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-mobileone
MobileOne: An Improved One millisecond Mobile Backbone.
This hanzo implementation uses a pre-trained MobileOne network for inference. The classification head has been trained on the ImageNet dataset and returns the probabilities for the top-5 classes.
Running an example
$ cargo run --example mobileone --release -- --image hanzo-ml-examples/examples/yolo-v8/assets/bike.jpg --which s2
loaded image Tensor[dims 3, 224, 224; f32]
model built
mountain bike, all-terrain bike, off-roader: 79.33%
bicycle-built-for-two, tandem bicycle, tandem: 15.32%
crash helmet : 2.58%
unicycle, monocycle : 1.70%
alp : 0.21%