newmindai/Mecellem-Qwen3-1.7B-TR
TEXT GENERATIONConcurrency Cost:1Model Size:2BQuant:BF16Ctx Length:32kPublished:Jan 16, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

Mecellem-Qwen3-1.7B-TR is a 1.7 billion parameter Turkish legal language model developed by newmindai. It is based on the Qwen3 decoder architecture and continually pre-trained on approximately 225 billion tokens of Turkish legal and official texts using a four-phase curriculum learning strategy. This model is specifically optimized for Turkish legal domain tasks, demonstrating a 36.2% perplexity reduction on legal text compared to its base model. Its primary strength lies in generating and understanding Turkish legal content.

Loading preview...