Pure C++ implementation of several models for real-time chatting on your computer (CPU)
active 2025-01-06 → 2026-08-04 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 576 days from 2025-01-06 to 2026-08-04. Pushes: 163 total, peak 5 in a day. Pull requests: 6 total, peak 4 in a day. Issues: 49 total, peak 3 in a day. Comments: 103 total, peak 8 in a day. Stars: 173 total, peak 5 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 comment#129ssechi22026-05-23 04:16Converting TranslateGemma to Q4_K takes an incredibly long time.
- Issue comment#127foldl2026-05-15 04:49Question: preferred channel for a security disclosure?
- Issue#125paulocoutinhox2026-04-16 05:38You this project as reference
- Issue#122ssechi22026-03-30 03:24Running the translategemma-4b-it (F16/q8_0) on a relatively long text with a context size greater than 1024 results in an error and chatllm crashes with CUDA / Vulkan backends, but it works fine on CPU.
- Issue comment#105nissansz2026-03-28 10:14dots ocr, result of some pages are wrong, some are correct.
- Issue#70foldl2026-03-27 11:23support new input dimension: batch
- Issue comment#121foldl2026-03-19 01:27About VL models (TAG: video)
- Issue#121dorpxam2026-03-18 23:56About VL models (TAG: video)
- Issue comment#119dorpxam2026-03-17 08:32About Thinking support of Qwen3-VL ?
- Issue#118foldl2026-03-17 02:18Support proposal : Nvidia's Music Flamingo (AudioFlamingo3ForConditionalGeneration)
- Issue#119foldl2026-03-17 02:18About Thinking support of Qwen3-VL ?
- Issue comment#119foldl2026-03-17 02:18About Thinking support of Qwen3-VL ?
- Issue comment#105foldl2026-03-12 04:09dots ocr, result of some pages are wrong, some are correct.
- Issue comment#105nissansz2026-03-12 03:18dots ocr, result of some pages are wrong, some are correct.
- Issue comment#105foldl2026-03-11 06:06dots ocr, result of some pages are wrong, some are correct.
- Issue comment#117kaituoxu2026-02-25 06:40[Feature Request] Any plans to integrate the FireRedASR2S model series?
- Issue#117foldl2026-02-25 01:30[Feature Request] Any plans to integrate the FireRedASR2S model series?
- Issue#116vieenrose2026-02-15 08:14ForcedAligner crashes when input contains punctuation-only sentences
- Issue comment#115vieenrose2026-02-14 06:06fix(memory): eliminate memory leaks in Python bindings and inference pipeline
- Issue#110bchtrue2026-02-13 19:39lack of examples, qwn3 asr
- Issue comment#111vieenrose2026-02-13 04:11Fix critical crash: 'free(): invalid pointer' in CoreAttention
- Pull request#114vieenrose2026-02-13 04:09
- Pull request#113vieenrose2026-02-13 04:08
- Pull request#112vieenrose2026-02-13 04:08
- Pull request#111vieenrose2026-02-13 04:08
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 173 stars here means stars gained during the window, not the repo's star count.