전체 검색
AI 모델 84768
- Qwen3.6-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking-i1-GGUF↗
mradermacher
- Rocinante-X-12B-v1-i1-GGUF↗
mradermacher
- SambaLingo-Hungarian-Chat-70B-i1-GGUF↗
mradermacher
- SAVER-0.8B-GGUF↗
mradermacher
- True-Qwen2.5-14B-Instruct-GGUF↗
mradermacher
- Video-Thinker-7B-GGUF↗
mradermacher
- XiYanSQL-QwenCoder-3B-2502-GGUF↗
mradermacher
- Yeet_51b_200k-i1-GGUF↗
mradermacher
- Qwen3-Coder-30B-APEX-GGUF↗
mudler
⚡ Each donation = another big MoE quantized I host 25+ free APEX MoE quantizations as independent research. My only local hardware is an NVIDIA DGX Spark (122 GB unified memory), enough for ~30-50B-class MoEs, but bigger ones (200B+) require rented compute on H100/H200/Blackwell, typically $20-100 per quant. If APEX quants are useful to you, your support directly funds those bigger runs. All measurements on NVIDIA DGX Spark (GB10, 128 GB VRAM). Perplexity on wikitext-2-raw, context 2048. Accuracy benchmarks via llama.cpp (400 tasks). APEX is a quantization strategy for Mixture-of-Experts (MoE) models. It classifies tensors by role (routed expert, shared expert, attention) and applies a layer-wise precision gradient -- edge layers get higher precision, middle layers get more aggressive compression. I-variants use diverse imatrix calibration (chat, code, reasoning, tool-calling -- no Wikipedia).
- C-RADIOv2-L↗
NVIDIA
