← Back to context Comment by algoth1 15 days ago Gemini 3.1 is still SOTA at weird handwriting recognition. If Argon is better it will be amazing 3 comments algoth1 Reply spwa4 15 days ago I bet that's one of those areas where a finetune of a 4B model will wipe the floor with anything else. matt123456789 15 days ago For readable or normal handwriting yes, big models are overkill. But weird handwriting is mostly edge cases. Bitter lesson territory. spwa4 15 days ago If you can detect -> finetune -> rollout it doesn't have to be a problem.
spwa4 15 days ago I bet that's one of those areas where a finetune of a 4B model will wipe the floor with anything else. matt123456789 15 days ago For readable or normal handwriting yes, big models are overkill. But weird handwriting is mostly edge cases. Bitter lesson territory. spwa4 15 days ago If you can detect -> finetune -> rollout it doesn't have to be a problem.
matt123456789 15 days ago For readable or normal handwriting yes, big models are overkill. But weird handwriting is mostly edge cases. Bitter lesson territory. spwa4 15 days ago If you can detect -> finetune -> rollout it doesn't have to be a problem.
I bet that's one of those areas where a finetune of a 4B model will wipe the floor with anything else.
For readable or normal handwriting yes, big models are overkill. But weird handwriting is mostly edge cases. Bitter lesson territory.
If you can detect -> finetune -> rollout it doesn't have to be a problem.