29 repos
Conversational large language models from the Qwen family, spanning multiple scales from 0.5B to 14B parameters optimized for instruction-following and text generation tasks. The cluster centers on the Qwen1.5 and Qwen2.5 series with their chat/instruct variants, which use the safetensors format for efficient model distribution and are compatible with standard text-generation endpoints. Repositories here provide model weights, inference implementations, and fine-tuning examples for deploying and adapting these models across different hardware configurations.