LFM2.5-24B-A2B-GPTQ
4-bit GPTQ build of LiquidAI/LFM2.5-24B-A2B, built to serve an office of agents from one RTX 4000 Ada 20...
4-bit GPTQ build of LiquidAI/LFM2.5-24B-A2B, built to serve an office of agents from one RTX 4000 Ada 20...
Standalone BF16 behaviorally modified derivative of Liquid AI’s LFM2.5-24B-A2B for reasoning, multilingual chat, and tool-enabled workflows. > Release...
This model LiquidAI/LFM2.5-24B-A2B-MLX-6bit was converted to MLX format from LiquidAI/LFM2.5-24B-A2B using mlx-lm version 0.32.0. Quantized to 6 bits...
A decensored variant of LiquidAI/LFM2.5-8B-A1B, produced with Heretic v1.4.0 (directional ablation / «abliteration»). Liquid’s hybrid conv+attention MoE keeps...
This is an LFM2.5-8B-A1B fine-tune, produced through P-E-W’s Heretic (v1.4.0) abliteration engine with Self-Organizing Maps & Magnitude-Preserving Orthogonal...
Quantized to Q5KM with importance matrices for higher accuracy. This model is a fine-tuned version of LiquidAI/LFM2.5-8B-A1B, adapted...
Tini-Cybersec-8B-A1B is a specialized fine-tuned model based on the LiquidAI/LFM2.5-8B-A1B architecture. It is customized to perform complex Cybersecurity...
Website cyan.kiwi Models Hugging Face · ModelScope Connect X · Reddit · Email Version cyankiwi 26.05.01 Calibration STEM...
I suddenly realized that the LFM2.5-Queen-8B-A1B is particularly suitable for role-playing Queen. Why be a coder when you...
LFM 8B-1A8 (MOE, 32 эксперта), обученный пользовательскому набору ДАННЫХ с использованием Qwen 3.6 35B-A3B (MOE, 256 экспертов, в...