`template_key` is deprecated. In 0.7.dev1, you don't have to specify the template for a task. Instead, task name implies the template it uses. When extracting tasks from a model, Relay op strategy will generate multiple tasks for different templates. For example, one conv2d op in a model may result in two tasks: one for direct and one for Winograd. You can check the task list extracted from the model to see if it has any Winograd tasks. If not, then it means the workloads in your model do not meet the requirement of using Winograd algorithm (see https://github.com/apache/incubator-tvm/blob/master/python/tvm/relay/op/strategy/cuda.py#L102).
--- [Visit Topic](https://discuss.tvm.ai/t/task-create-function-template-key-keywork/6084/2) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/188d3734177cbe0944a323ff7b20e8cab42176151c237868e30ed2da99b8787e).