Difference between deep stream close-source and open-source?

Hi all,
1- What’s exactly difference between deep stream close-source and open-source?

2- Deep stream only used TLT trained models?

3- For run TLT trained model with deep stream, I need to have all of deep stream codes? Is it possible to have simple python code for running of TLT models?

1 Like

1- What’s exactly difference between deep stream close-source and open-source ?

what does “close-source” mean?

2- Deep stream only used TLT trained models?

can support all models TRT support, such as caffe, onnx from TF or pythorch models, and UFF model from TF model
With DS5.0, can also support TF & pytorch model directly with Triton.

3- For run TLT trained model with deep stream, I need to have all of deep stream codes? Is it possible to have simple python code for running of TLT models?

yes, you can. many ds code are samples.
https://github.com/NVIDIA-AI-IOT/deepstream_tlt_apps is the sample to run TLT modes with some DS libs

1 Like