# Qairt Converter

The `qairt-converter` tool converts a model from one of ONNX/TensorFlow/TFLite/PyTorch framework to a DLC.
The DLC contains the model in a Qualcomm graph format to support inference on Qualcomm’s AI accelerator cores.
A new prefix `qairt` for `Qualcomm AI Runtime` signifies that this converter can be used with both the `Qualcomm Neural Processing SDK` API as well as the `Qualcomm AI Engine Direct`
API. The converter automatically detects the proper framework based on the source model extension.

Supported frameworks and file types are:

| Framework | File Type |
| --- | --- |
| Onnx | [\*](https://docs.qualcomm.com/doc/80-63442-10/topic/qairt_converter.html#id1).onnx |
| TensorFlow | [\*](https://docs.qualcomm.com/doc/80-63442-10/topic/qairt_converter.html#id3).pb |
| TFLite | [\*](https://docs.qualcomm.com/doc/80-63442-10/topic/qairt_converter.html#id5).tflite |
| PyTorch | [\*](https://docs.qualcomm.com/doc/80-63442-10/topic/qairt_converter.html#id7).pt |

## Basic Conversion

Basic conversion has only one required argument `--input_network`, which is the path to the source framework model.
The source model can either be float model or quantized model, `qairt-converter` will convert it to corresponding DLC, retaining the precision and datatype of the tensors.
Some frameworks may require additional arguments that are otherwise listed as optional. Please check the help text at general/tools:qairt-converter for more details.

> 
> 
> ![../images/qairt_basic_conversion.png](data:image/png;base64,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)

- Onnx Conversion

Current ONNX Conversion supports upto ONNX Opset 24.

> 
> 
> $ qairt-converter --input_network model.onnx
>     Copy to clipboard

- Tensorflow Conversion

Tensorflow additionally requires `--source_model_input_shape` and `--out_tensor_node` arguments.
`--source_model_input_shape` is for specifying the list of all the input names and dimensions to the network model.
`--out_tensor_node` is for specifying the network model’s output tensor name/s.

> 
> 
> $ qairt-converter \
>           --input_network inception_v3_2016_08_28_frozen.pb \
>           --source_model_input_shape input 1,299,299,3 \
>           --out_tensor_node InceptionV3/Predictions/Reshape_1
>     Copy to clipboard

In the above example, the model `inception_v3_2016_08_28_frozen.pb` has input named `input` with dimensions (1,299,299,3), and output tensor with
name `InceptionV3/Predictions/Reshape_1`.

## Input/Output Layouts

The default input and output layouts in the converted graph are the same as per the source model. This behavior differs
from the legacy converter which would modify the input and (optionally) the output layout to the spatial first format.
An example single layer Onnx model (spatial last) is shown below.

> 
> 
> ![../images/qairt-conversion-layout-comparison.png](data:image/png;base64,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)

## Input/Output Customization using YAML

> 
> 
> Note
> 
> 
> This feature allows user to specify their desired input/output tensor layout for the converted model.

Users can provide a YAML configuration file to simplify using different input and output configurations using the `--config` command-line option.
All configurations in the YAML are optional. If an option is provided in the YAML configuration and an equivalent
option is provided on the command line, the command line option takes precedence over the one provided in the configuration file.
The YAML configuration schema is shown below.

> 
> 
> 

- `Name:` Name of the input or output tensor present in the model that needs to be customized
- `Src Model Parameters`

    These are mandatory if a certain equivalent desired configuration is specified.

    - `DataType:` Data type of the tensor in source model.
    - `Layout:` Tensor layout in the source model. Valid values are:

        - NCDHW
        - NDHWC
        - NCHW
        - NHWC
        - NFC
        - NCF
        - NTF
        - TNF
        - NF
        - NC
        - F

        where

        - N = Batch
        - C = Channels
        - D = Depth
        - H = Height
        - W = Width
        - F = Feature
        - T = Time
- `Desired Model Parameters`

    - `DataType:` Desired data type of the tensor in the converted model. Valid values are float32, float16, uint8, int8 datatypes.
    - `Layout:` Desired tensor layout of the converted model. Same valid values as source layout.
    - `Shape:` Tensor shape/dimension in the converted model. Valid values are comma separated dimension values, i.e., (a,b,c,d).
    - `Color Conversion:` Tensor color encoding in the converted model. Valid values are BGR, RGB, RGBA, ARGB32, NV21, and NV12.
    - `QuantParams:` Required when the desired model data type is a quantized data type. Has two subfields: Scale and Offset.

        - `Scale:` Scale of the buffer as a float value.
        - `Offset:` Offset value as an integer.
    - `Optional:` During calls to graph execute, the client can use optional I/O tensors to signal to the backend which tensors to be optionally provided/produced. Valid values are True, False.

The `--dump_config_template` option of `qairt-converter` saves the IO configuration file for the user to update.
Pass the `--dump_config_template` option to the `qairt-converter` to save the IO configuration file at the specified location.

> 
> 
> ![../images/qairt_io_config.png](data:image/png;base64,UklGRuwRAABXRUJQVlA4TN8RAAAvbINIABXhYbZ/chtJChk6VKhQIUOGDBUqZMhQoUKFChUyVKhQoUKHChUym/vaO3S42WKrfv9/lagSi83aNTfS3KOA0wIMuAXY1ag5NBD6EmyQIGD3by8X9zIjbx9Cc28BfQmwiOmLkHYEQ23BHgJGq/ei9xAcsVCJXUAf2oNeAnaDe7ihOSzQarjJlgH2HlVzuRRZUiRJkiTJoMOACWPu/P9rZmavhAljb4N0JElyJElKxuL9/8v2cB520kkn+08JkiS5jRQJwgp9ALlHqXE0KL1/CdKaIv4T8Z+I/0Qq5hfQjTa0oBkONGC4qglueNGJDyKMTj9Gd3EtXsdc55zvXs+93nPDVd33uu79nVdWej5czih6xn2ffL9hr93n2uYakQ6P40KzmOijXz8Pe+l6/7lW795P4i4Yv4ipAezGeh+UXvSgPO7DbBfGxkqzs+Pl8nOKsj4/OZUIRdil1sFGH/v8NOzFBsdwMiS5ivdjniOLSrPj4+XnnlPW1yen9mD4pnZFKY+NtmLH3gVEdCCv2k7ncm7XwlqZh3lZuz1zYTvZ4tzs8t22mncwRMw3Q/OgHaXy+lTogOYdRBQUgigmQgam1sulNU+4w8lUZLSLNTK7OXOGK3PXWmqtNW+1rdpW89Y8tdasJXNnCytN0Rh5BxDREfbOp0fHxpUBDJ90r82Pr5238YKbJb32TKfxmR5f37z+ef758UHfJ328lRnw7Tbf65nnm9w+82Crs2y88OkOlCdz4k/J5ekgChKB/tsehQDng5XnOllrrDhwduGGJzvQvtq56lvtWelb6e/n7/fBfrzl3+nf2b3Su9K90rXasdq25k1fdmfN0RCqNupDPz4JgbdV1i7n8054bIftMf0+ma6f+fMrb8lUfXJ+XOw27bTVXrBUrRPrCTGn7BT7CSaLSzPifjN6NmljUzx5oKtybqBjUvXB53/4i25ud6d7de7kwtm8M0bLwmy6B7NJF1PPx3uNfx6vvP7pP7/yFq+MH5hXOON2+WaWfk60yfadCzZsOjAl2nVrybMrP7zW1/Hv8EnsDzz/w198En5gp7fTWb2Rz7egCx+EwFAeUb30+NPUvtoQiNq1z3SZpFg/X5p3if0El6VFY6LMADqZutBRvQXOU8nEDryoXScO3mhAyoeacOhyeP74B+eRbbUhFtn0nU7rYs3uORGmTZ33gg/d92wx/knrjPp/9XbYOrC/IRbZ1N+55s4a0MaEsh8XO99nYIkmGMGCOxfanfjeBfj2wPnOEGwWe110GXdgjqwTimDVxMvNMSUI7C5LLRFYgdd6ndiy1CjmE3Rwe+NiS3fc6N8x2TqhiQ3eawxB7KPJyWSJ4Ark/Q6Ib00w4uEZsWWle8WEJUIMbrQzAdCCpCzQj8aEalls9/rNYETXvZtiiyP6OwwCXcw5WwDMPLTbx24LMpJ9kyY2L7J0ov0KRvSec0tsiZ9kGwkwZjcb4GZC+D/TeqjD/hhUC67+nW27Ho4DFxZ6uqCl4EpMXwV/6IYQ+NLT150u/jSxL0G0hb375rXZ/E3ZxcJymO1VvPIS+WRPMHsxN95zZwvnerrCQJb7fDexxR1/h5ixAnc3xE/TL3bQy03u9tNyuA5zYm6hKdqq58CclZEh7EqdwdK/3rV9DnjSw6YpFMDTODdzoIldv91naLMwA5bu3PPEPjg/Cl9ys0ZnJ8ETw3AdbGJObNYd4Sze6ksbtEligUp3HhM0+zrvtaZT+ScPHmXMCwYiRZfvXHI6tO9pN8f7TIbP5OdATIJePrF66VkVXazD3MlAv2E7yBTt7Rx6Mp93xq8f+voEj8nPAZkDffTI757e9xUHpq50rUxAwwKCMttEzhR99EvXnc5qZG77YIddvXOiuP835UHL7xuYt3ter+/O6jGPXQGvKfT4KPM6NU0L30G7xDM3bs2dXNgVDnxl3+m5a18/NzHQmTUHq6MbfZ1bLx28enJhCes3Nq8dOUqbbiYsNE2jDpD7T3TSm0m3recRMW+vZO9ppduXzhy7vXr8hT//8xeOr94+dubS9h/ek12xWWNnnc7NPPbvol9+izoCa5oW7uNtsgpooHPuHzavuLOVpYcQ321iNfZ11R9unj508NpLX7912LdycO5qW81bPdlavTHpzKYqDyHmpxb23de+dpgchclzXkHB49JPJvTJefuLheixMzdvPrjtscnr593McjG0Wuz568lTD1662UbP3D5+4iLIbe7XFMJ98HtpgMj2/7t7ee4Lmwe81ZNXruy78eTkwoKrNvnkvleveKon7zuwefrg4d7uBmXIpV5cCo7B3G6RAyDOWzhhQQZzu8UMgLhv4QiUQbyXOnjRPIcRGdRNzsGL5lnQDOmkzLAkQzonM+I/e8UHfFmsZEYVzrmUyKjKBUQHiuQMZ3Z85ZijSfMhBVGViw4UZP+8ZIZ8q0CWZY63D2g+qtIQahF+z4KK6GIZeVktyHKHxcwyrjI2WDMFWU4iqstyVQoi42LqMvjE5QwmaaiYkTMYZ0fbOGbkZFJqgX/LJFtcYBJZLqjLFOpFWS7W2fzPYAFaqIZQWJYVrujGBHEHinGWYp2LgrosV2W2E6IaUmHpum6FKVYgjQuDXRENQyYsvXVOHTinW8IZXC3LuutYHI+D5GaOyBaD8qJFsKyC4iRmQjOEMLjSnIdiEN+8AAxQjQz/JuD5nly0EJNylT2AugoAJshu6GIZsd/wwI9BYEGRDmBekn/8T985B7rvMR44HHIYIBzQbHB70qMOML+on4dnUWa9EkhGKN8pxLajGRkZ5htcz/AIH6QJiqdQYPVIA7yo8MXSoSwLOrAAhBuATg6vXAZ1i+aZZpjbE0DnYdUSZTBdCRgj5mFktLRhmhtLh80RxsgRBxAXjpojw7sJx+wU0geHCYo30QcxXeGFpXfcQ1Q3VTaXY4g4rTNOmI0tNRpDl34bGWX957tuAYeLWUQHNlmDhBs9Ftt0gLWymAU2ph3YJMyxJrL3OYB/lkVGMV3ftChWo8CXKINMEDjYuisOHIEwPhgL5tWr5kiwfNsgBAD8P8YNTMwoYlRfzDJV0xjVCXCaY2XBTN50e4uLWcYXaKfH2MSNqVvAPTC2pWYXCfeobpErPYrTVKuivUIQBJAR88oVCNGV3Qj2rPgwz60xMVIKodKjHFj6lsqOoVFkc1kHNiAgWYli+lUfTFvCEWjLOmD9AgPAs76YjemgKXQLjFCsMq65i+0xHpoXIZUQiSBggH6QI1/ZYHcz8nbQ582UG9wSNbRLHMAsJvvpOkMMAJOfakiHGW1ZFjwAdBjLLrJMQwdRdQs4AS5Y9yTgZwDRLwbOhgOAJkJKIaFHAsiRrwAIG3ehXlp9BEaKMcZJFEnAVR4iHbqhGJGeDGCsXczSAbkcixHeETkA9UA0xHsDH1k3ArqiEPRW1SrVQggAiNgm/FfghHUoV1tqlNmCToC4V8T9UlGHiIJl8SA6kwfgMamFakH8ARELz+JpjIIOnHAVT3MX04B3N9UtSmrrFgtXsSX6gCiw3a4S3RhdDRLdqrIsV8kD0P3Q3jCoyvGkXBV9YGTw6RYjuoFcXcwiF1tqjOwGh9zNZm0RQ292kYOomo3ppNHYtE4Ax1viJ8i9m7qow0SBfr63cNRkgA/6rmJQeL4nJzEuV4kDUF8KZyF8hAYYBh2upQGYvY8dHqOI+HCWDriWxOwiZUGE4yMixpjhkc6WCu56QlPTcAWcT1Oe71mb4r76JIwhmblCGggWM1cyLlaHLxXU1bhMn8jCQPgIHRjMzBX6JSFxmkkZR3UeM1coB3jNotiEpiyaudc4nFtcdkKoOZkUA+SczOHPGSyAPE7GZTmD4C0euYDoYhNJOFUzLoc4vGZxXRJalKmVpCuLNErxRTFD2iRM0c1xOOeyE8q92hMEiMcxIy/XiwwZTMpyEjPLaoEpgC+MF9ViqMNrrxEMozfE79qxXAeviDMU68sym9UFzIANeIB46ANxThkmQU4mM1iQGQoqw3K9GCeIN4BV6ANxThk2oVhfVZdl/itZhoQ8XzQUPiGDDYjLLMtqHBQXHYDFRasgQSEXHCiykF9bCbol40Q3LEhEkBNWYDvyZXBeB2C2pKPfnkO0IwxUeR2gWpXYfi9cxH8i/hNRjEfWpTUa8Mi6pAYygZQGCKQYnhvKX8L4SFp6wY2h5NNSDOtDtvHpeWPcMxeM8F0ehZ2BkWEo7hGM+Ne1EJKXhMAHkX9LrgoyqFblE7YhkoCIodKkYESLNyu2bL9hjowI9GPuCkUQf4XShROyXBVaQZWJ5FUHxkQSNgLsAQYjnFLEFiV/2jRHhFYwwsg8tXLJEITK6LVvyaCgKpA6UCJfrP/FAIwgzVDKI98RbBZL/7d8Qmwxpt53RnfXBAUjwikxTdO/4+vmBXNqsgefxtRqv5GJquqQtiXq5G/Fz+KaYYgvyjeCD61aM0RYY3kHDtwyQdUQd2Vsw5LdnYOvqr/mhyEglTRM3byTkYmmgT5KlRSwvf8Vj63nt58zDDHGSG0Gl8XSN3HJEGXNj9qNceNQX8c0YdeAH4U0agJ9/YAzUBvfA8EN+0oHXKwBl790/Hv7KX9TOR+BlOYhmdQTf77qZXUHHhlbzxmGSFN2Ktjg0owh2pqcvbQrll6eO9zXOWqaxFH4iM1oPjIJDfTr7fhW3nBnDqTWykL9QZlTZrZTect2M6+9fedb+38s8xBIaR7KvPNlHz12ysWaWAPiz6bH5sHFiWhjdCgWTBZL/+1PG+IuP2ZH37fx7Hdebav+w93/N9Ax+YjNaB7y72z0dt+c81af/IoD+dRvzCi/AEPoylUUH9aSKRXzzc7FrnvpsVcK3+szxYtPP7HfP++tzI9l+ceZt5b3P/H0xeLI9wqvHHvwOtrvWvbO9sz4+pIBJe4Y2sO/vASJwLL0aAMSRiigqfnyzPaO7YAV/86Tf/mF9166+9y/Z/u6ExsbpwZ2jvp3TH/n6Kn+nY2JieyRt8/duvbGfa8u1PN4tjnOj3ZBqeSMYaZcZf45H9a8rsXtXE41o26rav5dC9F6N6+qdv1CysV2HknPzD43v5QzCIUGGNZv60wQFCK9c64BAzBCCuWmJpXxTox+2s12Uit1O67m82wWOSK/KxrDbopayhk773tirVRe9+MXYAQ7hRgYO83qPccGlmUN4zo26j/3zXs/CSPsqR0BYIziT+6Df/zucJXOaPc571yj0oYRFseYH/12izqv6OkPvp8+XHWo59znW4XpeSNcjmHUxkcfueAADlfZ3/Bq5oMwjHApIqGI/0T8J+I/e+8iFfGfiP9E/CfiPxIrCcmICgZE6wlJDSwnJDVwLCGpgWlpDfx0QjLi04OwtUZbdmqixIWwHHsQENBeksZ4ZIcIBpCQwphPaMSiJKQwDMoynpDCMHLksiaFQV08kMSgVe7kJC84FrsmfUGvxEpC2oJ7WU9IWvBYxhMSF1wdcEbSgk94ScqCV7gjVVGuUDWpInncnAT1as/gFKmMWY23LLHADi+S4y3DlnYIuBISELkKTTXRZD58RxlpKv0qMypjNE2KJ2HCJhHTqIwFRDkJaUcB0ZR0x5gIKPy5ITKMGZJaE2PvkUhqz/cMI9z5fE9i41d0bETIeb4Xcez5Xq0SECUktiZ7QSQXIElGrGsBUS0y2/O9iP/s3RktkQu4DCkHTGmBlyXVMGTauyTSmiL+E/GfcAlptGtSDxVLCZxDuwJcSmSqpEAs7VBLVQyjhIgK1YuNqFDclEppRAW4QLQqwBU4kc1pJdajYqQRtZyR+7Rio72koVWRNijZNUOxKhCYvLV0jnRTQoV1xq6Ay3SJHFxLWs5Q7FqpBKzmNLuW01Ax0lpOyiCnKWB45DpPpbgppUEPFsIlHIahmzS0m04Do8CkYtekDGo22yxNQbFZCDeQWqpCXVHSOF1ijeQ0RORCslnxIKfRLiwV7Tktx7SQVFZGWsvRi2s2KCbccFPTarAO3qOBPWq2hIKRRo5blxULUSHd8NhQ4BEVizVsVYBN69NcSEiP+ewKu++cQXaWskSMlyVITitJOXMugQY/SJaAN8roKq0p4j+hBaV0xH8i1iKtKeI/Ef+J+E/kLAwA)
> 
> 
> $ qairt-converter \
>           --input_network model.onnx \
>           --dump_config_template <output_folder>/io_config.yaml
>     Copy to clipboard

This is the sample output of the dumped IO configuration file:

> 
> 
> Converted Graph:
>     - Input Tensors:
>     - Output Tensors:
>     
>     Input Tensor Configuration:
>       # Input 1
>       - Name: 'input'
>         Src Model Parameters:
>             DataType:
>             Layout:
>             Shape:
>         Desired Model Parameters:
>             DataType:
>             Layout:
>             Color Conversion:
>             QuantParams:
>               Scale:
>               Offset:
>     
>     Output Tensor Configuration:
>       # Output 1
>       - Name: 'output'
>         Src Model Parameters:
>             DataType:
>             Layout:
>         Desired Model Parameters:
>             DataType:
>             Layout:
>             QuantParams:
>               Scale:
>               Offset:
>     Copy to clipboard

Consider a model with the Source model I/O and Desired model I/O configuration as shown in the table below:

> 
> 
> | Input/Output Name | Source Model I/O | Desired Model I/O |
> | --- | --- | --- |
> |  | Datatype / Layout | Datatype / Layout |
> | ‘input\_0’ | float32 / NCHW | uint8 / NHWC |
> | ‘output\_0’ | float32 / NCHW | uint8 / NHWC |

Here is an example io\_config.yaml, where:
Input and output tensor layouts are converted from `NCHW` format in source model to `NHWC` format in the converted model.
Also, the datatypes are converted from `float32` format in source model to `uint8` format in the converted model.

> 
> 
> Converted Graph:
>     - Output Tensors: ['output']
>     
>     Input Tensor Configuration:
>       # Input 1
>       - Name: 'input'
>         Src Model Parameters:
>             DataType: float32
>             Layout: NCHW
>             Shape:
>         Desired Model Parameters:
>             DataType: uint8
>             Layout: NHWC
>             Color Conversion:
>             QuantParams:
>               Scale:
>               Offset:
>     
>     Output Tensor Configuration:
>       # Output 1
>       - Name: 'output'
>         Src Model Parameters:
>             DataType: float32
>             Layout: NCHW
>         Desired Model Parameters:
>             DataType: uint8
>             Layout: NHWC
>             QuantParams:
>               Scale:
>               Offset:
>     Copy to clipboard
> 
> 
> $ qairt-converter \
>           --input_network model.onnx \
>           --config io_config.yaml
>     Copy to clipboard

## Disconnected Input Preservation

In deep learning framework models, computational graphs often contain multiple inputs. During graph optimization,
unused inputs may be removed through techniques like constant folding or dead code elimination. While this improves
performance and reduces memory usage, it can sometimes interfere with workflows that rely on the presence of all
inputs being present in the source framework model — especially in scenarios involving inference.
To address this, qairt-converter retains all the source framework model inputs, ensuring that all graph inputs remain
part of the graph regardless of their usage. This behavior is similar to the other open-source inference engines.
Unused graph input nodes can be removed by using `--remove_unused_inputs` command line argument while using
qairt-converter.

- Retaining unused or disconnected inputs provides the following benefits:
    - - Avoid unintended side effects during model conversion.
- Facilitate debugging and analysis by retaining all original inputs.

The following figure shows a source graph (left) with two inputs `i1` and `i2`. The input `i2` is disconnected post conversion, but it is preserved in the converter graph.

> 
> 
> ![../images/qairt_disconnected_input_nodes.png](data:image/png;base64,UklGRt4uAABXRUJQVlA4TNEuAAAv9wFoAFXpnfZ/lWTJSbPNa65ZZpttttlmmWWmuMw2U15ZijTTTDPNNNNMM80000wzzTTTTGsjvYo5PzgnT+b5nTyi22KWFktYFZEtXPwD7vYYE+0s91iinCoxLrXAGqaOWN61bkcZFSuLWUPpDfwBU1sxcaJkMcPwfDduRItx4LhiJqvEzAzTYlpsMUsd0f7y9hVaauGiJWoxt9ZgKzeiInrEjMsMP18MHQOWWFoaSJHHDPfeLjEz40/MrLrRU2JmVguHxAzDM2K46S3vETNYLNVVjHdvyGW6YmaWKkLLLKaMrthZHrTEUkWJmRnSEePIY2ZmBmvG7ynRUosZR+WF1vY8kqRtO5nF4mJ5sRQrsRYrQUFDQ8OAAQ0NAwZrQ0NDQ/e/oKCaqZmaqZiKKdjV80sF295abvTvdbTXEYtZxCpWsRoaGh54nqLAwsDCwIWFCwsDBxoauszQ0GEOc5jC+g8JkuS2zRwYxbCmAYICTrz/2f//J8Num/vP5hpx9v/7WallUjE3SnMpQKmpkW2UKwZuiAYA36iNS6w7Dkc45W3GKy0J2aBCtuEwX5HiRUua7hZjAm6EhvlLTNN0/PpZ3kzTXJ52tuGwSbHiRfOBG5+hG8syjt8+yxvKOIvSTByWpKHYPAeDkNLONNel1KrmprTgMD8NxXhWNTjDJpRp0sg23cDTIU3Fm4wT1aoiziENOHRTqFbxaoXOEsPDEoKqKQmoIpNFjeUq9mSCPVrK3aKMDYtwqqYjfJ3xMrNqIqtSWZ4W50DJqTOe2RFP0zTr9i+Ni9WdUdJsxAhxasaLo56mIzw2TTmqJTWsl2pWVaZqOkJtfiPkl+kI/4W1NQO2HXbuSKdZNWA4H6dnDqsd4XsCd6kBCeJmvGSC/+j11vAiM6BTatiulUy3P5r/gfc+DRkznodztdWO8M3nQ90V/ZwyQ+PDpm9ixtSHyNdlmjedxDXspiEjzDF5m45ymqaZGa53nX8qpmlWzbr9/5RcAtM0bFQziLTbuZ0m8bHqQjTimOcW0rGnYSNpN8ONeW5+m1KDN61uVo04fpOa8f5UTdM04AsefzT/8+6b6aqu6qrOttqsGm22tpFt2JexgsggWMdAZtqHsSNLZrwWiByirDdvkwoQWquY0TfRn4Er913VD3cP9ZifpuJFugMHffFd5bA8TeZn0N9jid7AYb5zjYtMJbZoedgE4Iy+jNJiejeLdG9R+unnw8F1N8x1VZnzatNsrE6bzKrSbFOVpmnSvcis6uWUw8xHNKsK/z1oxVNidVqkrACp4ozLYdMipYr1x52Tg99LmnVp0AZmzyqqMAnnezVc/hXaJtCd6fRXTpMZr8zsA8D8Tctxa1hi/RKdSyQUnEM3AFdF8TlsWi5rmyTJXG+GOBGlPYT8ki2XkgBw1v8wLwJyAwPUn+Vl7id9OnSusGFFrqt0oszAGd8Wgkj9Byq3g+jMz3BVaCKpmne4d4rrKpWIZtaXMJysyC16RpQ+XpojnxTXVWONa8ruO5Guq0abhiU1PJV0XTXabB37skjXVeMN6bpqvKG5rhrtBdJ11XhDc1012lBdVw02FNdVQ811VWVcV40010V3XTXC3BdEeq+dMqosS4KmXKfj17Oa1rDZeQJvyYHQJW/n2zSwRDWb1ary4d1/wo9jWGnhEHBimlXFvqufMGgFeItRpYXD+/6AplJbgo6SK8F8p9flN3Zvb4yU4J4Ve0u8kUNl19UYqBAHZKJCGFtaWijeqhhU31X9IWc5CwF5JTf3n7v7CJwbV8jOTWUyOAfg3JjCOflCLEUyOEBSIQEC4EYUDhDl6TKrMaVUA2A3F85lZUAK2+36FMYUYGrVSFLgxhMOKfRp857WtRHdeWUJALS9vIyAL/UxJSpDyNY6yonEwA0oKaywEvD7mwSkMKMQgBQ8CqLIiGlPNAZI0IqJQNCMYFGAFNqUJYRTBG5IF3DCttZxjCZZ43O1MaZkIwCMNW5fGmP8Xjs1uuoPEZCFbamjM5kIuCHdBIPD33hKboyGHN8kEDJguyZfG6M06uqJKSUADl53q5c9VeLJgBvSA5GscfvbFFgkGnCKjKTqU+c5I+VkIZT7orxCcJfpKBsaWfyuzKjJp1odYiwJZXTCRl/dUb56Y0ypCmzhhPVT9V/JyAJBd5vecqIwHHiqnMlMgLNd41TdIQ2f8I7dR0rJ9BhEqwKHpvGb+8iGlu0SYOBwO5RRgjCMpKxp8s9NIaU6IM6tbkQoRmO7gPbAV1PbDRI0na7VNZ6u1i/ypTjMBpSHk8vl+gUA+4wjCTIyOFpxuQngyACcCXNphJPL4foBicKqTAbHch8TBoBzWK4kLM+iJefQDTxn7f/VymbqRFRSFk1uiZSMq78Hh25TYTw5bIWUDCxpBtmUkkGFw6Y0X291DgaWJZT3llYVsdzAQn1dLfHeUpnYuMaXqvkfj0S8p1gmaRPOZWXkuDzIw75U8z96xhSncGn7MquaQ65JknMAzmVl5LQc3DOH1Y5Q1o4m6G6D9uBcpkmSAyQUkkIAnCrjyAA8l+U/nW31TftGWn0aTaHU6zWq2zS7/84UhMYUvxeQMkK2TdXz73xLb4zDwHNY0L6mqT24vJtrgnrdxMFokFTrjj1nAkJGnNuUOUeiwh+dNu5giDkvVY2h0Os17Gp9PO4FICGEjKiZHqgk41169mQwUe71ysHjn5vONEyTAQkaP/owBhPFXq8cYrrV2bgUSBkQdr1efCrRtx42mCj3eo3pPtVxOBpFBoc47xDg4zGV9aYMJkq9XuFnTB9XIBhwUgZmPl+QLiOJQq9XadgIQJHx468UOIRdqWsPqYycFDJrmWaV2E8zKPR63emr7ykhqNVqw4SMkKvTek5khqTaM305z+EclEWUVK2S+2l1AdsDfsrfQTYFeSkhY43X1SkcZjR/jaj5qCU3JR2Hj+zuF1Dm9crJGAgUGbdc+7a5afvw8jTZlOX9gMrcXEnfcVmv1xyU+WkyX8v9/jKZkaEw1sGrmt405/L3YOrHiuPJ4V4+uRKcgLFp0yJTmWHzuRIcCICxqZsST65hTc/txggFm7Cf6eRKcIgBgACosTZRrLfQS/9ZXg5FbRRR8wuAszqcXK9sOLkcaqEvKNhzodcLjLL6WmLQLlluRbngAQqFAvCWHAl9mRTuBlAYNY5wFQ3KEsDVNAMDl2yZmgYlaaqqTvSRMYGDZMOq2mhhQF2Z2Bc1IwKHWKNAgPsg10bZbZ6TgomSAuDGAw4Xn8/n7/qPl+y2fN69I1oAXKdh24CvyfqZaAVwYwGHjygNmki/aHgZNxj4LtII15gPZ4ChgC87cv4atcILMRhAvBMd4sFSTknW4DnIaBDvoJHPexcOXO4JWy5/5X0PiST4yswbXvKBVZPtatt2Qi1P3XnqFvngZM3v9yA1w0vef9dUvezkKzsffEqqXeI8XRflMbvG/EGb368uSwjiQ0wLpshwEt9Dj0e2lKEiszTjdUHetYXAHiKFws0/ed73ovAuze+3gII/7IPOCtLwHyGS3Jc2Ol7gFY4TrvuT512bBA73BAofFKwrbyzJu7aUzNfLzuO3E9emicDNeE8IdrvQp2Z0lxQuNsR5HXTWkWFu+57yfyq2S3st1S/J2/a07drO5r/f72FqQJSm0cS15VS6xHgS9Yf/bqnqQdXynjU1/4UIV3yh+Y0Iz0nXZsl4jX/OfGjUrfYkKYvyuNbmvxGn4YqUv0YDSj4R7mxbWvCsHL44UilS+TQ70FY/+MTHzJ9pYaTTb//tT2XFtf6WCN3yOdImnIAzciHv3yhw4Zc9mAUv7TRJ5dEOeR58avBEh2RRtXInOAAGAGfkJnmUyc6UvC1XJ+muI0dskx4XKC9ZzgSHcg2jXAbOxAN5y8o/4egh3Qs+fc+7+VORTZr/gcQWPp4UArM1f3+P0Jz8ubGDwzRzKGrrQ4BEBOgDzsT2otPYtt0RUWvS99APZY/vLVmvl533rDFVLvYKEikDpgaML/T3lUstFAtBoOAHCmUmco3ok5NjO98/X9hY49ATDpzwhoU1D6GGIkmEP/gk1lqUK1FD3s59wLRNFPX+UI5EFUX5LFfaDYoSZaahbv9vjoJiL/eaDyjsHhR7tOfwaCMUYggoFNOIksiNHxxgzwEhH0iheUBNQl6QUkF/eM8A3OjBISJwgZ3gTNhPiQqAMwCtmE7BIQQS4i7Lv5kwIcIAcM3TdJBmzHtiYBnXHTjcnDR3l/51zYRJM2XgGicKFLVjvaBHuMFCGjKThAesfXoHtGJjOgSHXjRkLN0OcK1ja8X0iZNQqVQq2H7qMWZk4DxtwgIK6BlRYE+KFdU4SU4P2WanUAbnAJyrRUUIb2OfwKlZCBFsIUhbVAwLisIIysiQDSOoDhURiHpE226/N9uKRLlbIYiKBjtSBbM/SOjQASJfi/uZCbqd03MkfGI4HCDGy8Nzqj5TCWQYQUpUQfkwgmpQEa527HYHF7kTGiq6Lqw7QqCtKtiRKkIQMnAdjCJBNJdGxIwmaKLe3XJKLmnQ7Zw8nu5twuKfzSlH83cwGw4prFd6MY1qT8FXCQC0PSWqoFwYQbXYOMtvmxCr4jXsIOEuIdhDgPa2oiKOFOocBvl4C+Fqsr1Tq7+GkHK3WkKa81XpP7OLXFSYxAOvcxSMkUbQlWkSY9Xrx+YOeW2a0y8Z0GNmz2hb9/h0UIaTwg5zSHWEXZUAPG4gJaqgXBhBtfDkdEVP181LSLYrOobBviWnZeEP1vzG+SaExqN6io1bJIRTA+a2h8wUq5a272PtLi7rosIi2uMKa+nEy32ckZSrbSdvHgfNlfhUcfBciQTh28pM91zP4qbpPA9EM344nk/Nf4LlC4BBxA0kowrKhxFUaXUo6zo7CQ/TdpFL9LjNEsFgzf+8+VwXsSJP10ksTpYapnPkjevCZs5zQpf2OGcOXHGrgBnCifO9RXVnE+c91VEo2lt2aHPKriN2VfkhHUEz63lM7aEeY2pJsQqna56QZgVsczg4gGnZolnapf3qQ/whKxpts6yZKT43mc//CoYd8M+BM7F8EwQljCAZVVA+jKBau7mahFpIdLY7fKaj1Xhl85n7+2MtdDsfjEnwxTjrxy9NDTtM1N//vVawC4cOnixBNafZ6Q3pcjbz25UthO00FRGjVudmU9IQfLIx5qjqD2C/Cxtsb4o5xLWOLFRCiEedCeFg5OcU8Am7D4RBCSOIRxWUDyOo2giZu9hpVswjcEL/VDPeY+QUnHhs/oawfo7X+9TgXZJFeLGLaZKuuGfmWJpCDpgyracrlhYkkIAEeQwcsMXBcFY0wqdxrw85gvkA6pP4e4h3thWPGA8ZRpAcKQkjqMrq0KEpLOedRUUsPfy3IPgzFxkyTYjwaRiYjmjWw38LfEeW0RSi35TpNS3A8ITot0USYLf7fXFQ8lYeGQlAjrhDE88zOxDDUZFE7BkSg1zNemhhBLGoggrCCKrEMZvfiOi5rfBV221MD0EhortwF7sQiXUxXi7trmZP5fDf4pizwnrOc8JimUNmsm10w2Gum1nxJ7ttt69oIcLtjKFg252yIaWw6mcMRhv/WOEeDjy9xQUrZvjhC2mc8qURB5+yHzxuIDkCBWEEVdvEtnd6BZTVkazjqQHtMcVPzuPnxlm3Y9sdfjOFExrOOj463uM1m3EkZTe5allWXc0SKdluVy3LAQukuJ3bN8uy3DHG1ay3xbKmJ4tlsSkTHHyqAwtoD9pquTCC6m3ihHo7+4TcaglRIRqJKuIGt99WYFnvSKV5SM4NblOax6iFjphCUq7J246B1rbHyePCkoKXuY+YouEY4XHYHRH2ICjWg22PPvSg4udr258qxAgJJ/vVJtmzY3rCchk2ZXC/gOIwghQZKoYR5NBP5iCyXfAPEo3CmI5KRUirlbcTRkC5fT8g4UutQAY+o0J/0M3hFYaSdn9yU3+wVPcfLxT3C8QFWqdc1gpQrLG3/1cFB3sOM8Gikm5vn1AqrCI+LHCtX/4xrTBQHNAl/677pwj53ZUSEtI8t1OsaYTymE758XqFMuPQWakILxi49q9/cUAT1Io1vfLb30aIjLhv74YB/wQOY5qY7S326tV7sDmc5DcI9e037Bk4C96MO9xbLPfWBrLZar1j2FH0iRYOyu06C73KNTHiCWsZ1Mpjxay2sbHrHNCzOBdphH26zkIvsC3EE+cAoLYL/jdVWSYALOPGJI6V2mGslmHxiHOmOFZ6EraOiD9uPOLWaeFK4qh6JTf3n80nhBNwow0HwADgxhoO5RpGuQjcYFNbHwIkiqVcbqTSK0bV9+rTTji5Yi6Haq8YBZUvuWbCyaFpjofsO0UVyDjfoF5OHv6QmJOKEsWcDy77TlF5GWHX6hScmaaZ8Vm18OSmsu95bfeFbwpcZctFkH2nKEVGTPv3AtjpViCm9w9Z5xxv2Q/noLaNFkZBbeM8B1lQ8E5RjDVuXxpj/F47Nbq6EQA0AAeIN4QyBSLdZ1XNF8gdyR+grsxHeD6PGIDnHpvIvVOU9gqyJl8bo9wTT0wpAXANwOFZ+S/ks9/8Ac8KeM5xINl3ilJkPO+p83y4yULANQCHF+vZoYV4i3nPgRcLPDdB/hWjcMI7/mxfvTGmewGawHuSNsJs5l9tBuAEXN+Qf8Xodq/0kodPeMfuIw1D1sPh+2olumr++6p+9TgABgDXNRS8YjRBGMalaADvRZrhZtWGQ3kUo1wGrnML2B6cKgOt1wL5a7xGrST1qa0flRjt06FR+q8YxdcDZPewbJhOPo/lPjkl2UQG2iNHCrsvlHWIdF8xShelFfJ5z8pv/gXL5a/8x48ECpSezyIy0wEhTXW8/1cr5P13+YG2bSfV9veRp26Rv9bfclXZRGY6IMq5L3nPah/qSwbayvufkizXXVyd8ti+15gPTvZgXJYw8nky9+H75bMVhR0QuwclL5vJOYku0ow3Pw2UyWYdIp/3byGSKFLdm1fe96LwLlt8qZNL+DeOBPjLRjnwpRwnXLcn7Hzev9HdfymRsl2Z8hkjA50To4Wiwk6GHJMHv+GsxI0ktof22uvf8n898rk2PN1r/S0DMLc9QiPv2kJsl/baEb/kA80/+IGtkmq5hLISDrBnP+g3HFfW3EIf1iMkL+U3nGoMgOeYrMCPnXdtx32NqJr/+Ml03YFYiNO4gT8Vv0LytlyNvvJWQRp515Yf9PnlH3yPt63sXOAQF9yZUM8O0gs81yR/JttuP0T+5z7R3/7bf/YK2uoz9ZziTKlhRzr99t/+VFagTJmVm3CICWYKqflOZLHvmwWeay487MCFwKXN394PcT4fqiPWXZ4H33GJkl1HFtMkJNR8B3pFHB14rrlJPn+tj+cQESfdHrXzfyoR7+T49D2nyEuWadIOeqBqCoGA55YHusr8zw0fOmtXSBkvtGrzGxJbPAEKgdm+3gNyXfRzD/syspejiwo9sIUKHAR4ztle1B7DchX6+InI9tyFT2XDWe9eqlzsFSR/6mlKUeMWXmX2Qg9qhEWzSptGcE7JNealT/zBJ/OSPeyT5/N5rLkYayQiREiaJAnXmIXwZcP0IGZH2t5FLoR8qlQqegAn4KwBz1j4CUgZSjRlORXhTfa9bNtO0PbqISrIkJoKJYYd++GAA8CZgLr9v9lPRbhaupAv6dVbeIs9tw1OzwKzChGlUIR3laIRV1gPhzIgoFYDnnsQCbqlx01REYHLZv14scpfQ/+lShWN+Mz9l/pFxHOOJO3xMXmUiwr7WY+olXKS0cddMQ+LSOsidyL+kPa4oiMu/C2rlrbv49QLvXbES5hrw/nJwuTUV3qdhZsrFHMRUBBq/CnCm2Kq+Y8XB7zIsKlOfLnIyxG7S1aK6EEssJ9ywQMUCgVgxSaaMdURNxhq3B8Gp9olxsOJ/LM7orS9jHlCGYyeFD57ocgGoBys0I6VgauykPhYTsP4wgPc7flGFCTEymgNa+0erxkZhTE/aSxLqP6TxuxFQ6YSr2pLrxsSwdOenM/1pdJBdvfedpZynvBUvZD452IXnIPaViqcBFQ2qZWB8UG0bbV4iheSvC2PcelAK27x5/S4vK862vjOH/CKwmYIJ9ICiR7nzEyCAyQF19mmrgUETcDNqazzZocBuA6gXnvRTm6zhCfbhdjjO7yHmJ6rolnGR8c7fKaLY66zd2bQbhwigjsTWW+Nws/2ZoHrIOhonsrbPC4hKkI858clsHNuU5onKqjVmD1wOJUQlay3uyx/C9EPXL9Q0B+EnYCIOC0E2UDMGDj4wShyetYbuvJ7Bs4K/sorImjGq3N6joRP6pJWP6CSrh/iSLTugT/IGDjcnMDmNDAjXOAs55bX0vjVTdDL6zhyGnS/0D9sExb/bJ/wilQlXTcf1jr6cNhGQiMJrcpiNtSPzR0yLb3llxcv7PalIaYLSXPaEdKjXucXY9u2e3w6aHYQEFKd9Wxg3/Ioqm3U8F2pqMJushr7b57apnLFnyohbNsdLzqJq8SnesBzJRKEbyszbcaPaZN6+3T4J8SqQ76YrKAmKhVVh30PLGMClOdbiqOpglKjRIqcVEyYGhw3q9nF6dqvPfbZeXc2n8rurBKp2hE0821N7eEMU0v6iB5fnCk/t189zXbLgRRItlXAtjNbb+iHrLhay5qZ4nOT+XzP3x3b92APMgto4dCv6hO5EBGQtqaqfOkl7P8EaaOLJ5NTPJ3myeYnOakV4WrixhIVBvBLZkYNiU82hhyh8Uu+QAjbaSoiLrvBl4akoQ63adNaDPf5APa78N5/iRDXOrIDnPs9/yrk5xTwiZqklWYKC/Vs32hgCRXxI21kh/8WUqeGwMsN/vhbERvXPWkh47jyB+lSz9evpxAfd5dvIaMMuwVRBGiGOhf+VbvPl/oc/VT3ecXSaomIGnU2/Cv7U335F/VfnBWN8Pf64j/EA6hP4u/h+dqZH6V3AQaCQnf2TxXYNVyDAikHuX8DwFtYw5Y+S9tss01pnsTu3tuTxu4f2FmEcDULd7wCQXwDJo5EleqT8CZ7/pbwFwnwFkLQlVFuQZh2LTGjJB/XxQtwwtTrjPwT2273++Kg5K08MvLojxjj2cTT4fCL7iOSiD1D4mRgdZpXQJGVCzWFTeUtzKF7R4FswYob9Imm6iWix/bHLc0TTqi3k4tciJfrhd/g9pzwWKmIxx4WutQvIid146wfL56zF47k/Z/Hz1GbWKhz/0MWwgnzdbgo/J+gAtJ7Ob5Okk4tEUI7oz/ZbZ8SLcRlTnfj1w4HbZ2qYqDPMwajjcd4hXvwuzK9xQUrZvhx8HHKl0YcfJoltCjqNsNQ1jXGMPzBUh32aAorBG3Gy2OW2rOjugsRz/c7SzmK8FgR4aTj2JEqVMLJJKemBSvE0k+VaoDxXV6BaFl3vs9qPKqn+DOv58ijUcZpIroJsXGL36ItEtW/oWVZdTVLJBsFXHNYWJY7pqS42zfLsl77b3rdljU92fuwKRMcbKoi6l80HEW5i0G0f5yurq6AroJYOODrCZoiolFgg9OxEw/3ImCrBOGxgjq2nt6Cs1CwAWDLoFkCY0uv3cKb0rqQuYnTPt154mSH/xZ/5iJjPLYVJ7u0WxRLV8zTBO3n+Dg6vvyp7QR5DDy3I6bwMvfx5df+zRjCuB3jHVQU7WGuPfrQn+Dzte1P9RYeTrJfbQn6MfBpTsCWYab+/n4fdxIndMHOTuQnEDSh+dubz/WNKVnOmSVwjxVc2w9BvdgUqWHw69/lFWBIav7MRT5pVBH1x1poxuuOmIRjbrFsd+/tY1NXM6hfQJMwa4FEvKoNPokP1jnrgqjuV3cW9ORAeKz8nPYX3dXVFVP7SZB07yiQ/xqVVX+qPUZO4ektiOq+sfn94ho9q+kd0wxjvTpHpVKh4Mm5+ftbTMLfEOFTd0m0u1tCEB4rP7LbNbnAd6Zxn09zOVT+zEV+bCEW4VSkPBurgiqjLCYhKGqGYldmXqwjDcwdSQ+CT+/y2s9PHPAie/aLIDxWDrjbGI9tV4WzfJt51NHLjVbZ0ue6DBHdhbvYhZAc8PXA+WU3A8WaRqgVB/QMvB1tR4lV8RykE4ir63nMYtUVr5iHID1WAtnWtc/tCJgpsKZOQmpHUC9BX73/8dH2z/DkLgqjIi6pxzmxDFn4wRhoAhgrA29RnxDAPoj8QLGKWOXt7IULJAQTU6F7rDhhz2j7bYWsYX0Jglx/g5gwXMaqK/7v0t+VzSwbGCuDFoby2MCyjFBiI3+wQjEs71WQEEzMH5T1WBEiUEI/ZgFagtIIjKukaMaVrnq3wapb+FyXIwsntbFibSDLGagVx9CDoU6R1g+GqdmROJL4U4BZlEeamivCieNC2sN6SNmZTjam4d5isTiWzVYsFnsBeGZ+3rCeuajQlUs5drrycGl0svPKwZF33auujY2Mqaxz1wPAeUtGKIzqCLJVkSry/mDW08KlGZWtXCir/6ODTNV8hTLr0IoRZOeM+r8yKmviR0YcyoUa49COYbDBXz1T4Qr15VgtHGpH15D5hWrAdZW+PrZRLmrJellBab0W4FArjDKOMmjH2MH6PuCaWWYbtnaMHRQLNeCaWM4xYMTiKHBNLBsQRjUw4VAqSYPhoIYm2Z+ly7kGOjnhgEWwYix7dkw4KI3uf+48jyvSKBq4+2d34tBXqrGXy5XJYX/FgeMP6xO17E4c+rB1jKUZPxxPPEjbpk9eUbbBmMSzeyPWcjDZG8vBHHh6MVkGS1KhDFl5Hbg0YOdiIbL3FQX3HV2ihQP0FUZrwLPz7ylKA3P5K9ZBlTxnMIEsvR+U+rJvKw41Yo+cY8Saa14qFCGrZjCnghIaGImS54io3hEJc+O8wwhrEx87pPg9ZOlqpmy1viCtBJ5dfweQA4t362kcy4yYEpVe4tm1FMsh4IqD47/CdtMvWji2cj1lJ54x+dQ/ogicMjC33UC6sXysOIL9qR7weEfX3OxsN2DOytFSidiJZ+ayEHI5cdOir80Z2gmhiP832LOyJuWTMpqBDD3/Y3I5QLFUKEI2Dwz/Jas9QPy2jUUryyXsHDCakcShPAoA5T6inmIntSJohtoYq0ohcZNYX4NyJhqNOdQKRaj1FfrK2Fp2smx4rKwVAP+EaTOj5UIfgTpe3QR9hdr6QonyPZSlE838hLk8dmT0CimWzawvFT57YZR4OFTByEfQUUlsYZQ+x9CPi2XQwkZjMluz6hzbFwrfdAzQ3O2oEP/hOm8HgAj1sj1tL9bOjhXLvbXstvJYsTzAvKE2WsAMOPQfXahiR+8HqGFSS+vLjE4DvWX1f5uqts5yL9YaxziKmAWA7cuwjRBqveDWB4by9p2jmF4m06L+Dxrx+KQqGwvrXQ6k7UaaU+tHqeJQ5GVldkBxlY2IQ626VnZefIgI1PxhckXsGTJ2WZked163rulB1MUrRB7XCCIN6b08pFKRSYYRDjUcucyFv2iIPAGdAeBGE5nMVRG+KGH4jYoKeQKjS0W8uSvaGb+7EC8RxU9ghDmZ3XMvsVwmcIQwxFSkt8RGNF3kLaFc5qkh3OFlPMXuHfETGGKu4sl5nF/xWeYJ6dWwMHe4e0f8BIaYS1oxL3Rp90sXx1x3YOGEvY/uHfETGF8q4oAX+bHFfZYitPzIZrw8Zux5Aj+BEeZVoXgvazc8iZ9z+G+BIZ3ACFMRT2/BTv39/f/oyflc/5HdLx0fdb90Y8wxZ71LIYQT+UL++weUIjO5i77JXYGfwABzMhS1STpSz98KFltnt3/5B7OjwE9ggFcLUcGPJC7nCl3kSAU6gbHtF5DpDCJCPSAV5AmMKcuOLCppd/4a0/7fFg5HV9ff486MJCGQuoQFuQucs9jBS01/zuNKg35DejMqk8E5AOfMZddHF69ELf/tOxtYpt9wTnovK5LBAW75dgDYm+DzNIKFKvY4AkIgDXoNB7g3z/TfphoAu7lwLisDtnO7PoXxdy7mn7A3cRiICr6FrHkFf1deyfM9MrqF6jUcvsdTntb1L75BCQC0vbyMPdTH9CYhZGsd5URi4DrwA5ZCQWF4Kx0nhfUGwO9vEnDCjwaw5uYoMmLavxfATrcCQTOCRQFSaFOW8OAY/NSl4BcqCIU/QdHzBZwP9pROtsbn6td5/QAYa9y+NMb4vXZqdHUjIPssd/QWdeBpG0Np/tLxTTA4/I2n5Mb4c28SCBmw3T1sjNKoqyemlAA4/IF62V/3ZMBb9ICRXIMWkjVuf3t/kWjAKTKe99R5Ptz7inJflFcI7jIdZUMdYCR3we/K//ep9k6WUkYnvOPP9tUbY7oX2MIJ66f6WY0s8MGnt/xU4MA3lMwEONs1TvWSh094x9eVknknzwM4NI3fXpahZbsEGDjcDmWUIAwjKWua/PMNpzrgjdWNOIyxXUB74Kup7QYJmk7X6hpPV+sXfethowlFBfYZRxJkZHC04nJrR0Y6jChYvpJta6FqwN+sYjBRr8WtpcVgouLl3zwSOA433HDAAW6w4VAGBIwCN9oUS7WRkVpfUe/hBFxHqfWVRnZfKOn+ADgAepqKBcn0fsqhDAio1YDr6aSQC0xrpZpEuU+frr/yAA9FNFUcyUFH0/rSqQqjhbIewTkotlqhkFbkFv2cFDxAodCnSwNAjGnTpuUKbcYSP9typeYbdT6Afqb1hc+uS1MO3XXH4QhnRmzRfOD6WNGiSWH3oD91LIdzMM1qRsxctAm4Xla0uyncHCizZZzpw3xpzvHb54xYVYdmOUCMOeCCd1RF5pq0YVTvKIrU+EYJWhkAzhnOjFcypSFjSWfgENXcsCa1q9jNzqNpNgDO7ikaMpf0ZcJh/pTrVL2iNc0lA8CZ/XlmOQc9gUO3NKh+4atmwxrZTeMSM8smfeHvm9WMXHq/F+B6Q7VaVaZFt+CQUCQVvIywRA+oYtmJ3M7EvlYq0qJbBK1k6IZbNdlPtWpi2YnczjU+NP/z8lJN05TTctNJXP1TkRToFUfOGOcPnO1UTbNhOf1evXPYu2neYeOiq02zYUkNT/1PnvFo/ue18WXHdoSvBzreD/aNYcZmpimdwDT/k3O1uuNL43JeamBakqqq3lKtVpWJ0QuqpiM8hnS044z7mRq8aXH8/rJ17MsK6pgaLnbyRpGv+3pnxbSkBlxGe3Dcq693luu5TDPMcYPr8ZBUZI8mb/ul6hfpV7SEDlwWqmj1Arc725+qeW4Rsg37IsDOcr3Nb4TdWm36TW7GW+PLCfallwtVJZUqpaYltJhbf2TTrOoEp0l8LEc+TbPJn1v9Bzn+omk6Qln7F+lUr0v3qJoznqfPm28K8pXqWSvTn4d3tJmLoIpXEnSHjYvVneeBYvQhYvdVHc+s6gNR+vhqzKp5KYq46Uu9OH2lar41O/6JOIV2791NmUp165gmuIzjxD93XYjY/eVStnI9cxJ0nNFP5VqXXhDitGFfpvmQt/x6Cqjbv6/zzVVXqZr1y9O4f16mGdCpp35nmUp16+RtTT/EbHh8nRscJ8F/0JWFxvlxK/+0V9ftj0O/x5HOw0mX2+2PcYnnV3+wROzLnr7LfzxNzx28sbUfebXz8bgW8cyl1brHBRz2pZpm3f48fvVyvBoT3YrpVHXjQF53pHAZPc6XbNjI/KURNtv+klGtr8Z8bWuP82qud11Ms213eB4wzaCJtr0wYnTDvs7t7do9OrYb/YaL7rDyYI9eegN+5e7wqo5tmlXTfBrHNk0Tydi7dNVRM0JVEnG21aZk5+Zc4rF1cOTT+bB1MZrSTt57k8c8t62jP2xTR9hE+dpEtP7i35GqSAa+2qwSmrDv0DKrz0GfeG1r3+Hq/3iaXS5eplLdeuFTl2pd/7TM37zh4r9TjKRdIvJ1MX/1fLMq3yBGOZL5xpKgqyfbRWhHoqNTuzWNp1TRLtyNXKW6deIDc237Ve5gusZzrt0ediKNgjrZo//BrDKdbrOaVqO3qUQVnW69ajf4jwc6m2ylKi2c243t3TSrkojjYbVvw5LOtpoGyxfTaBJPv9tfp9oJ5StVfLWJ1bpk7UsuDDCb7iUmlq3Ud3HBu7z1hPdNr2jlK1Xz7b4JakVLaKL1/7I6bcLKk/oubfi8rqSLSremrSrV07Amdl99DjB/+ZIM/FpjuQ45b7vuluYzVFWHXFo5h8wYGvQmTduyWa2q78/5lNh++XmGrEV/+DjTX1n1etY0bxN4i/EzDuCebZGpri0JuIAGI5jQD7Tui5p2/pCDDHr5iMVzj0E3H7FaDLtt7j93lzE45xnRl7t3GmZRhy0HAE75S2h/k0J9MioYmnSlOPEMuaAqzEabAVYRyefE30FRIK+vZDUAwOVUaLqkZUwvpoDlxYlDcSYwaA3OKQJX/5aiAMsKCyQ4tA1a1gwQpqCrDrBQbpLFDUKq6Cq0Cs+Qz7VMeVJYnIrWIHA5FUwaRizMhoDmoUNRgJUibD9ZTTBolQBXQdFGY9KSrBMkJSX8L5LRJ0HVomU4zEaloQhcfb3TAMuaaMOucKsln5vwyw5cTgWLhjbLWjwJXbNbgVM88oBUAIBdVVCiybJmEyoo2oDOoDUHMWEttoYU6JtN1aJd8MwiWxzI/WglSa6YESC5TQC6ckXLwnITp5YnCaoWhjHHsoqAZ6nOIcuaWNkFMGRNSnJau8CySoD+L7OhC1VCuKaRIWulhWw24CW1DWmesCY6qay0BrsAJq3B2QiYvdiyFpeAS/ota85KpIXUrFmwzDIbL2icLEzkftTSAECXQTJoTSDmWIsRHGTK0yRVC8OYsBZjhYJS0SxGECWw1RoC6LKszq7FuBLEoGW5AII5FrLBLui0kFGZHLRGAOZYcxCEPmslwGLyT6Jo1jCTltWJZSZKcYCiZRVRAWgFsAY7J6wiFFvxQtGJlE0CJ1iJLjF0DVoly6KWVA6zB60hTAuplF1Y1hBxFQetVvTPtzpRnupE+QFlwS4YsQahZC1uQ1euDeXFya4ufKFoDU5C24Q1AkPWRFtXK5051hD6fxQliBLdalldk0hB56A1RNOsYbosa3B2FwCnFAfsgiGGJFC5mixigubACKU44cxuteYAjFgTkxLUkrrSwg0oSlkHnkmKgHadLTEbsNzVZVkjgC7tEHkfkVhJWT2bUDMbq4gA6HRaVtuINQGI2dYgcLTrJKYAaaFp1vLqESwvdVGLgwwTbW3Qak20QWkODFolpGwOhRFrsAuGrBKCWlItSUFxsWUBRSm7GCQeqMjacIjKJLRarTBodVIoYpo4l6OE7ysDTFglSSOCfH7EGJKlRKLNj9HGi6nFQYZOAOLBCiYtolqkgEpim2W1SVBLKqYAJi2LppRdtFpWG3AAZaNJtFmnNQgScwAzWSbwrxWKRrDSmrCsNuUjmmYtwwHaWi2rkygOMtBLEKfQSgHmWK1InwS1pGIKZGhlF5P4gUZauyziLlmk0mVZE2jbldYgEjdBZSW0dWLZrHNxERZbE11drTIUsZoWo5O4gQ8CUoqvpmjW8gIAB5R3aMWhSxbqqI1oJCLptKwJawQUj2hKCdh0KMyGyGehOUAFWi3sRts2iD/1k2B6ZsNistYZwWXRAWk6gv/zSH0jABOYPqoW7cKhNNQJULKsIq04ALrsnYM0FuOru/Dnj9YSbQSDlmV1YdBKKlIA2GqKUgImpc4hC2sfhpHFlrV4NqA7JeW7VKeFCiHZiNyGKGE6FluDndC1csKyWifx702LJ1sJilghm401HZQQXaiNYKgTleYha3DlpDUEFM0SrVqFKF/U4oA/P1HpxCsJsizSmYMkY1BL6hxMN1ULw2jhIN8mwnEVtP3o4ug65NpJqDdgOeFcmT7Zv0STdJGNWNTi0DVkDc6ZtFoRReD4t//WNoAS0jhCXNZBawSIm2nRsoDTSmrXHCRz0AJOKpUYZBTyTf6SElwFZT8ZcVQdMu2i9F5k+cZWRfrk/hKNlzMZBfLlSq50EJdVtji1cEXFjqtTnHJ8f4c0igNNHHkCQrF8cWpRVuzoxcmw2+b+c7fP0QIA)

## QAT encodings

QAT encodings are quantization-aware training encodings which are present in the source network model. They can be present
in the following form in the source network model.

> 
> 
> - FakeQuant Nodes: There can be FakeQuant nodes in the source network model. This nodes simulate the quantize-dequantize operations
> and use parameters like scale and zero-points to map the floating point values to quantized values and back.
> During conversion this nodes will be removed and corresponding encodings are applied to generate a quantized or mixed precision DLC output.
> 
> 
> ![../images/qairt_fakequant.png](data:image/png;base64,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)
> - Quantization overrides: Tensor output encodings can be associated with the output tensors in the source network model via overrides.
> The quantization overrides for the tensors(output, weights, bias, activations) in the source network model can be provided to the `qairt-converter` with a JSON file
> using the `--quantization_overrides` command-line option. When the overrides option is specified, `qairt-converter` produces a fully quantized or mixed precision
> graph depending on the overrides by applying encoding overrides, propagate encodings across data invariant Ops and fallback the missing tensors in float datatype.
> - Quant-Dequant Nodes: There can be Quant-Dequant(QDQ) nodes present in the source network model. The Quant nodes convert floating-point values to
> lower precision values typically integers to reduce model’s memory footprint and improving inference time. The Dequant do the opposite and convert
> from lower precision values to floating-point values for getting higher precision for certain operations.
> During conversion this nodes will be removed and corresponding encodings are applied to generate a quantized or mixed precision DLC output.
> 
> 
> ![../images/qairt_qdq.png](data:image/png;base64,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)
> 
> 
> Note
> 
>     - Inference fails for CPU and DSP runtimes if QAT encodings contain 16-bit.

## Float model Usecases

> 
> 
> - Float bitwidth conversions
> 
> 
>     Users can convert float source model between float bitwidth 16 and 32 using the `--float_bitwidth` flag to the `qairt-converter` tool.
> 
> 
> ![../images/qairt_float_conversion.png](data:image/png;base64,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)
> 
> 
>     For converting a source model with all float32 tensors to float16 tensor use `--float_bitwidth 16`.
> 
> 
> Note
> 
>     - Float bitwidth 32 is the default bitwidth for float source model conversion.
>     - Float bitwidth 16 is the default bitwidth for source model with quantization encodings or overrides
> 
> 
> $ qairt-converter --input_network model.onnx \
>               --float_bitwidth 16
>         Copy to clipboard
> 
> 
>     For converting a source model with all float16 tensors to float32 tensor use `--float_bitwidth 32`.
> 
> 
> $ qairt-converter --input_network model.onnx \
>               --float_bitwidth 32
>         Copy to clipboard
> - Float16 Conversion with Float32 bias
> 
> 
>     To generate a float16 graph with the bias still in float32, an additional `--float_bias_bitwidth 32` flag can be
> passed.
> 
> 
> $ qairt-converter --input_network model.onnx \
>             --float_bitwidth 16 \
>             --float_bias_bitwidth 32
>         Copy to clipboard

## Quantization overrides Usecases

> 
> 
> - Float mixed precision conversion
> 
> 
>     User can provide overrides to `qairt-converter` to floating point source model to a mixed float precision (float16 and float32) model.
> For example, if the source model has all tensors with float32 precision and user wants to change precision of some tensors to float16, override file should contain
> names of the tensor with type as float16.
> 
> 
> ![../images/qairt_override_float_mp.png](data:image/png;base64,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)
> 
> 
> $ qairt-converter \
>           --input_network model.onnx \
>           --quantization_overrides <path to json>/overrides.json
>         Copy to clipboard
> - Quant conversion
> 
> 
>     User can also convert a float source model or mixed precision source model to a quantized model using quantization overrides. The `qairt-converter` will generate a fully quantized
> or mixed precision graph based on the overrides provided.
> 
> 
> ![../images/qairt_override_float_mp_quant.png](data:image/png;base64,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)
> 
> 
> $ qairt-converter \
>           --input_network model.onnx \
>           --quantization_overrides <path to json>/overrides.json
>         Copy to clipboard
> - Overrides to Float conversion
> 
> 
>     User can convert a source model with overrides to float to run on floating point runtimes i.e. QNN-GPU and QNN-CPU using the command-line option `--export_format=DLC_STRIP_QUANT`.
> 
> 
> Note
> 
>     - This might result in loss of accuracy.
> 
> 
> ![../images/qairt_override_strip_quant.png](data:image/png;base64,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)
> 
> 
> $ qairt-converter \
>               --input_network model.onnx \
>               --quantization_overrides <path to json>/overrides.json \
>               --export_format=DLC_STRIP_QUANT
>         Copy to clipboard

## Quantized model Usecases

> 
> 
> - Quant model conversion
> 
> 
>     User can now convert a quantized model in a single step using the `qairt-converter` without any additional steps.
> 
> 
> ![../images/qairt_quant_conversion.png](data:image/png;base64,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)
> 
> 
> $ qairt-converter \
>               --input_network quant_model.onnx
>         Copy to clipboard
> - Quant to Float conversion
> 
> 
>     User can convert a quantized source model to float to run on floating point runtimes i.e. QNN-GPU and QNN-CPU using the command-line option `--export_format=DLC_STRIP_QUANT`.
> 
> 
> Note
> 
>     - This might result in loss of accuracy.
> 
> 
> ![../images/qairt_quant_strip_quant.png](data:image/png;base64,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)
> 
> 
> $ qairt-converter --input_network quant_model.onnx \
>             --export_format=DLC_STRIP_QUANT
>         Copy to clipboard

## Quant-Dequant(QDQ) model Usecases

> 
> 
> - QDQ model conversion
> 
> 
>     User can now convert a Quant-Dequant source model to quantized model in a single step using the `qairt-converter` without any additional steps.
> 
> 
> ![../images/qairt_qdq_conversion.png](data:image/png;base64,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)
> 
> 
> $ qairt-converter \
>               --input_network quant_dequant_model.onnx
>         Copy to clipboard
> - QDQ to Float conversion
> 
> 
>     User can convert a Quant-Dequant source model to float to run on floating point runtimes i.e. QNN-GPU and QNN-CPU using the command-line option `--export_format=DLC_STRIP_QUANT`.
> 
> 
> ![../images/qairt_qdq_strip_quant.png](data:image/png;base64,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)
> 
> 
> Note
> 
>     - This might result in loss of accuracy.
> 
> 
> $ qairt-converter --input_network model.onnx \
>             --export_format=DLC_STRIP_QUANT
>         Copy to clipboard

## DryRun

Use the `--dry_run` option to evaluate the model without actually converting any ops. This returns unsupported ops/attributes
and unused inputs/outputs.

## FAQs

- How is QAIRT Converter different from Legacy Converters?

    - Single converter vs independent framework converters

        The qairt-converter is a single converter tool supporting conversion for all supported frameworks based
on the model extension while legacy converters had different framework specific tools.
    - Changed some optional arguments as default behavior

        The default input and output layouts in the Converted graph will be same as in the Source graph. The legacy ONNX and
Pytorch converters may not always retain the input and output layouts from Source graph.
    - Removed deprecated arguments

        Deprecated arguments on the legacy converters are not enabled on the new converter.
    - Renamed some arguments for clarity

        The –input\_encoding argument is renamed to –input\_color\_encoding. Framework-specific arguments have the
framework name present. eg- –define\_symbol is renamed to –onnx\_define\_symbol, –show\_unconsumed\_nodes is
renamed to –tf\_show\_unconsumed\_nodes, –signature\_name is renamed to –tflite\_signature\_name.
    - DLC as the Converter output file format

        The QAIRT Converter uses DLC as output format. The .cpp/.bin & .json format used by `qnn-<framework>-converter`
Converter are not supported by QAIRT Converter.  In order to generate the .cpp/.bin and .json output, continue to use
the legacy converter.
    - HTP as Default Backend in QAIRT vs Legacy Converters

        HTP is set as the default backend in the QAIRT converter, which may enable certain HTP-specific behaviors that
wouldn’t be triggered by default in legacy converters where the backend is left empty. This difference can affect
how some backend-dependent features behave during conversion/quantization.

        - For example, during quantization, an optimization called `IntBiasUpdates` is applied to the FullyConnected op if
the backend is set to `HTP` in SNPE, whereas it is always applied in QAIRT.
    - Quantizer functionality is separated from Conversion functionality

        - `qnn-<framework>-converter` invokes the quantizer as part of the converter tool when `--input_list` or `--float_fallback`
is passed.
        - `qairt-quantizer` however is a standalone tool for quantization like `snpe-dlc-quant`.
        - Please refer to [qairt-quantizer](https://docs.qualcomm.com/doc/80-63442-10/topic/SNPE_general_tools.html#qairt-quantizer) for more information and usage details.
    - QAIRT Converter preserves the original output order from ONNX models, while legacy converters may reorder outputs.

> 
> 
> To maintain output order in the legacy converter (`qnn-onnx-converter`), use the `--preserve_onnx_output_order` flag.
- Will the Converted model be any different with QAIRT converter compared to Legacy Converter?

    - The result of the QAIRT Converter will be different from the result of Legacy Converters in terms of the input/output layout.
    - Legacy converters will by default modify the input tensors to Spatial First (e.g. NHWC) layout. This means for Frameworks
like ONNX, where the predominant layout is Spatial Last (e.g. NCHW), the input/output layout is different between the
source model and the converted model.
    - Since QAIRT Converter preserves the source layouts be default, the QAIRT-converted graphs in case of many ONNX/Pytorch
models will be different from the Legacy-converted graphs.
    - QAIRT Converter preserves the original output order from ONNX models.

Last Published: Jun 04, 2026

[Previous Topic
Offline Graph Caching for DSP Runtime on HTP](https://docs.qualcomm.com/bundle/publicresource/80-63442-10/topics/offline_graph_caching.md) [Next Topic
Qairt Quantizer](https://docs.qualcomm.com/bundle/publicresource/80-63442-10/topics/qairt_quantizer.md)