Compressed Qwen3 (8B) models. Compression methods: SparseGPT, Wanda
ar-llm-browser
community
AI & ML interests
Model compression and optimization methods, such as quantization, pruning, distillation, and fine-tuning.
Models (Gemma3 1b, Qwen3 1.7b) pruned using SparseGPT and Wanda.
-
ar-llm-browser/Qwen3-1.7B-SparseGPT-50
Text Generation • 2B • Updated • 11 -
ar-llm-browser/gemma-3-1b-it-SparseGPT-50
Text Generation • 1B • Updated • 7 -
ar-llm-browser/Qwen3-1.7B-SparseGPT-50-2of4
Text Generation • 2B • Updated • 12 -
ar-llm-browser/gemma-3-1b-it-SparseGPT-50-2of4
Text Generation • 1B • Updated • 6
Compressed Qwen3 (8B) models. Compression methods: SparseGPT, Wanda
Collection of compressed ALLaM (by humain-ai) models. Methods include multiple quantization formats and pruning with SparseGPT and Wanda.
Models (Gemma3 1b, Qwen3 1.7b) pruned using SparseGPT and Wanda.
-
ar-llm-browser/Qwen3-1.7B-SparseGPT-50
Text Generation • 2B • Updated • 11 -
ar-llm-browser/gemma-3-1b-it-SparseGPT-50
Text Generation • 1B • Updated • 7 -
ar-llm-browser/Qwen3-1.7B-SparseGPT-50-2of4
Text Generation • 2B • Updated • 12 -
ar-llm-browser/gemma-3-1b-it-SparseGPT-50-2of4
Text Generation • 1B • Updated • 6