Update MODEL_CARD.md

MODEL_CARD.md

@@ -7,7 +7,7 @@
 **Model Version**: 1.0.0  
 **Release Date**: September 2025  
 **License**: Apache 2.0  
-**Base Model**: mistralai/Mistral-7B-Instruct-v0.2  
+**Base Model**: mistralai/Mistral-7B-Instruct-v0.3  
 **Model Type**: Causal Language Model with LoRA Adapters  
 
 ## Intended Use
@@ -61,7 +61,7 @@
 
 ### Training Configuration
 - **Method**: Supervised Fine-tuning with LoRA
-- **Base Model**: mistralai/Mistral-7B-Instruct-v0.2
+- **Base Model**: mistralai/Mistral-7B-Instruct-v0.3
 - **Training Framework**: Transformers + PEFT
 - **Hardware**: NVIDIA RTX 3060 (12GB)
 - **Training Duration**: Multiple epochs until convergence
@@ -241,10 +241,9 @@ The VRRE evaluation framework used to assess Apollo V1 7B demonstrates:
 
 ## Contact and Support
 
-- **Primary Contact**: research@vanta.ai
+- **Primary Contact**: tyler [at] alignmentstack [dot] xyz
 - **GitHub Issues**: [vanta-research/apollo-v1-7b](https://github.com/vanta-research/apollo-v1-7b/issues)
-- **Documentation**: [vanta.ai/models/apollo-v1-7b](https://vanta.ai/models/apollo-v1-7b)
-- **Community**: [Discord Server](https://discord.gg/vanta-research)
+- **Community**: [Find VANTA Research on X!](https://x.com/vanta_research)
 
 ## Acknowledgments