FAQs
Some most common questions about various Gathr related features have been answered in this section.
AWS Compute Environments FAQs
Q: How can I connect from the AWS region which is not listed in the Compute Setup?
In case if the preferred region where you want to host the compute environments is not supported by Gathr, please write to our support team at:
Alternatively, you can use VPC peering for clusters in this region.
Q: Can I switch an Ingestion or CDC Application configured to run in Gathr’s account to run in my account?
Yes. A pipeline created with tunnel/upload file feature for source or target, configured to run in Gathr’s account can be reconfigured to run in your account.
Prerequisite: Source/target and EMR compute should be configured in the same region, else (optional) you can create VPC peering between regions in your account.
Q: Can I switch an Ingestion or CDC Application configured to run in my account to run in Gathr’s account?
Yes. A pipeline created with tunnel for source or target, configured to run in your account can be reconfigured to run in Gathr’s account in the same region.
No. A pipeline created with upload file feature for source or target configured to run in your account during creation cannot be reconfigured to run in Gathr’s account.
Q: Will I be charged differently if I am running pipelines in my Account?
Yes. Please contact our support team for more details.
Azure Databricks Compute Environments FAQs
Q: What costs are associated with using Databricks in Gathr?
The costs primarily consist of the infrastructure and cloud-related expenses incurred on your Databricks account. This includes the charges levied by Databricks for the resources consumed during the active period of your clusters. Gathr, however, does not impose any additional charges for designing or deploying pipelines within the platform.Q: Are there any additional charges from Gathr for using Databricks?
You will be only charged for the credit points consumed during application execution. Gathr does not impose any supplementary charges for designing or deploying applications on Databricks.Q: How can I register my Azure Databricks account with Gathr?
To register your Azure Databricks account with Gathr, navigate to your User Profile > Compute Setup tab, click on ADD CLOUD SERVICE ACCOUNT, select Azure Databricks as the Account Type, and provide the required details, including Account Name, Instance URL, Access Token, and DBFS Gathr Metadata Repository Path. Then simply save the account details.Q: How do I link a compute environment to a project after registering a Databricks account?
Administrators can link registered compute accounts to specific projects within Gathr. This allows project users to utilize the registered clusters for deploying their applications. For detailed steps, refer to Link Compute Environment to Project →Q: What is the difference between the Gathr and Databricks engines in Gathr?
The Gathr engine is the default computing environment for designing and deploying applications within the Gathr platform. The Databricks engine, when registered, enables users to use clusters that can be launched from Gathr on their Databricks workspaces.Q: Can I switch between the Gathr and Databricks engines at any time?
Yes, users can switch between the Gathr and Databricks engines based on their specific requirements. However, switching is not allowed while actively designing applications within the interface.OpenAI Image Processor FAQs
Q: Can I fine-tune the image capabilities in gpt-4?
No, you cannot fine-tune the image capabilities of gpt-4 at this time.Q: Can I use gpt-4 to generate images?
No, you cannot generate images but use gpt-4-vision-preview to understand images.Q: What type of files can I upload?
You can upload PNG (.png), JPEG (.jpeg and .jpg), WEBP (.webp), and non-animated GIF (.gif).Q: Is there a limit to the size of the image I can upload?
Yes, image uploads are restricted to 20MB per image.Q: Can I delete an image I uploaded?
No, the image will automatically get deleted after it has been processed by the model.Q: How do rate limits for GPT-4 with Vision work?
Images are processed at the token level, so each image processed counts towards your tokens per minute (TPM) limit.Q: Can GPT-4 with Vision understand image metadata?
No, the model does not receive image metadata.Q: What happens if my image is unclear?
If an image is ambiguous or unclear, the model will do its best to interpret it. However, the results may be less accurate. A good rule of thumb is that if an average human cannot see the info in an image at the resolutions used in low/high res mode, then the model cannot either.If you have any feedback on Gathr documentation, please email us!