Popular repositories Loading

1. whisper-transcriber whisper-transcriber Public

   One-click local transcription for Windows: provisions its own environment and picks the best available runtime (CUDA / Intel NPU / iGPU / CPU).

   Python 1

2. Simple_Code_LLM Simple_Code_LLM Public

   A LLM designed for python coding, trained from scratch on a single RTX 5090.

   Python

3. Memory-Overclocking-Lab Memory-Overclocking-Lab Public

   DDR5 memory overclocking lab notebook: latency tuning, stability testing, and hardware limits on AMD Ryzen 9800X3D

4. local-llm-serving-bench local-llm-serving-bench Public

   Single-stream measurements of three local inference engines (vLLM, NInfer, LM Studio) on one RTX 5090: context depth 0-96K, KV cache dtypes, CUDA graphs, speculative decoding.

   Python