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

Rename backend name from “torch_distributed” to “ray” in Orca Pytorch Estimator (#4388)

Browse source 
* change backend name from torch_distributed to ray

* update in examples and docs
This commit is contained in:
Shan Yu 2022-04-12 09:29:21 +08:00 committed by GitHub
parent 0c0e99fee5
commit 71cd125837
3 changed files with 4 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
docs/readthedocs/source/doc/Orca/Overview/distributed-training-inference.md
View file

@ -181,7 +181,7 @@ View the related [Python API doc](https://bigdl.readthedocs.io/en/latest/doc/Pyt
**Using `torch.distributed` or *Horovod* backend** **Using `torch.distributed` or *Horovod* backend**
Alternatively, users can create a PyTorch `Estimator` using `torch.distributed` or *Horovod* backend by specifying the `backend` argument to be "torch_distributed" or "horovod". In this case, the `model` and `optimizer` should be wrapped in _Creater Functions_. For example: Alternatively, users can create a PyTorch `Estimator` using `torch.distributed` or *Horovod* backend by specifying the `backend` argument to be "ray" or "horovod". In this case, the `model` and `optimizer` should be wrapped in _Creater Functions_. For example:
```python ```python
def model_creator(config): def model_creator(config):
@ -196,7 +196,7 @@ est = Estimator.from_torch(model=model,
optimizer=optimizer_creator, optimizer=optimizer_creator,
loss=nn.NLLLoss(), loss=nn.NLLLoss(),
config={"lr": 1e-2}, config={"lr": 1e-2},
backend="torch_distributed") # or backend="horovod" backend="ray") # or backend="horovod"
``` ```
Then users can perform distributed model training and inference as follows: Then users can perform distributed model training and inference as follows:

2
docs/readthedocs/source/doc/Orca/QuickStart/orca-pytorch-distributed-quickstart.md
View file

@ -118,7 +118,7 @@ from bigdl.orca.learn.pytorch import Estimator
from bigdl.orca.learn.metrics import Accuracy from bigdl.orca.learn.metrics import Accuracy
est = Estimator.from_torch(model=model_creator, optimizer=optim_creator, loss=criterion, metrics=[Accuracy()], est = Estimator.from_torch(model=model_creator, optimizer=optim_creator, loss=criterion, metrics=[Accuracy()],
backend="torch_distributed") backend="ray")
``` ```
Next, fit and evaluate using the Estimator Next, fit and evaluate using the Estimator

2
docs/readthedocs/source/doc/PythonAPI/Orca/orca.rst
View file

@ -62,7 +62,7 @@ orca.learn.pytorch.estimator
orca.learn.pytorch.pytorch_ray_estimator orca.learn.pytorch.pytorch_ray_estimator
^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^^^^^^^^^^
Orca Pytorch Estimator with backend of "horovod" or "torch_distributed". Orca Pytorch Estimator with backend of "horovod" or "ray".
.. autoclass:: bigdl.orca.learn.pytorch.pytorch_ray_estimator.PyTorchRayEstimator .. autoclass:: bigdl.orca.learn.pytorch.pytorch_ray_estimator.PyTorchRayEstimator
:members: :members: