L3arbi ChatBot: A Technical deep dive

In this episode we discuss more of L3arbi, that was resented by the DevoxxMA team at the vent, from a technical point of view with Nouamane and Faissal.

Guests

• Faissal Boutaounte

• Nouamane Tazi

Notes

0:00:00 - Introduction and welcoming

0:01:21 - Who/what is l3arbi?

0:16:24 - Is there possibilities that L3arbi will learn other arabic dialects other than darija?

0:26:35 - Challenges of audio transcription

0:27:31 - Live demo of L3arbi

0:32:56 - Data training and Whisper and how many hours of darija data is used ?

0:34:30 - Format of data that is used for training, and architecture of web application presented in DevoxxMA

0:39:37 - The use case of DevoxxMA

0:44:30 - Fine-tuning of an LLM, models and details

0:56:20 - Evaluation set for different dialects per regions in morocco

1:03:31 - Did you use manual transcription for the audio sets ?

1:12:09 - The future plans of data sources for L3arbi solution.

1:15:12 - Plans to open-source? Are there APIs available for developers to extend its functionality?

1:25:00 - QA & Giveaway

1:59:40 - Conclusion and goodbye.

Links

• Huggingface
• Translation demo for/from any language
• Common Voice
• Whisper API

Prepared and Presented by

• Meriem Zaid