Skip to content

issues Search Results · repo:Tencent/PatrickStar language:Python

Filter by

56 results
 (87 ms)

56 results

inTencent/PatrickStar (press backspace or delete to remove)

Thank you for the great work.OVO
  • HongLouyemeng
  • Opened 
    on Sep 21, 2023
  • #311

In the readme document, the installation seems wrong. Could you clarify the correct way of installation? Thanks
  • ys0923
  • 2
  • Opened 
    on Jul 12, 2022
  • #309

While training a GPT3_6B model on 4x v100, the program stop because of runtime error at step 47. The exception show like this: RuntimeError: chunk move failed. cpu has not 385.875968 MB memory space. ...
  • ouyangliqi
  • 3
  • Opened 
    on Jun 30, 2022
  • #308

Currently, PatrickStart could train the largest pretrained model with the lowest hardware requirement, i.e. gpu and cpu memory. However, this comes with a price, as we are specializing the design to naive ...
  • zhuzilin
  • 1
  • Opened 
    on Feb 28, 2022
  • #304

有办法让 BatchNorm2d 之类的层保持 float32 进行训练吗?用 half 可能导致 loss 不好收敛
  • liaojianjin
  • 5
  • Opened 
    on Feb 21, 2022
  • #302

Failed to run the vit models. from timm.models import vit_base_patch16_224, vit_large_patch16_224
bug
  • feifeibear
  • Opened 
    on Jan 27, 2022
  • #299

pip install . --user met below error, because future is not supported on python3.6. comment patrickstar/core/eviction_policy.py#L30 solved this, seems importing future is not necessary? ERROR: Command ...
  • Jack47
  • 1
  • Opened 
    on Dec 27, 2021
  • #296

PatrickStar is awesome, it helps reduce memory used by model state! currently trends show using MegatronDeepSpeed as framework to train transformer based NLP models, both pretrain and finetuing. so will ...
  • Jack47
  • 3
  • Opened 
    on Dec 22, 2021
  • #293

The chunk list construction in PreprocessCtx not including memory copy is very time-consuming. It prevents us from quickly testing large models and makes the interaction frequency lower.
  • feifeibear
  • 1
  • Opened 
    on Dec 10, 2021
  • #274

Chunk size is a critical hyperparameter in PatrickStar. An appropriate chunk size setting is able 1. reduce fragments in chunks. 2. improve memory utilization. We intend to develop a script to choose ...
  • feifeibear
  • 1
  • Opened 
    on Dec 8, 2021
  • #272
Issue origami icon

Learn how you can use GitHub Issues to plan and track your work.

Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub Issues
ProTip! 
Press the
/
key to activate the search input again and adjust your query.
Issue origami icon

Learn how you can use GitHub Issues to plan and track your work.

Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub Issues
ProTip! 
Press the
/
key to activate the search input again and adjust your query.
Issue search results · GitHub