Skip to content

Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding

active 2025-12-022026-05-25 (UTC)

Partial coverage11,680 / 12,902 hourly files (91%) · 2 absent upstream · 1,219 failed, retryable2025-02-192026-08-10 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
58
Pushes
0
Pull requests
0
Issues
4
Stars
50
Forks
3

Activity over time

Daily event counts in the loaded window

Line chart, 175 days from 2025-12-02 to 2026-05-25. Pushes: 0 total, peak 0 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 4 total, peak 2 in a day. Comments: 1 total, peak 1 in a day. Stars: 50 total, peak 6 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
yechenzhi4001
ChenZH9939991000

Recent activity

Latest issues, pull requests and releases

  • Issue#4ChenZH9939992026-02-09 12:57
    [Performance Issue] Pillar mode is slower than Full mode on H200 with Qwen3-8B (AIME24)
  • Issue#2yechenzhi2025-12-23 00:29
    [Build Error] FlashInfer JIT compilation fails with recent RAFT versions (missing select_k.cuh & linking librmm)
  • Issue comment#2yechenzhi2025-12-18 02:20
    [Build Error] FlashInfer JIT compilation fails with recent RAFT versions (missing select_k.cuh & linking librmm)
  • Issue#2yechenzhi2025-12-13 03:40
    [Build Error] FlashInfer JIT compilation fails with recent RAFT versions (missing select_k.cuh & linking librmm)
  • Issue#1yechenzhi2025-12-13 02:07
    High CPU Memory Usage During KV Pool Initialization on H800 with 100GB RAM

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 50 stars here means stars gained during the window, not the repo's star count.