flint-int4
An EigenLabs mirror of Recursive’s official MLX INT4 release of Flint, a mixture-of-experts language model for coding agents...
An EigenLabs mirror of Recursive’s official MLX INT4 release of Flint, a mixture-of-experts language model for coding agents...
4-bit GPTQ build of LiquidAI/LFM2.5-24B-A2B, built to serve an office of agents from one RTX 4000 Ada 20...
This is UkisAI’s Swift 1.5 Qwen3.8-Flash-Next AutoRound W4A16 checkpoint, repackaged to match the layout of albucino/Qwen3.8-Flash-Next-W4A16-FP8PLE. That layout...
Unofficial 4-bit GPTQ quantization of Altworld/Hemmingway-1 for low-VRAM inference, with the MTP (multi-token-prediction) draft head deliberately kept unquantized....
INVALID LANGUAGE PAIR SPECIFIED. EXAMPLE: LANGPAIR=EN|IT USING 2 LETTER ISO OR RFC3066 LIKE ZH-CN. ALMOST ALL LANGUAGES SUPPORTED...
A GPTQ 4-bit (group size 128, symmetric) quantization of IFM/K2-Horizon-MoVA-36B-A4B, MBZUAI IFM’s sparse frontier model with MoVA (Mixture-of-Values)...
Квантование INT4 (W4A16) Qwen/Qwen3.8-27B построено специально для того, чтобы модель работала на графических процессорах Pre-Hopper, включая Volta (V100,...
> Work in progress. This card and the checkpoint are still being validated. Routed experts of DeepSeek-V4-Flash-0731 transcoded...
AWQ (W4A16) quantization of empero-ai/Qwythos-9B-v2 — a 9.65B reasoning / uncensored model from Empero AI (qwen35`: dense hybrid...
We reduce Qwen3.5-4B inference latency on a single A10G by quantizing both the target model and its speculative-decoding...