Open LouisCastricato opened 1 year ago
Would this task not dilute the model? It seems to me that it is neither chat or code oriented. I can see the tradeoff being worth it if the model can surpass the open source SOTA but I am not sure it can
Could be rephrased as translating code comments I think then. We're ok with going a bit out of distribution, we don't need to use all tasks for version 1. It's just hard to spin annotation back up when necessary
Okok, I hadn't though about the issue of spinning annotations back up
🚀 The Task
Ask a chatbot to translate an utterance or set of utterances. This can be augmented with existing aligned datasets
Example
User: Here are two sentences, please translate them to [TARGET LANGUAGE] Bot: [Sentences in target language]
Additional Notes
This can trivially be instructed from a plethora of existing datasets, and probably does not require prompt collection.