Docs: Update docs to add ORT AMD benchmarks based on docker (#237)

Update docs to add ORT AMD benchmarks based on docker.

docs/superbench-config.mdx

@@ -148,8 +148,8 @@ superbench:
 
 Mappings of `${benchmark_name}: Benchmark`.
 
-There are two types of benchmarks, micro-benchmark and model-benchmark.
-For micro-benchmark, `${benchmark_name}` should be the exact same as provided micro-benchmarks' name.
+There are three types of benchmarks, micro-benchmark, model-benchmark and docker-benchmark.
+For micro-benchmark and docker-benchmark, `${benchmark_name}` should be the exact same as provided benchmarks' name.
 For model-benchmark, `${benchmark_name}` should be in `${name}_models` format,
 each model-benchmark can have a customized name while ending with `_models`.
 
@@ -269,7 +269,7 @@ See [`Mode` Schema](#mode-schema) for mode definition.
 A list of frameworks in which the benchmark runs.
 Some benchmarks can support multiple frameworks while others only support one.
 
-* accepted values: `[ onnx | pytorch | tf1 | tf2 | none ]`
+* accepted values: `[ onnxruntime | pytorch | tf1 | tf2 | none ]`
 * default value: `[ none ]`
 
 ### `models`

docs/user-tutorial/benchmarks/docker-benchmarks.md

+---
+id: docker-benchmarks
+---
+
+# Docker Benchmarks
+
+## ROCm ONNXRuntime Model Benchmarks
+
+### `ort-models`
+
+#### Introduction
+
+Run the rocm onnxruntime model training benchmarks packaged in docker `superbench/benchmark:rocm4.3.1-onnxruntime1.9.0` which includes Bert-large, Distilbert-base, GPT-2, facebook/Bart-large and Roberta-large.
+
+#### Metrics
+
+| Name                                                  | Unit                   | Description                                               |
+|-------------------------------------------------------|------------------------|-----------------------------------------------------------|
+| onnxruntime-ort-models/bert_large_uncased_ngpu_1      | throughput (samples/s) | The throughput of bert large uncased model on 1 GPU.      |
+| onnxruntime-ort-models/bert_large_uncased_ngpu_8      | throughput (samples/s) | The throughput of bert large uncased model on 8 GPU.      |
+| onnxruntime-ort-models/distilbert_base_uncased_ngpu_1 | throughput (samples/s) | The throughput of distilbert base uncased model on 1 GPU. |
+| onnxruntime-ort-models/distilbert_base_uncased_ngpu_8 | throughput (samples/s) | The throughput of distilbert base uncased model on 8 GPU. |
+| onnxruntime-ort-models/gpt2_ngpu_1                    | throughput (samples/s) | The throughput of gpt2 model on 1 GPU.                    |
+| onnxruntime-ort-models/gpt2_ngpu_8                    | throughput (samples/s) | The throughput of gpt2 model on 8 GPU.                    |
+| onnxruntime-ort-models/facebook_bart_large_ngpu_1     | throughput (samples/s) | The throughput of facebook bart large model on 1 GPU.     |
+| onnxruntime-ort-models/facebook_bart_large_ngpu_8     | throughput (samples/s) | The throughput of facebook bart large model on 8 GPU.     |
+| onnxruntime-ort-models/roberta_large_ngpu_1           | throughput (samples/s) | The throughput of roberta large model on 1 GPU.           |
+| onnxruntime-ort-models/roberta_large_ngpu_8           | throughput (samples/s) | The throughput of roberta large model on 8 GPU.           |

website/sidebars.js

@@ -27,6 +27,7 @@ module.exports = {
           items: [
             'user-tutorial/benchmarks/micro-benchmarks',
             'user-tutorial/benchmarks/model-benchmarks',
+            'user-tutorial/benchmarks/docker-benchmarks',
           ],
         },
         'user-tutorial/system-config',