tphuang(@tphuang):Eddie Wu said Qwen team had success in progressing w/ RSI. plan for 5-10T param model. expecting significant increase in data consumption for reasoning the in the future -> bullish for memory. Qwen-4 to use new model architecture for training. Qwen-4.5 & 5 are moving steadily. Qwen-Image-3.1 to be unveiled. Same w/ Happy Shrimp 1.1 & HappOyster 2.0 Preview. What's up w/ all this happy creature names for models?

tphuang

@tphuang

My random thoughts on EVs, clean energy, chips, aerospace and other tech. Find more extended pieces at substack

加入 May 2009

407 正在關注 37.4K 粉絲

tphuang@tphuang

2026.09.22 19:30

Eddie Wu said Qwen team had success in progressing w/ RSI. plan for 5-10T param model. expecting significant increase in data consumption for reasoning the in the future -> bullish for memory. Qwen-4 to use new model architecture for training. Qwen-4.5 & 5 are moving steadily. Qwen-Image-3.1 to be unveiled. Same w/ Happy Shrimp 1.1 & HappOyster 2.0 Preview. What's up w/ all this happy creature names for models?