ganarajpr
Joined 14 years ago
Joined 14 years ago
Planning to release this into the wild for people to use. I currently am struggling on how to fund this properly as the server costs are going to be huge for such an effort (search engine + LLM).
I have seen videos where a layman, who earns around 20-30k a month is slapped with a 32k electricity bill. I am not sure if a layman with means of this level is adopting Solar, but surely anyone else with the money to afford it will. Mostly because it makes economic sense.
Feels very similiar to DSPy except you dont have optimizations yet. But I like your API and the programming model your are enforcing through this.
We have rules in govts that companies should not discriminate against employees based on sex, religion, sexual orientation etc etc.. But it is fair to discriminate the salary of an employee based on location? For ex: I know a few friends who have moved from Europe to Asia with the same role and are getting paid less compared to what they were getting paid in Europe. Its the same role, its the same person, but getting paid less just because of location ?
When it produces a set of images for a given prompt, wouldnt it be better if we could remove a set of images from the possible selection ? Does it not work this way? Another idea would be to provide a few different kinds of prompts and based on that select all the images that matter for the given "class".
Some other things that would be good to know:
1. Can we keep adding items to the classifier? and getting newer versions of the classifier with the newly added item ? 2. How to deploy and host this kind of models? Is there any guidelines on how to deploy this in AWS or GCS for production use cases ?
That is why I created Dhee Desktop / Dhee Core ( Open Source https://github.com/dheeai/dhee-desktop https://github.com/dheeai/dhee-core ) - which allows anyone to author and run their own workflows for creating AI videos. Its local first, fully customizable, private runner for your workflows or any kind of community provided workflows(bundles) too (Comes with a few built in bundles to generate a full multi-scene segment just from story! ).
Its not just a workflow runner, but rather comes with an AI agent to manipulate the generations. AI images and video models often make mistakes ( extra limbs, wrong face etc ), You can ask the Agent to make changes to the prompts, trigger a rerun, Completely change the character and only trigger the exact shots in which he/she appears etc. It knows and tracks the dependencies.
My mission is to make it super easy for anyone and everyone to be able to create AI videos that they want to watch. I am envisioning a future where you make your own movies and watch it on your PC / TV rather than wait for someone else to make a movie that you want.
You can check out how it works on https://dhee.studio