Skip to content

Cap Mistral's context length at 2k#495

Merged
alan-cooney merged 3 commits intoTransformerLensOrg:mainfrom
collingray:cap_mistral_n_ctx
Jan 28, 2024
Merged

Cap Mistral's context length at 2k#495
alan-cooney merged 3 commits intoTransformerLensOrg:mainfrom
collingray:cap_mistral_n_ctx

Conversation

@collingray
Copy link
Copy Markdown
Contributor

@collingray collingray commented Jan 25, 2024

Description

Caps Mistral's context length at 2k, as its default length of 32k uses a very large amount of memory.

Discussed in #490 and #491

This change will not be needed after #493, and is just to allow for Mistral to be easily used in the short term while #493 is in review.

Fixes #490

Type of change

Please delete options that are not relevant.

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to not work as expected)
  • This change requires a documentation update

Checklist:

  • I have commented my code, particularly in hard-to-understand areas
  • I have made corresponding changes to the documentation
  • My changes generate no new warnings
  • I have added tests that prove my fix is effective or that my feature works
  • New and existing unit tests pass locally with my changes
  • I have not rewritten tests relating to key interfaces which would affect backward compatibility

Copy link
Copy Markdown
Collaborator

@alan-cooney alan-cooney left a comment

Choose a reason for hiding this comment

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

Yep thanks makes sense - approving now and we'll revert once we have the full fix

@alan-cooney alan-cooney merged commit ba3fb3b into TransformerLensOrg:main Jan 28, 2024
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.

[Proposal] Change Mistral's config to reduce context size from 32k to 4k

2 participants