jerry/OpenBB
1
0
Fork 0
mirror of https://github.com/OpenBB-finance/OpenBB.git synced 2026-11-07 09:52:13 +00:00
Code Issues Packages Projects Releases Wiki Activity

codespell

Browse Source
This commit is contained in:
Jamie
2021-05-14 15:41:10 -04:00
parent 461f705694
commit 0032d794ed
2 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
gamestonk_terminal/prediction_techniques/FUNE-TUNING-NN-models.md
View File

@@ -23,7 +23,7 @@ To this end, we allow users to specify how much validation data to use and if it
into 10% validation data with a random shuffle. This will allow the user to see how the model performs at various times in your sequence.
```
-v/--valid : validation split of data. Deafult 0.1 (10%)
-v/--valid : validation split of data. Default 0.1 (10%)
--no_shuffle : Flag that will order the validation data so that the last (-v) percent of data is the validation.
```
To try "backtesting" to a certain data, you can specify the end date. This will forecast and compare the true data from that time.
@@ -103,7 +103,7 @@ This model is set in [config_neural_network_models.py](/config_neural_network_mo
**Optimizer technique**:Adaptive Moment Estimation (_adam_) is usually the default choice for this type of problems. See https://www.tensorflow.org/api_docs/python/tf/keras/optimizers.
The optimizer selection has been moved to teh config_neural_network_models.py file and the new argument is the learning rate.
The optimizer selection has been moved to the config_neural_network_models.py file and the new argument is the learning rate.
This tells your model how "fast" to train by adjusting the factor that gets multiplied to the loss function for updating weights
```
--lr : learning rate. Default 0.01.

2
gamestonk_terminal/prediction_techniques/pred_helper.py
View File

@@ -473,7 +473,7 @@ def plot_data_predictions(
alpha=0.2,
)
# Leave this one out of the loop so that the legend doesnt get overpopulated with "Predictions"
# Leave this one out of the loop so that the legend doesn't get overpopulated with "Predictions"
if scaler:
final_pred = scaler.inverse_transform(preds[-1].reshape(-1, 1)).ravel()
final_valid = scaler.inverse_transform(y_valid[-1].reshape(-1, 1)).ravel()