Merge branch 'iwslt_22' of github.com:NVIDIA/NeMo into iwslt_22

CONTRIBUTING.md

@@ -29,7 +29,7 @@ pytest --with_downloads
 1. For changes to NeMo's core: @ericharper, @titu1994, @blisc, or @okuchaiev  
 1. For changes to NeMo's ASR collection: @titu1994, @redoctopus, @jbalam-nv, or @okuchaiev
 1. For changes to NeMo's NLP collection: @MaximumEntropy, @ericharper, @ekmb, @yzhang123, @VahidooX, @vladgets, or @okuchaiev 
-1. For changes to NeMo's TTS collection: @blisc or @Oktai15, or @okuchaiev
+1. For changes to NeMo's TTS collection: @blisc, or @okuchaiev
 
 Note that some people may self-assign to review your PR - in which case, please wait for them to add a review.
 
@@ -44,6 +44,17 @@ Your  pull requests must pass all checks and peer-review before they can be merg
 1. **Legal**: if you copy even one line of code from the Internet, make sure that the code allows the license that NeMo supports. Give credit and link back to the code.
 1. **Sensible**: code should make sense. If you think a piece of code might be confusing, write comments.
 
+## Class naming conventions
+* No “I”, “Interface”, “NM” nor “NeMo” pre/postfixes anywhere
+* Core interfaces have simple names: Typing, Cloud, Serialization, FileIO*
+* Core classes have the simplest names ever: NeuralModule, Model, Graph, Dataset, Loss, Module*
+* Abstract classes in the Model hierarchy have Model postfix
+* A config class for MyModel should be called MyModelConfig
+* Leaf Neural Module classes have simple names without any postfixes (e.g. AudioPreprocess)
+* Leaf Datasets have Dataset postfix (e.g. MNISTDataset)
+* Leaf Losses have Loss postfix (e.g. CTCLoss)
+* Leaf Models do not have any postfix, just name (e.g. QuartzNet)
+
 ## Python style
 We use ``black`` as our style guide. To check whether your code will pass style check (from the NeMo's repo folder) run:
 ``python setup.py style`` and if it does not pass run ``python setup.py style --fix``.

Dockerfile

@@ -14,7 +14,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-ARG BASE_IMAGE=nvcr.io/nvidia/pytorch:22.01-py3
+ARG BASE_IMAGE=nvcr.io/nvidia/pytorch:22.02-py3
 
 
 # build an image that includes only the nemo dependencies, ensures that dependencies
@@ -60,13 +60,17 @@ RUN for f in $(ls requirements*.txt); do pip install --disable-pip-version-check
 COPY nemo_text_processing /tmp/nemo/nemo_text_processing/
 RUN /bin/bash /tmp/nemo/nemo_text_processing/setup.sh
 
+# install k2, skip if installation fails
+COPY scripts /tmp/nemo/scripts/
+RUN /bin/bash /tmp/nemo/scripts/speech_recognition/k2/setup.sh; exit 0
+
 # copy nemo source into a scratch image
 FROM scratch as nemo-src
 COPY . .
 
 # start building the final container
 FROM nemo-deps as nemo
-ARG NEMO_VERSION=1.7.0
+ARG NEMO_VERSION=1.8.0
 
 # Check that NEMO_VERSION is set. Build will fail without this. Expose NEMO and base container
 # version information as runtime environment variable for introspection purposes
@@ -81,7 +85,7 @@ RUN --mount=from=nemo-src,target=/tmp/nemo cd /tmp/nemo && pip install ".[all]"
     python -c "import nemo.collections.tts as nemo_tts" && \
     python -c "import nemo_text_processing.text_normalization as text_normalization"
 
-# TODO: Update to newer numba 0.56.0RC1 for 22.02 container
+# TODO: Update to newer numba 0.56.0RC1 for 22.03 container if possible
 # install pinned numba version
 # RUN conda install -c conda-forge numba==0.54.1