Explore GitLab
Discover projects, groups and snippets. Share your projects with others
-
Mini-Omni2是个视觉-音频助理，能同时处理视觉、听觉和文本三种多模态，实时提供端到端的语音对话响应。
-
-
-
image classification by transformer
-
-
-
Masked Image Training for Generalizable Deep Image Denoising
-
通过结合CoT微调、MCTS算法和推理行动策略，Marco-o1超越部分闭源产品。
-
-
Mamba2提出了一种基于半可分离矩阵块分解的 SSD 算法,以显著提高训练效率
-
-
-
-
LM Compiler: Specializing Code Llama for compiler optimization
-
LLama model inferce by TGI
-
-
-
-
-