Closed benfred closed 2 years ago
In certain cases we could get an OOM on inference on the GPU models. Fix by calculating the temp memory appropriately.
In certain cases we could get an OOM on inference on the GPU models. Fix by calculating the temp memory appropriately.