GGUF quantization of ressl/Laguna-S-2.1-Uncensored, derived from poolside/Laguna-S-2.1.

Original uncensoring work by Robert Ressl — full model card and evaluation details are at the source repository.

All variants use the oscar llama.cpp build (v4310aa4f8) with native LLM_ARCH_LAGUNA support. The BF16 baseline was converted with convert_hf_to_gguf.py from the oscar repo, which handles Laguna's per-layer alternating head counts, sigmoid MoE routing, mixed full/sliding attention RoPE, and the shared-expert topology.

All quantizations use --leave-output-tensor to preserve output.weight in BF16 for generation quality.

Mixed-precision recipes use per-tensor --tensor-type overrides for 814 tensors via --tensor-type-file. The default fallback type for unlisted tensors is the top-level quant type passed to llama-quantize.

For GPU offloading:

Full evaluation details, datasets, and methodology are in the source repository.

Original uncensoring: Robert Ressl (ressl.ch)

GGUF conversion and quantization: converted with oscar convert_hf_to_gguf.py (commit 4310aa4f8) and llama-quantize from the same build, using the registered LagunaForCausalLM conversion plugin.