Do I need to retrain the model once I load it from the transformers library ? I ask this question because I got poor results from the default real time speech recognition pipeline. Even after adding a sleep of 2 seconds as mentioned in your shared tutorial.
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Does the model load on the memory? | 2 | 501 | January 10, 2024 | |
Converting DeepSpeech model to Transformers | 0 | 304 | September 17, 2021 | |
Pretrain a model on a very specific language for NER | 0 | 371 | September 28, 2023 | |
How can I pretrain a new model re-initializing with my own vocab? | 0 | 287 | May 25, 2021 | |
How is memory managed when loading a model? | 2 | 5939 | July 4, 2023 |