feature: quantization + remote client/server mode + retriever storage_qdrant option - #2

Open

random-plm wants to merge 7 commits into

mainfrom

feature/client-server

Commits on Apr 16, 2026

Add 4-bit and 8-bit model quantization support
random-plm
committed
c46c483
feat(architecture): introduce client-server split to offload inference to remote server
random-plm
committed
b7149c3
Adding storage qdrant option to MultiModalRetrieverModel from_pretrained signature; defaulting to false as it's an extra dependency
random-plm
committed
a3dc287

Commits on Apr 17, 2026

Apply query token filtering in both local and remote modes
random-plm
committed
1ab28ef

Commits on Apr 19, 2026

cap transformers lib version to <5; no support for qwen2.5 when using >= 5
random-plm
committed
75cb82d
Adding verify ssl option support
random-plm
committed
8810c14

Commits on Apr 20, 2026

Multiple features:
random-plm
committed
7b274b8