Post
68
I'm working on an open-source dataset with 1M and 10M diverse, calibrated decisions. Get your encoders ready! 🔥
The dataset will be filtered, and there will be a normal LLM which will analyze it for accuracy purposes. I will release the cleanest samples I could (alongside the unfiltered one ofc)
The dataset will be filtered, and there will be a normal LLM which will analyze it for accuracy purposes. I will release the cleanest samples I could (alongside the unfiltered one ofc)