~/system-one-bench / models/fastino-glide
Fastino · Fastino API
GLiDEGLiDE by Fastino scores 84.1 on the System One Index (3rd of 39) and 82.5 on the Vision Index (8th of 14). Calibration error 3.6%.
| Dataset | Accuracy | Rank | Best on the board | Calibration error |
|---|---|---|---|---|
| AG News English | 92.0 | 2nd of 39 | 92.5 | 6.2% |
| SST-2 English | 96.0 | 5th of 39 | 98.0 | 3.5% |
| SST-5 English | 61.5 | 3rd of 39 | 62.5 | 7.2% |
| DAIR Emotion English | 61.5 | 11th of 39 | 91.5 | 11.3% |
| Banking77 English | 91.5 | 9th of 39 | 98.5 | 3.8% |
| BoolQ English | 92.0 | 3rd of 39 | 93.0 | 2.8% |
| XNLI English English | 94.7 | 2nd of 39 | 95.3 | 4.2% |
| Typed decisions English | 75.0 | 6th of 39 | 80.0 | 6.3% |
| XNLI French French | 88.0 | 6th of 39 | 91.3 | 3.9% |
| XNLI Arabic Arabic | 83.3 | 5th of 41 | 90.0 | 2.9% |
| MASSIVE scenario Arabic | 73.5 | 9th of 41 | 89.5 | 13.6% |
| MASSIVE intent Arabic | 91.0 | 3rd of 41 | 91.5 | 2.9% |
| OSACT4 offensive Arabic | 95.0 | 4th of 41 | 97.0 | 3.3% |
| Tweet sentiment Arabic | 73.5 | 8th of 41 | 80.0 | 5.7% |
| MIRACL rerank Arabic | 87.0 | 17th of 41 | 90.5 | 5.4% |
| MIRACL pairwise Arabic | 90.5 | 9th of 41 | 91.5 | 2.4% |
| Dataset | Accuracy | Rank | Best on the board | Calibration error |
|---|---|---|---|---|
| SST-2, as an image | 94.0 | 8th of 14 | 97.0 | 5.1% |
| Arabic sentiment, as an image | 51.5 | 9th of 14 | 75.0 | 13.0% |
| Oxford-IIIT Pet | 84.5 | 10th of 14 | 98.0 | 8.5% |
| VQAv2 yes/no | 89.0 | 3rd of 14 | 89.5 | 2.7% |
| CountBench | 93.5 | 3rd of 14 | 95.5 | 3.2% |
fastino/GLiDERan on Fastino's own API, which reports the model as glide_v2. The open weights are V1.2 (fastino/GLiDE, 26B, Fastino's community licence). Cost is input tokens at the $0.04 per 1M list price, since the API reports no cost. Doesn't answer Box questions.
Ranks count the models charted on the leaderboard. Every model gets the same cases and the same request, under the benchmark rules.