Play-with-LLMs icon indicating copy to clipboard operation
Play-with-LLMs copied to clipboard

Tutorial on training, evaluating LLM, as well as utilizing RAG, Agent, Chain to build entertaining applications with LLMs.分享如何训练、评估LLMs,如何基于RAG、Agent、Chain构建有趣的LLMs应用。

Results 14 Play-with-LLMs issues
Sort by recently updated
recently updated
newest added

when i try to train the model,It was all normal at first,but some times later ,it get a strange mistake,how can i solve it ![image](https://github.com/EvilPsyCHo/train_custom_LLM/assets/41049893/3ce327dc-23ea-49e2-91e7-d38a73773497)

请问多轮对话微调和推理具体如何实现呀,发现训练的时候dataset.py实际没有用到

后面p-tuning能讲讲吗

model.enable_input_require_grads() 这块报错。如何解决呀

在微调的时候,200个迭代之后,开始eval,显存会直接飙升一截,并且不再下降。不知道这个泄露是什么引起的

我有一点疑惑,请问为什么没有mask将instruction+input 的损失mask掉。看代码的意思是输入为instruction+input+output,然后预测目标也是instruction+input+output。

对于CUDA大于等于12.0的,需要直接安装bitsandbytes最新版本(我这边显示不兼容):pip install -U bitsandbytes 除了requirements.txt里的,还要安装其他包:datasets、sentencepiece、scipy

This repo with videos is helpful to me. Nevetheless, I still met some issues. May the following tips save your time. 1. [**Linux:Yes, Windows:No**] Bitsandbytes is officially maintained on Linux....

哈喽,当前拉的最新peft 没有0.4.0dev ![image](https://github.com/EvilPsyCHo/train_custom_LLM/assets/81786651/bc9394ca-08c9-45b4-930a-e2846feaa0fb)