AISatoshi (@AiXsatoshi)Mistral의 Mistral-Medium-3.5-128B-EAGLE에 대해 speculative decoding 가속 테스트 결과를 공유했다. 비디오 코딩 기준 수용률이 약 25~30% 수준이며, MoE 모델이 빠르지만 Dense 모델과 전용 speculative decoding 모델 조합도 유용하다고 언급했다.https://x.com/AiXsatoshi/status/2049543302530355622#mistral #speculativedecoding #llm #aigeneration #moe
Related
Google is working on a fix for those annoying Pixel Watch permissions errorsThere's no fix ready just yet, but Google sa...
Google is working on a fix for those annoying Pixel Watch permissions errorsThere's no fix ready just yet, but Google says that it's aware of the problem and working on one.https:/...
📰 Salad Chains Are Seeing Foot Traffic Drop Over Cyclosporiasis FearsFoot traffic to leafy green chains is falling, data...
📰 Salad Chains Are Seeing Foot Traffic Drop Over Cyclosporiasis FearsFoot traffic to leafy green chains is falling, data shows. Still, a few brave souls who spoke to WIRED were det...
@GhostOnTheHalfShell Whenever people become over reliant on a technology system, human factors #psych calls it “over-rel...
@GhostOnTheHalfShell Whenever people become over reliant on a technology system, human factors #psych calls it “over-reliance.” It means trusting the system to do what you expect i...