int8 repositories

Trending repositories tagged int8, ranked by trending score. Updated daily.

  1. 1
    Video-Rebirth/hyperflow-h3-flash

    HyperFlow H3 Flash: Video Rebirth's 13.7B, 4- to 8-step compressed MiniMax-H3 for joint video and audio generation (INT8 and GGUF)

    GitHub · Python · 241 stars · 17 forks · +3 stars in 24 hours · Trending score 1.65

  2. 2
    IxMxAMAR/SageAttention-RDNA4

    Drop-in SageAttention for AMD RX 9070 / 9070 XT (RDNA4, gfx1201): custom HIP attention kernels for fp16 and bf16, 1.06-2.2x faster than PR #368's gfx12 port over the full call. Prebuilt Windows wheels; ComfyUI uses it through --use-sage-attention.

    GitHub · Python · 15 stars · 1 forks · +3 stars in 24 hours · Trending score 0.93

  3. 3
    BennyDaBall930/comfyui_hyperflow_int8_nodes

    Unofficial ComfyUI loader for HyperFlow H3 Flash INT8 checkpoints (fl2va and ref2va) on ComfyUI's native INT8 convrot kernels. Needs the official ComfyUI-HyperFlow-Flash pack.

    GitHub · Python · 17 stars · 2 forks · 0 stars in 24 hours · Trending score 0.45

  4. 4
    LaoBro/ComfyUI-GPU-Tuning

    Measured runtime tuning for ComfyUI: per-architecture kernel selection, int8 linear tiles, butterfly convrot rotation, attention kernels (gfx906 first-class)

    GitHub · Python · 10 stars · 1 forks · 0 stars in 24 hours · Trending score 0.07

  5. 5
    intel/neural-compressor

    SOTA low-bit LLM quantization (INT8/FP8/MXFP8/INT4/MXFP4/NVFP4) & sparsity; leading model compression techniques on PyTorch, TensorFlow, and ONNX Runtime

    GitHub · Python · 2,730 stars · 333 forks · 0 stars in 24 hours · Trending score 0.03

  6. 6
    awesome-pro/hiqcache

    INT8-quantized hierarchical KV cache for SGLang HiCache. Compresses KV evicted from GPU L1 into the CPU L2 tier, holding 1.78x more tokens in the same host memory. Measured on Qwen3-8B against a BF16 baseline.

    GitHub · Python · 11 stars · 0 forks · 0 stars in 24 hours · Trending score 0.02

  7. 7
    dansupergameprogrammer/SuperSLM

    Deterministic, integer-only SLM inference runtime — no floating point on the inference path, so the same model and prompt produce bit-identical output across CPUs, GPUs, compilers, and vendors.

    GitHub · C++ · 8 stars · 1 forks · 0 stars in 24 hours · Trending score 0.00

  8. 8
    yzfly/edgejev

    离线可用的本地类型化决策:4 核 CPU 单题 15.6ms。Local & offline Jev / System One inference on CPU — ONNX + INT8, no torch at runtime. 支持 laya / kev / PlayJev

    GitHub · Python · 34 stars · 4 forks · 0 stars in 24 hours · Trending score 0.00

  9. 9
    yeungtuzi/Xiaotu-LLM

    XTU = X Transformers Unity — hybrid (CPU-expert + GPU) inference acceleration for vLLM: mainline plugin + CPU MoE engine (AVX512-VNNI/BF16 MXFP4/FP8), long-prefill GPU streaming, SM80 (A100) port.

    GitHub · Python · 16 stars · 0 forks · 0 stars in 24 hours · Trending score 0.00

  10. 10
    T8mars/Comfyui-Kandinsky-T8

    Native Kandinsky 6 video/audio nodes for ComfyUI, with single-file INT8 ConvRot and PiFlow support.

    GitHub · Python · 7 stars · 1 forks · 0 stars in 24 hours · Trending score 0.00

  11. 11
    StartLuxLabs/StartLux-Prism

    Fast, exact inference for StartLux-Decision models on NVIDIA GPUs: the released GGUF files on int8 and 4-bit tensor cores, behind the /v1/systemone API, with up to 5.3× lower latency than vLLM in FP8 and bit-identical answers.

    GitHub · Python · 459 stars · 5 forks · 0 stars in 24 hours · Trending score 0.00

  12. 12
    tuanpmt/vtts

    Tiny multilingual text-to-speech that runs fully on ESP32-P4 and ESP32-S3: C99 int8 runtime, per-language front-ends, single-language and bilingual voices

    GitHub · C · 25 stars · 1 forks · Trending score 0.10

  13. 13
    mgolcu00/freeimage-local

    Z-Image Turbo at full speed on a 6 GB laptop GPU. int8 + SageAttention on stable-diffusion.cpp, the engine behind freeimage.dev.

    GitHub · Shell · 72 stars · 6 forks · Trending score 0.09