NVIDIA NIM for MSA Search# MSA-Search Overview Advantages of NIMs Release Notes Release 1.0.0 Summary Model Variants Notes and Limitations Getting Started Prerequisites Installing curl, jq, and the Python requests module NGC Authentication Generate an API key Export the API key Docker Login to NGC Starting the NIM Container Runtime Parameters for the Container Caching Models Stopping the Container Next Steps Deploy Using Docker Advanced Usage Configuration GPU Selection Environment Variables Volumes Support Matrix Sequence Databases and Search Types Supported Hardware Minimum System Hardware Requirements Supported NVIDIA GPUs Testing Locally Available Hardware Optimization Automatic Profile Selection Deploying the NIM on a multi-GPU System Using the MMSeqs2 GPU Server with Compatible Databases Using a Compatible Database NIM Behavior without the GPU Server Enabling the GPU Server Querying the NIM Repeatedly or with Batches of Inputs Running Repeated Queries Against the NIM Serially Using Client-Side Batching to Improve Throughput Performance NIM Accuracy Factors Affecting NIM Performance Tools and Testing Practices API Reference OpenAPI Specification Predict a Multiple Sequence Alignment Search Input parameters Outputs Example Get the configured databases Input parameters Outputs Example Readiness check Input parameters Outputs Example
