
OpenAI ライブストリーム
OpenAIがライブストリーム配信イベントを開催します。放送中に具体的な発表、新製品発表、またはデモンストレーションが明かされる予定です。
OpenAIが前回予告なしのライブストリームをやった時、GPT-4 Turboをドロップして、一晩で価格設定を完全に変えました

This is a follow-up update to my previous post comparing Qwen 3.6 35B vs Gemma 4 26B. I wanted to particularly follow-up with the following: 1. Gemma 4 26B could've suffered the quantization tax and perform drastically better with an 8-bit quant. So I wanted to put that to the test with UD's Q8_K_XL this time 2. A lot of people (including myself) were curious to see how the Qwen 3.5 27B dense would perform in these tests. 3. Speaking of dense models, I also wanted to include the Gemma 4 31B to