Installation verification is divided into three key steps:
- Basic environmental inspection::
- Make sure Python ≥ 3.10 and install git
- (of a computer) run
import torch; print(torch.cuda.is_available())Verify CUDA support
- Project deployment::
- pass (a bill or inspection etc)
git clone https://github.com/jingyaogong/minimind.gitGet source code - Use Tsinghua Source to accelerate the installation of dependencies:
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
- pass (a bill or inspection etc)
- Functional verification::
- Download the 1.6GB
pretrain_hq.jsonlDataset to . /dataset - (of a computer) run
python eval_model.py --load 1Testing basic reasoning skills
- Download the 1.6GB
run intoflash_attnIf the installation fails, you can temporarily comment out the relevant code for a test run.
This answer comes from the articleMiniMind: 2 hours from scratch training 26M parameters GPT open source toolsThe
May not be reproduced without permission:AI productivity tools " How do I install and verify the MiniMind runtime environment on my own computer?
