Open demetrii opened 6 years ago
Which part needs clarification - training a new model or setting up and using a pre-trained one?
Training a new model, with a language different from English.
You can check example directory. In script run.sh
, change line 1 (wget
command) and line 5 to your own dataset file. The step by step usage is described in README file in the example directory.
I suppose the README doesn't do a good job of explaining the model file, its extension, its path or how to train a model step by step. If someone can explain step by step the preprocessing to get the required data, ways to train the preprocessed file, I'd really appreciate it. Right now, I get some gibberish output using the README commands which is of no use. Thanks
The README is terribly confusing. How do you use the pcl model files? The run.sh script doesn't reference them at all.
Can someone explain me how to use it, step by step?