Qwen3-0.6B-int4
On-device INT4 builds of Qwen/Qwen3-0.6B in the .litertlm format for Google’s LiteRT-LM runtime (Android / iOS / desktop...
On-device INT4 builds of Qwen/Qwen3-0.6B in the .litertlm format for Google’s LiteRT-LM runtime (Android / iOS / desktop...
> Quantized release > This repository is an AutoRound W4A16 quantized version of deepreinforce-ai/Ornith-1.0-397B. > It is not...
AWQ 4-bit (W4A16) quantization of empero-ai/Qwythos-9B-Claude-Mythos-5-1M — a Claude-Mythos/Fable-trace reasoning fine-tune of Qwen3.5-9B (qwen35: a dense hybrid GatedDeltaNet...
INT4 weight-only (W4A16) quantization of Qwen/Qwen3.6-35B-A3B — a hybrid Mixture-of-Experts model (256 experts, top-8, ~3B active) with GatedDeltaNet...
Qwen 3 Reranker 4B в INT4. Сопряжение с набором Embedding для полностью локального стека поиска. ~ 2,7 ГБ...
———— ——————————————————————————————- | Следующие слои не квантуются для сохранения качества модели: > Зачем декантировать mtp.fc? Загрузчик Qwen35MTP от...
———— | ———————————————————————————————- | The following layers are not quantized to preserve model quality: Tested on a single...
> [!IMPORTANT] > Naming notice (2026-04-10). The «HLWQ» technique used in this model is being rebranded to HLWQ...
— Model Architecture: Qwen3ForCausalLM — Input: Text — Output: Text — Model Optimizations: — Weight quantization: INT4 —...
This repository contains the official 4-bit quantized version of the Kimi-Linear-48B model using the SINQ (Sinkhorn-Normalized Quantization) method...