From bbf6ff02548795921fbf9cd1fb7ea7b0eb66b2e1 Mon Sep 17 00:00:00 2001 From: rachfop Date: Mon, 11 Mar 2024 11:21:28 -0700 Subject: [PATCH 1/2] Priority numbers --- docs/serverless/endpoints/manage-endpoints.md | 9 +++++++++ docs/serverless/workers/overview.md | 2 +- 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/docs/serverless/endpoints/manage-endpoints.md b/docs/serverless/endpoints/manage-endpoints.md index c09e885a..c255fe95 100644 --- a/docs/serverless/endpoints/manage-endpoints.md +++ b/docs/serverless/endpoints/manage-endpoints.md @@ -37,6 +37,15 @@ You can edit a running Endpoint in the Web interface after you've deployed it. 3. Select **Edit Endpoint** and make your changes. 4. Choose **Update**. +## Set GPU prioritization an Endpoint + +When creating or modifying a Worker Endpoint, specify your GPU preferences in descending order of priority. + +This allows you to configure the desired GPU models for your Worker Endpoints. + +RunPod attempts to allocate your first choice if it's available. +If your preferred GPU is not available, the system will automatically default to the next available GPU in your priority list. + :::note You can force a configuration update by setting **Max Workers** to 0, selecting **Update**, then updating your max workers back to your needed value. diff --git a/docs/serverless/workers/overview.md b/docs/serverless/workers/overview.md index 1a35d9eb..f37359ac 100644 --- a/docs/serverless/workers/overview.md +++ b/docs/serverless/workers/overview.md @@ -16,6 +16,6 @@ Workers run your code in the cloud. To start using RunPod Workers: -1. **Write Your Function**: Code your Handler Functions in a supported language. +1. **Write your function**: Code your Handler Functions in a supported language. 2. **Deploy to RunPod**: Upload your Handler Functions to RunPod. 3. **Integrate and Execute**: Use the provided Endpoint to integrate with your application. From ba05d32e48f5dc477b9fb11ea4c8676998817d6a Mon Sep 17 00:00:00 2001 From: rachfop Date: Mon, 11 Mar 2024 11:23:42 -0700 Subject: [PATCH 2/2] U'Update GPU prioritization for Worker Endpoints and provide clear instructions (#58) --- docs/serverless/endpoints/manage-endpoints.md | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/docs/serverless/endpoints/manage-endpoints.md b/docs/serverless/endpoints/manage-endpoints.md index c255fe95..d14284b5 100644 --- a/docs/serverless/endpoints/manage-endpoints.md +++ b/docs/serverless/endpoints/manage-endpoints.md @@ -40,11 +40,15 @@ You can edit a running Endpoint in the Web interface after you've deployed it. ## Set GPU prioritization an Endpoint When creating or modifying a Worker Endpoint, specify your GPU preferences in descending order of priority. - This allows you to configure the desired GPU models for your Worker Endpoints. RunPod attempts to allocate your first choice if it's available. -If your preferred GPU is not available, the system will automatically default to the next available GPU in your priority list. +If your preferred GPU isn't available, the system automatically defaults to the next available GPU in your priority list. + +1. Navigate to [Serverless Endpoints](https://www.dev.runpod.io/console/serverless). +2. Select the Endpoint you'd like to update. +3. Select the priority of the GPUs you'd like to use. +4. Choose **Update**. :::note