Repository navigation

Issues

New issue

One-click Windows launcher for Bonsai-demo (community tool, not a PR)
#114 · Basearchio opened on Jul 19, 2026
1
POC: From-scratch Bonsai Q1_0 inference engine running on WebGPU compute shaders
#72 · ruihe774 opened on May 10, 2026
3

Bonsai image 2
PrismML-Eng/Bonsai-demo#272 · opened on Oct 21, 2026
(Benchmark] Bonsai 2 27B PTQ1_0 vs Qwen3.8-27B: Coding, Tool Safety and Completion Regressions in Local Agent Workflows
PrismML-Eng/Bonsai-demo#269 · opened on Oct 19, 2026
2
Benchmark: Ternary Bonsai 2 27B (PQ2_0) on RTX PRO 6000 Blackwell Max-Q + RTX PRO 4000 Blackwell SFF — builds clean for sm_120
PrismML-Eng/Bonsai-demo#267 · opened on Oct 16, 2026
1
[Feedback/Demo] Successfully running Ternary-Bonsai-2-27B natively on Android via custom C++ Hadamard Runtime (5.6 tok/s)
PrismML-Eng/Bonsai-demo#260 · opened on Oct 12, 2026
[Documentation/Bug] Fix for CUDA 13.x / RTX 50-series initialization failure on Windows
PrismML-Eng/Bonsai-demo#248 · opened on Sep 28, 2026
2
Bug: Rigid Jinja chat template causes 500 error ("No user query found") with VS Code Continue extension
PrismML-Eng/Bonsai-demo#241 · opened on Sep 25, 2026
Feature Request: Native MTP Support for Bonsai-2-27B
PrismML-Eng/Bonsai-demo#235 · opened on Sep 24, 2026
1
Chat template raises on mid-conversation system messages — breaks Claude Code / Anthropic-protocol clients
PrismML-Eng/Bonsai-demo#213 · opened on Sep 21, 2026
cuda version bug in download_binaries.sh?
PrismML-Eng/Bonsai-demo#206 · opened on Sep 20, 2026
4
Three CPU/AVX2 + recurrent-state PRs for Ternary Bonsai-2 27B (3.9 → 8.3 tok/s bit-exact on a laptop CPU)
PrismML-Eng/Bonsai-demo#196 · opened on Sep 19, 2026
8
scripts: context tier sizes by system RAM only - default -c doesn't fit on mid-size GPUs
PrismML-Eng/Bonsai-demo#193 · opened on Sep 19, 2026
1
[27B / agentic] Harness works, code quality unusable — feedback wanted on setup and approach (Apple M5 Max)
PrismML-Eng/Bonsai-demo#183 · opened on Sep 18, 2026
8