mirror of
https://github.com/hanzoai/ml.git
synced 2026-11-20 22:24:53 +00:00
Three-way per file through the rename: base 39355c6c91, theirs the upstream
merge of main with 35d7ae7c (via 7ef6a694d), ours origin/main.
Lands the barrier thread pool, runtime CPU feature dispatch, aarch64 and x86
repacked quantized kernels (indexed_gemv, gemv_fused_shared_lhs), quantized
embedding on CPU/Metal/CUDA, paged flash-attn kernels, cutile fused MoE, the
in-place op aliasing API, parking_lot storage locks, Metal buffer builders and
debug labels, safetensors 0.8, hf-hub 1.0, tokenizers 0.23.
Ours keeps ROCm, Vulkan, wgpu, the IQ/MXFP4/NVFP4/TQ dtypes, and the i-quant
weight/activation block split in matmul. Our backends ride InplaceOpN and the
quantized embedding; Metal get_rows gains the IQ kernels. ug leaves, as upstream.
Co-authored-by: Hanzo Dev <[email protected]>
hanzo-ml-mimi
Mimi is a state of the art audio compression model using an encoder/decoder architecture with residual vector quantization. The hanzo implementation supports streaming meaning that it's possible to encode or decode a stream of audio tokens on the flight to provide low latency interaction with an audio model.
Running one example
Generating some audio tokens from an audio files.
wget https://github.com/metavoiceio/metavoice-src/raw/main/assets/bria.mp3
cargo run --example mimi --features mimi --release -- audio-to-code bria.mp3 bria.safetensors
And decoding the audio tokens back into a sound file.
cargo run --example mimi --features mimi --release -- code-to-audio bria.safetensors bria.wav