NEXS Qwen3 32b Russian T Tech Lora

NEXS Qwen3 32b Russian T Tech Lora is a LoRA adapter designed for use with the Qwen3-32B base model. The adapter is created using mergekit, with tensors extracted from t-tech/T-pro-it-2.0, and is intended for integration with the Qwen/Qwen3-32B model. The model card indicates that the adapter is vLLM-ready and has undergone sanitization for vLLM serving. It is implemented as a Rank-128 LoRA adapter in bf16 format. The evidence also notes compatibility with PEFT, as usage instructions are provided for loading the adapter with the PEFT library in conjunction with the transformers library. The model is distributed via Hugging Face, where it can be accessed for use in libraries, inference providers, notebooks, and local applications. No information is provided regarding pricing, licensing, or specific intended users beyond its technical integration details. The evidence does not specify the particular tasks or domains for which the adapter is optimized, nor does it describe any additional features beyond its technical specifications and compatibility.