Llama Impact Hackathon

Start

FT LLama to improve ASR. Then apply it to low-resource language

AILabs Founder Solo

AILabs

[​Task 3. Expanding Low-Resource Languages]
Today, even the best production-level real-time speech Recognition systems use old n-gram language models on top of the Speech Recognition models (CTC ASR). 
CTC means that the ASR model outputs character probabilities (logits), not real words. Once logits are received, an n-gram language model is applied on top of it to maximize the accuracy.
The problem with n-gram model is that it's able to take into account only n words (usually n=3) in the sequence, which is a huge limitation. 
Normally, In order to properly understand the speech, we need: 1) the whole context of the conversation, 2) prior context (something we talked about before)

In this project, I address the first problem by training LLM like Llama3.2 on top of the ASR generated logits.

(ONSITE) Today, even the best real-time speech recognition systems use old n-gram language models on top of the ASR models. 
In this project, I address the n-gram LM limitation by replacing it with a Large language model like LLama3.2.

"AILabs Founder Solo team on Llama Impact Hackathon Hackathon"

Team Idea

Submission