diffuse-cpp

company

https://github.com/iafiscal1212/diffuse-cpp

AI & ML interests

Campo: Interés en IA y ML Valor: High-performance CPU inference for Diffusion Language Models (LLaDA, Dream). GGML-based runtime with quantization and semantic scheduling.

Carmenest

in diffuse-cpp/LLaDA-8B-Instruct-GGUF 3 months ago

Performance notes: entropy_exit + inter-step cache benchmarks

#1 opened 3 months ago by

Carmenest

in diffuse-cpp/Dream-v0-Instruct-7B-GGUF 3 months ago

diffuse-cpp is now Apache-2.0 — Dream-7B excels at math and code on CPU

#1 opened 3 months ago by

Carmenest

in diffuse-cpp/LLaDA-8B-Instruct-GGUF 3 months ago

diffuse-cpp is now Apache-2.0 — 3.3x faster than llama.cpp on CPU

#2 opened 3 months ago by

Carmenest

updated 2 models 3 months ago

diffuse-cpp/Dream-v0-Instruct-7B-GGUF

Text Generation • 8B • Updated Mar 31 • 49 • 1

diffuse-cpp/LLaDA-8B-Instruct-GGUF

Text Generation • 8B • Updated Mar 31 • 128 • 6

Carmenest

published 2 models 3 months ago

diffuse-cpp/Dream-v0-Instruct-7B-GGUF

Text Generation • 8B • Updated Mar 31 • 49 • 1

diffuse-cpp/LLaDA-8B-Instruct-GGUF

Text Generation • 8B • Updated Mar 31 • 128 • 6

Carmenest

updated a model 3 months ago

diffuse-cpp/LLaDA-8B-Instruct-GGUF

Text Generation • 8B • Updated Mar 31 • 128 • 6

AI & ML interests

Team members 1

diffuse-cpp's activity

Performance notes: entropy_exit + inter-step cache benchmarks

diffuse-cpp is now Apache-2.0 — Dream-7B excels at math and code on CPU

diffuse-cpp is now Apache-2.0 — 3.3x faster than llama.cpp on CPU