Kalkidan Model V1 is the original fine-tuned version of Qwen 1.5 using MLX + LoRA adapters.
It is designed for basic chat tasks and handling questions about data activism.
This model is ideal for:
Students learning about data activism
Simple chat-based demonstrations
Experiments with fine-tuned small models
Differences from V2 and V3
V1: Original fine-tuning with basic performance
V2: Improved training with more examples and better response accuracy
V3: Further fine-tuning with enhanced conversational style
