be logged in to vote

  🙏

  如何从断点继续训练

  Lookey-Luo asked Sep 13, 2025 in Q&A · Unanswered

  

  

  0