| What is the language balance of the model validation data? |
en-US: 25.63%, es-ES: 10.50%, pt-PT: 5.26%, hi-IN: 5.10%, fr-FR: 4.56%, de-DE: 4.28%, pt-BR: 3.14%, mt-MT: 2.83%, hu-HU: 2.82%, ro-RO: 2.81%, bg-BG: 2.78%, el-GR: 2.77%, lt-LT: 2.26%, fi-FI: 2.26%, it-IT: 2.14%, zh-CN: 2.14%, lv-LV: 2.06%, ja-JP: 1.98%, sk-SK: 1.90%, ru-RU: 1.76%, ko-KR: 1.72%, et-EE: 1.67%, sl-SI: 1.59%, ar-AR: 1.57%, es-US: 1.43%, he-IL: 0.85%, vi-VN: 0.83%, fr-CA: 0.32%, nl-NL: 0.21%, tr-TR: 0.19%, en-GB: 0.14%, pl-PL: 0.14%, uk-UA: 0.11%, th-TH: 0.09%, hr-HR: 0.04%, cs-CZ: 0.03%, da-DK: 0.03%, nb-NO: 0.03%, sv-SE: 0.03%, nn-NO: 0.00% |
| What is the geographic origin language balance of the model validation data? |
Europe: 55.01%, North America: 27.38%, Asia: 11.86%, South America: 3.14%, Middle East: 2.61% |
| What is the accent balance of the model validation data? |
en-US: 25.63%, es-ES: 10.50%, pt-PT: 5.26%, hi-IN: 5.10%, fr-FR: 4.56%, de-DE: 4.28%, pt-BR: 3.14%, mt-MT: 2.83%, hu-HU: 2.82%, ro-RO: 2.81%, bg-BG: 2.78%, el-GR: 2.77%, lt-LT: 2.26%, fi-FI: 2.26%, it-IT: 2.14%, zh-CN: 2.14%, lv-LV: 2.06%, ja-JP: 1.98%, sk-SK: 1.90%, ru-RU: 1.76%, ko-KR: 1.72%, et-EE: 1.67%, sl-SI: 1.59%, ar-AR: 1.57%, es-US: 1.43%, he-IL: 0.85%, vi-VN: 0.83%, fr-CA: 0.32%, nl-NL: 0.21%, tr-TR: 0.19%, en-GB: 0.14%, pl-PL: 0.14%, uk-UA: 0.11%, th-TH: 0.09%, hr-HR: 0.04%, cs-CZ: 0.03%, da-DK: 0.03%, nb-NO: 0.03%, sv-SE: 0.03%, nn-NO: 0.00% |
| Participation considerations from adversely impacted groups (protected classes) in model design and testing: |
Age, Gender, Linguistic Background |
| Measures taken to mitigate against unwanted bias: |
Used a custom dataset to evaluate model performance across genders, age groups, and linguistic backgrounds. |
