20+ models
pulled as BF16, vocab slice (combined_en_huge) requant to IQ4_NL (combined_en_huge) -- (English + Coding focus) - ~20% faster than Q4_K_M
Revolutionary model with unique thinking/non-thinking modes, delivering superior reasoning performance with seamless mode switching for any task.
This model is Qwen3-32b-Instruct fine-tuned on ~13k Climsight data which are pairs of prompts/completions extracted from climsight pipeline. Below are all the technical specs of the fine-tuning process along with the evaluation used.
Zhi-Create-Qwen3-32B is a fine-tuned model derived from Qwen/Qwen3-32B, with a focus on enhancing creative writing capabilities. Through careful optimization, the model shows promising improvements in creative writing performance, as evaluated using the W
包含2个量化版本GGUF：Qwen3-32B-Q8_0，Qwen3-32B-Q5_K_M
这是unsloth的Q4动态量化版本，精度第一的量化版本！Unsloth Dynamic 2.0 实现了卓越的准确性，并超越了其他领先的量化模型。
这是unsloth的Q8动态量化版本，精度第一的量化版本！Unsloth Dynamic 2.0 实现了卓越的准确性，并超越了其他领先的量化模型。
GGUF Ruadapt версии модели Qwen/Qwen3-32B (квантизованная версия Q4_K_M)
Model with tweaked params optimized for agent use
Qwen3.8-27B in Q4_K_M quantization (32GB+ VRAM required). Dense 27.8B parameters with hybrid attention for long context (256K tokens). Apache 2.0 license. Ideal for high-VRAM setups (RTX 5090/4090 dual, M4 Ultra, etc.).
32k token limit, for running on mac book pro with 16GB (~ 10 GB available for AI without anything else running)
Recommended for transcribing and summarizing text from screenshots.
Qwen2.5 coder tools model can work with Cline (prev. Claude Dev). Update 0.5b, 1.5b, 3b, 7b, 14b, 32b coder models.