Commit History

Yesterday

  1. fix(app): attach FastAPI routes to server_app post-launch Running

    Verified signature
    d2c36d8
  2. fix(app): resilient demo.launch with direct demo.app route attachment

    Verified signature
    7e9d0ad
  3. deploy: update to glibc 2.35 ubuntu-22.04 sm_120 native wheel

    Verified signature
    76ee3eb
  4. ci: use Blackwell sm_120 native llama-cpp-python wheel

    Verified signature
    1f93d93

2 days ago

  1. Configure n_batch=256 and n_ubatch=256 micro-batch optimization

    Verified signature
    4ca471f
  2. Finalize calibrated ZeroGPU reservation duration

    Verified signature
    a9fad2a
  3. Test FlashAttention with FP16 KV cache on ZeroGPU

    Verified signature
    0871892
  4. Fix llm_mmproj resolution and healthz status

    Verified signature
    fc7afc1
  5. Fix Blackwell sm_120 KV cache stability with FP16 and disable fattn

    Verified signature
    c10d604
  6. Deploy DocNeedle updates

    Verified signature
    4fbd330
  7. Deploy DocNeedle updates

    Verified signature
    12aac4b
  8. Deploy DocNeedle updates

    Verified signature
    8e0f826
  9. Deploy DocNeedle updates

    Verified signature
    b1a6c16
  10. Deploy DocNeedle updates

    Verified signature
    5411b9b
  11. Deploy DocNeedle updates

    Verified signature
    e08bc8c

7 days ago

  1. Deploy DocNeedle updates

    Verified signature
    2077976
  2. Deploy DocNeedle updates

    Verified signature
    1d035d2
  3. Deploy DocNeedle updates

    Verified signature
    b408ef1
  4. initial commit

    Verified signature
    16f1a6e