| Model Type | | text-to-text, decoder-only, large language model |
|
| Use Cases |
| Areas: | | Research, Commercial applications |
|
| Applications: | | Chatbots and Conversational AI, Text Generation, Text Summarization |
|
| Primary Use Cases: | | Question answering, Summarization, Reasoning |
|
| Limitations: | | Potential bias in responses, Might generate incorrect or outdated factual statements |
|
| Considerations: | | Models are suitable for environments with limited resources. |
|
|
| Additional Notes | | Models help democratize access to state-of-the-art AI technology. |
|
| Supported Languages | |
| Training Details |
| Data Sources: | | Web Documents, Code, Mathematics |
|
| Data Volume: | |
| Methodology: | |
| Hardware Used: | |
| Model Architecture: | | Large language model, text-to-text, decoder-only |
|
|
| Safety Evaluation |
| Methodologies: | | Red-teaming, Human evaluation |
|
| Risk Categories: | | Text-to-Text Content Safety, Text-to-Text Representational Harms, Memorization, Large-scale harm |
|
| Ethical Considerations: | | Models were filtered for sensitive data and personal information. |
|
|
| Responsible Ai Considerations |
| Fairness: | | Models underwent input data pre-processing for bias control. |
|
| Transparency: | | Model card provides architecture and evaluation details. |
|
| Accountability: | | Google is responsible for model outputs. |
|
| Mitigation Strategies: | | Data filtering and safety guidelines provided. |
|
|
| Input Output |
| Input Format: | |
| Accepted Modalities: | |
| Output Format: | |
| Performance Tips: | | Use longer context for better outputs. |
|
|