Files
zeekayandHanzo Dev 6a363f7ee3 merge: upstream main ddf1b879 and the aarch64 repack branch 35d7ae7c
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]>
2026-09-15 15:52:28 -07:00
..

hanzo-ml-quantized-glm4

Hanzo implementation of various quantized GLM4-0414 models.

Running an example

Run local gguf file (with local tokenizer.json)

$ cargo run --example quantized-glm4 --release --features cuda -- --tokenizer /home/data/GLM-4-9B-0414/tokenizer.json --model /home/data/GLM-4-9B-0414-Q4_K_M.gguf  --prompt "How are you today?"

Run local gguf file with tokenizer.json downloaded form huggingface

$ cargo run --example quantized-glm4 --release --features cuda -- --which q4k9b --model /home/data/GLM-4-9B-0414-Q4_K_M.gguf  --prompt "How are you today?"

Run with model-id (download from huggingface)

$ cargo run --example quantized-glm4 --release --features cuda -- --which q4k9b  --prompt "How are you today?"

Options for which [q2k9b, q2k32b, q4k9b, q4k32b]

Example output:

avx: true, neon: false, simd128: false, f16c: true
temp: 0.80 repeat-penalty: 1.10 repeat-last-n: 64
loaded 523 tensors (6.16GB) in 0.86s
model built

I'm just a computer program, so I don't have feelings or emotions. However, I'm functioning well and ready to assist you with any questions or tasks you might have. How can I help you today?

  10 prompt tokens processed: 67.12 token/s
  44 tokens generated: 45.28 token/s