3 ms·
How does one go about minimizing costs for certain situations? For example if I share a db schema and then ask it to generate some sql, I need to share that en
by fareesh 4y ago
How does one go about minimizing costs for certain situations?
For example if I share a db schema and then ask it to generate some sql, I need to share that entire db schema for every single question that follows, is that right?
Or is it possible for me to somehow pay and have it "retain" that schema knowledge for all subsequent queries without having to send the schema along with every single question?
- mirekrusin 4y agoIt's called fine tuning [0] but it's not yet available for chatgpt, only older models. [0] https://platform.openai.com/docs/guides/fine-tuning https://platform.openai.com/docs/guides/fine-tuning
- fareesh 4y agoYeah that seems to be the trouble - the pricing is great with ChatGPT but not so much with the older ones
- mirekrusin 4y agoThat may be partially because it's not possible to do pre-training on such a large model.
- fareesh 4y agoOn the web version is it doing that under the hood? I.e. if I have a 10+ follow up conversation does it start from the beginning each time?
- mirekrusin 4y agoIt keeps context of 8k tokens. In pre-training you're using much more examples and network is tuned around them.