Update README.md

README.md

@@ -18,6 +18,15 @@ At the very core of the development of this model is the desire to make sure tha
 Test this out and see if you find anything interesting or intriguing. I will keep iterating more versions but this one seems like a fun and useful way to start.
 
 
+# Phased Training Methodology
+Leveraging this dataset, we've adopted a phased training methodology that focuses sequentially on different dataset components, namely "System" and "Instruction," across separate training phases. This approach allows models to build layered understandings from general systemic insights to specific instructional cues, enriching their generative output with both broad contextual awareness and detailed, topic-specific knowledge.
+
+**Phase 1: System Focus**
+In the initial phase, the model concentrates on the "System" component, absorbing overarching guidelines and objectives. This phase lays the foundational understanding, enabling the model to grasp the contextual framework and systemic knowledge encapsulated in the dataset.
+
+**Phase 2: Instruction Focus**
+Building upon the systemic knowledge, the second phase shifts the model's focus to the "Instructions" component. This phase sharpens the model's ability to interpret specific prompts and generate responses that are not only informed by the broader context but also precisely tailored to the instructional cues.
+
 ## GGUF Q8 Version: https://huggingface.co/Severian/Nexus-IKM-Mistral-7B-GGUF