Boltz-1安装使用指南

安装

首先先看官方GitHub：https://github.com/jwohlwend/boltz
最好新建环境安装：conda create --name boltz1 python=3.9
默认安装流程是pip install boltz -U，但是国内安装速度巨慢，所以这边加一个镜像参数pip install boltz -U -i https://pypi.tuna.tsinghua.edu.cn/simple

安装成功以后使用命令boltz查看有无问题

我这里有报错

(boltz1) ┌─[chunhui@Degron-ChunhuiXu] - [~] - [2024-12-20 04:41:58]
└─[0] <> boltz
Traceback (most recent call last):
File "/home/chunhui/miniconda3/envs/boltz1/bin/boltz", line 5, in <module>
 from boltz.main import cli
File "/home/chunhui/boltz/src/boltz/main.py", line 8, in <module>
 import torch
File "/home/chunhui/miniconda3/envs/boltz1/lib/python3.9/site-packages/torch/__init__.py", line 367, in <module>
 from torch._C import *  # noqa: F403
ImportError: /home/chunhui/miniconda3/envs/boltz1/lib/python3.9/site-packages/torch/lib/../../nvidia/cusparse/lib/libcusparse.so.12: undefined symbol: __nvJitLinkComplete_12_4, version libnvJitLink.so.12

(boltz1) ┌─[chunhui@Degron-ChunhuiXu] - [~] - [2024-12-23 11:18:23]
└─[0] <> python -c "import torch; print(torch.__version__); print(torch.__path__)"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/home/chunhui/miniconda3/envs/boltz1/lib/python3.9/site-packages/torch/__init__.py", line 367, in <module>
    from torch._C import *  # noqa: F403
ImportError: /home/chunhui/miniconda3/envs/boltz1/lib/python3.9/site-packages/torch/lib/../../nvidia/cusparse/lib/libcusparse.so.12: undefined symbol: __nvJitLinkComplete_12_4, version libnvJitLink.so.12

Torch包无法被正常加载，通过 pip install torch torchvision torchaudio -i https://pypi.tuna.tsinghua.edu.cn/simple解决

此时继续测试boltz命令

(boltz1) ┌─[chunhui@Degron-ChunhuiXu] - [~/boltz] - [2024-12-23 01:51:54]
└─[130] <git:(main 327a0a6) > boltz
Usage: boltz [OPTIONS] COMMAND [ARGS]...
      
  Boltz1.
      
Options:
  --help  Show this message and exit.
      
Commands:
  predict  Run predictions with Boltz-1.

加载CCD文件和训练权重文件

查阅GitHub repo可知，CCD文件和ckpt文件都需要从huggingface下载。

国内目前ban掉了huggingface

从VPS上下载好后转移到环境里使用

wget https://huggingface.co/boltz-community/boltz-1/resolve/main/ccd.pkl
wget https://huggingface.co/boltz-community/boltz-1/resolve/main/boltz1_conf.ckpt

测试示例文件

使用boltz predict examples/ligand.yaml --use_msa_server来测试使用

(boltz1) ┌─[chunhui@Degron-ChunhuiXu] - [~/boltz] - [2024-12-23 02:04:30] └─[130] <git:(main 327a0a6) > boltz predict examples/ligand.yaml –use_msa_server Checking input data. Running predictions for 1 structure Processing input data. 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 8.65it/s] GPU available: True (cuda), used: True TPU available: False, using: 0 TPU cores HPU available: False, using: 0 HPUs /home/chunhui/miniconda3/envs/boltz1/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/logger_connector/logger_connector.py:75: Starting from v1.9.0, tensorboardX has been removed as a dependency of the pytorch_lightning package, due to potential conflicts with other packages in the ML ecosystem. For this reason, logger=True will use CSVLogger as the default logger, unless the tensorboard or tensorboardX packages are found. Please pip install lightning[extra] or one of them to enable TensorBoard support by default You are using a CUDA device (‘NVIDIA RTX A4000’) that has Tensor Cores. To properly utilize them, you should set torch.set_float32_matmul_precision('medium' | 'high') which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0] Predicting DataLoader 0: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [06:18<00:00, 0.00it/s]Number of failed examples: 0 Predicting DataLoader 0: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [06:18<00:00, 0.00it/s]

Boltz-1安装使用指南

行前定，则不疚。道前定，则不穷。

Boltz-1安装使用指南

安装

加载CCD文件和训练权重文件

测试示例文件

Catalog

FEATURED TAGS

FRIENDS