ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

Update Orca 5min (#6524)

Browse source
This commit is contained in:
Kai Huang 2022-11-10 10:57:19 +08:00 committed by GitHub
parent 7da102243e
commit 9c99751288
1 changed files with 5 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
docs/readthedocs/source/doc/Orca/Overview/orca.md
View file

@ -50,6 +50,7 @@ Finally, use [sklearn-style Estimator APIs in Orca](distributed-training-inferen
```python ```python
from bigdl.orca.learn.tf2.estimator import Estimator from bigdl.orca.learn.tf2.estimator import Estimator
# Define the NCF model in standard TensorFlow API
def model_creator(config): def model_creator(config):
from tensorflow import keras from tensorflow import keras
@ -79,6 +80,8 @@ val_steps = int(test_df.count() / batch_size)
est = Estimator.from_keras(model_creator=model_creator, backend="spark", est = Estimator.from_keras(model_creator=model_creator, backend="spark",
config={"embed_dim": 8, "num_users": num_users, "num_items": num_items}) config={"embed_dim": 8, "num_users": num_users, "num_items": num_items})
# Distributed training
est.fit(data=train_df, est.fit(data=train_df,
batch_size=batch_size, batch_size=batch_size,
epochs=4, epochs=4,
@ -87,6 +90,8 @@ est.fit(data=train_df,
steps_per_epoch=train_steps, steps_per_epoch=train_steps,
validation_data=test_df, validation_data=test_df,
validation_steps=val_steps) validation_steps=val_steps)
# Distributed inference
prediction_df = est.predict(test_df, prediction_df = est.predict(test_df,
batch_size=batch_size, batch_size=batch_size,
feature_cols=['user', 'item'], feature_cols=['user', 'item'],