mlcommons / mlcommons/algorithmic-efficiency

LM1B: Jax

Open
#621 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

✨ Feature Request Future Version
Dominant language
Python
Stars
425
Forks
78
PR merge metrics
No merged PRs in 30d

Description

Workload

LM1B

Task

Transformer language model training, resembling https://github.com/google/flax/tree/main/examples/lm1b .

Dataset

LM1B dataset. https://www.tensorflow.org/datasets/catalog/lm1b

Model

decoder-only Transformer, encoder+decoder based Transformer.

Reference Implementation

https://github.com/google/flax/tree/main/examples/lm1b


ToDo
  • Implement data input pipeline
    • Document specific dataset version in workload-specific README
  • Add model
    • Document model in workload-specific README
  • Provide sample submission (and sample tuning search space)
    • Document results of sample submission in workload-specific README (how long did it take to reach the target performance)

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the Google Flax LM1B reference implementation and the linked TensorFlow Datasets LM1B specification. Define the data pipeline, decoder-only and encoder-decoder Transformer models, and workload-specific README documentation. Done means the workload runs, includes a sample submission and tuning search space, and records how long it takes to reach the target performance.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, tensorflow
Domain
data, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.