nguyenthilaitrieulong/

Llama-4-Maverick-17B-128E-Instruct-Projected-Abliterated-GGUF

Experimental abliterated model using improved https://huggingface.co/blog/grimjim/projected-abliteration technique. Abliteration tries to remove refusals from model's behaviour without fine-tuning.

For non-abliterated GGUF quantized version I recommend https://huggingface.co/unsloth/Llama-4-Maverick-17B-128E-Instruct-GGUF quants.

Warning: Safety guardrails and refusal mechanisms have been broken through abliteration. This model may generate harmful content and shall not be used in production, user-facing applications, etc. You will be solely responsible for its outputs.

Warning 2: Neither removal of refusals nor preservation of original model's capabilies is guaranteed.