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 LLMARCHLAGUNA support. The BF16 baseline was converted with converthftogguf.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. Full evaluation details, datasets, and methodology are in the source repository. GGUF conversion and quantization: converted with oscar converthftogguf.py (commit 4310aa4f8) and llama-quantize from the same build, using the registered LagunaForCausalLM` conversion plugin.
Модальности:
Генерация текста
Области применения:
Генерация кода Диалог / чат
Задача: Генерация текста
Автор: jabbatheduck
Теги: llama.cpp, gguf, laguna, moe, uncensored, quantized, bilingual, code
Лайков: 3 | Загрузок: 519
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
