Update main README with VRAM Matrix

README.md

@@ -22,6 +22,7 @@ The model weights are stored in separate branches. **Please switch to a branch t
 | 4.0 | h6 | [4.0bpw_h6](https://huggingface.co/DBMe/gemma-3-12b-it-ultra-uncensored-heretic-exl3/tree/4.0bpw_h6) | 9.7746 | ~9.93 GB | ~10.68 GB | ~12.18 GB | ~15.18 GB | ~21.18 GB |
 | 5.0 | h6 | [5.0bpw_h6](https://huggingface.co/DBMe/gemma-3-12b-it-ultra-uncensored-heretic-exl3/tree/5.0bpw_h6) | 9.7767 | ~11.19 GB | ~11.94 GB | ~13.44 GB | ~16.44 GB | ~22.44 GB |
 | 6.0 | h6 | [6.0bpw_h6](https://huggingface.co/DBMe/gemma-3-12b-it-ultra-uncensored-heretic-exl3/tree/6.0bpw_h6) | 9.7294 | ~12.44 GB | ~13.19 GB | ~14.69 GB | ~17.69 GB | ~23.69 GB |
+| 8.0 | h8 | [8.0bpw_h8](https://huggingface.co/DBMe/gemma-3-12b-it-ultra-uncensored-heretic-exl3/tree/8.0bpw_h8) | 9.7336 | ~15.18 GB | ~15.93 GB | ~17.43 GB | ~20.43 GB | ~26.43 GB |
 
 ¹ *Evaluated against WikiText-2 with ExLlamaV3 using a strided 512-token context window (-c 512) in llama.cpp parity mode (-g). Lower is better.*
 *(Higher BPW = higher quality, lower BPW = fits in less VRAM).*
@@ -48,7 +49,7 @@ These models are highly optimized for modern GPUs and can be run using:
 
 ### 📈 Perplexity Degradation Curve
 *(Lower is better)*
-![Perplexity Graph](https://huggingface.co/DBMe/gemma-3-12b-it-ultra-uncensored-heretic-exl3/resolve/main/metrics_graph.png?v=1776285160)
+![Perplexity Graph](https://huggingface.co/DBMe/gemma-3-12b-it-ultra-uncensored-heretic-exl3/resolve/main/metrics_graph.png?v=1776286443)
 
 <details>
 <summary><b>⚙️ Advanced: Quantization Environment & Settings</b></summary>