I’ve been in text generation for a while, but I want to mess about with video. I have some old 4:3 videos that I want to upscale. I suspect Topaz would be a good way to go, but I also want to infill to stretch them to 16:9. 1. How would I best go about fine-tuning a video model with the current corpus of 4:3 video to infill the space? 2. What base model should I use? 3. Is there a model I should be using to do my upscaling? 4. Is there a tool that can do all of the above and I just missed it?
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Does a model which will level, center, and stabilize an existing video exist? | 0 | 271 | February 5, 2024 | |
Tokenizer effect on the fine-tuning | 0 | 360 | October 6, 2023 | |
Finetuning model with smaller sequence size and Dmodel | 0 | 334 | April 15, 2021 | |
Context window decreased after finetuning? | 0 | 186 | November 4, 2023 | |
Best Affordable Way to Generate Videos from Text? | 2 | 37 | February 19, 2025 |