Speculative Decoding in vLLM on AMD GPUs(vllm.ai)
130 points by ankitg12 | 47 comments
130 points by ankitg12 | 47 comments
24 points by rajveerb | 4 comments
406 points by simedw | 177 comments
Blog post introducing Spegel, a proof-of-concept terminal web browser that passes HTML through an LLM and renders the result as markdown in the terminal.