Caenorst commented on issue #14040: Reformat of TensorRT to use subgraph API
URL: https://github.com/apache/incubator-mxnet/pull/14040#issuecomment-499975995
 
 
   @ThomasDelteil the PR for updating it is on its way: 
https://mxnet.incubator.apache.org/versions/master/tutorials/tensorrt/inference_with_trt.html
   
   Meanwhile here is the new way to use TensorRT:
   ```
   sym, arg_params, aux_params = mx.model.load_checkpoint('my_net', epochs)
   trt_sym = sym.get_backend_symbol('TensorRT')
   remaining_arg_params, remaining_aux_params = 
mx.contrib.tensorrt.init_tensorrt_params(trt_sym, arg_params, aux_params)
   ```
   And then you can use `trt_sym` as your usual symbol, no more `tensorrt_bind` 
or `MXNET_USE_TENSORRT` environment variable. Also now FP16 computation is by 
default if you want to disable it use the function 
`mx.contrib.tensorrt.set_use_fp16(False)`

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

Reply via email to