4 ms·
I'd be interested in reading more of your findings! Are you able to accomplish this with prompt-engineering, or are you doing fine-tuning of LLMs / custom-trai
by HanClinto 2y ago
I'd be interested in reading more of your findings!
Are you able to accomplish this with prompt-engineering, or are you doing fine-tuning of LLMs / custom-trained models?
- seveibar 2y agoNo fine tuning needed, as long as the target language/DSL is fairly natural, just give eg a couple examples of tscircuit React, atopile JotX etc and it can generate compliant circuits. It can hallucinate imports, but if you give it an import list you can improve that a lot.
- DHaldane 2y agoI've found the same thing - a little syntax example, some counter examples and generative AI does well generating syntactically correct code for PCB design. A lot of the netlists are electrically nonsense when it's doing synthesis for me. Have you found otherwise?
- seveibar 2y agoNetlists, footprint diagrams, constraint diagrams etc. are mostly nonsense. I’m working on finetuning Phi3 and I’m hopeful it’ll get better. I’m also working on synthesized datasets and mini-DSLs to make that tuning possible eg https://text-to-footprint.tscircuit.com https://text-to-footprint.tscircuit.com My impression is that synthetic datasets and finetuning will basically completely solve the problem, but eventually it’ll be available in general purpose models- so it’s not clear if its worth it to build a dedicated model. Overall the article’s analysis is great. I’m very optimistic that this will be solved in the next 2 years.