READ2016 dataset results
| Model | CER | WER | WER BoW |
|---|---|---|---|
| Gpt-4o-mini-2024-07-18 | 78.22% | 97.23% | 95.69% |
| Gpt-4o-2024-08-06 | 77.42% | 97.59% | 94.19% |
| Claude-3-5-sonnet-20240620 | 71.08% | 95.53% | 92.17% |
| MinicpmV-2 6 | 80.74% | 99.74% | 99.24% |
| Qwen2-VL-7B | 76.25% | 97.77% | 96.78% |
| Pixtral-12B | 79% | 99% | 98.03% |
| InternVL2-8B | 82.34% | 99.89% | 99.03% |
| Phi-3-mini-128k-instruct | 82.31% | 99.77% | 99.42% |
| Model | CER | WER | WER BoW |
|---|---|---|---|
| Gpt-4o-mini-2024-07-18 | 78.22% | 97.23% | 95.69% |
| Gpt-4o-2024-08-06 | 77.42% | 97.59% | 94.19% |
| Claude-3-5-sonnet-20240620 | |||
| MinicpmV-2 6 | 80.74% | 99.74% | 99.24% |
| Qwen2-VL-7B | 76.25% | 97.77% | 96.78% |
| Pixtral-12B | 79% | 99% | 98.03% |
| InternVL2-8B | 82.34% | 99.89% | 99.03% |
| Phi-3-mini-128k-instruct | 82.31% | 99.77% | 99.42% |