LLM inference with 7x longer context. Pure C, zero dependencies. Lossless KV cache compression + single-header library.
C · active 2026-04-03 → 2026-06-06 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 65 days from 2026-04-03 to 2026-06-06. Pushes: 157 total, peak 19 in a day. Pull requests: 22 total, peak 9 in a day. Issues: 26 total, peak 23 in a day. Comments: 17 total, peak 8 in a day. Stars: 50 total, peak 11 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
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| unamedkr | 218 | 157 | 22 | 14 |
| dpblnt | 1 | 0 | 0 | 0 |
| Krysalide | 1 | 0 | 0 | 1 |
| PierreChatelier | 1 | 0 | 0 | 1 |
| consigcody94 | 1 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#83unamedkr2026-04-13 04:51KV cache pre-build for RLV: question latency 35s → 5s
- Issue comment#93unamedkr2026-04-13 04:01feat(qwen3): full Qwen3-4B architecture support
- Issue comment#86unamedkr2026-04-13 03:50Auto-detect chat template from GGUF metadata
- Issue comment#90unamedkr2026-04-13 03:48Confidence score API: quantify answer reliability [0.0-1.0]
- Issue comment#82unamedkr2026-04-12 23:23Qwen3-4B architecture support — GGUF loads but inference produces garbage
- Issue#90unamedkr2026-04-12 21:42Confidence score API: quantify answer reliability [0.0-1.0]
- Issue#90unamedkr2026-04-12 21:42Confidence score API: quantify answer reliability [0.0-1.0]
- Issue#77unamedkr2026-04-12 07:48SmolLM2-1.7B server inference regression after 91814d4 (Phi-3 CPU fallback)
- Pull request#79unamedkr2026-04-12 07:48
- Issue#78unamedkr2026-04-12 07:43Make quant.h the single source of truth — auto-generate split sources
- Issue#78unamedkr2026-04-12 07:43Make quant.h the single source of truth — auto-generate split sources
- Issue comment#77unamedkr2026-04-12 07:39SmolLM2-1.7B server inference regression after 91814d4 (Phi-3 CPU fallback)
- Issue#77unamedkr2026-04-12 07:29SmolLM2-1.7B server inference regression after 91814d4 (Phi-3 CPU fallback)
- Issue#77unamedkr2026-04-12 07:29SmolLM2-1.7B server inference regression after 91814d4 (Phi-3 CPU fallback)
- Issue#75unamedkr2026-04-12 07:17quantcpp serve crashes with SIGABRT on port conflict instead of graceful error
- Issue#73unamedkr2026-04-12 06:53Server crashes when assistant message contains "!" in multi-turn chat
- Issue#75unamedkr2026-04-12 06:38quantcpp serve crashes with SIGABRT on port conflict instead of graceful error
- Issue#75unamedkr2026-04-12 06:38quantcpp serve crashes with SIGABRT on port conflict instead of graceful error
- Issue comment#71unamedkr2026-04-12 04:27Port Phi-3 architecture support to libturboquant + Qwen3.5 issues
- Pull request#71unamedkr2026-04-12 04:16
- Pull request#71unamedkr2026-04-12 04:01
- Issue comment#69unamedkr2026-04-12 03:58Qwen3.5-4B: quant_generate works but quant_ask produces empty/garbage output
- Issue#70unamedkr2026-04-12 03:51Qwen3.5-4B DeltaNet layers: FP32 dequant bottleneck causes 0.7 tok/s
- Issue#70unamedkr2026-04-12 03:51Qwen3.5-4B DeltaNet layers: FP32 dequant bottleneck causes 0.7 tok/s
- Issue#69unamedkr2026-04-12 03:50Qwen3.5-4B: quant_generate works but quant_ask produces empty/garbage output
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 50 stars here means stars gained during the window, not the repo's star count.