Process long prompts 3x faster with close to full prefill quality
AI & ML interests
LLM speculative inference, built for specific consumer hardware & heterogeneous computing
Recent Activity
View all activity
models 17
Lucebox/pflash-scorer-head
2.62M • Updated • 25
Lucebox/pflash-request-probe
265k • Updated • 26
Lucebox/pflash-segment-probe
265k • Updated • 41
Lucebox/Qwen3.8-Flash-Next-GGUF
177B • Updated • 96 • 1
Lucebox/DeepSeek-V4.1-Flash-DSpark-GGUF
14B • Updated • 1.41k • 1
Lucebox/DeepSeek-V4.1-Flash-ROCMFP23-GGUF
755B • Updated • 42.7k • 9
Lucebox/Qwen3.8-27B-IQ4_XS-fast-GGUF
Image-Text-to-Text • 27B • Updated • 672 • 3
Lucebox/DeepSeek-V4-Flash-0731-ROCmFP3
284B • Updated • 6.27k • 25
Lucebox/Qwen3.8-27B-DFlash2-GGUF
2B • Updated • 559
Lucebox/DeepSeek-V4-Flash-0731-DSpark-GGUF
20B • Updated • 682 • 7
datasets 0
None public yet