Skip to content

[CVPR 2025] This repository contains the official implementation of "PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models"

active 2025-03-222026-04-09 (UTC)

Complete coverage26,536 / 26,536 hourly files (100%) · 2 absent upstream2023-08-152026-08-24 (UTC)
Events
147
Pushes
61
Pull requests
0
Issues
9
Stars
49
Forks
2

Activity over time

Daily event counts in the loaded window

Line chart, 384 days from 2025-03-22 to 2026-04-09. Pushes: 61 total, peak 18 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 9 total, peak 4 in a day. Comments: 23 total, peak 12 in a day. Stars: 49 total, peak 3 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
Mohamed-Dhouib7961014
Fuzm76004
Dog-Yang2002
Danielxu02082002
Zizzzzzzz2001
qgh1223qgh1000
wolkendolf1000

Recent activity

Latest issues, pull requests and releases

  • Issue comment#7Mohamed-Dhouib2025-12-03 13:52
    Unstable Inference Results and Progressive Pruning Configuration
  • Issue#6Mohamed-Dhouib2025-08-13 11:40
    Support for Qwen2.5, InternVL2.5, InternVL3 models
  • Issue#6wolkendolf2025-08-06 07:58
    Support for Qwen2.5, InternVL2.5, InternVL3 models
  • Issue#4Mohamed-Dhouib2025-08-03 14:15
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Mohamed-Dhouib2025-07-15 15:36
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Mohamed-Dhouib2025-07-15 14:47
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#5Mohamed-Dhouib2025-07-15 14:37
    Only a batch size of 1 is supported.
  • Issue#5Mohamed-Dhouib2025-07-15 14:37
    Only a batch size of 1 is supported.
  • Issue comment#4Mohamed-Dhouib2025-07-15 14:35
    Wrong mask preparing pipeline in llava_arch.py
  • Issue#4Mohamed-Dhouib2025-07-15 14:34
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Mohamed-Dhouib2025-07-15 14:34
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Fuzm72025-07-15 14:30
    Wrong mask preparing pipeline in llava_arch.py
  • Issue#4Fuzm72025-07-15 14:30
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Fuzm72025-07-15 14:29
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Fuzm72025-07-15 14:23
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Mohamed-Dhouib2025-07-15 14:21
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Mohamed-Dhouib2025-07-15 14:19
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Fuzm72025-07-15 14:17
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#4Mohamed-Dhouib2025-07-15 12:56
    Wrong mask preparing pipeline in llava_arch.py
  • Issue#4Fuzm72025-07-15 07:30
    Wrong mask preparing pipeline in llava_arch.py
  • Issue comment#3Danielxu02082025-06-28 04:20
    Cannot change the token remained even though I change “pruning_filter_with_percentage”
  • Issue comment#3Mohamed-Dhouib2025-06-27 22:37
    Cannot change the token remained even though I change “pruning_filter_with_percentage”
  • Issue comment#3Danielxu02082025-06-27 20:31
    Cannot change the token remained even though I change “pruning_filter_with_percentage”
  • Issue comment#3Mohamed-Dhouib2025-06-27 19:32
    Cannot change the token remained even though I change “pruning_filter_with_percentage”
  • Issue comment#2Zizzzzzzz2025-06-26 09:14
    the core code of PACT

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 49 stars here means stars gained during the window, not the repo's star count.