Popular repositories Loading

1. nanoserve nanoserve Public

   From-scratch LLM inference server (Python/PyTorch, no custom CUDA): naive to continuous batching to paged KV cache, plus an honest audit of published optimizations against a measured noise floor.

   Python 1

2. TimeBank TimeBank Public

   JavaScript

3. Mate Mate Public

   TEAM SHRIMP 🍤

   Python

4. StrategyBuilder StrategyBuilder Public

   HTML

5. PennAIR PennAIR Public

   Python

6. eeg-microstate-pipeline eeg-microstate-pipeline Public

   Harmonised EEG microstate pipeline across eight public resting-state datasets (504 participants): one preprocessing recipe, one global microstate model, and classification of neurodegenerative and …

   Python