YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Kokoro-82M Fast CPU GGUF
Project Overview
This is an optimized version of the Kokoro-82M TTS text encoder, designed specifically for high-performance CPU inference.
Key Optimizations
- Architecture: Unrolled the 12-shared-layer ALBERT encoder into 6 independent layers.
- Speed: Reduced depth by 50% for approximately 2x faster execution on CPU.
- Format: Converted to GGUF for seamless integration with C++ runtimes and efficient memory mapping.
- Weights: Pre-projected the 128-dim embeddings into a 768-dim space to eliminate runtime computation bottlenecks.
File Details
kokoro-fast-cpu.gguf: The optimized model binary.
Usage
Target runtime should support ALBERT-style GGUF models with 6 layers and a hidden dimension of 768.
- Downloads last month
- 133
Hardware compatibility
Log In to add your hardware
We're not able to determine the quantization variants.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support