Model Overview

The yibba/GemMaroc-Qwen2.5-7B-Instruct-Therapy-Darija is a 7.6 billion parameter instruction-tuned language model developed by yibba. It is fine-tuned from the GemMaroc/Qwen2.5-7B-Instruct-darija base model, building upon the Qwen2 architecture. A key aspect of its development is the utilization of Unsloth and Huggingface's TRL library, which enabled a 2x faster training process.

Key Capabilities

• Darija Language Proficiency: Specialized in understanding and generating text in Darija, a Moroccan Arabic dialect.
• Therapeutic Instruction Following: Fine-tuned for conversational interactions, likely focusing on therapeutic or supportive dialogue.
• Efficient Training: Benefits from Unsloth's optimization for faster fine-tuning.
• Extended Context Window: Features a 32768 token context length, allowing for more comprehensive and longer-form conversations.

Use Cases

This model is particularly well-suited for applications requiring:

• Therapeutic Chatbots: Engaging in supportive and therapeutic conversations in Darija.
• Darija-specific AI Assistants: Developing instruction-following agents for Moroccan users.
• Content Generation in Darija: Creating nuanced and contextually relevant text in the Darija dialect.