LMFlow

LMFlow copied to clipboard

训练时是将多个instruction合并起来训练还是一个instruct训练一次 ？

1

我看训练文本是text only，在训练时好像没有区分instruction和output,是直接concat在一起训练的吗？

mengjiexu

License for the non-LLaMA based models

1

Hi, I can see that you released models that are not based on LLaMA. What datasets were they trained on, though? I can see in `data/download.sh` you download datasets such...

Acrobot

为什么加载数据，一直无法访问

3

 像这样我试了好久

Moon-creat

请问可以用这个框架进行LLM的自监督训练吗？

1

如果我有大量的领域文本，能用LMFlow直接在纯文本上训练吗？也就是直接把纯文本喂给模型。

calvinzhan

window GPU选择最小规格

2

window 10 成功安装lmflow 目前GPU RTX2060s 能运行那个模型，怎么运行？

NingBoHao

请问有没有关于LMFlow的结构阐述

2

比如论文之类的，想知道里面的训练结构，模型细节

ylz123

What's the differences between RAFT and InstructGPT's PPO?

1

Greetings! I am very interested in your work RAFT. When I read the paper, you still use the RL methods to score the model's output. It isn't very clear to...

Unrealluver

no config.json

4

Why is there no “config.json” after training with the file “run_finetune_with_lora_save_aggregated_weights.sh”? 

w-JiqQian

How to evaluate RAFT?

1

请问repo中提供了raft fine-tuning后测试模型效果的代码吗？比如在imdb测试集上计算reward或perplexity

yyhycx

Unable to pre-compile ops without torch installed. Please install torch before attempting to pre-compile ops.

1

在执行 pip install -e . 命令报错 Preparing metadata (setup.py) ... error error: subprocess-exited-with-error × python setup.py egg_info did not run successfully. │ exit code: 1 ╰─> [9 lines of output]...

ChenWenKai8023ni