Initial Readme

README.md

@@ -0,0 +1,58 @@
+---
+license: apache-2.0
+datasets:
+- ccibeekeoc42/TinyStories_yoruba
+- ccibeekeoc42/TinyStories_igbo
+- ccibeekeoc42/DollyHHRLHF_igbo
+- ccibeekeoc42/DollyHHRLHF_yoruba
+language:
+- ig
+- yo
+- en
+metrics:
+- accuracy
+library_name: adapter-transformers
+pipeline_tag: text-generation
+---
+
+## Model Name: Multilingual LLM for English, Igbo, and Yoruba Translation
+
+This model specializes in understanding and generating text across English, Igbo, and Yoruba, trained on curated datasets including "tiny_stories" and "dolly_hhrlhf." Utilizing advanced techniques to ensure efficiency and accuracy, the model supports a wide range of applications from translation to content creation.
+
+## Key Features:
+- **Trilingual Support:** Seamlessly processes and generates content in English, Igbo, and Yoruba, promoting linguistic diversity and accessibility.
+- **Custom Training Approach:** Employs a tailored training setup, leveraging specific prompts and responses to enhance model performance on relevant tasks.
+- **High Sequence Capability:** Handles extensive text inputs (up to 3453 tokens), making it suitable for detailed narratives and complex translation tasks.
+- **Efficiency Optimizations:** Incorporates strategies such as dataset packing, reducing computational demands while maintaining high-quality output.
+
+## Applications:
+- **Translation Services:** Offers precise, context-aware translations, bridging communication gaps between English, Igbo, and Yoruba speakers.
+- **Content Generation:** Generates culturally and linguistically nuanced content, catering to a diverse audience.
+- **Educational Tools:** Assists in language learning and preservation, providing resources in underrepresented languages.
+
+## Future Directions:
+- Further refinement with diverse text sources to enhance understanding and generation capabilities.
+- Expansion to additional languages, supporting broader multilingual communication.
+
+This model represents a step towards more inclusive language technologies, recognizing the importance of language diversity in global communication.
+
+## Training Loss:
+- Below is the training loss.
+  
+| Step | Training Loss |
+|------|---------------|
+| 200  | 1.490900      |
+| 400  | 1.375600      |
+| 600  | 1.304100      |
+| 800  | 1.198700      |
+| 1000 | 1.228200      |
+| 1125 | 1.226600      |
+
+
+## About the Creators
+[Christopher Ibe](https://www.linkedin.com/in/christopher-ibe-ekeocha/) and [Okezie Okoye](https://www.linkedin.com/in/okezie-okoye-43432b62/) continue to lead Hypa AI towards new frontiers in AI translation. Their dedication to leveraging advanced AI for genuine understanding and connection across language barriers is what sets Hypa AI apart in the field of artificial intelligence.
+
+*Hypa AI* remains steadfast in its mission to pioneer intelligent solutions that are not just technologically advanced but are also culturally aware, ensuring that the future of AI is as diverse and inclusive as the world it serves.
+
+*AfroVoices*, a subsidiary of Hypa AI, is dedicated to amplifying African voices, languages, and cultures in the intelligence age. Focused on bridging the digital representation gap, AfroVoices curates datasets and resources for African languages, promoting inclusivity and cultural appreciation in AI technologies. Their mission goes beyond technological innovation, aiming to celebrate the richness of African linguistic diversity on a global stage.
+