GithubHelp home page GithubHelp logo

Comments (5)

jeffra avatar jeffra commented on July 1, 2024 2

Thanks for reporting this bug. We will take a look at this as soon as possible. I just created two test cases that reproduce the error (one with ZeRO and one with FP16 but no ZeRO).

https://github.com/microsoft/DeepSpeed/blob/jeffra/onecycle_bug/tests/unit/test_fp16.py#L147-L246

from deepspeed.

astariul avatar astariul commented on July 1, 2024 1

Note : I have the same error when I try the same configuration and LRRangeTest as scheduler.

from deepspeed.

jeffra avatar jeffra commented on July 1, 2024 1

Hi @colanim, it should be up to date. Can you tell us this info from inside your docker container?

python -c 'import deepspeed; print("deepspeed info:", deepspeed.__version__, deepspeed.__git_branch__, deepspeed.__git_hash__)'

Also I just looked at the lasted docker build, it prints this same version info and it looks to be aligned with the latest March 12th commit (3d3f8d3): https://dev.azure.com/DeepSpeedMSFT/DeepSpeed/_build/results?buildId=416&view=logs&j=3dc8fd7e-4368-5a92-293e-d53cefc8c4b3&t=a1aa9649-a94b-5ac4-3f5e-9bb6223edb04&l=1717

** info: 0.1.0 master 3d3f8d3

from deepspeed.

astariul avatar astariul commented on July 1, 2024

I'm still meeting this issue in deepspeed/deepspeed:latest. How should I update the docker image to pull latest code from source ?

from deepspeed.

astariul avatar astariul commented on July 1, 2024

My bad, I didn't pull the latest image..

After doing docker pull deepspeed/deepspeed:latest, it's working 👍

from deepspeed.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.