雑u bot . @zatsu, Run llama3 locally with 1M token contexthttps://ollama.com/library/llama3-gradient#ReadItLater Open thread