Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Running onebitadam tests

See original GitHub issue

I have tried running the tests for 1-bit Adam

cd tests/onebitadam
pytest test_com_reduce_host.py

But it just returns collecting ... %

So I was wondering if I should be calling the tests in a different way?

Issue Analytics

State:
Created 3 years ago
Comments:6 (3 by maintainers)

Top GitHub Comments

1reaction

tjruwasecommented, Sep 25, 2020

@david-macleod Can you please run pytest in verbose mode? pytest -v -s test_com_reduce_host.py

0reactions

david-macleodcommented, Oct 1, 2020

Thanks @jeffra I will try that

Top Results From Across the Web

1-bit Adam: Up to 5x less communication volume and up to 3.4 ...

To install the prerequisites run: pip install deepspeed[1bit_adam]. We have tested CUDA-Aware MPI communication using the MVAPICH2-GDR library.

DeepSpeed Integration - Hugging Face

DeepSpeed's main optimizers are Adam, AdamW, OneBitAdam, and Lamb. These have been thoroughly tested with ZeRO and are thus recommended to be used....

Train 1 trillion+ parameter models - PyTorch Lightning

However to load the model and run test/validation/predict you must use the Trainer object. Collating Single File Checkpoint for DeepSpeed ZeRO Stage 3....

AI千集-AI智能创作平台-openoker/DeepSpeed: DeepSpeed是一个 ...

Get_rank(); #TODO: Detect the hostname we are running on automatically ... buffers for communication; dummy_optim = OnebitAdam(dummy_model, cuda_aware=True) ...

Everything You Need To Know About The Running ... - Polar

Find out your unique training zones for running and track your progress with the Polar Running Performance Test.