microsoft · natke · Jan 13, 2021 · Jan 5, 2021 · Jan 5, 2021 · Jan 5, 2021
diff --git a/docs/reference/execution-providers/ACL-ExecutionProvider.md b/docs/reference/execution-providers/ACL-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: ARM Compute Library (ACL)
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 7
+nav_order: 8
 ---
 
 # ACL Execution Provider

diff --git a/docs/reference/execution-providers/ArmNN-ExecutionProvider.md b/docs/reference/execution-providers/ArmNN-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: ARM NN
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 8
+nav_order: 9
 ---
 
 ## ArmNN Execution Provider

diff --git a/docs/reference/execution-providers/CUDA-ExecutionProvider.md b/docs/reference/execution-providers/CUDA-ExecutionProvider.md
@@ -0,0 +1,90 @@
+---
+title: CUDA
+parent: Execution Providers
+grand_parent: Reference
+nav_order: 1
+---
+
+# CUDA Execution Provider
+
+The CUDA Execution Provider enables hardware accelerated computation on Nvidia CUDA-enabled GPUs.
+
+## Build
+For build instructions, please see the [BUILD page](../../how-to/build.md#CUDA).
+
+## Configuration Options
+The CUDA Execution Provider supports the following configuration options.
+
+### device_id
+The device ID.
+
+Default value: 0
+
+### cuda_mem_limit
+The size limit of the device memory arena in bytes. This size limit is only for the execution provider's arena. The total device memory usage may be higher.
+
+Default value: max value of C++ size_t type (effectively unlimited)
+
+### arena_extend_strategy
+The strategy for extending the device memory arena.
+
+Value                   | Description
+-|-
+kNextPowerOfTwo (0)     | subsequent extensions extend by larger amounts (multiplied by powers of two)
+kSameAsRequested (1)    | extend by the requested amount
+
+Default value: kNextPowerOfTwo
+
+### cudnn_conv_algo_search
+The type of search done for cuDNN convolution algorithms.
+
+Value           | Description
+-|-
+EXHAUSTIVE (0)  | expensive exhaustive benchmarking using cudnnFindConvolutionForwardAlgorithmEx
+HEURISTIC (1)   | lightweight heuristic based search using cudnnGetConvolutionForwardAlgorithm_v7
+DEFAULT (2)     | default algorithm using CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_PRECOMP_GEMM
+
+Default value: EXHAUSTIVE
+
+### do_copy_in_default_stream
+Whether to do copies in the default stream or use separate streams. The recommended setting is true. If false, there are race conditions and possibly better performance.
+
+Default value: true
+
+## Example Usage
+
+### Python
+
+```python
+import onnxruntime as ort
+
+model_path = '<path to model>'
+
+providers = [
+    ('CUDAExecutionProvider', {
+        'device_id': 0,
+        'arena_extend_strategy': 'kNextPowerOfTwo',
+        'cuda_mem_limit': 2 * 1024 * 1024 * 1024,
+        'cudnn_conv_algo_search': 'EXHAUSTIVE',
+        'do_copy_in_default_stream': True,
+    }),
+    'CPUExecutionProvider',
+]
+
+session = ort.InferenceSession(model_path, providers=providers)
+```
+
+### C/C++
+
+```c++
+OrtSessionOptions* session_options = /* ... */;
+
+OrtCUDAProviderOptions options;
+options.device_id = 0;
+options.arena_extend_strategy = 0;
+options.cuda_mem_limit = 2 * 1024 * 1024 * 1024;
+options.cudnn_conv_algo_search = OrtCudnnConvAlgoSearch::EXHAUSTIVE;
+options.do_copy_in_default_stream = 1;
+
+SessionOptionsAppendExecutionProvider_CUDA(session_options, &options);
+```
diff --git a/docs/reference/execution-providers/DNNL-ExecutionProvider.md b/docs/reference/execution-providers/DNNL-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: Intel DNNL
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 4
+nav_order: 5
 ---
 
 # DNNL Execution Provider

diff --git a/docs/reference/execution-providers/DirectML-ExecutionProvider.md b/docs/reference/execution-providers/DirectML-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: Direct ML
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 3
+nav_order: 4
 ---
 
 # DirectML Execution Provider

diff --git a/docs/reference/execution-providers/MIGraphX-ExecutionProvider.md b/docs/reference/execution-providers/MIGraphX-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: AMD MI GraphX
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 6
+nav_order: 7
 ---
 
 # MIGraphX Execution Provider

diff --git a/docs/reference/execution-providers/NNAPI-ExecutionProvider.md b/docs/reference/execution-providers/NNAPI-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: NNAPI
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 5
+nav_order: 6
 ---
 
 

diff --git a/docs/reference/execution-providers/Nuphar-ExecutionProvider.md b/docs/reference/execution-providers/Nuphar-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: Nuphar
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 9
+nav_order: 10
 ---
 
 # Nuphar Execution Provider (preview)
@@ -33,7 +33,13 @@ Ort::Session session(env, model_path, sf);
 
 ### Python
 
-You can use the Nuphar execution provider via the python wheel from the ONNX Runtime build. The Nuphar execution provider will be automatically prioritized over the default CPU execution providers, thus no need to separately register the execution provider. Python APIs details are [here](/python/api_summary).
+```python
+import onnxruntime as ort
+
+model_path = '<path to model>'
+providers = ['NupharExecutionProvider', 'CPUExecutionProvider']
+session = ort.InferenceSession(model_path, providers=providers)
+```
 
 ## Performance and Accuracy Testing
 
@@ -161,12 +167,12 @@ SessionOptions.MakeSessionOptionWithNupharProvider("nuphar_cache_path:/path/to/c
 
 * Using in Python
 
-Settings string should be passed in before InferenceSession is created, as providers are not currently exposed yet. Here's an example in Python to set cache path and model checksum:
+Settings string can be set as an execution provider-specific option. Here's an example in Python to set cache path and model checksum:
 
 ```python
 nuphar_settings = 'nuphar_cache_path:{}, nuphar_cache_model_checksum:{}'.format(cache_dir, model_checksum)
-onnxruntime.capi._pybind_state.set_nuphar_settings(nuphar_settings)
-sess = onnxruntime.InferenceSession(model_path)
+providers = [('NupharExecutionProvider', {'nuphar_settings': nuphar_settings}), 'CPUExecutionProvider']
+sess = onnxruntime.InferenceSession(model_path, providers=providers)
 ```
 
 ## Known issues

diff --git a/docs/reference/execution-providers/OpenVINO-ExecutionProvider.md b/docs/reference/execution-providers/OpenVINO-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: OpenVINO
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 2
+nav_order: 3
 ---
 
 # OpenVINO Execution Provider

diff --git a/docs/reference/execution-providers/RKNPU-ExecutionProvider.md b/docs/reference/execution-providers/RKNPU-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: RKNPU
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 10
+nav_order: 11
 ---
 
 # RKNPU Execution Provider (preview)

diff --git a/docs/reference/execution-providers/TensorRT-ExecutionProvider.md b/docs/reference/execution-providers/TensorRT-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: TensorRT
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 1
+nav_order: 2
 ---
 
 # TensorRT Execution Provider

diff --git a/docs/reference/execution-providers/Vitis-AI-ExecutionProvider.md b/docs/reference/execution-providers/Vitis-AI-ExecutionProvider.md
@@ -2,7 +2,7 @@
 title: Vitis AI
 parent: Execution Providers
 grand_parent: Reference
-nav_order: 11
+nav_order: 12
 ---
 
 # Vitis-AI Execution Provider