# MLPerf Inference v0.7 - Image Classification - TensorRT

- Set up [`program:image-classification-tensorrt-loadgen-py`](https://github.com/ctuning/ck-mlperf/blob/master/program/image-classification-tensorrt-loadgen-py/README.md) on your SUT.
- Customize the examples below for your SUT.

<a name="resnet50"></a>
## ResNet50

<a name="singlestream"></a>
### SingleStream

#### Performance

```bash
$ ck run cmdgen:benchmark.tensorrt-loadgen --verbose \
--model=resnet50 --mode=performance --target_latency=2.04 \
--scenario=singlestream --batch_size=1
```

#### Accuracy

```bash
$ ck run cmdgen:benchmark.tensorrt-loadgen --verbose \
--model=resnet50 --mode=accuracy --dataset_size=50000 \
--scenario=singlestream --batch_size=1
```

#### Compliance **TODO**


<a name="offline"></a>
### Offline

#### Performance

```bash
$ ck run cmdgen:benchmark.tensorrt-loadgen --verbose \
--model=resnet50 --mode=performance --target_qps=1300 \
--scenario=offline --batch_size=64
```

#### Accuracy

```bash
$ ck run cmdgen:benchmark.tensorrt-loadgen --verbose \
--model=resnet50 --mode=accuracy --dataset_size=50000 \
--scenario=offline --batch_size=64
```

#### Compliance **TODO**


<a name="multistream"></a>
### MultiStream

#### Performance ((272,160 == 9! * 3/4) > 270,336)

```bash
$ ck run cmdgen:benchmark.tensorrt-loadgen --verbose \
--model=resnet50 --mode=performance --max_query_count=272160 \
--scenario=multistream --batch_size=68 --nstreams={{{batch_size}}}
```

#### Accuracy

```bash
$ ck run cmdgen:benchmark.tensorrt-loadgen --verbose \
--model=resnet50 --mode=accuracy --dataset_size=50000 \
--scenario=multistream --batch_size=70 --nstreams={{{batch_size}}}
```

#### Compliance **TODO**