| Model Type | | text generation, decoder-only, large language model |
|
| Use Cases |
| Areas: | | Content Creation and Communication, Research and Education |
|
| Applications: | | Text Generation, Chatbots and Conversational AI, Text Summarization, NLP Research, Language Learning Tools, Knowledge Exploration |
|
| Primary Use Cases: | | Text generation tasks such as question answering, summarization, reasoning |
|
| Limitations: | | Biases or gaps in data, open-ended tasks may be challenging, accuracy on factual information |
|
| Considerations: | | Developers should be mindful of content safety and privacy issues. |
|
|
| Additional Notes | | Models require adequate safety safeguards based on application use cases. |
|
| Supported Languages | |
| Training Details |
| Data Sources: | | Web Documents, Code, Mathematics |
|
| Data Volume: | | 9B model with 8 trillion tokens |
|
| Hardware Used: | |
| Model Architecture: | | text-to-text, decoder-only |
|
|
| Safety Evaluation |
| Methodologies: | | internal red-teaming, structured evaluations |
|
| Findings: | | Meets internal policies for child safety, content safety, representational harms, memorization, large-scale harms |
|
| Risk Categories: | | child sexual abuse and exploitation, harassment, violence and gore, hate speech |
|
|
| Responsible Ai Considerations |
| Fairness: | | Models undergo scrutiny for socio-cultural biases. |
|
| Transparency: | | Model details are shared in the model card. |
|
| Accountability: | | Guidelines for responsible use are provided. |
|
| Mitigation Strategies: | | Continuous monitoring and de-biasing encouraged. |
|
|
| Input Output |
| Input Format: | |
| Accepted Modalities: | |
| Output Format: | | Generated English-language text |
|
|