storytime-with-grandma

Ever heard the proverb "Every picture tells it's story" ? Well, now you've heard and there is something else that you can hear; the story of the image. For the technical aspect, this project takes an image as input an uses the image captioning model 'Blip' to caption it. This caption is passed to a LLM named 'Mixtral-8x7B' which generates a 30 word story. This story is then translated to the desired language and this story is converted to an audio file which can be played by the user. Click the link below to take a sneak peek.

grandma.streamlit.app

Also, you can run this locally by following this steps:

1. Clone this repository

2. Install the dependencies

3. Enter your credentials

Visit huggingface.co and log in to your account or create a new one.

Then, create a new api token by visiting huggingface.co/settings/tokens

Copy this api token and make sure to keep it private

Now, open .env file and replace 'YOUR_API_TOKEN_HERE' with your huggingface api token.

4. Now that everything is ready, let's run the app

a. If you have a dedicated GPU and a fast internet connection, use this as it utilizes a lot of processing power and bandwidth because it downloads all the models and runs them.

b. If you don't have a dedicated GPU, use this