do you all see the use cases being similar to what you might use Fastino's Gliner models for? i see similar differentiation from general purpose LLMs in the sense that they can take natural-language input and return outputs adherent to a user-defined schema.
im thinking about how well Jev could be used to replace a current LLM-as-Judge evaluation workflows, specifically on chat transcript data (think ~1,500 tokens) i wonder if the reasoning usually required pushes it a bit out of scope. didnt see anything published about constraints on the state size, so would be curious to hear about that.
definitely seems like a modified version of GLiNER2 or 2.5:
- encoder-based (no text generation)
- multiple tasks in a single forward pass
- deterministic outputs
- constraint-based classification
flowerboy-t · · focus · HN ↗
<a href="https://fastino.ai/blog/gliner2-5-span-free-information-extraction" rel="nofollow">https://fastino.ai/blog/gliner2-5-span-free-information-extr...
im thinking about how well Jev could be used to replace a current LLM-as-Judge evaluation workflows, specifically on chat transcript data (think ~1,500 tokens) i wonder if the reasoning usually required pushes it a bit out of scope. didnt see anything published about constraints on the state size, so would be curious to hear about that.
mary776 · · focus · HN ↗