BROADCASTS.com

In this eisode 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

\n

• Faissal Boutaounte

  \n
\n

• Nouamane Tazi

  \n
\n

Notes

0:00:00\xa0- Introduction and welcoming

0:01:21\xa0- Who/what is l3arbi?

0:16:24\xa0- Is there possibilties that L3arbi will learn other arabic dialects other than darija?

0:26:35\xa0- Challenges of audio transcription

0:27:31\xa0- Live demo of L3arbi

0:32:56\xa0- Data training and Whisper and how many hours of darija data is used ?

0:34:30\xa0- Format of data that is used for training, and arcitecture of web application presented in DevoxxMA

0:39:37\xa0- The use case of DevoxxMA

0:44:30\xa0- Finetuning of an LLM, models and details

0:56:20\xa0- Evaluation set for different dialects per regions in morocco

1:03:31\xa0- Did you use manual transcription for the audio sets ?

1:12:09\xa0- The future plans of data sources for L3arbi solution.

1:15:12\xa0- Plans to open-source? Are there APIs available for developers to extend its functionality?

1:25:00\xa0- QA & Giveaway

1:59:40\xa0- Conclusion and goodbye.

Links

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

Prepared and Presented by

\n
• Meriem Zaid
\n