add index bitwidth=10 (#181)add index bitwidth=10

fix(csrc): Remove strong dependency on specific Torch version. (#166)If the built wheel is created with Torch version A, it cannot be used ina local environment that has Torch version B installed, where A and Bare different. This PR addresses and fixes this issue.

Update model_base.py (#124)fix quantization config for layers

release v0.0.3

add gpu monitor at web app (#50)Thank you for taking the time to review this PR. Please handle it in theway you see fit. I would appreciate any suggestions to help me furtherimprove my coding skills.https://github.com/user-attachments/assets/4289dee8-894b-42b6-bcad-1b0653a6bde4

bump to 0.0.2 (#39)

memory and bf16 (#23)- shrink memory- support bf16