Benchmark
Pashto ASR FLEURS ps_af script Ar/Da/Ur (zero-shot)
Scores are percentages.
- Models measured
- 7
- Scores recorded
- 7
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 63.6%
- Mean
- 71.5%
- Median
- 35%
- Std dev
- 29.5%
- 10th percentile
- 98.8%
- 90th percentile
- 6.4%
- Min
Results
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model | Score |
|---|---|
| Whisper Basetable: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 42.6% |
| Whisper Large v2table: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 71.5% |
| Whisper Large V3table: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 99.2% |
| Whisper Large V3 Turbotable: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 98.8% |
| Whisper Mediumtable: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 29.5% |
| Whisper Smalltable: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 97.5% |
| Whisper Tinytable: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot· Pashto ASR paper data | 6.4% |
Data credits: Hanif Rahman. Benchmarking Multilingual Speech Models on Pashto: Zero-Shot ASR, Script Failure, and Cross-Domain Evaluation, arXiv:2604.04598v1. Numeric/protocol extract from Tables II, III, V and VI; CC BY 4.0 article grant verified. (licence)