ClearML Dockerfile fix #9876

glenn-jocher · 2022-10-20T18:13:16Z

Signed-off-by: Glenn Jocher [email protected]

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Update to YOLOv5 Dockerfile, optimizing Python package installations.

📊 Key Changes

Removed torchtext and torchvision from the pip uninstall command.
Updated pip install command to exclude clearml and fix OpenCV version constraint.

🎯 Purpose & Impact

🔁 The Dockerfile is simplified by removing unnecessary uninstalls.
📈 Users will experience more stable builds due to the pinned version of OpenCV.
❌ Removing clearml could hint at a streamlining of dependencies for specific use-cases, reducing image size and build time.

Signed-off-by: Glenn Jocher <[email protected]>

glenn-jocher · 2022-10-20T18:16:25Z

@thepycoder ran into a ValueError with ClearML default install in Dockerfile on DDP training. Occurs when ClearML is installed and a training command is run (no auth or other steps taken).

root@44816c00311e:/usr/src/app# python -m torch.distributed.run --nproc_per_node 2 --master_port 1 train.py --data coco128.yaml --weights yolov5s.pt --img 640 --device 2,3
train: weights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=300, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=None, image_weights=False, device=2,3, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest
remote: Enumerating objects: 29, done.
remote: Counting objects: 100% (29/29), done.
remote: Compressing objects: 100% (16/16), done.
remote: Total 29 (delta 17), reused 22 (delta 13), pack-reused 0
Unpacking objects: 100% (29/29), 19.13 KiB | 2.13 MiB/s, done.
From https://github.com/ultralytics/yolov5
   6371de8..3b1a9d2  master     -> origin/master
   319b395..c6e9ea5  exp8       -> origin/exp8
github: ⚠️ YOLOv5 is out of date by 1 commit. Use `git pull` or `git clone https://github.com/ultralytics/yolov5` to update.
YOLOv5 🚀 v6.2-203-g6371de8 Python-3.8.13 torch-1.12.1+cu113 CUDA:2 (A100-SXM-80GB, 81251MiB)
                                                             CUDA:3 (A100-SXM-80GB, 81251MiB)

Added key: store_based_barrier_key:1 to store for rank: 0
Rank 0: Completed store-based barrier for key:store_based_barrier_key:1 with 2 nodes.
hyperparameters: lr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0
Comet: run 'pip install comet_ml' to automatically track and visualize YOLOv5 🚀 runs in Comet
TensorBoard: Start with 'tensorboard --logdir runs/train', view at http:https://localhost:6006/
Traceback (most recent call last):
  File "train.py", line 630, in <module>
    main(opt)
  File "train.py", line 524, in main
    train(opt.hyp, opt, device, callbacks)
  File "train.py", line 93, in train
    loggers = Loggers(save_dir, weights, opt, hyp, LOGGER)  # loggers instance
  File "/usr/src/app/utils/loggers/__init__.py", line 121, in __init__
    self.clearml = ClearmlLogger(self.opt, self.hyp)
  File "/usr/src/app/utils/loggers/clearml/clearml_utils.py", line 87, in __init__
    self.task = Task.init(
  File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 601, in init
    task = cls._create_dev_task(
  File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 3122, in _create_dev_task
    task = cls(
  File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 199, in __init__
    super(Task, self).__init__(**kwargs)
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/task/task.py", line 155, in __init__
    super(Task, self).__init__(id=task_id, session=session, log=log)
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/base.py", line 145, in __init__
    super(IdObjectBase, self).__init__(session, log, **kwargs)
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/base.py", line 39, in __init__
    self._session = session or self._get_default_session()
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/base.py", line 115, in _get_default_session
    InterfaceBase._default_session = Session(
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_api/session/session.py", line 186, in __init__
    raise ValueError(
ValueError: ClearML configuration could not be found (missing `~/clearml.conf` or Environment CLEARML_API_HOST)
To get started with ClearML: setup your own `clearml-server`, or create a free account at https://app.clear.ml
WARNING:torch.distributed.elastic.multiprocessing.api:Sending process 418 closing signal SIGTERM
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 417) of binary: /opt/conda/bin/python
Traceback (most recent call last):
  File "/opt/conda/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/opt/conda/lib/python3.8/site-packages/torch/distributed/run.py", line 765, in <module>
    main()
  File "/opt/conda/lib/python3.8/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 345, in wrapper
    return f(*args, **kwargs)
  File "/opt/conda/lib/python3.8/site-packages/torch/distributed/run.py", line 761, in main
    run(args)
  File "/opt/conda/lib/python3.8/site-packages/torch/distributed/run.py", line 752, in run
    elastic_launch(
  File "/opt/conda/lib/python3.8/site-packages/torch/distributed/launcher/api.py", line 131, in __call__
    return launch_agent(self._config, self._entrypoint, list(args))
  File "/opt/conda/lib/python3.8/site-packages/torch/distributed/launcher/api.py", line 245, in launch_agent
    raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
============================================================
train.py FAILED
------------------------------------------------------------
Failures:
  <NO_OTHER_FAILURES>
------------------------------------------------------------
Root Cause (first observed failure):
[0]:
  time      : 2022-10-20_18:10:27
  host      : 44816c00311e
  rank      : 0 (local_rank: 0)
  exitcode  : 1 (pid: 417)
  error_file: <N/A>
  traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
============================================================
root@44816c00311e:/usr/src/app#

Also occurs with basic single-GPU training:

root@44816c00311e:/usr/src/app# python train.py --data coco128.yaml --weights yolov5s.pt --img 640 --device 2
train: weights=yolov5s.pt, cfg=, data=coco128.yaml, hyp=data/hyps/hyp.scratch-low.yaml, epochs=300, batch_size=16, imgsz=640, rect=False, resume=False, nosave=False, noval=False, noautoanchor=False, noplots=False, evolve=None, bucket=, cache=None, image_weights=False, device=2, multi_scale=False, single_cls=False, optimizer=SGD, sync_bn=False, workers=8, project=runs/train, name=exp, exist_ok=False, quad=False, cos_lr=False, label_smoothing=0.0, patience=100, freeze=[0], save_period=-1, seed=0, local_rank=-1, entity=None, upload_dataset=False, bbox_interval=-1, artifact_alias=latest
remote: Enumerating objects: 5, done.
remote: Counting objects: 100% (5/5), done.
remote: Compressing objects: 100% (5/5), done.
remote: Total 5 (delta 0), reused 0 (delta 0), pack-reused 0
Unpacking objects: 100% (5/5), 3.12 KiB | 3.13 MiB/s, done.
From https://github.com/ultralytics/yolov5
 * [new branch]      glenn-jocher-patch-2 -> origin/glenn-jocher-patch-2
github: ⚠️ YOLOv5 is out of date by 1 commit. Use `git pull` or `git clone https://github.com/ultralytics/yolov5` to update.
YOLOv5 🚀 v6.2-203-g6371de8 Python-3.8.13 torch-1.12.1+cu113 CUDA:2 (A100-SXM-80GB, 81251MiB)

hyperparameters: lr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=0.05, cls=0.5, cls_pw=1.0, obj=1.0, obj_pw=1.0, iou_t=0.2, anchor_t=4.0, fl_gamma=0.0, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0
Comet: run 'pip install comet_ml' to automatically track and visualize YOLOv5 🚀 runs in Comet
TensorBoard: Start with 'tensorboard --logdir runs/train', view at http:https://localhost:6006/
Traceback (most recent call last):
  File "train.py", line 630, in <module>
    main(opt)
  File "train.py", line 524, in main
    train(opt.hyp, opt, device, callbacks)
  File "train.py", line 93, in train
    loggers = Loggers(save_dir, weights, opt, hyp, LOGGER)  # loggers instance
  File "/usr/src/app/utils/loggers/__init__.py", line 121, in __init__
    self.clearml = ClearmlLogger(self.opt, self.hyp)
  File "/usr/src/app/utils/loggers/clearml/clearml_utils.py", line 87, in __init__
    self.task = Task.init(
  File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 601, in init
    task = cls._create_dev_task(
  File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 3122, in _create_dev_task
    task = cls(
  File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 199, in __init__
    super(Task, self).__init__(**kwargs)
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/task/task.py", line 155, in __init__
    super(Task, self).__init__(id=task_id, session=session, log=log)
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/base.py", line 145, in __init__
    super(IdObjectBase, self).__init__(session, log, **kwargs)
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/base.py", line 39, in __init__
    self._session = session or self._get_default_session()
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_interface/base.py", line 115, in _get_default_session
    InterfaceBase._default_session = Session(
  File "/opt/conda/lib/python3.8/site-packages/clearml/backend_api/session/session.py", line 186, in __init__
    raise ValueError(
ValueError: ClearML configuration could not be found (missing `~/clearml.conf` or Environment CLEARML_API_HOST)
To get started with ClearML: setup your own `clearml-server`, or create a free account at https://app.clear.ml
root@44816c00311e:/usr/src/app#

glenn-jocher · 2022-10-20T18:18:48Z

@thepycoder seems unrelated to Docker. I'll raise a bug report.

EDIT: raised in #9877

ClearML Dockerfile fix

7588cfd

Signed-off-by: Glenn Jocher <[email protected]>

glenn-jocher mentioned this pull request Oct 20, 2022

Multi-GPU distributed error #9504

Closed

1 task

glenn-jocher merged commit eef9057 into master Oct 20, 2022

glenn-jocher deleted the glenn-jocher-patch-2 branch October 20, 2022 18:17

glenn-jocher linked an issue Oct 20, 2022 that may be closed by this pull request

Multi-GPU distributed error #9504

Closed

1 task

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ClearML Dockerfile fix #9876

ClearML Dockerfile fix #9876

glenn-jocher commented Oct 20, 2022 •

edited by UltralyticsAssistant

Loading

glenn-jocher commented Oct 20, 2022

glenn-jocher commented Oct 20, 2022 •

edited

Loading

ClearML Dockerfile fix #9876

ClearML Dockerfile fix #9876

Conversation

glenn-jocher commented Oct 20, 2022 • edited by UltralyticsAssistant Loading

🛠️ PR Summary

🌟 Summary

📊 Key Changes

🎯 Purpose & Impact

glenn-jocher commented Oct 20, 2022

glenn-jocher commented Oct 20, 2022 • edited Loading

glenn-jocher commented Oct 20, 2022 •

edited by UltralyticsAssistant

Loading

glenn-jocher commented Oct 20, 2022 •

edited

Loading