Views
No views yet
TQ3_4S is a 3.5-bit Walsh-Hadamard-transform weight format with four per-8 scales per 32-weight block.TQ3_4S GGUF quantization of croll83/Qwopus3.5-27B-v3-Abliterated, derived from the Qwen3.5-27B family.croll83/Qwopus3.5-27B-v3-AbliteratedQwen/Qwen3.5-27BQwopus3.5-27B-v3-Abliterated-f16.gguf1./build/bin/llama-quantize --pure \
2 /path/to/Qwopus3.5-27B-v3-Abliterated-f16.gguf \
3 /path/to/Qwopus3.5-27B-v3-Abliterated-TQ3_4S.gguf \
4 TQ3_4S \
5 16llama.cpp-tq3:turbo-tan/llama.cpp-tq3llama-server --reasoning offWrite ONLY the word ok.ok1./build/bin/llama-simple-chat \
2 -m /path/to/Qwopus3.5-27B-v3-Abliterated-TQ3_4S.gguf \
3 -ngl 99 -c 20481./build/bin/llama-server \
2 -m /path/to/Qwopus3.5-27B-v3-Abliterated-TQ3_4S.gguf \
3 --host 127.0.0.1 --port 8080 \
4 -ngl 99 -c 8192 -np 1 \
5 -ctk q8_0 -ctv q8_0 -fa on \
6 --no-warmup --jinjaQwopus3.5-27B-v3-Abliterated-mmproj.gguf.mmproj.gguf alongside this model if you want to preserve the same multimodal packaging.TQ3_4S runtime support from turbo-tan/llama.cpp-tq3.