Skip to content

Conversation

@jeffra
Copy link
Contributor

@jeffra jeffra commented Jul 25, 2022

  • Task specific request_dict parsing, ensures required keys are given for each task type.
  • Support appending model_path and aml model path in aml deployments
  • Pass checkpoint dict to ds-inference instead of writing to file
  • Add flake8 pre-commit and associated fixes
  • Big refactor of overall code structure

Requires DeepSpeed version that includes this: deepspeedai/DeepSpeed#2132

Copy link
Contributor

@mrwyattii mrwyattii left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good so far

@jeffra jeffra changed the title task specific parsing in score file Bloom custom ckpt files, improved model path support, MII refactor, flake8, etc. Jul 28, 2022
@jeffra jeffra merged commit 7a1544d into aml Jul 29, 2022
@jeffra jeffra deleted the task-specific-score branch July 29, 2022 18:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants