Skip to content
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.

Fix AMP Tutorial warnings and failures #15526

Merged
merged 1 commit into from
Jul 18, 2019
Merged

Fix AMP Tutorial warnings and failures #15526

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 9 additions & 9 deletions docs/tutorials/amp/amp_tutorial.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -262,19 +262,20 @@ Below, we demonstrate for a gluon model and a symbolic model:
```python
with mx.Context(mx.gpu(0)):
# Below is an example of converting a gluon hybrid block to a mixed precision block
model = get_model("resnet50_v1")
model.collect_params().initialize(ctx=mx.current_context())
model.hybridize()
model(mx.nd.zeros((1, 3, 224, 224)))
converted_model = amp.convert_hybrid_block(model)
with warnings.catch_warnings(record=True) as w:
sandeep-krishnamurthy marked this conversation as resolved.
Show resolved Hide resolved
warnings.simplefilter("ignore")
model = get_model("resnet50_v1")
model.collect_params().initialize(ctx=mx.current_context())
model.hybridize()
model(mx.nd.zeros((1, 3, 224, 224)))
converted_model = amp.convert_hybrid_block(model)

# Run dummy inference with the converted gluon model
result = converted_model.forward(mx.nd.random.uniform(shape=(1, 3, 224, 224),
dtype=np.float32))

# Below is an example of converting a symbolic model to a mixed precision model
dir_path = os.path.dirname(os.path.realpath(__file__))
model_path = os.path.join(dir_path, 'model')
model_path = "model"
sandeep-krishnamurthy marked this conversation as resolved.
Show resolved Hide resolved
if not os.path.isdir(model_path):
os.mkdir(model_path)
prefix, epoch = mx.test_utils.download_model("imagenet1k-resnet-18", dst_dir=model_path)
Expand All @@ -301,8 +302,7 @@ for symbolic model. You can do the same for gluon hybrid block with `amp.convert
with mx.Context(mx.gpu(0)):
# Below is an example of converting a symbolic model to a mixed precision model
# with only Convolution op being force casted to FP16.
dir_path = os.path.dirname(os.path.realpath(__file__))
model_path = os.path.join(dir_path, 'model')
model_path = "model"
if not os.path.isdir(model_path):
os.mkdir(model_path)
prefix, epoch = mx.test_utils.download_model("imagenet1k-resnet-18", dst_dir=model_path)
Expand Down