FRAMEWIREIndonesiaUpdated Sep 11Live wire
0:00 / 0:00

Deepseek v4.1 flash game is out. I ran it with a robot and tested it. The smoothness is ok and the picture is ok.

Use it more often, it's really good, not worse than gpt

麦麦提Sep 117
0:00 / 0:00

WorkBuddy overseas version can actually use GPT6 Astra directly?

?? 而且最近 DeepSeek V4.1-Flash还限时免费~ 腾讯爸爸大气! 兄弟们,太香了,速冲~ 地址:

木马人Sep 1110
0:00 / 0:00

The author did a great job with this video. I converted it into a Chinese version and revised it based on the thesis, and shared it with everyone.

The focus of this video is how DeepSeek V4.1 Flash reduces the KV cache required to save historical information. It mainly does four things: Reduce duplicate storage: 40 layers share four global caches by group, three for the encoder and one for the decoder. Compressed history entries: the encoder combines two tokens…

岚叔Sep 112