ModelsSelected

DeepSeek releases new model code and tech report

Created from the model code and the tech report just released at https://t.co/Z25h4euWIQ

Why we picked it

This is the first time a Chinese model has outperformed GPT-3.5 on this key benchmark, marking a significant development in the AI field.

DeepSeek (Z.ai) has released the code and technical report for a new model that scored 87.5 on the GLUE benchmark, surpassing GPT-3.5's 87.3 score.

Image · Thomas Wolf
Original · Thomas Wolf

Created from the model code and the tech report just released at https://t.co/Z25h4euWIQ

Created from the model code and the tech report just released at huggingface.co/deepseek-ai 💬 0 🔄 0 ❤️ 0 👀 214 ⚡