Comment by garciasn
4 days ago
I am in no way trying to sell Jev here as some panacea of the modern world; I'm only responding to your questions:
> But their example is classification but that would also be possible and faster with a classic BERT model.
With BERT, you need a large, labeled dataset, and you have to train/fine-tune the model. Jev is pitched as a zero- or 'few-shot' model. You define the schema in code, give it instructions, and it works without a traditional training pipeline.
> So their pitch is a task specific smaller model or am I completely misunderstanding the whole thing?
Yup; that about sums it up: it is more or less an optimized, task-specific small model with the flexible understanding of a traditional LLM.
> With BERT, you need a large, labeled dataset, and you have to train/fine-tune the model.
BERT requires a huge corpus, but it isn't labeled. BERT is trained through self-supervised learning using mask tokens and next sentence prediction. Fine-tuning is useful for specific tasks, but isn't absolutely essential for the model to function.
If something is task-specific (well understood) wouldn't this be a good candidate for a computer program?
Not necessarily. It may be well-understood but still require judgment calls.
couldn't be more wrong - there are so many zero shot classifiers available on HF which do the same thing.
I think you're possibly arguing a point I wasn't making? I'm not saying Jev invented zero-shot classification, or that there aren't already zero-shot classifiers on HF that can do classification without fine-tuning; I was responding to questions asked in a silo.
>With BERT, you need a large, labeled dataset, and you have to train/fine-tune the model.
I think they were responding to this. You can use BERT to provide zero shot classification predictions.
1 reply →