May 19

1. 17:58

   some experimental

   0xarchit
   57426fb
2. 15:32

   fix flash attention

   0xarchit
   3a9ad3f
3. 15:26

   max optimisation

   0xarchit
   0c43530
4. 15:14

   optimised cpu

   0xarchit
   d83933c
5. 14:52

   minor fixes

   0xarchit
   edf2ef0
6. 14:49

   experimental boosting changes

   0xarchit
   434da24
7. 14:15

   remove parallel support for faster inference

   0xarchit
   e22f356
8. 14:05

   Fix llama-server startup and enable LangSearch via shell tool

   0xarchit
   1420bd1
9. 13:57

   Optimize inference performance: add continuous-batching, NO_WARMUP support, improve thread/batch recommendations

   0xarchit
   392efee
10. 13:45

    Add LangSearch web search tool for small model knowledge augmentation

    0xarchit
    c459436
11. 13:32

    Add REASONING env toggle for llama-server reasoning mode

    0xarchit
    a6b5235
12. 12:07

    Add optimized llama.cpp CPU inference backend

    0xarchit
    8ebf335
13. 11:43

    initial commit

    zrxarchit
    Verified signature

    957406c