Installation

trlX is a pure Python library that supports two optional distributed backends: Huggingface 🤗 Accelerate and NVIDIA NeMo, the latter is optional and can be installed separately.

Requirements

• OS: Linux

• Python: 3.9-3.11

Install with pip

You can install trlX using pip:

$ pip install -U git+https://github.com/CarperAI/trlx.git

Install from source

You can also install trlX from source:

$ git clone https://github.com/CarperAI/trlx.git
$ cd trlx
$ pip install torch --extra-index-url https://download.pytorch.org/whl/cu118
$ pip install -e .

Install NeMo

Install NeMo version v1.17.0:

$ git clone https://github.com/NVIDIA/NeMo/
$ cd NeMo
$ git checkout d3017e4
$ pip install -e '.[all]'

Install Apex:

$ git clone https://github.com/NVIDIA/apex
$ cd apex
$ # if pip >= 23.1 (ref: https://pip.pypa.io/en/stable/news/#v23-1) which supports multiple `--config-settings` with the same key...
$ pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" ./