Skip to content

LongRoPE is a novel method that can extends the context window of pre-trained LLMs to an impressive 2048k tokens.

active 2024-07-142026-02-21 (UTC)

Complete coverage26,426 / 26,426 hourly files (100%) · 2 absent upstream2023-08-152026-08-20 (UTC)
Events
284
Pushes
3
Pull requests
2
Issues
20
Stars
231
Forks
17

Activity over time

Daily event counts in the loaded window

Line chart, 588 days from 2024-07-14 to 2026-02-21. Pushes: 3 total, peak 1 in a day. Pull requests: 2 total, peak 1 in a day. Issues: 20 total, peak 2 in a day. Comments: 8 total, peak 2 in a day. Stars: 231 total, peak 13 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

Recent activity

Latest issues, pull requests and releases

  • Issue#26L-z-Chen2026-01-22 00:02
    Request for LongRoPE2 Reproduction Data
  • Issue comment#22J-shang2025-10-28 06:58
    Request for LongRoPE2 Reproduction Code and Details
  • Pull request#24J-shang2025-10-28 06:45
  • Issue#22L-z-Chen2025-10-25 16:16
    Request for LongRoPE2 Reproduction Code and Details
  • Issue#21brianchmiel2025-06-25 12:38
    attention scaling
  • Issue comment#20J-shang2025-03-27 08:54
    Release LongRoPE2 artifacts on Hugging Face
  • Issue comment#20blap2025-03-26 14:02
    Release LongRoPE2 artifacts on Hugging Face
  • Issue#20NielsRogge2025-03-01 01:37
    Release LongRoPE2 artifacts on Hugging Face
  • Issue#19PizzaTowerFanGD2025-02-28 21:45
    !!NOT A BUG!! Intruiging.
  • Issue comment#16sherlcok3141592025-01-20 12:27
    Can I extend gemma2-9b and llama3.1-8b models to 128k tokens?
  • Issue comment#18chqy992024-12-11 10:10
    In perplexity.py#L84, Why are the inputs and labels the same sequence (disregarding the first character)?
  • Issue#18chqy992024-12-11 09:58
    In perplexity.py#L84, Why are the inputs and labels the same sequence (disregarding the first character)?
  • Issue#17zhimin-z2024-12-04 23:00
    Any pip package release plan?
  • Issue#16hahmad20082024-11-22 11:17
    Can I extend gemma2-9b and llama3.1-8b models to 128k tokens?
  • Issue#14yangqqq-yq2024-10-31 02:48
    Can you open source your training code
  • Issue#15sysuyy2024-10-30 16:47
    A little code modify to support newer version of transformer
  • Issue#14yangqqq-yq2024-10-15 08:47
    Can you open source your training code
  • Issue#13momandai2024-09-18 09:21
    2M rescale factor
  • Issue comment#8wang997111232024-09-09 13:02
    Evolutionary search parameters
  • Issue#12momandai2024-09-05 06:56
    how can I finetune llama2-7B with 128K seqence length by using only 8 A100 GPUS?
  • Issue#11Mooler04102024-08-16 23:36
    Why doesn't Phi-3 adopt the design of "initial tokens"?
  • Issue#12momandai2024-08-06 13:21
    how can I finetune 128K seqence length use 8 A100 GPUS?
  • Issue#10momandai2024-07-30 01:48
    why target_ids is input_ids' clone?
  • Issue#11Mooler04102024-07-29 23:14
    Why doesn't Phi-3 adopt the design of "initial tokens"?
  • Issue#10momandai2024-07-29 12:35
    why target_ids is input_ids' clone?

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 231 stars here means stars gained during the window, not the repo's star count.