Commit History

  1. Keep extraction chunks free of context prefixes

    0a62245
  2. Harden provider response handling

    b5e99b3
  3. Use Nemotron 3 Nano Omni reasoning model

    a70f2fd
  4. Allow longer NVIDIA pipeline runs

    d679523
  5. Use NVIDIA hosted Mistral fallback

    4bd5dc7
  6. Support NVIDIA and Mistral provider setup

    2fe3d02
  7. Fallback to Mistral when Groq is unavailable

    15f6b5f
  8. Prevent UI from sticking while running

    aef38b9
  9. Fix query stream startup deadlock

    02b94ab
  10. Mark preloaded corpus ready on startup

    9999955
  11. Add Pluto memory layer and pipeline fixes

    23cdeed

  1. Restoring .env for the Space to run - fixing missing API key error

    691e458
  2. Update codebase with latest fixes and improvements

    5cf9bb7

  1. Add application file

    66ad25b