GitHub: https://github.com/DeepFriedYeti
e-mail: ayushrajbvb@gmail.com
| Dates | Progress |
|---|---|
| 05/05/2023 - 08/05/2023 | → Used Lexica to generate images using prompts. → Read about stable diffusion from the provided github link (https://jalammar.github.io/illustrated-stable-diffusion/) |
| 09/05/2023 | → Tried to understand the contents of the presentation, but understood the Maths partially only, rest was fine. |
| 10/05/2023-11/05/2023 | Wasted a lot of time in configuring VS code to use gpu for training, as I faced timeout issues on Collab (due to long training time probably) while using it for my DS project 🙃. But it works now :) |
| 13/05/2023-14/05/2023 | → Completed the assignment 1 along with the reading material mentioned on the notion page for the assignment. |
| 18/05/2023 | → Read the given material (https://towardsdatascience.com/stable-diffusion-using-hugging-face-501d8dbdd8, https://towardsdatascience.com/stable-diffusion-using-hugging-face-variations-of-stable-diffusion-56fd2ab7a265) and also gained knowledge on how stable diffusion is used to generate videos by referring to the code of Pix2Pix and ControlNet |
| 24/05/2023-24/06/2023 | Worked on using Cross Attention control and other similar models that work on attention in prompts to edit the contents of image |
| 24/06/2023-15/07/2023 | Worked on Modifying the prompts using LLM model and then generate prompts for stable diffusion models |
| 15/07/2023-Present | Working on adjusting the attention for the prompts using LLM, to achieve better results and continuity, and also making use of LORAs |