Skip to content

A PyTorch native library for training speculative decoding models on npu

active 2026-02-262026-04-15 (UTC)

Complete coverage26,648 / 26,648 hourly files (100%) · 2 absent upstream2023-08-152026-08-29 (UTC)
Events
266
Pushes
83
Pull requests
21
Issues
8
Stars
17
Forks
2

Activity over time

Daily event counts in the loaded window

Line chart, 49 days from 2026-02-26 to 2026-04-15. Pushes: 83 total, peak 12 in a day. Pull requests: 21 total, peak 4 in a day. Issues: 8 total, peak 2 in a day. Comments: 58 total, peak 15 in a day. Stars: 17 total, peak 4 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
yubofredwang8662181
Copilot550044
chatgpt-codex-connector[bot]270013
cicirori221900
torchspec-bot4110
FlamingoPg3120

Recent activity

Latest issues, pull requests and releases

  • Pull request#76yubofredwang2026-04-15 06:46
  • Pull request#76yubofredwang2026-04-15 06:38
  • Pull request#75yubofredwang2026-04-13 22:42
  • Pull request#75yubofredwang2026-04-13 22:27
  • Pull request#60yubofredwang2026-04-06 07:15
  • Pull request#59yubofredwang2026-04-05 09:57
  • Pull request#58yubofredwang2026-04-05 00:08
  • Pull request#56yubofredwang2026-03-30 05:00
  • Issue#53torchspec-bot2026-03-28 05:36
    [Bug]: Failed to run vllm backend training example with `vllm_qwen3_8b.yaml` config
  • Issue#53torchspec-bot2026-03-28 05:36
    [Bug]: Failed to run vllm backend training example with `vllm_qwen3_8b.yaml` config
  • Pull request#43yubofredwang2026-03-19 06:28
  • Pull request#40yubofredwang2026-03-17 23:32
  • Pull request#36torchspec-bot2026-03-07 22:10
  • Issue#34yubofredwang2026-03-07 01:51
    Development Roadmap (2026 Q2)
  • Issue#34yubofredwang2026-03-07 01:51
    Development Roadmap (2026 Q2)
  • Pull request#31FlamingoPg2026-03-07 00:07
  • Pull request#32yubofredwang2026-03-06 21:59
  • Pull request#31FlamingoPg2026-03-06 07:46
  • Issue#30cicirori2026-03-06 00:29
    [Feature] Adopt FlexAttention native FA4 backend (BACKEND="FLASH") to replace manual CuTeDSL integration
  • Pull request#23yubofredwang2026-03-03 10:58
  • Pull request#22yubofredwang2026-03-02 11:00
  • Issue#15yubofredwang2026-03-01 15:08
    [Feature] Support async hidden states offloading from engine to mooncake store
  • Issue#15yubofredwang2026-03-01 15:08
    [Feature] Support async hidden states offloading from engine to mooncake store
  • Pull request#21yubofredwang2026-03-01 14:08
  • Pull request#19yubofredwang2026-03-01 10:01

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 17 stars here means stars gained during the window, not the repo's star count.