4 commits

Author	SHA1	Message	Date
Claude	ef22946f35	debug: add shape logging and stderr error handler for inference debugging	2026-02-16 02:16:30 +00:00
Claude	c8ec0f9e49	chore: target macOS 26.0, fix duplicate -lstdc++ linker warning ... Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-16 01:55:49 +00:00
Claude	f89e80732a	fix: resolve CGo type conflict in error handler ... Use pure C callback instead of //export to avoid const char* vs GoString type mismatch in cgo-generated headers. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-16 01:53:36 +00:00
Claude	9d664c055a	feat: add native MLX backend for Apple Silicon inference (pkg/mlx) ... CGo wrapper for mlx-c providing zero-Python Metal GPU inference. Includes Gemma 3 model architecture, BPE tokenizer, KV cache, composable sampling, and OpenAI-compatible serve command. Build-tagged (darwin && arm64 && mlx) with stubs for cross-platform. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-16 01:19:04 +00:00