search

microsoft / AzureML-BERT

End-to-End recipes for pre-training and fine-tuning BERT using Azure Machine Learning Service
https://azure.microsoft.com/en-us/blog/microsoft-makes-it-easier-to-build-popular-language-representation-model-bert-at-large-scale/
MIT License
394 stars 127 forks source link

broken modified run_classifier.py link for hyperdrive #51

Open sriramjcet opened 4 years ago

sriramjcet commented 4 years ago

Hi, I am trying to follow the tutorial to finetune BERT and perform hyperparamter tuning. The instruction refers to adding logging of metrics to run_classifier.py and refers to the modified run_classifier.py in hyperlink. But that is broken. Can you please share the modified run_classifier.py training script that has logging included for hyperdrive? Appreciate the help. Copy pasted the instructions from the tutorial below

Further within run_classifier.py, we log the learning rate, and the epoch training and eval loss the model achieves:

run.log('lr', np.float(args.learning_rate))

run.log('train_mean_loss', mean_loss) run.log('eval_mean_loss', mean_loss) run.log('train_example_loss', mean_loss) run.log('eval_example_loss', mean_loss) These run metrics will become particularly important when we begin hyperparameter tuning our model in the "Tune model hyperparameters" section.

Let's first copy the modified run_classifier.py into our local project directory this link is broken Thanks, Sriram