Skip to content

Actions: hiyouga/LLaMA-Factory

Actions

label_issue

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
label_issue.yml
1,900 workflow runs
1,900 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

How to set save_steps? I set a 15,000, and mine has more than 34,000 steps. I just trained 1.28epoch at first sight, and directly generated 57GB checkpoints for me to dry up the hard disk. label_issue #1998: Issue #6584 opened by yuanllong
January 9, 2025 14:00 14s
January 9, 2025 14:00 14s
qwen2vl全量微调后进行评估报错 label_issue #1997: Issue #6582 opened by gsfsdv
January 9, 2025 12:59 13s
January 9, 2025 12:59 13s
请问在sft多个数据集时，max_steps应该怎么算 label_issue #1996: Issue #6581 opened by 128Ghe980
January 9, 2025 10:27 10s
January 9, 2025 10:27 10s
增量）预训练、（多模态）指令监督微调、奖励模型训练、PPO 训练、DPO 训练、KTO 训练、ORPO 训练，经常提到的rlhf指的是哪几种训练方法 label_issue #1995: Issue #6580 opened by qkkcoolmax
January 9, 2025 09:19 10s
January 9, 2025 09:19 10s
如何使用CPU进行推理？ label_issue #1994: Issue #6579 opened by starryzwh
January 9, 2025 07:44 11s
January 9, 2025 07:44 11s
lora微调模型合并时报错 label_issue #1993: Issue #6578 opened by xiao-liya
January 9, 2025 07:11 13s
January 9, 2025 07:11 13s
llamafactory 和 trl 训练的DPO 存在差异 label_issue #1992: Issue #6577 opened by suanflower
January 9, 2025 06:33 14s
January 9, 2025 06:33 14s
api推理，正常打印info，无报错但无响应 label_issue #1991: Issue #6576 opened by BreezeHavana
January 9, 2025 06:20 11s
January 9, 2025 06:20 11s
多机训练时，由于不同机器pip初始化环境下载速度不同，导致部分worker进入训练进程等待通信时超时，如何增加timeout等待时间呢？ label_issue #1990: Issue #6575 opened by Mr-lonely0
January 9, 2025 03:13 13s
January 9, 2025 03:13 13s
考虑支持TTT训练吗 label_issue #1989: Issue #6574 opened by af-74413592
January 9, 2025 02:52 14s
January 9, 2025 02:52 14s
如何对Qwen的量化版本模型做sft及dpo label_issue #1988: Issue #6573 opened by gmm41
January 9, 2025 02:18 12s
January 9, 2025 02:18 12s
模型测评求助 label_issue #1987: Issue #6570 opened by kuang1216
January 8, 2025 14:08 12s
January 8, 2025 14:08 12s
The meaning of the eval parameter in qwen-full-sft.yml in the llama factory library label_issue #1986: Issue #6568 opened by yuanllong
January 8, 2025 13:35 11s
January 8, 2025 13:35 11s
Qwen2vl如何修改视频采样频率？ label_issue #1985: Issue #6567 opened by Imbernoulli
January 8, 2025 12:14 13s
January 8, 2025 12:14 13s
如何在训练的中途进行predict计算ROUGE label_issue #1984: Issue #6566 opened by gsfsdv
January 8, 2025 12:07 12s
January 8, 2025 12:07 12s
Batch Inference Error for qwen2vl Model After Full Fine-Tuning label_issue #1983: Issue #6563 opened by Bravo5542
January 8, 2025 09:24 16s
January 8, 2025 09:24 16s
Batch Inference Error for qwen2vl Model After Full Fine-Tuning label_issue #1982: Issue #6562 opened by Bravo5542
January 8, 2025 08:35 14s
January 8, 2025 08:35 14s
可以不进行 merge lora权重而多次进行SFT训练过程吗 label_issue #1981: Issue #6560 opened by Maydaytyh
January 8, 2025 07:54 11s
January 8, 2025 07:54 11s
qwen2-vl训练有bug，ValueError: Image features and image tokens do not match: tokens: 1468, features 1936,有人能回一下吗 label_issue #1980: Issue #6559 opened by gouqi666
January 8, 2025 06:05 12s
January 8, 2025 06:05 12s
【DPO相关】请问DPO训练结束后，应该如何进行推理呢？能使用类似SFT训练后推理的脚本吗 label_issue #1979: Issue #6558 opened by Maydaytyh
January 8, 2025 02:36 10s
January 8, 2025 02:36 10s
mistral-of-depth微调报错:forward() takes from 7 to 8 positional arguments but 9 were given label_issue #1978: Issue #6557 opened by jingtian11
January 7, 2025 15:53 12s
January 7, 2025 15:53 12s
【预训练】请问一下预训练时数据集需要注意什么？ label_issue #1977: Issue #6555 opened by leeburt
January 7, 2025 15:02 13s
January 7, 2025 15:02 13s
docker下内核和pip兼容问题 label_issue #1976: Issue #6554 opened by SDAIer
January 7, 2025 12:58 14s
January 7, 2025 12:58 14s
无效issue label_issue #1975: Issue #6553 opened by ranwithyou
January 7, 2025 11:35 11s
January 7, 2025 11:35 11s
How to Qwen2-vl dataset batched inference label_issue #1974: Issue #6552 opened by oyly16
January 7, 2025 10:23 11s
January 7, 2025 10:23 11s