This folder provides different notebooks to run Jasper inference step by step.
- Jasper Jupyter Notebook for TensorRT
- Jasper Colab Notebook for TensorRT
- Jasper Jupyter Notebook for TensorRT Inference Server
./trt/ contains a Dockerfile which extends the PyTorch 19.09-py3 NGC container and encapsulates some dependencies. Aside from these dependencies, ensure you have the following components:
- NVIDIA Turing or Volta based GPU
- NVIDIA Docker
- PyTorch 19.09-py3 NGC container
- NVIDIA machine learning repository and NVIDIA cuda repository for NVIDIA TensorRT 6
- NVIDIA Volta or Turing based GPU
- Pretrained Jasper Model Checkpoint
Running the following scripts will build and launch the container containing all required dependencies for both TensorRT as well as native PyTorch. This is necessary for using inference with TensorRT and can also be used for data download, processing and training of the model.
git clone https://github.com/NVIDIA/DeepLearningExamples
cd DeepLearningExamples/PyTorch/SpeechRecognition/Jasper
bash trt/scripts/docker/build.sh
Prepare to start a detached session in the NGC container. Create three directories on your local machine for dataset, checkpoint, and result, respectively, naming "data" "checkpoint" "result":
mkdir data checkpoint result
Download the checkpoint file jasperpyt_fp16 from NGC Model Repository:
to the directory: checkpoint
The Jasper PyTorch container will be launched in the Jupyter notebook. Within the container, the contents of the root repository will be copied to the /workspace/jasper directory.
The /datasets, /checkpoints, /results directories are mounted as volumes and mapped to the corresponding directories "data" "checkpoint" "result" on the host.
For running the notebook on your local machine, run:
jupyter notebook -- notebooks/JasperTRT.ipynb
For running the notebook on another machine remotely, run:
jupyter notebook --ip=0.0.0.0 --allow-root
And navigate a web browser to the IP address or hostname of the host machine at port 8888: http://[host machine]:8888
Use the token listed in the output from running the jupyter command to log in, for example: http://[host machine]:8888/?token=aae96ae9387cd28151868fee318c3b3581a2d794f3b25c6b
./trt/ contains a Dockerfile which extends the PyTorch 19.09-py3 NGC container and encapsulates some dependencies. Aside from these dependencies, ensure you have the following components:
- NVIDIA Turing or Volta based GPU
- NVIDIA Docker
- PyTorch 19.09-py3 NGC container
- NVIDIA machine learning repository and NVIDIA cuda repository for NVIDIA TensorRT 6
- NVIDIA Volta or Turing based GPU
- Pretrained Jasper Model Checkpoint
Running the following scripts will build and launch the container containing all required dependencies for both TensorRT as well as native PyTorch. This is necessary for using inference with TensorRT and can also be used for data download, processing and training of the model.
git clone https://github.com/NVIDIA/DeepLearningExamples
cd DeepLearningExamples/PyTorch/SpeechRecognition/Jasper
bash trt/scripts/docker/build.sh
Prepare to start a detached session in the NGC container. Create three directories on your local machine for dataset, checkpoint, and result, respectively, naming "data" "checkpoint" "result":
mkdir data checkpoint result
Download the checkpoint file jasperpyt_fp16 from NGC Model Repository:
to the directory: checkpoint
The Jasper PyTorch container will be launched in the Jupyter notebook. Within the container, the contents of the root repository will be copied to the /workspace/jasper directory.
The /datasets, /checkpoints, /results directories are mounted as volumes and mapped to the corresponding directories "data" "checkpoint" "result" on the host.
2deaddbc2ea58d5318b06203ae30ace2dd576ecb For running the notebook on your local machine, run:
jupyter notebook -- notebooks/Colab_Jasper_TRT_inference_demo.ipynb
For running the notebook on another machine remotely, run:
jupyter notebook --ip=0.0.0.0 --allow-root
And navigate a web browser to the IP address or hostname of the host machine at port 8888: http://[host machine]:8888
Use the token listed in the output from running the jupyter command to log in, for example: http://[host machine]:8888/?token=aae96ae9387cd28151868fee318c3b3581a2d794f3b25c6b
./trtis/ contains a Dockerfile which extends the PyTorch 19.09-py3 NGC container and encapsulates some dependencies. Aside from these dependencies, ensure you have the following components:
- NVIDIA Turing or Volta based GPU
- NVIDIA Docker
- PyTorch 19.09-py3 NGC container
- TensorRT Inference Server 19.09 NGC container
- NVIDIA machine learning repository and NVIDIA cuda repository for NVIDIA TensorRT 6
- NVIDIA Volta or Turing based GPU
- Pretrained Jasper Model Checkpoint
git clone https://github.com/NVIDIA/DeepLearningExamples
cd DeepLearningExamples/PyTorch/SpeechRecognition/Jasper
2. Build a container that extends NGC PyTorch 19.09, TensorRT, TensorRT Inference Server, and TensorRT Inference Client.
bash trtis/scripts/docker/build.sh
Download the checkpoint file jasper_fp16.pt from NGC Model Repository:
to an user specified directory CHECKPOINT_DIR
For running the notebook on your local machine, run:
jupyter notebook -- notebooks/JasperTRTIS.ipynb
For running the notebook on another machine remotely, run:
jupyter notebook --ip=0.0.0.0 --allow-root
And navigate a web browser to the IP address or hostname of the host machine at port 8888: http://[host machine]:8888
Use the token listed in the output from running the jupyter command to log in, for example: http://[host machine]:8888/?token=aae96ae9387cd28151868fee318c3b3581a2d794f3b25c6b