Giters
SpongebBob
/
Finetune-ChatGLM2-6B
ChatGLM2-6B 全参数微调,支持多轮对话的高效微调。
Geek Repo:
Geek Repo
Github PK Tool:
Github PK Tool
Stargazers:
389
Watchers:
8
Issues:
22
Forks:
39
SpongebBob/Finetune-ChatGLM2-6B Issues
请问可以训练chatGLM2-6B-32K的模型嘛
Updated
6 months ago
基于school_math_0.25M.json数据集进行微调训练后得到的模型推理效果很差,是什么原因?
Updated
7 months ago
全量微调显存占用
Updated
7 months ago
Comments count
8
ChatGLM数据处理中input_ids的padding问题
Updated
7 months ago
微调时在runing tokenizer on train dataset这一步卡住
Closed
7 months ago
Comments count
1
我只有200多条多轮对话的数据,去做全参微调能有效果吗?
Updated
9 months ago
Comments count
2
微调出来的模型可以直接用chatglm2官网发版的web_demo2.py来测吗,会有影响吗?
Updated
9 months ago
Comments count
1
模型训练,出现deepspeed_init() got an unexpected keyword argument 'resume_from_checkpoint'
Updated
9 months ago
Comments count
4
max_len 768 这个长度在多轮对话中,是单次输入的长度,还是history+这次输入的长度? 如果是单次输入的长度,那么history的长度在哪里?
Closed
9 months ago
--max_length 762,我想问一下这个是输入的最大长度吗,如果是,那么如果是多轮对话,有长度限制吗?
Closed
9 months ago
timed out
Updated
9 months ago
Comments count
1
有考虑 加入lora训练吗
Updated
9 months ago
Comments count
1
请问这个问题怎么解决OSError: Can't get source for <function apply_rotary_pos_emb at 0x7fef8c15f790>. TorchScript requires source access in order to carry out compilation, make sure original .py files are available.
Updated
10 months ago
请问全参数微调需要什么配置呢?
Updated
10 months ago
Comments count
9
请问这个项目的依赖是什么,安装的deepspeed版本好像跑不了。
Closed
10 months ago
Comments count
1
训练数据过多时报错Socket Timeout
Closed
10 months ago
模型训练,input_ids出现None type
Updated
10 months ago
Comments count
2
为什么不用新版本的transformer么?
Closed
10 months ago
CUDA out of memory. Tried to allocate 11.63 GiB (GPU 0; 23.69 GiB total capacity; 11.63 GiB already allocated; 11.28 GiB free
Updated
10 months ago
Comments count
1
多轮数据下载后要如何处理?
Updated
10 months ago
do_eval时报错
Updated
10 months ago
Comments count
1
RuntimeError: CUDA error: invalid device ordinal
Updated
10 months ago
Comments count
1