Update README.md

README.md

@@ -5,32 +5,18 @@ pipeline_tag: image-text-to-text
 base_model: google/gemma-3-27b-it
 ---
 
-# 💎 Gemma 3 27B IT Abliterated
+💎 Gemma 3 27B IT Abliterated
 
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/WjFfc8hhj20r5XK07Yny9.png)
-<center><a href="https://huggingface.co/mlabonne/gemma-3-4b-it-abliterated">Gemma 3 4B Abliterated</a> • <a href="https://huggingface.co/mlabonne/gemma-3-12b-it-abliterated">Gemma 3 12B Abliterated</a></center>
 
-This is an uncensored version of [google/gemma-3-27b-it](https://huggingface.co/google/gemma-3-27b-it) created with a new abliteration technique.
-See [this article](https://huggingface.co/blog/mlabonne/abliteration) to know more about abliteration.
+This is an experimental, uncensored version of google/gemma-3-27b-it created using a new “abliteration” technique. It removes refusals while keeping most of the model’s capabilities intact.
 
-I was playing with model weights and noticed that Gemma 3 was much more resilient to abliteration than other models like Qwen 2.5. 
-I experimented with a few recipes to remove refusals while preserving most of the model capabilities.
+Key Features:
+	•	Tool Calling Enabled: This version includes tool support, making it more flexible for various tasks.
+	•	Minimal Fine-tuning: The model has been minimally fine-tuned, focusing on improving output coherence and handling requests effectively.
 
-Note that this is fairly experimental, so it might not turn out as well as expected.
+Recommended Generation Parameters:
+	•	temperature=1.0
+	•	top_k=64
+	•	top_p=0.95
 
-I recommend using these generation parameters: `temperature=1.0`, `top_k=64`, `top_p=0.95`.
-
-## ✂️ Layerwise abliteration
-
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/HnBRigUfoQaCnpz96jnun.png)
-
-In the original technique, a refusal direction is computed by comparing the residual streams between target (harmful) and baseline (harmless) samples.
-
-Here, the model was abliterated by computing a refusal direction based on hidden states (inspired by [Sumandora's repo](https://github.com/Sumandora/remove-refusals-with-transformers/)) for each layer, independently.
-This is combined with a refusal weight of 1.5 to upscale the importance of this refusal direction in each layer.
-
-This created a very high acceptance rate (>90%) and still produced coherent outputs.
-
-## ⚡️ Quantization
-
-TBD.
+Abliteration - The model is abliterated by computing refusal directions based on hidden states for each layer independently.