newmindai/Mursit-Base-TR-Retrieval
Sentence Similarity
•
0.2B
•
Updated
•
135
Mecellem Models
Project: Turkish Embeddings from Scratch and CPT Decoders Infrastructure: MareNostrum 5 (BSC)
-
-
newmindai/Mursit-Large-TR-Retrieval
Sentence Similarity • 0.4B • Updated • 70 -
newmindai/Mursit-Base
Fill-Mask • Updated • 82 -
newmindai/Mursit-Large
Fill-Mask • Updated • 57 -
newmindai/Mursit-Embed-Qwen3-4B-TR
Sentence Similarity • 4B • Updated • 62 -
newmindai/Mursit-Embed-Qwen3-1.7B-TR
Sentence Similarity • 2B • Updated • 91 -
newmindai/Mecellem-Qwen3-4B-TR
Text Generation • 4B • Updated • 39 -
newmindai/Mecellem-Qwen3-1.7B-TR
Text Generation • 2B • Updated • 38 -
newmindai/Muhakim
Text Classification • Updated • 29 -
Mecellem Models: Turkish Models Trained from Scratch and Continually Pre-trained for the Legal Domain
Paper • 2601.16018 • Published