deep_learning/ultralytics
1
0
Fork 0
mirror of https://github.com/ultralytics/ultralytics.git synced 2025-09-15 15:48:41 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

ultralytics 8.2.71 Multinode DDP training (#14879)

Browse Source 
Co-authored-by: Haris Rehman <haris.rehman.cowlar@gmail.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
This commit is contained in:
Haris Rehman 2024-08-01 20:31:03 +05:00 committed by GitHub
parent 16fc325308
commit 9c5d1a2451
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 9 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
.gitignore vendored
View File

@ -26,6 +26,9 @@ share/python-wheels/
.installed.cfg .installed.cfg
*.egg *.egg
MANIFEST MANIFEST
requirements.txt
setup.py
ultralytics.egg-info
# PyInstaller # PyInstaller
# Usually these files are written by a python script from a template # Usually these files are written by a python script from a template

2
ultralytics/__init__.py
View File

@ -1,6 +1,6 @@
# Ultralytics YOLO 🚀, AGPL-3.0 license # Ultralytics YOLO 🚀, AGPL-3.0 license
__version__ = "8.2.70" __version__ = "8.2.71"
import os import os

5
ultralytics/engine/trainer.py
View File

@ -26,6 +26,7 @@ from ultralytics.data.utils import check_cls_dataset, check_det_dataset
from ultralytics.nn.tasks import attempt_load_one_weight, attempt_load_weights from ultralytics.nn.tasks import attempt_load_one_weight, attempt_load_weights
from ultralytics.utils import ( from ultralytics.utils import (
DEFAULT_CFG, DEFAULT_CFG,
LOCAL_RANK,
LOGGER, LOGGER,
RANK, RANK,
TQDM, TQDM,
@ -129,7 +130,7 @@ class BaseTrainer:
# Model and Dataset # Model and Dataset
self.model = check_model_file_from_stem(self.args.model) # add suffix, i.e. yolov8n -> yolov8n.pt self.model = check_model_file_from_stem(self.args.model) # add suffix, i.e. yolov8n -> yolov8n.pt
with torch_distributed_zero_first(RANK): # avoid auto-downloading dataset multiple times with torch_distributed_zero_first(LOCAL_RANK): # avoid auto-downloading dataset multiple times
self.trainset, self.testset = self.get_dataset() self.trainset, self.testset = self.get_dataset()
self.ema = None self.ema = None
@ -285,7 +286,7 @@ class BaseTrainer:
# Dataloaders # Dataloaders
batch_size = self.batch_size // max(world_size, 1) batch_size = self.batch_size // max(world_size, 1)
self.train_loader = self.get_dataloader(self.trainset, batch_size=batch_size, rank=RANK, mode="train") self.train_loader = self.get_dataloader(self.trainset, batch_size=batch_size, rank=LOCAL_RANK, mode="train")
if RANK in {-1, 0}: if RANK in {-1, 0}:
# Note: When training DOTA dataset, double batch size could get OOM on images with >2000 objects. # Note: When training DOTA dataset, double batch size could get OOM on images with >2000 objects.
self.test_loader = self.get_dataloader( self.test_loader = self.get_dataloader(

3
ultralytics/utils/torch_utils.py
View File

@ -48,11 +48,12 @@ TORCHVISION_0_18 = check_version(TORCHVISION_VERSION, "0.18.0")
def torch_distributed_zero_first(local_rank: int): def torch_distributed_zero_first(local_rank: int):
"""Ensures all processes in distributed training wait for the local master (rank 0) to complete a task first.""" """Ensures all processes in distributed training wait for the local master (rank 0) to complete a task first."""
initialized = dist.is_available() and dist.is_initialized() initialized = dist.is_available() and dist.is_initialized()
if initialized and local_rank not in {-1, 0}: if initialized and local_rank not in {-1, 0}:
dist.barrier(device_ids=[local_rank]) dist.barrier(device_ids=[local_rank])
yield yield
if initialized and local_rank == 0: if initialized and local_rank == 0:
dist.barrier(device_ids=[0]) dist.barrier(device_ids=[local_rank])
def smart_inference_mode(): def smart_inference_mode():