Multigpu precompute dataset map function and share between processes

DealayLomoi · July 8, 2024, 11:54am

Here’ codes that precompute data using main_process_first() which use 1st gpu to caculate all precomputation and shares results across all processes. However, using one gpu to precompute on big data is slow, can we use all gpu multi-processing-ly and finally merge data in one piece which shares across all processes.

github.com

huggingface/diffusers/blob/main/examples/text_to_image/train_text_to_image_sdxl.py

#!/usr/bin/env python
# coding=utf-8
# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Fine-tuning script for Stable Diffusion XL for text2image."""

import argparse
import functools
import gc

This file has been truncated. show original

 with accelerator.main_process_first():
        from datasets.fingerprint import Hasher

        # fingerprint used by the cache for the other processes to load the result
        # details: https://github.com/huggingface/diffusers/pull/4038#discussion_r1266078401
        new_fingerprint = Hasher.hash(args)
        new_fingerprint_for_vae = Hasher.hash(vae_path)
        train_dataset_with_embeddings = train_dataset.map(
            compute_embeddings_fn, batched=True, new_fingerprint=new_fingerprint

Topic		Replies	Views
Cannot use Datasets.map on multi-gpu during evaluation Beginners	3	2909	July 1, 2024
Stable diffusion `train_text_to_image.py` only on one gpu 🧨 Diffusers	5	1182	May 2, 2023
How to ensure GPU utilisation when preprocessing huggingface datasets Beginners	1	683	April 27, 2024
Can a diffuser pipeline run on multiple GPUs? Amazon SageMaker	2	1151	May 31, 2023
Fine tunning llama2 with multiple GPUs and Hugging face trainer 🤗Transformers	1	3419	November 3, 2023

Multigpu precompute dataset map function and share between processes

Related topics