Code

History

Commits on fix-git-https-support to crates/collab/src/llm.rs35+
  1. collab: Adjust maximum spending limit check (#25596)Marshall Bowers committed 1 year ago7f16629
  2. collab: Limit free tier usage across all models (#25573)Marshall Bowers committed 1 year ago3a3621f
  3. Add Anthropic Claude 3.7 support (#25497)Peter Tripp committed 1 year ago10a4760
  4. collab: Remove unused `POST /predict_edits` endpoint from LLM service (#23997)Marshall Bowers committed 1 year ago8be73bf
  5. Implement simpler logic for edit predictions prompt byte limits (#23983)Michael Sloan committed 1 year ago87b0f62
  6. zeta: Onboarding and title bar banner (#23797)Agus Zubiaga committed 1 year agoe23e035
  7. llm: Sample ~10% of staff members inputs/outputs to LLM (#23537)Thorsten Ball committed 1 year ago0cb4175
  8. Timeout if completion takes longer than 2s (#23215)Antonio Scandurra committed 1 year ago880f3ff
  9. Check for `predict-edits` feature flag, remove `is_staff` check (#23165)Agus Zubiaga committed 1 year ago4a76302
  10. Add more metrics for Fireworks Completion Requested (#23062)Antonio Scandurra committed 1 year agoc26553d
  11. zeta: Report Fireworks request data to Snowflake (#22973)Thorsten Ball committed 1 year ago1fcc9b3
  12. Log errors when a prediction fails (#22961)Antonio Scandurra committed 1 year agoc330107
  13. Include outline when predicting edits with Zeta (#22895)Antonio Scandurra committed 1 year agoa8ef0f2
  14. Stop sending data to Clickhouse (#21763)Conrad Irwin committed 1 year ago03efd0d
  15. collab: Attach additional properties to `Language Model Used` event (#21770)Marshall Bowers committed 1 year ago158cdc3
  16. Introduce staff-only inline completion provider (#21739)Antonio Scandurra committed 1 year ago77b8296
  17. Send llm events to snowflake too (#21091)Conrad Irwin committed 1 year ago984bb19
  18. assistant: Remove `low_speed_timeout` (#20681)Thorsten Ball committed 1 year agoaee01f2
  19. collab: Exempt staff from LLM usage limits (#19836)Marshall Bowers committed 2 years agoa451bcc
  20. collab: Add support for a custom monthly allowance for LLM usage (#19525)Marshall Bowers committed 2 years ago1a4b253
  21. collab: Unconditionally execute billing checks (#19432)Marshall Bowers committed 2 years agob44bed0
  22. Subtract FREE_TIER_MONTHLY_SPENDING_LIMIT from reported monthly spend (#19358)Antonio Scandurra committed 2 years ago8c91054
  23. collab: Remove lifetime spending limit in favor of LLM usage billing (#19321)Marshall Bowers committed 2 years agof6fad3b
  24. Increase monthly free tier spend from 5 dollars to 10 dollars (#19291)Antonio Scandurra committed 2 years ago474e670
  25. Restore HTTP client transition, but use reqwest everywhere (#19055)Mikayla Maki committed 2 years ago22ac178
  26. collab: Don't record billing events if billing is not enabled (#19102)Marshall Bowers committed 2 years agoc709b66
  27. collab: Add usage-based billing for LLM interactions (#19081)Marshall Bowers committed 2 years ago22ea7ce
  28. collab: Make LLM billing fields required in `LlmTokenClaims` (#18959)Marshall Bowers committed 2 years ago6971166
  29. collab: Add billing preferences for maximum LLM monthly spend (#18948)Marshall Bowers committed 2 years agod316577
  30. collab: Clarify naming around free tier spending limits (#18936)Marshall Bowers committed 2 years agof1053ff
  31. collab: Add a `Cents` type (#18935)Marshall Bowers committed 2 years ago817a41c
  32. Revert http client changes (#18892)Mikayla Maki committed 2 years ago5d5c4b6
  33. collab: Update billing code for LLM usage billing (#18879)Marshall Bowers committed 2 years agof861479
  34. collab: Track cache writes/reads in LLM usage (#18834)Marshall Bowers committed 2 years agod55f025
  35. Replace isahc with async ureq (#18414)Conrad Irwin committed 2 years ago3a5deb5