Shenzhen

Jitai Hao

Intermediate
@CURRENTF

https://currentf.github.io/

currentf.github.io/

Sparse-vLLM. A sparse-first inference engine (sparsevllm). It also contains DeltaKV compressor training + evaluation tooling (deltakv).

60

LowRankClone. [NeurIPS 2025 Spotlight] A Token is Worth over 1,000 Tokens: Efficient Knowledge Distillation through Low-Rank Clone.

48

MEFT. Python

22

HRN. Code of Multi-Defendant Legal Judgment Prediction via Hierarchical Reasoning

15

Uni-X. [ICLR 2026] Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal Models

10

qydemo. qy

1

UGen-Pro. Python

1

Codex-Skills. Python

1
8
Apply