shibing624/lmft issues and pull requests

#18 - RuntimeError: No executable batch size found, reached zero.

Issue - State: open - Opened by km1994 almost 2 years ago - 1 comment

#17 - lora微调版本只能在chatglm-6b基础上使用吗，4位量化模型或8位量化，加载会报错

Issue - State: open - Opened by liuyunrui123 almost 2 years ago - 1 comment

#16 - 使用lora方式微调，是否对模型原先回答能力造成混乱，以及多轮对话造成影响呢。

Issue - State: open - Opened by huangding1535 almost 2 years ago - 2 comments

#15 - 如何开启多卡训练？

Issue - State: open - Opened by Zarc98 almost 2 years ago - 2 comments

#14 - 触发到训练后的数据时，回答完成后会继续产生超长的不相关内容

Issue - State: closed - Opened by xioozq almost 2 years ago - 2 comments

#13 - 先膜拜一下

Issue - State: closed - Opened by buhu2k almost 2 years ago - 1 comment

#12 - 针对中文纠错，运行gradio_demo.py文件报错问题

Issue - State: closed - Opened by huangding1535 almost 2 years ago - 5 comments

#11 - 对Finetune后的效果有点疑问

Issue - State: open - Opened by CoderAnn almost 2 years ago - 11 comments

#10 - Load预训练参数后发现效果不太对

Issue - State: closed - Opened by CoderAnn almost 2 years ago - 1 comment

#9 - 貌似没有看到finetune Belle和llama的相关代码

Issue - State: closed - Opened by AItechnology almost 2 years ago - 2 comments

#8 - RuntimeError: The size of tensor a (72) must match the size of tensor b (44) at non-singleton dimension 0

Issue - State: closed - Opened by archwolf118 almost 2 years ago - 9 comments

#7 - RuntimeError: Tensors must have same number of dimensions: got 4 and 2

Issue - State: open - Opened by zbbwss almost 2 years ago - 1 comment

#6 - P-tuning方法训练完以后模型直接没用了。

Issue - State: open - Opened by dragononly almost 2 years ago - 3 comments

#5 - AttributeError: 'ChatGLMForConditionalGeneration' object has no attribute 'enable_input_require_grads'

Issue - State: closed - Opened by zbbwss almost 2 years ago - 2 comments

#4 - Please specify `target_modules` in `peft_config`

Issue - State: closed - Opened by zbbwss almost 2 years ago - 3 comments

#3 - 训练完成最后一次保存会报错，之前每个epoch的保存没有问题

Issue - State: closed - Opened by xiaoranchenwai almost 2 years ago - 2 comments

#2 - Int4 qe

Pull Request - State: closed - Opened by shibing624 almost 2 years ago

#1 - update int4qe.