Modulus Release Update (#123)

* Updating for new modulus release

* update to cuda12 jax in toml

README.md

@@ -39,18 +39,10 @@ for more details and other options.
 
 ## Getting Started
 
-The [notebooks](./examples/notebooks/) folder contains reference Jupyter
-notebooks for a few common use cases.
-These are the best ways to start exploring Earth-2 MIP:
-
-1. [Ensemble Inference](./examples/notebooks/01_ensemble_inference.ipynb) - This
-notebook will demonstrate how to perturb the initial conditions to generate ensemble of
-inference outputs for a given pre-trained model and generate some standard metrics to
-evaluate the skill of the model.
-
-2. [Model Inter Comparison](./examples/notebooks/02_model_comparison.ipynb) - This
-notebook will showcase how to generate forecasts from multiple models with Earth-2 MIP
-and how to compare their predictions.
+Earth-2 MIP provides a set of examples which can be viewed on the [examples documentation](https://nvidia.github.io/earth2mip/examples/index.html)
+page which can be used to get started with various workflows.
+These examples can be downloaded both as Jupyer Notebooks and Python scripts.
+The source Python scripts can be found in the [examples](./examples/) folders.
 
 ### Basic Inference
 
@@ -139,11 +131,8 @@ table below.
 \* = coming soon
 
 Some models require additional dependencies not installed by default.
-To install all optional model dependencies use:
-
-```python
-pip install .[pangu,graphcast]
-```
+Refer to the [installation instructions](https://nvidia.github.io/earth2mip/userguide/install.html)
+for details.
 
 We want to integrate your model into the scoreboard to show the community!
 The best way to do this is via [NVIDIA Modulus](https://github.com/NVIDIA/modulus).

pyproject.toml

@@ -29,7 +29,7 @@ dependencies = [
     "joblib>=1.1.0",
     "netCDF4>=1.6.4",
     "numpy",
-    "nvidia-modulus >= 0.3.0",
+    "nvidia-modulus >= 0.4.0",
     "pandas>=1.5.3",
     "properscoring>=0.1",
     "pydantic >= 1.10, <=1.10.11",
@@ -59,7 +59,7 @@ pangu = [
     "onnxruntime-gpu>=1.15.1",
 ]
 graphcast = [
-    "jax>=0.4.16",
+    "jax[cuda12_pip]>=0.4.16",
     "graphcast @ https://github.com/deepmind/graphcast/archive/e622a15c1b9742d78f9b662f2af262604d58c204.tar.gz",
     "gcsfs>=2023.6.0",
 ]

requirements.txt

@@ -1,6 +1,2 @@
 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
-jax[cuda12_local]==0.4.16
-https://github.com/deepmind/graphcast/archive/e622a15c1b9742d78f9b662f2af262604d58c204.tar.gz
-gcsfs
-# TODO point to modulus main once https://github.com/NVIDIA/modulus/pull/194 is merged
-nvidia-modulus@git+https://github.com/nvidia/modulus.git#egg=ea5d2c35998ad0c1c00407b0ecd2bf6049aef5e7
+jax[cuda12_local]==0.4.16