modelconv 0.5.4__tar.gz → 0.5.6__tar.gz

{modelconv-0.5.4 → modelconv-0.5.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: modelconv
-Version: 0.5.4
+Version: 0.5.6
 Summary: Converter for neural models into various formats.
 Author-email: Luxonis <support@luxonis.com>
 Maintainer-email: Luxonis <support@luxonis.com>
@@ -17,10 +17,12 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: Pillow
-Requires-Dist: luxonis-ml[data,gcs,nn_archive,s3]>=0.8.4
-Requires-Dist: cyclopts==4.8.0
+Requires-Dist: luxonis-ml[data,gcs,nn-archive,s3]>=0.8.5
+Requires-Dist: cyclopts~=4.16
 Requires-Dist: onnx==1.21.0
 Requires-Dist: onnxruntime
+Requires-Dist: onnxsim~=0.6
+Requires-Dist: onnxslim~=0.1
 Requires-Dist: onnxsim
 Requires-Dist: docker
 Requires-Dist: keyring
@@ -52,8 +54,6 @@ Requires-Dist: polars; extra == "rvc4"
 Requires-Dist: pytest; extra == "rvc4"
 Requires-Dist: onnx==1.18.0; extra == "rvc4"
 Provides-Extra: hailo
-Requires-Dist: nvidia-dali-cuda120==1.49.0; extra == "hailo"
-Requires-Dist: nvidia-dali-tf-plugin-cuda120==1.49.0; extra == "hailo"
 Requires-Dist: protobuf==3.20.3; extra == "hailo"
 Requires-Dist: matplotlib==3.10.6; extra == "hailo"
 Requires-Dist: pyparsing==2.4.7; extra == "hailo"
@@ -125,10 +125,10 @@ Run `modelconverter --help` to see the available commands and options.
 
 There are two main ways to execute configure the conversion process:
 
-1. **YAML Configuration File (Primary Method)**:
-   The primary way to configure the conversion is through a YAML configuration file. For reference, you can check [defaults.yaml](shared_with_container/configs/defaults.yaml) and other examples located in the [shared_with_container/configs](shared_with_container/configs) directory.
 1. **NN Archive**:
    Alternatively, you can use an [NN Archive](https://docs.luxonis.com/software-v3/ai-inference/nn-archive/#NN%20Archive) as input. An NN Archive includes a model in one of the supported formats—ONNX (.onnx), OpenVINO IR (.xml and .bin), or TensorFlow Lite (.tflite)—alongside a `config.json` file. The config.json file follows a specific configuration format as described under the `Configuration` section.
+2. **YAML Configuration File (Legacy)**:
+   An alternative way to configure the conversion is through a YAML configuration file. For reference, you can check [defaults.yaml](shared_with_container/configs/defaults.yaml) and other examples located in the [shared_with_container/configs](shared_with_container/configs) directory.
 
 **Modifying Settings with Command-Line Arguments**:
 In addition to these two configuration methods, you have the flexibility to override specific settings directly via command-line arguments. By supplying `key-value` pairs in the CLI, you can adjust particular settings without explicitly altering the config files (YAML or NN Archive). For further details, refer to the [Examples](#examples) section.
@@ -259,21 +259,21 @@ Only the version `2022.3.0` of `OpenVino` is supported for `RVC3`. Follow the sa
 
 **RVC4**
 
-Requires `snpe-<version>.zip` archive to be present in `docker/extra_packages`. When building locally via the CLI, the tool will attempt to download the archive automatically if it is missing, but only when a **full SNPE build version** is provided (e.g. `2.32.6.250402`) and that version exists in the Qualcomm catalog. If you pass only the short version (e.g. `2.32.6`), the CLI expects either the image or archive to already be present. After the first download though you can use the local image with the short or long version.
+Requires `snpe-<version>.zip` archive to be present in `docker/extra_packages`. When building locally via the CLI, the tool will attempt to download the archive automatically if it is missing, but only when a **full SNPE build version** is provided (e.g. `2.41.0.251128`) and that version exists in the Qualcomm catalog. If you pass only the short version (e.g. `2.41.0`), the CLI expects either the image or archive to already be present. After the first download though you can use the local image with the short or long version.
 You can also download different SNPE versions manually from [here](https://softwarecenter.qualcomm.com/catalog/item/Qualcomm_AI_Runtime_Community). After downloading, rename the archive according to the version number and place it in the `docker/extra_packages` directory.
 
-Example (auto-download only on first build, afterwards available as `2.32.6.250402` or `2.32.6` ):
+Example (auto-download only on first build, afterwards available as `2.41.0.251128` or `2.41.0` ):
 
 ```bash
-modelconverter convert rvc4 --tool-version 2.32.6.250402 --path <config_or_archive>
+modelconverter convert rvc4 --tool-version 2.41.0.251128 --path <config_or_archive>
 ```
 
 Example (short version, assumes archive or image already present):
 
 ```bash
 # Place the archive ahead of time (or run the command above beforehand):
-# docker/extra_packages/snpe-2.32.6.zip
-modelconverter convert rvc4 --tool-version 2.32.6 --path <config_or_archive>
+# docker/extra_packages/snpe-2.41.0.zip
+modelconverter convert rvc4 --tool-version 2.41.0 --path <config_or_archive>
 ```
 
 **HAILO**
@@ -305,7 +305,7 @@ docker build -f docker/rvc2/Dockerfile \
 > When the `--build-arg` flag is not provided, the Dockerfiles (`docker/<target>/Dockerfile`) fall back to their built-in default tool versions:
 >
 > - RVC2: `2022.3.0`
-> - RVC4: `2.32.6`
+> - RVC4: `2.41.0`
 
 #### Local Image Tagging and CLI Usage
 
@@ -361,10 +361,10 @@ modelconverter convert rvc4 \
 Example letting the CLI construct the tag automatically:
 
 ```bash
-# Resolves to: luxonis/modelconverter-rvc4:2.32.6-latest
+# Resolves to: luxonis/modelconverter-rvc4:2.41.0-latest
 modelconverter convert rvc4 \
 --image luxonis/modelconverter-rvc4 \
---tool-version 2.32.6 \
+--tool-version 2.41.0 \
 --path <config_or_archive>
 ```
 
@@ -421,11 +421,11 @@ You can run the built image either manually using the `docker run` command or us
    export AWS_S3_ENDPOINT_URL=<your_aws_s3_endpoint_url>
    ```
 
-1. If `shared_with_container` directory doesn't exist on your host, create it.
+2. If `shared_with_container` directory doesn't exist on your host, create it.
 
-1. Without remote files, place the model, config, and calibration data in the respective directories (refer [Sharing Files](#sharing-files)).
+3. Without remote files, place the model, config, and calibration data in the respective directories (refer [Sharing Files](#sharing-files)).
 
-1. Execute the conversion:
+4. Execute the conversion:
 
 - If using the `modelconverter` CLI:
 
@@ -526,7 +526,7 @@ modelconverter convert rvc3 --path configs/resnet18.yaml \
 Specify all options via the command line without a config file:
 
 ```bash
-modelconverter convert rvc2 input_model models/yolov6n.onnx \
+modelconverter convert rvc2 --path models/yolov6n.onnx \
                         scale_values "[255,255,255]" \
                         inputs.0.encoding.from RGB \
                         inputs.0.encoding.to BGR \
@@ -542,7 +542,7 @@ modelconverter convert rvc2 input_model models/yolov6n.onnx \
 ## Multi-Stage Conversion
 
 The converter supports multi-stage conversion. This means conversion of multiple
-models where the output of one model is the input to another model. For mulit-stage
+models where the output of one model is the input to another model. For multi-stage
 conversion you must specify the `stages` section in the config file, see [defaults.yaml](shared_with_container/configs/defaults.yaml)
 and [multistage.yaml](shared_with_container/configs/multistage.yaml) for reference.
 
@@ -667,37 +667,52 @@ For usage instructions, see `modelconverter benchmark --help`.
 **Example:**
 
 ```bash
-modelconverter benchmark rvc4 --model-path <path_to_model.xml>
+modelconverter benchmark rvc4 --model-path <path_to_model.tar.xz>
 ```
 
 The command prints a table with the benchmark results to the console and
-optionally saves the results to a `.csv` file.
+kptionally saves the results to a `.csv` file.
 
-> [!NOTE]
-> For **RVC2** and **RVC4**: The `--model-path` can be a path to a local .blob file, an NN Archive file (.tar.xz), or a name of a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai). To access models from different teams in Luxonis HubAI, remember to update the HUBAI_API_KEY environment variable respectively.
+> [!IMPORTANT]
+> **Device Connection Requirements for RVC4**: The device must be connected and accessible either using the [Android Debug Bridge (ADB)](https://developer.android.com/tools/adb) or via SSH for the benchmarking to work in the following cases:
+>
+> - When `--device-monitor` is enabled (requires ADB or SSH connection to the device to calculate power consumption)
+> - When benchmarking is conducted using the SNPE tools (with `--no-dai-benchmark`; default is `--dai-benchmark`)
+>
+> The tool can find the correct device automatically but you can also specify it with the `--device-id` flag.
 
 > [!NOTE]
-> **Benchmark Duration Control (RVC2/RVC4)**: Two flags can affect the duration of benchmarking:
+> **Duration Control**:
+> There are two ways how to specify the duration of the benchmark:
+>
+> `depthai` backend (default):
+>
+> - `--benchmark-time`: The duration of the benchmark in seconds (default is 20 seconds).
 >
-> - `--benchmark-time`: Duration in seconds for time-based benchmarking (default: 20)
-> - `--repetitions`: Number of iterations to perform (default: 10)
+> `SNPE` backend (only **RVC4** with `--no-dai-benchmark`):
 >
-> By default, the benchmarking uses `--benchmark-time` (20 seconds) which takes precedence over `--repetitions`. To use `--repetitions` instead, you must explicitly set `--benchmark-time` to a negative value (e.g., `--benchmark-time -1`).
+> - `--num-images`: The number of images to run through the model (default is 500). The total benchmark time will depend on the model size and complexity, as well as the device performance. Too many images may lead to an out-of-memory error, while too few images may lead to less accurate results. Adjust this parameter based on your specific situation and requirements.
 
-> [!IMPORTANT]
-> **ADB Connection Requirements for RVC4**: The device must be connected and accessible using the [Android Debug Bridge (ADB)](https://developer.android.com/tools/adb) in the following cases:
+> [!NOTE]
+> Benchmark input support depends on the target and backend:
+>
+> - **RVC2**: `--model-path` can be a local `.blob`, an NN Archive (`.tar.xz`), or a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai).
+> - **RVC4** with `--dai-benchmark` (default): `--model-path` can be an NN Archive (`.tar.xz`) or a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai).
+> - **RVC4** with `--no-dai-benchmark`: `--model-path` can be an NN Archive (`.tar.xz`), a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai) or a `.dlc` file.
+>   In the case of a `.dlc` file, the tool `snpe-dlc-info` must be available in `PATH` either on the host machine or on the device.
 >
-> - When `--power-benchmark` is enabled (requires ADB connection to calculate power consumption)
-> - When `--dsp-benchmark` is enabled (requires ADB connection to calculate DSP utilization)
-> - When benchmarking is conducted using the SNPE tools (by setting `--dai-benchmark` to `False`, default is `True`)
+> To access models from different teams in Luxonis HubAI, remember to update the `HUBAI_API_KEY` environment variable accordingly when using a model slug as an input.
+
+> [!NOTE]
+> If you experience Segmentation Faults (core dumps) during benchmarking on RVC4 using the SNPE backend (`--no-dai-benchmark`), it may be due to the device running out of memory. Try decreasing the value of `--num-images` (the default is 500) to reduce RAM usage.
 
 ## [RVC4] DLC model analysis
 
 ModelConverter offers additional analysis tools for the RVC4 platform. The tools provide an in-depth look at the following:
 
 1. The outputs of all layers in comparison to the ground truth ONNX model,
-1. The cycle usage of each layer on an RVC4 device.
-1. Visualizations for fast and easy comparison of multiple models.
+2. The cycle usage of each layer on an RVC4 device.
+3. Visualizations for fast and easy comparison of multiple models.
 
 This gives the user better insight into the successful quantization of a model, helps discover potential speed bottleneck layers, and allows for the comparison of different quantization parameters.
 
@@ -710,18 +725,21 @@ pip install modelconv[analysis]
 There are several options to run the tools. The most general approach is:
 
 ```bash
-modelconverter analyze
-              <dlc_model>
-              <onnx_model>
-              <input_name_1> <path_to_input_images_1>
-              ...
-              <input_name_n> <path_to_input_images_n>
+modelconverter analyze \
+  --dlc-model-path <dlc_model> \
+  --onnx-model-path <onnx_model> \
+  --image-dirs <input_name_1> <path_to_input_images_1> \
+               ...
+               <input_name_n> <path_to_input_images_n>
 ```
 
 If the model accepts only one input, there is no need to specify the input name and the tools can simply be ran as:
 
 ```bash
-modelconverter analyze <dlc_model> <onnx_model> <path_to_input_images>
+modelconverter analyze \
+  --dlc-model-path <dlc_model> \
+  --onnx-model-path <onnx_model> \
+  --image-dirs <path_to_input_images>
 ```
 
 For other usage instructions run `modelconverter analyze --help`

{modelconv-0.5.4 → modelconv-0.5.6}/README.md

@@ -63,10 +63,10 @@ Run `modelconverter --help` to see the available commands and options.
 
 There are two main ways to execute configure the conversion process:
 
-1. **YAML Configuration File (Primary Method)**:
-   The primary way to configure the conversion is through a YAML configuration file. For reference, you can check [defaults.yaml](shared_with_container/configs/defaults.yaml) and other examples located in the [shared_with_container/configs](shared_with_container/configs) directory.
 1. **NN Archive**:
    Alternatively, you can use an [NN Archive](https://docs.luxonis.com/software-v3/ai-inference/nn-archive/#NN%20Archive) as input. An NN Archive includes a model in one of the supported formats—ONNX (.onnx), OpenVINO IR (.xml and .bin), or TensorFlow Lite (.tflite)—alongside a `config.json` file. The config.json file follows a specific configuration format as described under the `Configuration` section.
+2. **YAML Configuration File (Legacy)**:
+   An alternative way to configure the conversion is through a YAML configuration file. For reference, you can check [defaults.yaml](shared_with_container/configs/defaults.yaml) and other examples located in the [shared_with_container/configs](shared_with_container/configs) directory.
 
 **Modifying Settings with Command-Line Arguments**:
 In addition to these two configuration methods, you have the flexibility to override specific settings directly via command-line arguments. By supplying `key-value` pairs in the CLI, you can adjust particular settings without explicitly altering the config files (YAML or NN Archive). For further details, refer to the [Examples](#examples) section.
@@ -197,21 +197,21 @@ Only the version `2022.3.0` of `OpenVino` is supported for `RVC3`. Follow the sa
 
 **RVC4**
 
-Requires `snpe-<version>.zip` archive to be present in `docker/extra_packages`. When building locally via the CLI, the tool will attempt to download the archive automatically if it is missing, but only when a **full SNPE build version** is provided (e.g. `2.32.6.250402`) and that version exists in the Qualcomm catalog. If you pass only the short version (e.g. `2.32.6`), the CLI expects either the image or archive to already be present. After the first download though you can use the local image with the short or long version.
+Requires `snpe-<version>.zip` archive to be present in `docker/extra_packages`. When building locally via the CLI, the tool will attempt to download the archive automatically if it is missing, but only when a **full SNPE build version** is provided (e.g. `2.41.0.251128`) and that version exists in the Qualcomm catalog. If you pass only the short version (e.g. `2.41.0`), the CLI expects either the image or archive to already be present. After the first download though you can use the local image with the short or long version.
 You can also download different SNPE versions manually from [here](https://softwarecenter.qualcomm.com/catalog/item/Qualcomm_AI_Runtime_Community). After downloading, rename the archive according to the version number and place it in the `docker/extra_packages` directory.
 
-Example (auto-download only on first build, afterwards available as `2.32.6.250402` or `2.32.6` ):
+Example (auto-download only on first build, afterwards available as `2.41.0.251128` or `2.41.0` ):
 
 ```bash
-modelconverter convert rvc4 --tool-version 2.32.6.250402 --path <config_or_archive>
+modelconverter convert rvc4 --tool-version 2.41.0.251128 --path <config_or_archive>
 ```
 
 Example (short version, assumes archive or image already present):
 
 ```bash
 # Place the archive ahead of time (or run the command above beforehand):
-# docker/extra_packages/snpe-2.32.6.zip
-modelconverter convert rvc4 --tool-version 2.32.6 --path <config_or_archive>
+# docker/extra_packages/snpe-2.41.0.zip
+modelconverter convert rvc4 --tool-version 2.41.0 --path <config_or_archive>
 ```
 
 **HAILO**
@@ -243,7 +243,7 @@ docker build -f docker/rvc2/Dockerfile \
 > When the `--build-arg` flag is not provided, the Dockerfiles (`docker/<target>/Dockerfile`) fall back to their built-in default tool versions:
 >
 > - RVC2: `2022.3.0`
-> - RVC4: `2.32.6`
+> - RVC4: `2.41.0`
 
 #### Local Image Tagging and CLI Usage
 
@@ -299,10 +299,10 @@ modelconverter convert rvc4 \
 Example letting the CLI construct the tag automatically:
 
 ```bash
-# Resolves to: luxonis/modelconverter-rvc4:2.32.6-latest
+# Resolves to: luxonis/modelconverter-rvc4:2.41.0-latest
 modelconverter convert rvc4 \
 --image luxonis/modelconverter-rvc4 \
---tool-version 2.32.6 \
+--tool-version 2.41.0 \
 --path <config_or_archive>
 ```
 
@@ -359,11 +359,11 @@ You can run the built image either manually using the `docker run` command or us
    export AWS_S3_ENDPOINT_URL=<your_aws_s3_endpoint_url>
    ```
 
-1. If `shared_with_container` directory doesn't exist on your host, create it.
+2. If `shared_with_container` directory doesn't exist on your host, create it.
 
-1. Without remote files, place the model, config, and calibration data in the respective directories (refer [Sharing Files](#sharing-files)).
+3. Without remote files, place the model, config, and calibration data in the respective directories (refer [Sharing Files](#sharing-files)).
 
-1. Execute the conversion:
+4. Execute the conversion:
 
 - If using the `modelconverter` CLI:
 
@@ -464,7 +464,7 @@ modelconverter convert rvc3 --path configs/resnet18.yaml \
 Specify all options via the command line without a config file:
 
 ```bash
-modelconverter convert rvc2 input_model models/yolov6n.onnx \
+modelconverter convert rvc2 --path models/yolov6n.onnx \
                         scale_values "[255,255,255]" \
                         inputs.0.encoding.from RGB \
                         inputs.0.encoding.to BGR \
@@ -480,7 +480,7 @@ modelconverter convert rvc2 input_model models/yolov6n.onnx \
 ## Multi-Stage Conversion
 
 The converter supports multi-stage conversion. This means conversion of multiple
-models where the output of one model is the input to another model. For mulit-stage
+models where the output of one model is the input to another model. For multi-stage
 conversion you must specify the `stages` section in the config file, see [defaults.yaml](shared_with_container/configs/defaults.yaml)
 and [multistage.yaml](shared_with_container/configs/multistage.yaml) for reference.
 
@@ -605,37 +605,52 @@ For usage instructions, see `modelconverter benchmark --help`.
 **Example:**
 
 ```bash
-modelconverter benchmark rvc4 --model-path <path_to_model.xml>
+modelconverter benchmark rvc4 --model-path <path_to_model.tar.xz>
 ```
 
 The command prints a table with the benchmark results to the console and
-optionally saves the results to a `.csv` file.
+kptionally saves the results to a `.csv` file.
 
-> [!NOTE]
-> For **RVC2** and **RVC4**: The `--model-path` can be a path to a local .blob file, an NN Archive file (.tar.xz), or a name of a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai). To access models from different teams in Luxonis HubAI, remember to update the HUBAI_API_KEY environment variable respectively.
+> [!IMPORTANT]
+> **Device Connection Requirements for RVC4**: The device must be connected and accessible either using the [Android Debug Bridge (ADB)](https://developer.android.com/tools/adb) or via SSH for the benchmarking to work in the following cases:
+>
+> - When `--device-monitor` is enabled (requires ADB or SSH connection to the device to calculate power consumption)
+> - When benchmarking is conducted using the SNPE tools (with `--no-dai-benchmark`; default is `--dai-benchmark`)
+>
+> The tool can find the correct device automatically but you can also specify it with the `--device-id` flag.
 
 > [!NOTE]
-> **Benchmark Duration Control (RVC2/RVC4)**: Two flags can affect the duration of benchmarking:
+> **Duration Control**:
+> There are two ways how to specify the duration of the benchmark:
+>
+> `depthai` backend (default):
+>
+> - `--benchmark-time`: The duration of the benchmark in seconds (default is 20 seconds).
 >
-> - `--benchmark-time`: Duration in seconds for time-based benchmarking (default: 20)
-> - `--repetitions`: Number of iterations to perform (default: 10)
+> `SNPE` backend (only **RVC4** with `--no-dai-benchmark`):
 >
-> By default, the benchmarking uses `--benchmark-time` (20 seconds) which takes precedence over `--repetitions`. To use `--repetitions` instead, you must explicitly set `--benchmark-time` to a negative value (e.g., `--benchmark-time -1`).
+> - `--num-images`: The number of images to run through the model (default is 500). The total benchmark time will depend on the model size and complexity, as well as the device performance. Too many images may lead to an out-of-memory error, while too few images may lead to less accurate results. Adjust this parameter based on your specific situation and requirements.
 
-> [!IMPORTANT]
-> **ADB Connection Requirements for RVC4**: The device must be connected and accessible using the [Android Debug Bridge (ADB)](https://developer.android.com/tools/adb) in the following cases:
+> [!NOTE]
+> Benchmark input support depends on the target and backend:
+>
+> - **RVC2**: `--model-path` can be a local `.blob`, an NN Archive (`.tar.xz`), or a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai).
+> - **RVC4** with `--dai-benchmark` (default): `--model-path` can be an NN Archive (`.tar.xz`) or a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai).
+> - **RVC4** with `--no-dai-benchmark`: `--model-path` can be an NN Archive (`.tar.xz`), a model slug from [Luxonis HubAI](https://hub.luxonis.com/ai) or a `.dlc` file.
+>   In the case of a `.dlc` file, the tool `snpe-dlc-info` must be available in `PATH` either on the host machine or on the device.
 >
-> - When `--power-benchmark` is enabled (requires ADB connection to calculate power consumption)
-> - When `--dsp-benchmark` is enabled (requires ADB connection to calculate DSP utilization)
-> - When benchmarking is conducted using the SNPE tools (by setting `--dai-benchmark` to `False`, default is `True`)
+> To access models from different teams in Luxonis HubAI, remember to update the `HUBAI_API_KEY` environment variable accordingly when using a model slug as an input.
+
+> [!NOTE]
+> If you experience Segmentation Faults (core dumps) during benchmarking on RVC4 using the SNPE backend (`--no-dai-benchmark`), it may be due to the device running out of memory. Try decreasing the value of `--num-images` (the default is 500) to reduce RAM usage.
 
 ## [RVC4] DLC model analysis
 
 ModelConverter offers additional analysis tools for the RVC4 platform. The tools provide an in-depth look at the following:
 
 1. The outputs of all layers in comparison to the ground truth ONNX model,
-1. The cycle usage of each layer on an RVC4 device.
-1. Visualizations for fast and easy comparison of multiple models.
+2. The cycle usage of each layer on an RVC4 device.
+3. Visualizations for fast and easy comparison of multiple models.
 
 This gives the user better insight into the successful quantization of a model, helps discover potential speed bottleneck layers, and allows for the comparison of different quantization parameters.
 
@@ -648,18 +663,21 @@ pip install modelconv[analysis]
 There are several options to run the tools. The most general approach is:
 
 ```bash
-modelconverter analyze
-              <dlc_model>
-              <onnx_model>
-              <input_name_1> <path_to_input_images_1>
-              ...
-              <input_name_n> <path_to_input_images_n>
+modelconverter analyze \
+  --dlc-model-path <dlc_model> \
+  --onnx-model-path <onnx_model> \
+  --image-dirs <input_name_1> <path_to_input_images_1> \
+               ...
+               <input_name_n> <path_to_input_images_n>
 ```
 
 If the model accepts only one input, there is no need to specify the input name and the tools can simply be ran as:
 
 ```bash
-modelconverter analyze <dlc_model> <onnx_model> <path_to_input_images>
+modelconverter analyze \
+  --dlc-model-path <dlc_model> \
+  --onnx-model-path <onnx_model> \
+  --image-dirs <path_to_input_images>
 ```
 
 For other usage instructions run `modelconverter analyze --help`