MCPcopy Create free account

hub / github.com/RizhaoCai/PyTorch_ONNX_TensorRT / functions

Functions33 in github.com/RizhaoCai/PyTorch_ONNX_TensorRT

⨍ Functions 33◇ Types & classes 6
↓ 1 callersMethod__init__
For calibiration, you need to implement your 'next_batch' and 'reset' functions
helpers/trt_int8_calibration_helper.py:56
↓ 1 callersMethod__str__
(self)
dynamic_shape_example.py:50
↓ 1 callersMethod__str__
(self)
helpers/trt_helper.py:17
↓ 1 callersFunctionallocate_buffers
(engine)
dynamic_shape_example.py:56
↓ 1 callersFunctionbuild_engine
Takes an ONNX file and creates a TensorRT engine to run inference with
dynamic_shape_example.py:95
↓ 1 callersFunctionbuild_engine
Takes an ONNX file and creates a TensorRT engine to run inference with
helpers/trt_helper.py:47
↓ 1 callersFunctiondo_inference
(context, bindings, inputs, outputs, stream, batch_size=1)
dynamic_shape_example.py:79
↓ 1 callersMethodforward_default
(self, X_in)
trt_int8_demo.py:27
↓ 1 callersMethodforward_onnx
(self, X_in)
trt_int8_demo.py:43
↓ 1 callersFunctiongenerate_onnx_model
(onnx_model_path, img_size, batch_size)
trt_int8_demo.py:63
↓ 1 callersFunctionget_engine
Attempts to load a serialized engine if available, otherwise builds a new TensorRT engine and saves it.
dynamic_shape_example.py:92
↓ 1 callersFunctionmain
()
trt_int8_demo.py:77
↓ 1 callersMethodnext_batch
Return a batch of data every time called
helpers/trt_int8_calibration_helper.py:73
↓ 1 callersMethodreset
(self)
helpers/trt_int8_calibration_helper.py:70
↓ 1 callersFunctiontest_dynamic_shape
()
dynamic_shape_example.py:147
↓ 1 callersFunctionto_onnx
()
dynamic_shape_example.py:29
Method__call__
(self, *args,**kargs)
trt_int8_demo.py:57
Method__init__
(self, num_classes=10,)
trt_int8_demo.py:11
Method__init__
(self, num_clasess=2)
dynamic_shape_example.py:17
Method__init__
(self, host_mem, device_mem)
dynamic_shape_example.py:46
Method__init__
Within this context, host_mom means the cpu memory and device means the GPU memory
helpers/trt_helper.py:12
Method__init__
(self, input_layers, stream, cache_file='calibration_cache.bin')
helpers/trt_int8_calibration_helper.py:12
Method__repr__
(self)
dynamic_shape_example.py:53
Method__repr__
(self)
helpers/trt_helper.py:20
Functionallocate_buffers
(engine)
helpers/trt_helper.py:23
Functiondo_inference
(context, bindings, inputs, outputs, stream, batch_size=1)
helpers/trt_helper.py:97
Methodforward
(self, x)
dynamic_shape_example.py:22
Methodget_batch
(self, bindings, names)
helpers/trt_int8_calibration_helper.py:23
Methodget_batch_size
(self)
helpers/trt_int8_calibration_helper.py:20
Functionget_engine
Attempts to load a serialized engine if available, otherwise builds a new TensorRT engine and saves it.
helpers/trt_helper.py:43
Functionpostprocess_the_outputs
(h_outputs, shape_of_output)
helpers/trt_helper.py:109
Methodread_calibration_cache
(self, length)
helpers/trt_int8_calibration_helper.py:35
Methodwrite_calibration_cache
(self, ptr, size)
helpers/trt_int8_calibration_helper.py:41