hii guysp? , first of all let congratulates all perticipants of this challenge also thanks to all for being supportive. My question is how the final renk esp... See more

Hi Everyone, It's my first game here, and the whole process is fun. The game is around the end now, and I'd like to have some open discussions even after the ga... See more

Hi all — I’m sharing a Jupyter notebook for Swahili Text-to-Speech (TTS) that I built. The notebook walks through data loading and preprocessing, model training... See more

it's done. i cannot delete the post.

In today’s stream, we built the training notebook for our models. Since training takes quite a while, we didn’t get to see the results yet, but tomorrow’s strea... See more

Over the weekend, I discovered a Whisper baseline that performed surprisingly well — so in today’s stream we focused on testing multiple Whisper models.The big... See more

I am not able to access the https://huggingface.co/datasets/mozilla-foundation/common_voice_17_0 dataset on my colab, even I am using my token of huggin face, c... See more

We just evaluated our trained Wav2Vec models (1 epoch), and they gave a small but noticeable improvement over our baseline. Even more importantly — we now have... See more

We finally wrote the code to train our ASR model! 🥳 The training takes a while, so results will be shared in tomorrow’s stream. Still, this is a huge step as w... See more

Today’s focus was data preparation for Wav2Vec finetuning — making sure our audio and transcripts are clean, properly formatted, and ready for the model.We went... See more

Hello fellow ZindiansI’m seeking clarification about the following requirement *"To reflect the realities of deploying AI in low-resource environments, your sol... See more

Today’s stream was a reality check. We focused on evaluating our finetuned Kiswahili ASR models to see if they actually improved performance — and unfortunately... See more

Today’s stream was a big one we finally got our Kiswahili ASR model training with NeMo! 🎉What we covered: Loaded and customized the training configs Launched o... See more

Today’s stream was shorter than usual but no less important. We spent the session working through data preparation for CTC Finetuning, which ended up being more... See more

A starter guide for audio transcription using the SeamlessM4T model, tailored to the Sartify ITU Zindi Test Dataset. It demonstrates how to load, play, and proc... See more

I'm currently unable to load the Common Voice 17.0 dataset (Swahili split) from Hugging Face using the datasets library. Here's the code I'm using:from datasets... See more

Day 2 of our Kiswahili ASR challenge stream focused on Conformer + CTC-based models and wow, the progress we made was huge!✅ What we did: Explained CTC models a... See more

In this stream, we kicked off a brand new challenge: building a Kiswahili speech-to-text (ASR) system for real-world, low-resource environments.🧠 Here's what w... See more

I just tried out a Converse TTS train to get some Swahili text read out.link to audio--->>https://drive.google.com/file/d/1PoYq1fvSgl0NS1cmH80wu6rDiIbw8Ps... See more