Welcome! Type "help" for available commands.

$

Loading terminal interface...

Experience Education

Projects Bookmarks Investments Contact Blog

Welcome! Type "help" for available commands.

$

Loading terminal interface...

Speculative Decoding Bookmarks | William Callahan - Bookmarks

~/bookmarks

Similar Content

A collection of bookmarks filtered by the tag "Speculative Decoding".

Filter by:

Showing 1-14 of 14 bookmarks tagged with "Speculative Decoding"

All Stories

GitHub - bilikaz/qwen38-flash-next-recipe: qwen38-flash-next recipe for solo GB10

GitHub - bilikaz/qwen38-flash-next-recipe: qwen38-flash-next recipe for solo GB10

Qwen LLM Inference Speculative Decoding

Sep 14, 2026

GitHub - sxuff/qwen38-27b-nvfp4-dflash2-dgx-spark: Pinned stock Qwen3.8-27B NVFP4 + DFlash2 recipe and measured evidence for one NVIDIA DGX Spark

GitHub - sxuff/qwen38-27b-nvfp4-dflash2-dgx-spark: Pinned stock Qwen3.8-27B NVFP4 + DFlash2 recipe...

Speculative Decoding NVFP4 Qwen3.8-27B

Sep 12, 2026

GitHub - quimmedes/cafe-llama.cpp: A experimental fork of llama.cpp with LLM inference in C/C++

GitHub - quimmedes/cafe-llama.cpp: A experimental fork of llama.cpp with LLM...

Mixture-of-Experts LLM Inference Speculative Decoding

Aug 30, 2026

GitHub - LaurentZuijdwijk/llama.cpp: llama.cpp with adaptive speculative decoding (--spec-draft-adaptive) and a Vulkan backend tuned for AMD Strix Halo. 4.7x on structured output, 1.9x mainline prefill on MoE.

GitHub - LaurentZuijdwijk/llama.cpp: llama.cpp with adaptive speculative decoding (--spec-draft-adaptive) and...

LLM Inference Vulkan Speculative Decoding

Aug 29, 2026

GitHub - sxuff/qwen38-27b-nvfp4-dflash2-dgx-spark: Pinned stock Qwen3.8-27B NVFP4 + DFlash2 recipe and measured evidence for one NVIDIA DGX Spark

GitHub - sxuff/qwen38-27b-nvfp4-dflash2-dgx-spark: Pinned stock Qwen3.8-27B NVFP4 + DFlash2 recipe...

Speculative Decoding Qwen3.8-27B DFlash2

Aug 29, 2026

GitHub - MiaAI-Lab/GLM-5.3-Flash-EXL3-2x-DGX-Sparks: GLM-5.3 Flash EXL3 for 2x DGX Sparks

GitHub - MiaAI-Lab/GLM-5.3-Flash-EXL3-2x-DGX-Sparks: GLM-5.3 Flash EXL3 for 2x DGX Sparks

EXL3 Quantization Speculative Decoding GLM-5.3-Flash

Aug 29, 2026

GitHub - julianmb/q38rocm: Qwen 3.8 27B ROCmFP4 on AMD Strix Halo (Ryzen AI Max+ 395). Up to 36 tok/s via MTP Speculation, TurboQuant & Mesa RADV Wave64.

GitHub - julianmb/q38rocm: Qwen 3.8 27B ROCmFP4 on AMD Strix...

Speculative Decoding AMD Strix Halo ROCmFP4

Aug 19, 2026

GitHub - MiaAI-Lab/Qwen3.8-27B-DGX-Spark-RTX-6000: Qwen3.8 27B for DGX Spark / RTX 6000 PRO

GitHub - MiaAI-Lab/Qwen3.8-27B-DGX-Spark-RTX-6000: Qwen3.8 27B for DGX Spark / RTX...

Speculative Decoding Docker Deployment NVFP4 Quantization

Aug 16, 2026

Introducing Muse Glimmer: An Open Agentic Model That Runs on Your Device

research.meta.ai

Introducing Muse Glimmer: An Open Agentic Model That Runs on...

Speculative Decoding On-Device LLM Model Quantization

Aug 10, 2026

RedHatAI/Qwen3.6-35B-A3B-speculator.dspark · Hugging Face

RedHatAI/Qwen3.6-35B-A3B-speculator.dspark · Hugging Face

Qwen DSpark Speculative Decoding

Aug 9, 2026

z-lab/Qwen3.6-27B-DFlash · Hugging Face

z-lab/Qwen3.6-27B-DFlash · Hugging Face

LLM Inference Qwen3.6-27B DFlash

Aug 8, 2026

unsloth/gemma-4-26B-A4B-it-qat-GGUF · Hugging Face

unsloth/gemma-4-26B-A4B-it-qat-GGUF · Hugging Face

Quantization-Aware Training Speculative Decoding Gemma 4

Aug 1, 2026

GitHub - Luce-Org/lucebox-hub: Lucebox optimization hub: hand-tuned LLM inference, built for specific consumer hardware.

GitHub - Luce-Org/lucebox-hub: Lucebox optimization hub: hand-tuned LLM inference, built...

LLM Inference RTX 3090 Quantization Kernels

May 3, 2026

RedHatAI/gemma-4-31B-it-speculator.eagle3 · Hugging Face

RedHatAI/gemma-4-31B-it-speculator.eagle3 · Hugging Face

vLLM Integration Gemma Model Eagle-3 Algorithm

Apr 11, 2026

Discover More

github.com

September 19, 2026

GitHub - MiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark: Qwen3.8-Flash-Next on ONE DGX Spark (TP=1)

GitHub - MiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark: Qwen3.8-Flash-Next on ONE DGX Spark (TP=1)

Qwen3.8-Flash-Next on ONE DGX Spark (TP=1). Contribute to MiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark development by creating an account on GitHub.

LLM Inference Qwen3.8-Flash-Next Vision-Language Model vLLM DGX Spark

LINK