NVIDIA / NVIDIA/GenerativeAIExamples

ImportError: Apex was not found. Please see the NeMo README for installation instructions: https://github.com/NVIDIA/NeMo#megatron-gpt.

Open
#101 3 comments 0 reactions 1 assignee View on GitHub

@shubhadeepd is already working on this.

Since Apr 22, 2024.

bug
Dominant language
Jupyter Notebook
Stars
4.2k
Forks
1.1k
Avg merge
10h 15m
Merged PRs (30d)
1

Description

I am trying the below Lora training notebook
https://github.com/NVIDIA/GenerativeAIExamples/blob/main/models/Gemma/lora.ipynb
When running the below code segment

from nemo.collections.nlp.parts.megatron_trainer_builder import MegatronLMPPTrainerBuilder
from nemo.utils.exp_manager import exp_manager

trainer = MegatronLMPPTrainerBuilder(cfg).create_trainer()
exp_manager(trainer, cfg.exp_manager)

I am getting the error

ImportError: Apex was not found. Please see the NeMo README for installation instructions: https://github.com/NVIDIA/NeMo#megatron-gpt.

So as per the official documentation when I try to install Apex,

git clone https://github.com/NVIDIA/apex.git
cd apex
git checkout b496d85fb88a801d8e680872a12822de310951fd
pip install -v --no-build-isolation --disable-pip-version-check --no-cache-dir --config-settings "--build-option=--cpp_ext --cuda_ext --fast_layer_norm --distributed_adam --deprecated_fused_adam" ./

this installation raises the given error


Usage:
  pip install [options] <requirement specifier> [package-index-options] ...
  pip install [options] -r <requirements file> [package-index-options] ...
  pip install [options] [-e] <vcs project url> ...
  pip install [options] [-e] <local project path> ...
  pip install [options] <archive url/path> ...
no such option: --config-settings

Screenshot from 2024-04-04 10-51-03

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.