Hi Ronan, Sorry couldn't join the live stream but would love to see a video on image/3d model diffusion for e:g train the model on images and when we ask a query get the image or 3d model (sldprt) model.
@TrelisResearch
3 ай бұрын
all grantd, thanks for the tip, yeah lemme think a bit more and dig around on 3d
@reagansenoron6763
3 ай бұрын
Hi Ronan, thanks for your videos. It is really helpful. We have a task to create an app to streamline the process of creating CENELEC documentation. CENELEC is responsible for European standardization in the area of electrical engineering. The idea is to generate or fill in data, for example, the concept section of the EN 50126 standards. The goal: To ask the AI to fill the concept section(just one of the many sections in EN 50126 standard) based on some proprietary data. So if there is a new project and we already collected enough knowledgebase, then we can ask AI to fill the documentation. I know a bit about fine tuning but what I am not sure at this point is how I format the dataset to be used for training. Because it is not in a question/answer format but rather filling a section with relevant data. Thanks for answering.
@TrelisResearch
3 ай бұрын
Take a look through the Trelis Research fine-tuning playlist, you'll see a video on data preparation. Also check the supervised fine-tuning one. The memorization one is also relevant.
@gileneusz
3 ай бұрын
where will you travel to?
@TrelisResearch
3 ай бұрын
will be between Dublin, New York, San Fran, Boston, Estonia, Latvia and Finland!
@gileneusz
3 ай бұрын
@@TrelisResearch that's a lot of travel. We can meet in person in Latvia if you have time 😁
Пікірлер: 7