Click to see the answer A token is 3.7 English characters. The token count includes the prompt and the answer that is provided by Ask Sage.
How does Ask Sage keep the data from being used to train the models?
Click to see the answer Ask Sage utilizes fire and forget APIs. Ask Sage ensures that data is not used to train models by leveraging the Azure OpenAI APIs, which operate on the Azure commercial FedRAMP high cloud, fully managed by Microsoft. We do not fine-tune models; instead, we embed information to maintain model agnosticism. This approach allows us to use our data with any model in the future, avoiding dependency on a single model. Our vector database, which ingests all government data and stores Controlled Unclassified Information (CUI), is hosted in Ask Sage's Azure GovCloud at IL5. At query time, we only pass 3 to 6 relevant results to the Azure OpenAI model. These are "fire and forget" queries, meaning that once a response is generated, the query data is not retained. By utilizing these "fire and forget" APIs, we ensure that the data is not stored or used for training purposes. This method, combined with our secure infrastructure, guarantees that user data remains protected and is not used to train the models.
Is the data/IP I ingest into Ask Sage accessible to other people?
Click to see the answer No, Ask Sage utilizes label-based access control. Only you have access to the data you ingest unless you give other people within your org access to it.
How can I use Ask Sage to write proposals?
Click to see the answer You can use Ask Sage to write proposals in response to a Request for quote by training Ask Sage on a previous proposal and Ask Sage can then create a new proposal that covers the requirements in the new solicitation.
What are some examples of documents I can ingest for ATO In a Box?
Click to see the answer Anything! If there is a file or document that you have an issue with, let us know and we can help you ingest it into Ask Sage.
When do Ask Sage tokens replenish?
Click to see the answer The 1st of every month.
What is a Persona?
Click to see the answer A persona is a set of characteristics that define the behavior and specific knowledge you want the model to focus on. It’s like choosing a character for the model to play.
What Models are used to generate images?
Click to see the answer DALL-E(2), DALL-E(3)
What is Temperature?
Click to see the answer The temperature parameter controls the randomness of the model. Lower temperatures make the model more deterministic and higher temperatures make the model more random. Which means that the higher the temperature, the more creative the model will be, however, it may also produce less coherent results. - focus on creativity
How do I sign up for Ask Sage?
Click to see the answer https://chat.asksage.ai/login or contact our sales team
What is a dataset?
Click to see the answer A dataset is equivalent to a folder where you can store all the data you want to ingest into Ask Sage. You can create multiple datasets to organize your data based on specific use cases or projects.
What does the Live/Live+ function do?
Click to see the answer The live feature is a powerful feature that allows users to generate a real-time web search and return up-to-date information in the inferences generated by the model.
What is an Inference and Training Token?
Click to see the answer Ask Sage uses tokens to manage the usage of the platform, which is split into two types: Inference and Training tokens. Inference tokens are used to generate inferences from the model, while training tokens are used when ingesting data into a dataset.
Can I interface with Ask Sage through an API?
Click to see the answer Yes, Ask Sage has an API that allows you to interact with the platform programmatically. You can use the API to ingest data, generate inferences, and more. Look at the API section for more information.
Can I share my datasets with other users?
Click to see the answer Yes, you can share your datasets with other users within your organization. However, note that only the owner of the dataset can modify (ingest data, delete data, etc.) the dataset.
