Skip to content

REST: Retrieval-Based Speculative Decoding, NAACL 2024

active 2023-11-152026-03-05 (UTC)

Partial coverage19,983 / 25,698 hourly files (78%) · 2 absent upstream · 5,712 failed, retryable2023-09-072026-08-12 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
209
Pushes
5
Pull requests
7
Issues
22
Stars
123
Forks
12

Activity over time

Daily event counts in the loaded window

Line chart, 842 days from 2023-11-15 to 2026-03-05. Pushes: 5 total, peak 1 in a day. Pull requests: 7 total, peak 2 in a day. Issues: 22 total, peak 2 in a day. Comments: 39 total, peak 7 in a day. Stars: 123 total, peak 34 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
zhenyuhe00295219
yangbohust9005
csAugust5003
Siegfried-qgf5003
wangpatrick574040
reflectionie3001
preminstrel3002
zomss2001
leiwen832000
liranringel2001
jiogenes1000
Lihui-Gu1001
gdy1111001
xbfs1010
tim-pan1001
mrlife2191000
YouthLiuYS1000
issueraiser20011001
20650374211000

Recent activity

Latest issues, pull requests and releases

  • Pull request#35xbfs2026-03-05 08:59
  • Issue#3320650374212026-01-02 07:36
    [BUG?] There appears to be a bug in DraftRetriever's core search code.
  • Issue comment#31Lihui-Gu2025-05-29 12:34
    thread '<unnamed>' panicked at 'called `Option::unwrap()` on a `None` value'
  • Issue#30YouthLiuYS2025-04-27 11:21
    question about token generation
  • Issue comment#13gdy1112025-03-13 02:40
    ValueError: draft_choices was not cut enough / draft_len should not exceed 65
  • Issue comment#28zhenyuhe002025-02-21 05:28
    Questions about the construction of c_i, t_i pairs in the datastore
  • Issue#28jiogenes2025-01-08 07:17
    Questions about the construction of c_i, t_i pairs in the datastore
  • Issue comment#17liranringel2024-12-07 09:48
    LLama3 8B is not supported
  • Issue#17liranringel2024-12-07 09:48
    LLama3 8B is not supported
  • Issue#27mrlife2192024-12-05 07:45
    Multi-GPU Processing Problem
  • Issue comment#26csAugust2024-12-03 02:07
    Llama3 Branch Still Suffers Segmentation Fault When Generating Datastore Using Qwen2.5
  • Issue#26csAugust2024-12-03 02:07
    Llama3 Branch Still Suffers Segmentation Fault When Generating Datastore Using Qwen2.5
  • Issue comment#26csAugust2024-11-30 08:08
    Llama3 Branch Still Suffers Segmentation Fault When Generating Datastore Using Qwen2.5
  • Issue comment#26csAugust2024-11-30 07:22
    Llama3 Branch Still Suffers Segmentation Fault When Generating Datastore Using Qwen2.5
  • Issue#26csAugust2024-11-30 06:20
    Segmentation Fault When Generating Datastore Using Qwen2.5
  • Issue comment#23zhenyuhe002024-11-20 01:32
    Support DraftRetriever datastore read/write for large tokenizers and vocabulary sizes (i.e. llama3+)
  • Issue#20Siegfried-qgf2024-09-27 08:24
    about retrieve sequence length
  • Issue comment#18Siegfried-qgf2024-09-26 06:39
    How to adjust the maximum token lengths when drafting?
  • Issue comment#19zhenyuhe002024-09-25 05:16
    OSError: failed to fill whole buffer
  • Issue comment#19Siegfried-qgf2024-09-25 04:50
    OSError: failed to fill whole buffer
  • Issue comment#19Siegfried-qgf2024-09-25 04:00
    OSError: failed to fill whole buffer
  • Issue comment#19zhenyuhe002024-09-25 03:43
    OSError: failed to fill whole buffer
  • Issue#19Siegfried-qgf2024-09-25 03:35
    OSError: failed to fill whole buffer
  • Issue comment#18zomss2024-09-07 11:30
    How to adjust the maximum token lengths when drafting?
  • Issue comment#18zhenyuhe002024-09-07 08:05
    How to adjust the maximum token lengths when drafting?

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 123 stars here means stars gained during the window, not the repo's star count.