GLiNER
Distilling a 287M model from Claude worked — except for that last 5% of accuracy
A 287M model distilled from Claude handled 5M court rulings — but missed the last 5% of accuracy. The real question: can AI be made affordable?
Back to home
GLiNER
2 articles tagged with this topic
OpenAI
OpenAI Privacy Filter Wins on Overlap F1, Fails Strict Match Due to Tokenizer Offset
On 600 PII samples, OpenAI privacy-filter beats GLiNER on overlap F1 (0.498 vs 0.416) but fails strict match (0.155) due to tokenizer offset. Choose b