Add files via upload

Create LICENSE.md

README.md

@@ -2,23 +2,103 @@
 
 This is the official repository for the Pangu-Weather paper.
 
-Resources including pseudocode and pre-trained models will be updated. Stay tuned!
+[Pangu-Weather: A 3D High-Resolution Model for Fast and Accurate Global Weather Forecast](https://arxiv.org/abs/2211.02556), arXiv preprint: 2211.02556, 2022.
 
-#### Policy of using the contents
+by Kaifeng Bi, Lingxi Xie, Hengheng Zhang, Xin Chen, Xiaotao Gu and Qi Tian
 
-All models are trained using the ERA5 dataset provided by ECMWF. Please follow their policy and note that commercial use of these models are forbidden.
+Resources including pseudocode, pre-trained models, and inference code are released.
 
-More policy to be updated.
+## Installation
 
-#### Pseudocode and how to use
+The downloaded files shall be organized as the following hierarchy:
 
-To be updated.
+```plain
+├── root
+│   ├── input_data
+│   │   ├── input_surface.npy
+│   │   ├── input_upper.npy
+│   ├── output_data
+│   ├── model_jit_cpu_1.onnx
+│   ├── model_jit_cpu_3.onnx
+│   ├── model_jit_cpu_6.onnx
+│   ├── model_jit_cpu_24.onnx
+│   ├── inference_cpu.py
+│   ├── inference_gpu.py
+│   ├── inference_iterative.py
+```
 
-#### Pre-trained models
+If you use a CPU environment, please run:
+```
+pip install -r requirement_cpu.txt
+```
 
-To be updated.
+If you use a GPU environment, please first confirm that the cuda version is 11.6 and the cudnn version is the 8.2.4 for Linux and 8.5.0.96 for Windows (please see [this page](https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html) for details). Then, please run:
+```
+pip install -r requirement_gpu.txt
+```
 
-#### References
+## Global weather forecasting (inference) using the trained models
+
+#### Downloading trained models
+
+Please download the four pre-trained models (~1.1GB each) from the Google drive:
+
+The 1-hour model: [model_jit_cpu_1.onnx](https://drive.google.com/file/d/1fg5jkiN_5dHzKb-5H9Aw4MOmfILmeY-S/view?usp=share_link)
+
+The 3-hour model: [model_jit_cpu_3.onnx](https://drive.google.com/file/d/1EdoLlAXqE9iZLt9Ej9i-JW9LTJ9Jtewt/view?usp=share_link)
+
+The 6-hour model: [model_jit_cpu_6.onnx](https://drive.google.com/file/d/1a4XTktkZa5GCtjQxDJb_fNaqTAUiEJu4/view?usp=share_link)
+
+The 24-hour model: [model_jit_cpu_24.onnx](https://drive.google.com/file/d/1lweQlxcn9fG0zKNW8ne1Khr9ehRTI6HP/view?usp=share_link)
+
+These models are stored using the ONNX format, and thus can be used via different languages such as Python, C++, C#, Java, etc.
+
+#### Input data preparation using Python
+
+Please prepare the input data using [numpy](https://numpy.org/). There are two files that shall be put under the `input_data` folder, namely, `input_surface.npy` and `input_upper.npy`.
+
+`input_surface.npy` stores the input surface variables. It is a numpy array shaped (4,721,1440) where the first dimension represents the 4 surface variables (MSLP, U10, V10, T2M **in the exact order**).
+
+`input_upper.npy` stores the upper-air variables. It is a numpy array shaped (5,13,721,1440) where the first dimension represents the 5 surface variables (Z, Q, T, U and V **in the exact order**), and the second dimension represents the 13 pressure levels (1000hPa, 925hPa, 850hPa, 700hPa, 600hPa, 500hPa, 400hPa, 300hPa, 250hPa, 200hPa, 150hPa, 100hPa and 50hPa **in the exact order**).
+
+In both cases, the dimensions of 721 and 1440 represent the size along the latitude and longitude, where the numerical range is [90,-90] degree and [0,359.75] degree, respectively, and the spacing is 0.25 degrees. For each 721x1440 slice, the data format is exactly the same as the `.nc` file download from the ERA5 official website.
+
+Note that the numpy arrays should be in single precision (`.astype(np.float32)`), not in double precision.
+
+We support ERA5 initial fields and ECMWF initial fields (e.g., the initial fields of the HRES forecast), where the latter often leads to a slight accuracy drop (mainly for T2M because the two fields are quite different in temperature). A `.nc` file of ERA5 can be transformed into a `.npy` file using the netCDF4 package, and a `.grib` file of the ECMWF initial fields can be transformed into a `.npy` file using the pygrib package. Note that Z represents geopotential, not geopotential height, so a factor of 9.80665 should be multiplied if the raw data contains the geopotential height.
+
+We temporarily do not support other kinds of initial fields due to the possibly dramatic differences in the fields when Z<0.
+
+We provide an example of transferred input files, `input_surface.npy` and `input_upper.npy`, which correspond to the ERA5 initial fields of at 12:00UTC, 2018/09/27. Please download them using Google drive:
+
+[`input_surface.npy`](https://drive.google.com/file/d/1pj8QEVNpC1FyJfUabDpV4oU3NpSe0BkD/view?usp=share_link)
+
+[`input_upper.npy`](https://drive.google.com/file/d/1--7xEBJt79E3oixizr8oFmK_haDE77SS/view?usp=share_link)
+
+#### Inference
+
+After the above steps are finished, please check `inference_cpu.py` for an example of making a 24-hour weather forecast on CPU with the 24-hour model, and `inference_gpu.py` for the GPU version.
+
+For example, running the following command, one can get the 24-hour forecast in the `output_data` folder:
+```
+python inference_cpu.py # python inference_gpu.py for gpu environment
+```
+
+Also, `inference_iterative.py` shows an example to generate per-6-hour forecast within a week.
+
+## Pseudocode and how to use
+
+`pseudocode.py` contains the pseudocode that elaborates our main algorithm. It is written in Python and can be implemented using any deep learning library, e.g. PyTorch and TensorFlow.
+
+Note that one needs to download about 60TB of ERA5 data and prepare for computational resource of 3000 GPU-days (in V100) to train each model.
+
+## License
+
+Pangu-Weather is released by Huawei Cloud. **The commercial use of these models is forbidden.**
+
+Also, please note that all models were trained using the ERA5 dataset provided by ECMWF. Please do follow [their policy](https://apps.ecmwf.int/datasets/licences/copernicus/).
+
+## References
 
 If you use the resource in your research, please cite our paper:
 

inference_cpu.py

@@ -0,0 +1,36 @@
+import os
+import numpy as np
+import onnx
+import onnxruntime as ort
+
+
+# The directory of your input and output data
+input_data_dir = 'input_data'
+output_data_dir = 'output_data'
+model_24 = onnx.load('pangu_weather_24.onnx')
+
+# Set the behavier of onnxruntime
+options = ort.SessionOptions()
+options.enable_cpu_mem_arena=False
+options.enable_mem_pattern = False
+options.enable_mem_reuse = False
+# Increase the number for faster inference and more memory consumption
+options.intra_op_num_threads = 1
+
+# Set the behavier of cuda provider
+cuda_provider_options = {'arena_extend_strategy':'kSameAsRequested',}
+
+# Initialize onnxruntime session for Pangu-Weather Models
+ort_session_24 = ort.InferenceSession('pangu_weather_24.onnx', sess_options=options, provider=['CPUExecutionProvider'])
+
+# Load the upper-air numpy arrays
+input = np.load(os.path.join(input_data_dir, 'input_upper.npy')).astype(np.float32)
+# Load the surface numpy arrays
+input_surface = np.load(os.path.join(input_data_dir, 'input_surface.npy')).astype(np.float32)
+
+# Run the inference session
+output, output_surface = ort_session_24.run(None, {'input':input, 'input_surface':input_surface})
+
+# Save the results
+np.save(os.path.join(output_data_dir, 'output_upper'), output)
+np.save(os.path.join(output_data_dir, 'output_surface'), output_surface)

inference_gpu.py

@@ -0,0 +1,35 @@
+import os
+import numpy as np
+import onnx
+import onnxruntime as ort
+
+
+# The directory of your input and output data
+input_data_dir = 'input_data'
+output_data_dir = 'output_data'
+model_24 = onnx.load('pangu_weather_24.onnx')
+
+# Set the behavier of onnxruntime
+options = ort.SessionOptions()
+options.enable_cpu_mem_arena=False
+options.enable_mem_pattern = False
+options.enable_mem_reuse = False
+# Increase the number for faster inference and more memory consumption
+options.intra_op_num_threads = 1
+
+# Set the behavier of cuda provider
+cuda_provider_options = {'arena_extend_strategy':'kSameAsRequested',}
+
+# Initialize onnxruntime session for Pangu-Weather Models
+ort_session_24 = ort.InferenceSession('pangu_weather_24.onnx', sess_options=options, provider=[('CUDAExecutionProvider', cuda_provider_options)])
+
+# Load the upper-air numpy arrays
+input = np.load(os.path.join(input_data_dir, 'input_upper.npy')).astype(np.float32)
+# Load the surface numpy arrays
+input_surface = np.load(os.path.join(input_data_dir, 'input_surface.npy')).astype(np.float32)
+
+# Run the inference session
+output, output_surface = ort_session_24.run(None, {'input':input, 'input_surface':input_surface})
+# Save the results
+np.save(os.path.join(output_data_dir, 'output_upper'), output)
+np.save(os.path.join(output_data_dir, 'output_surface'), output_surface)

inference_iterative.py

@@ -0,0 +1,42 @@
+import os
+import numpy as np
+import onnx
+import onnxruntime as ort
+
+
+# The directory of your input and output data
+input_data_dir = 'input_data'
+output_data_dir = 'output_data'
+model_24 = onnx.load('pangu_weather_24.onnx')
+model_6 = onnx.load('pangu_weather_6.onnx')
+
+# Set the behavier of onnxruntime
+options = ort.SessionOptions()
+options.enable_cpu_mem_arena=False
+options.enable_mem_pattern = False
+options.enable_mem_reuse = False
+# Increase the number for faster inference and more memory consumption
+options.intra_op_num_threads = 1
+
+# Set the behavier of cuda provider
+cuda_provider_options = {'arena_extend_strategy':'kSameAsRequested',}
+
+# Initialize onnxruntime session for Pangu-Weather Models
+ort_session_24 = ort.InferenceSession('pangu_weather_24.onnx', sess_options=options, provider=[('CUDAExecutionProvider', cuda_provider_options)])
+ort_session_6 = ort.InferenceSession('pangu_weather_6.onnx', sess_options=options, provider=[('CUDAExecutionProvider', cuda_provider_options)])
+
+# Load the upper-air numpy arrays
+input = np.load(os.path.join(input_data_dir, 'input_upper.npy')).astype(np.float32)
+# Load the surface numpy arrays
+input_surface = np.load(os.path.join(input_data_dir, 'input_surface.npy')).astype(np.float32)
+
+# Run the inference session
+input_24, input_surface_24 = input, input_surface
+for i in range(28):
+  if (i+1) % 4 == 0:
+    output, output_surface = ort_session_24.run(None, {'input':input_24, 'input_surface':input_surface_24})
+    input_24, input_surface_24 = output, output_surface
+  else:
+    output, output_surface = ort_session_6.run(None, {'input':input, 'input_surface':input_surface})
+  input, input_surface = output, output_surface
+  # Your can save the results here

pseudocode.py

@@ -0,0 +1,598 @@
+'''
+Pseudocode of Pangu-Weather
+'''
+# The pseudocode can be implemented using deep learning libraries, e.g., Pytorch and Tensorflow or other high-level APIs
+
+# Basic operations used in our model, namely Linear, Conv3d, Conv2d, ConvTranspose3d and ConvTranspose2d
+# Linear: Linear transformation, available in all deep learning libraries
+# Conv3d and Con2d: Convolution with 2 or 3 dimensions, available in all deep learning libraries
+# ConvTranspose3d, ConvTranspose2d: transposed convolution with 2 or 3 dimensions, see Pytorch API or Tensorflow API
+from Your_AI_Library import Linear, Conv3d, Conv2d, ConvTranspose3d, ConvTranspose2d
+
+# Functions in the networks, namely GeLU, DropOut, DropPath, LayerNorm, and SoftMax
+# GeLU: the GeLU activation function, see Pytorch API or Tensorflow API
+# DropOut: the dropout function, available in all deep learning libraries
+# DropPath: the DropPath function, see the implementation of vision-transformer, see timm pakage of Pytorch
+# LayerNorm: the layer normalization function, see Pytorch API or Tensorflow API
+# Softmax: softmax function, see Pytorch API or Tensorflow API
+from Your_AI_Library import GeLU, DropOut, DropPath, LayerNorm, SoftMax
+
+# Common functions for roll, pad, and crop, depends on the data structure of your software environment
+from Your_AI_Library import roll3D, pad3D, pad2D, Crop3D, Crop2D
+
+# Common functions for reshaping and changing the order of dimensions
+# reshape: change the shape of the data with the order unchanged, see Pytorch API or Tensorflow API
+# TransposeDimensions: change the order of the dimensions, see Pytorch API or Tensorflow API
+from Your_AI_Library import reshape, TransposeDimensions
+
+# Common functions for creating new tensors
+# ConstructTensor: create a new tensor with an arbitrary shape
+# TruncatedNormalInit: Initialize the tensor with Truncate Normalization distribution
+# RangeTensor: create a new tensor like range(a, b)
+from Your_AI_Library import ConstructTensor, TruncatedNormalInit, RangeTensor
+
+# Common operations for the data, you may design it or simply use deep learning APIs default operations
+# LinearSpace: a tensor version of numpy.linspace
+# MeshGrid: a tensor version of numpy.meshgrid
+# Stack: a tensor version of numpy.stack
+# Flatten: a tensor version of numpy.ndarray.flatten
+# TensorSum: a tensor version of numpy.sum
+# TensorAbs: a tensor version of numpy.abs
+# Concatenate: a tensor version of numpy.concatenate
+from Your_AI_Library import LinearSpace, MeshGrid, Stack, Flatten, TensorSum, TensorAbs, Concatenate
+
+# Common functions for training models
+# LoadModel and SaveModel: Load and save the model, some APIs may require further adaptation to hardwares
+# Backward: Gradient backward to calculate the gratitude of each parameters
+# UpdateModelParametersWithAdam: Use Adam to update parameters, e.g., torch.optim.Adam
+from Your_AI_Library import LoadModel, Backward, UpdateModelParametersWithAdam, SaveModel
+
+# Custom functions to read your data from the disc
+# LoadData: Load the ERA5 data
+# LoadConstantMask: Load constant masks, e.g., soil type
+# LoadStatic: Load mean and std of the ERA5 training data, every fields such as T850 is treated as an image and calculate the mean and std
+from Your_Data_Code import LoadData, LoadConstantMask, LoadStatic
+
+
+def Inference(input, input_surface, forecast_range):
+  '''Inference code, describing the algorithm of inference using models with different lead times. 
+  PanguModel24, PanguModel6, PanguModel3 and PanguModel1 share the same training algorithm but differ in lead times.
+  Args:
+    input: input tensor, need to be normalized to N(0, 1) in practice
+    input_surface: target tensor, need to be normalized to N(0, 1) in practice
+    forecast_range: iteration numbers when roll out the forecast model
+  '''
+
+  # Load 4 pre-trained models with different lead times
+  PanguModel24 = LoadModel(ModelPath24)
+  PanguModel6 = LoadModel(ModelPath6)
+  PanguModel3 = LoadModel(ModelPath3)
+  PanguModel1 = LoadModel(ModelPath1)
+
+  # Load mean and std of the weather data
+  weather_mean, weather_std, weather_surface_mean, weather_surface_std = LoadStatic()
+
+  # Store initial input for different models
+  input_24, input_surface_24 = input, input_surface
+  input_6, input_surface_6 = input, input_surface
+  input_3, input_surface_3 = input, input_surface
+
+  # Using a list to store output
+  output_list = []
+
+  # Note: the following code is implemented for fast inference of [1,forecast_range]-hour forecasts -- if only one lead time is requested, the inference can be much faster.
+  for i in range(forecast_range):
+    # switch to the 24-hour model if the forecast time is 24 hours, 48 hours, ..., 24*N hours
+    if (i+1) % 24 == 0:
+      # Switch the input back to the stored input
+      input, input_surface = input_24, input_surface_24
+
+      # Call the model pretrained for 24 hours forecast
+      output, output_surface = PanguModel24(input, input_surface)
+
+      # Restore from uniformed output
+      output = output * weather_std + weather_mean
+      output_surface = output_surface * weather_surface_std + weather_surface_mean
+
+      # Stored the output for next round forecast
+      input_24, input_surface_24 = output, output_surface
+      input_6, input_surface_6 = output, output_surface
+      input_3, input_surface_3 = output, output_surface
+
+    # switch to the 6-hour model if the forecast time is 30 hours, 36 hours, ..., 24*N + 6/12/18 hours
+    elif (i+1) % 6 == 0:
+      # Switch the input back to the stored input
+      input, input_surface = input_6, input_surface_6
+
+      # Call the model pretrained for 6 hours forecast
+      output, output_surface = PanguModel6(input, input_surface)
+
+      # Restore from uniformed output
+      output = output * weather_std + weather_mean
+      output_surface = output_surface * weather_surface_std + weather_surface_mean
+      
+      # Stored the output for next round forecast
+      input_6, input_surface_6 = output, output_surface
+      input_3, input_surface_3 = output, output_surface
+
+    # switch to the 3-hour model if the forecast time is 3 hours, 9 hours, ..., 6*N + 3 hours
+    elif (i+1) % 3 ==0:
+      # Switch the input back to the stored input
+      input, input_surface = input_3, input_surface_3
+
+      # Call the model pretrained for 3 hours forecast
+      output, output_surface = PanguModel3(input, input_surface)
+
+      # Restore from uniformed output
+      output = output * weather_std + weather_mean
+      output_surface = output_surface * weather_surface_std + weather_surface_mean
+      
+      # Stored the output for next round forecast
+      input_3, input_surface_3 = output, output_surface
+
+    # switch to the 1-hour model
+    else:
+      # Call the model pretrained for 1 hours forecast
+      output, output_surface = PanguModel1(input, input_surface)
+
+      # Restore from uniformed output
+      output = output * weather_std + weather_mean
+      output_surface = output_surface * weather_surface_std + weather_surface_mean
+
+    # Stored the output for next round forecast
+    input, input_surface = output, output_surface
+
+    # Save the output
+    output_list.append((output, output_surface))
+  return output_list
+
+
+def Train():
+  '''Training code'''
+  # Initialize the model, for some APIs some adaptation is needed to fit hardwares
+  model = PanguModel()
+
+  # Train single Pangu-Weather model
+  epochs = 100
+  for i in range(epochs):
+    # For each epoch, we iterate from 1979 to 2017
+    # dataset_length is the length of your training data, e.g., the sample between 1979 and 2017
+    for step in range(dataset_length):
+      # Load weather data at time t as the input; load weather data at time t+1/3/6/24 as the output
+      # Note the data need to be randomly shuffled
+      # Note the input and target need to be normalized, see Inference() for details
+      input, input_surface, target, target_surface = LoadData(step)
+
+      # Call the model and get the output
+      output, output_surface = model(input, input_surface)
+
+      # We use the MAE loss to train the model
+      # The weight of surface loss is 0.25
+      # Different weight can be applied for differen fields if needed
+      loss = TensorAbs(output-target) + TensorAbs(output_surface-target_surface) * 0.25
+
+      # Call the backward algorithm and calculate the gratitude of parameters
+      Backward(loss)
+
+      # Update model parameters with Adam optimizer
+      # The learning rate is 5e-4 as in the paper, while the weight decay is 3e-6
+      # A example solution is using torch.optim.adam
+      UpdateModelParametersWithAdam()
+
+  # Save the model at the end of the training stage
+  SaveModel()
+
+class PanguModel:
+  def __init__(self):
+    # Drop path rate is linearly increased as the depth increases
+    drop_path_list = LinearSpace(0, 0.2, 8)
+
+    # Patch embedding
+    self._input_layer = PatchEmbedding((2, 4, 4), 192)
+
+    # Four basic layers
+    self.layer1 = EarthSpecificLayer(2, 192, drop_list[:2], 6)
+    self.layer2 = EarthSpecificLayer(6, 384, drop_list[6:], 12)
+    self.layer3 = EarthSpecificLayer(6, 384, drop_list[6:], 12)
+    self.layer4 = EarthSpecificLayer(2, 192, drop_list[:2], 6)
+
+    # Upsample and downsample
+    self.upsample = UpSample(384, 192)
+    self.downsample = DownSample(192)
+
+    # Patch Recovery
+    self._output_layer = PatchRecovery(384)
+    
+  def forward(self, input, input_surface):
+    '''Backbone architecture'''
+    # Embed the input fields into patches
+    x = self._input_layer(input, input_surface)
+
+    # Encoder, composed of two layers
+    # Layer 1, shape (8, 360, 181, C), C = 192 as in the original paper
+    x = self.layer1(x, 8, 360, 181) 
+
+    # Store the tensor for skip-connection
+    skip = x
+
+    # Downsample from (8, 360, 181) to (8, 180, 91)
+    x = self.downsample(x, 8, 360, 181)
+
+    # Layer 2, shape (8, 180, 91, 2C), C = 192 as in the original paper
+    x = self.layer2(x, 8, 180, 91) 
+
+    # Decoder, composed of two layers
+    # Layer 3, shape (8, 180, 91, 2C), C = 192 as in the original paper
+    x = self.layer3(x, 8, 180, 91) 
+
+    # Upsample from (8, 180, 91) to (8, 360, 181)
+    x = self.upsample(x)
+
+    # Layer 4, shape (8, 360, 181, 2C), C = 192 as in the original paper
+    x = self.layer4(x, 8, 360, 181) 
+
+    # Skip connect, in last dimension(C from 192 to 384)
+    x = Concatenate(skip, x)
+
+    # Recover the output fields from patches
+    output, output_surface = self._output_layer(x)
+    return output, output_surface
+
+class PatchEmbedding:
+  def __init__(self, patch_size, dim):
+    '''Patch embedding operation'''
+    # Here we use convolution to partition data into cubes
+    self.conv = Conv3d(input_dims=5, output_dims=dim, kernel_size=patch_size, stride=patch_size)
+    self.conv_surface = Conv2d(input_dims=7, output_dims=dim, kernel_size=patch_size[1:], stride=patch_size[1:])
+
+    # Load constant masks from the disc
+    self.land_mask, self.soil_type, self.topography = LoadConstantMask()
+    
+  def forward(self, input, input_surface):
+    # Zero-pad the input
+    input = Pad3D(input)
+    input_surface = Pad2D(input_surface)
+
+    # Apply a linear projection for patch_size[0]*patch_size[1]*patch_size[2] patches, patch_size = (2, 4, 4) as in the original paper
+    input = self.conv(input)
+
+    # Add three constant fields to the surface fields
+    input_surface =  Concatenate(input_surface, self.land_mask, self.soil_type, self.topography)
+
+    # Apply a linear projection for patch_size[1]*patch_size[2] patches
+    input_surface = self.conv_surface(input_surface)
+
+    # Concatenate the input in the pressure level, i.e., in Z dimension
+    x = Concatenate(input, input_surface)
+
+    # Reshape x for calculation of linear projections
+    x = TransposeDimensions(x, (0, 2, 3, 4, 1))
+    x = reshape(x, target_shape=(x.shape[0], 8*360*181, x.shape[-1]))
+    return x
+    
+class PatchRecovery:
+  def __init__(self, dim):
+    '''Patch recovery operation'''
+    # Hear we use two transposed convolutions to recover data
+    self.conv = ConvTranspose3d(input_dims=dim, output_dims=5, kernel_size=patch_size, stride=patch_size)
+    self.conv_surface = ConvTranspose2d(input_dims=dim, output_dims=4, kernel_size=patch_size[1:], stride=patch_size[1:])
+    
+  def forward(self, x, Z, H, W):
+    # The inverse operation of the patch embedding operation, patch_size = (2, 4, 4) as in the original paper
+    # Reshape x back to three dimensions
+    x = TransposeDimensions(x, (0, 2, 1))
+    x = reshape(x, target_shape=(x.shape[0], x.shape[1], Z, H, W))
+
+    # Call the transposed convolution
+    output = self.conv(x[:, :, 1:, :, :])
+    output_surface = self.conv_surface(x[:, :, 0, :, :])
+
+    # Crop the output to remove zero-paddings
+    output = Crop3D(output)
+    output_surface = Crop2D(output_surface)
+    return output, output_surface
+
+class DownSample:
+  def __init__(self, dim):
+    '''Down-sampling operation'''
+    # A linear function and a layer normalization
+    self.linear = Linear(4*dim, 2*dim, bias=Fasle)
+    self.norm = LayerNorm(4*dim)
+  
+  def forward(self, x, Z, H, W):
+    # Reshape x to three dimensions for downsampling
+    x = reshape(x, target_shape=(x.shape[0], Z, H, W, x.shape[-1]))
+
+    # Padding the input to facilitate downsampling
+    x = Pad3D(x)
+
+    # Reorganize x to reduce the resolution: simply change the order and downsample from (8, 360, 182) to (8, 180, 91)
+    Z, H, W = x.shape
+    # Reshape x to facilitate downsampling
+    x = reshape(x, target_shape=(x.shape[0], Z, H//2, 2, W//2, 2, x.shape[-1]))
+    # Change the order of x
+    x = TransposeDimensions(x, (0,1,2,4,3,5,6))
+    # Reshape to get a tensor of resolution (8, 180, 91)
+    x = reshape(x, target_shape=(x.shape[0], Z*(H//2)*(W//2), 4 * x.shape[-1]))
+
+    # Call the layer normalization
+    x = self.norm(x)
+
+    # Decrease the channels of the data to reduce computation cost
+    x = self.linear(x)
+    return x
+
+class UpSample:
+  def __init__(self, input_dim, output_dim):
+    '''Up-sampling operation'''
+    # Linear layers without bias to increase channels of the data
+    self.linear1 = Linear(input_dim, output_dim*4, bias=False)
+
+    # Linear layers without bias to mix the data up
+    self.linear2 = Linear(output_dim, output_dim, bias=False)
+
+    # Normalization
+    self.norm = LayerNorm(output_dim)
+  
+  def forward(self, x):
+    # Call the linear functions to increase channels of the data
+    x = self.linear1(x)
+
+    # Reorganize x to increase the resolution: simply change the order and upsample from (8, 180, 91) to (8, 360, 182)
+    # Reshape x to facilitate upsampling.
+    x = reshape(x, target_shape=(x.shape[0], 8, 180, 91, 2, 2, x.shape[-1]//4))
+    # Change the order of x
+    x = TransposeDimensions(x, (0,1,2,4,3,5,6))
+    # Reshape to get Tensor with a resolution of (8, 360, 182)
+    x = reshape(x, target_shape=(x.shape[0], 8, 360, 182, x.shape[-1]))
+
+    # Crop the output to the input shape of the network
+    x = Crop3D(x)
+
+    # Reshape x back
+    x = reshape(x, target_shape=(x.shape[0], x.shape[1]*x.shape[2]*x.shape[3], x.shape[-1]))
+
+    # Call the layer normalization
+    x = self.norm(x)
+
+    # Mixup normalized tensors
+    x = self.linear2(x)
+    return x
+  
+class EarthSpecificLayer:
+  def __init__(self, depth, dim, drop_path_ratio_list, heads):
+    '''Basic layer of our network, contains 2 or 6 blocks'''
+    self.depth = depth
+    self.blocks = []
+
+    # Construct basic blocks
+    for i in range(depth):
+      self.blocks.append(EarthSpecificBlock(dim, drop_path_ratio_list[i], heads))
+      
+  def forward(self, x, Z, H, W):
+    for i in range(self.depth):
+      # Roll the input every two blocks
+      if i % 2 == 0:
+        self.blocks[i](x, Z, H, W, roll=False)
+      else:
+        self.blocks[i](x, Z, H, W, roll=True)
+    return x
+
+class EarthSpecificBlock:
+  def __init__(self, dim, drop_path_ratio, heads):
+    '''
+    3D transformer block with Earth-Specific bias and window attention, 
+    see https://github.com/microsoft/Swin-Transformer for the official implementation of 2D window attention.
+    The major difference is that we expand the dimensions to 3 and replace the relative position bias with Earth-Specific bias.
+    '''
+    # Define the window size of the neural network 
+    self.window_size = (2, 6, 12)
+
+    # Initialize serveral operations
+    self.drop_path = DropPath(drop_rate=drop_path_ratio)
+    self.norm1 = LayerNorm(dim)
+    self.norm2 = LayerNorm(dim)
+    self.linear = MLP(dim, 0)
+    self.attention = EarthAttention3D(dim, heads, 0, self.window_size)
+
+  def forward(self, x, Z, H, W, roll):
+    # Save the shortcut for skip-connection
+    shortcut = x
+
+    # Reshape input to three dimensions to calculate window attention
+    reshape(x, target_shape=(x.shape[0], Z, H, W, x.shape[2]))
+
+    # Zero-pad input if needed
+    x = pad3D(x)
+
+    # Store the shape of the input for restoration
+    ori_shape = x.shape
+
+    if roll:
+      # Roll x for half of the window for 3 dimensions
+      x = roll3D(x, shift=[self.window_size[0]//2, self.window_size[1]//2, self.window_size[2]//2])
+      # Generate mask of attention masks
+      # If two pixels are not adjacent, then mask the attention between them
+      # Your can set the matrix element to -1000 when it is not adjacent, then add it to the attention
+      mask = gen_mask(x)
+    else:
+      # e.g., zero matrix when you add mask to attention
+      mask = no_mask
+
+    # Reorganize data to calculate window attention
+    x_window = reshape(x, target_shape=(x.shape[0], Z//window_size[0], window_size[0], H // window_size[1], window_size[1], W // window_size[2], window_size[2], x.shape[-1]))
+    x_window = TransposeDimensions(x_window, (0, 1, 3, 5, 2, 4, 6, 7))
+
+    # Get data stacked in 3D cubes, which will further be used to calculated attention among each cube
+    x_window = reshape(x_window, target_shape=(-1, window_size[0]* window_size[1]*window_size[2], x.shape[-1]))
+
+    # Apply 3D window attention with Earth-Specific bias
+    x_window = self.attention(x, mask)
+
+    # Reorganize data to original shapes
+    x = reshape(x_window, target_shape=((-1, Z // window_size[0], H // window_size[1], W // window_size[2], window_size[0], window_size[1], window_size[2], x_window.shape[-1])))
+    x = TransposeDimensions(x, (0, 1, 4, 2, 5, 3, 6, 7))
+
+    # Reshape the tensor back to its original shape
+    x = reshape(x_window, target_shape=ori_shape)
+
+    if roll:
+      # Roll x back for half of the window
+      x = roll3D(x, shift=[-self.window_size[0]//2, -self.window_size[1]//2, -self.window_size[2]//2])
+
+    # Crop the zero-padding
+    x = Crop3D(x)
+
+    # Reshape the tensor back to the input shape
+    x = reshape(x, target_shape=(x.shape[0], x.shape[1]*x.shape[2]*x.shape[3], x.shape[4]))
+
+    # Main calculation stages
+    x = shortcut + self.drop_path(self.norm1(x))
+    x = x + self.drop_path(self.norm2(self.linear(x)))
+    return x
+    
+class EarthAttention3D:
+  def __init__(self, dim, heads, dropout_rate, window_size):
+    '''
+    3D window attention with the Earth-Specific bias, 
+    see https://github.com/microsoft/Swin-Transformer for the official implementation of 2D window attention.
+    '''
+    # Initialize several operations
+    self.linear1 = Linear(dim, dim=3, bias=True)
+    self.linear2 = Linear(dim, dim)
+    self.softmax = SoftMax(dim=-1)
+    self.dropout = DropOut(dropout_rate)
+
+    # Store several attributes
+    self.head_number = heads
+    self.dim = dim
+    self.scale = (dim//heads)**-0.5
+    self.window_size = window_size
+
+    # input_shape is current shape of the self.forward function
+    # You can run your code to record it, modify the code and rerun it
+    # Record the number of different window types
+    self.type_of_windows = (input_shape[0]//window_size[0])*(input_shape[1]//window_size[1])
+
+    # For each type of window, we will construct a set of parameters according to the paper
+    self.earth_specific_bias = ConstructTensor(shape=((2 * window_size[2] - 1) * window_size[1] * window_size[1] * window_size[0] * window_size[0], self.type_of_windows, heads))
+
+    # Making these tensors to be learnable parameters
+    self.earth_specific_bias = Parameters(self.earth_specific_bias)
+
+    # Initialize the tensors using Truncated normal distribution
+    TruncatedNormalInit(self.earth_specific_bias, std=0.02) 
+
+    # Construct position index to reuse self.earth_specific_bias
+    self.position_index = self._construct_index()
+    
+  def _construct_index(self):
+    ''' This function construct the position index to reuse symmetrical parameters of the position bias'''
+    # Index in the pressure level of query matrix
+    coords_zi = RangeTensor(self.window_size[0])
+    # Index in the pressure level of key matrix
+    coords_zj = -RangeTensor(self.window_size[0])*self.window_size[0]
+
+    # Index in the latitude of query matrix
+    coords_hi = RangeTensor(self.window_size[1])
+    # Index in the latitude of key matrix
+    coords_hj = -RangeTensor(self.window_size[1])*self.window_size[1]
+
+    # Index in the longitude of the key-value pair
+    coords_w = RangeTensor(self.window_size[2])
+
+    # Change the order of the index to calculate the index in total
+    coords_1 = Stack(MeshGrid([coords_zi, coords_hi, coords_w]))
+    coords_2 = Stack(MeshGrid([coords_zj, coords_hj, coords_w]))
+    coords_flatten_1 = Flatten(coords_1, start_dimension=1) 
+    coords_flatten_2 = Flatten(coords_2, start_dimension=1)
+    coords = coords_flatten_1[:, :, None] - coords_flatten_2[:, None, :]
+    coords = TransposeDimensions(coords, (1, 2, 0))
+
+    # Shift the index for each dimension to start from 0
+    coords[:, :, 2] += self.window_size[2] - 1
+    coords[:, :, 1] *= 2 * self.window_size[2] - 1
+    coords[:, :, 0] *= (2 * self.window_size[2] - 1)*self.window_size[1]*self.window_size[1]
+
+    # Sum up the indexes in three dimensions
+    self.position_index = TensorSum(coords, dim=-1)
+
+    # Flatten the position index to facilitate further indexing
+    self.position_index = Flatten(self.position_index)
+    
+  def forward(self, x, mask):
+    # Linear layer to create query, key and value
+    x = self.linear1(x)
+
+    # Record the original shape of the input
+    original_shape = x.shape
+
+    # reshape the data to calculate multi-head attention
+    qkv = reshape(x, target_shape=(x.shape[0], x.shape[1], 3, self.head_number, self.dim // self.head_number)) 
+    query, key, value = TransposeDimensions(qkv, (2, 0, 3, 1, 4))
+
+    # Scale the attention
+    query = query * self.scale
+
+    # Calculated the attention, a learnable bias is added to fix the nonuniformity of the grid.
+    attention = query @ key.T # @ denotes matrix multiplication
+
+    # self.earth_specific_bias is a set of neural network parameters to optimize. 
+    EarthSpecificBias = self.earth_specific_bias[self.position_index]
+
+    # Reshape the learnable bias to the same shape as the attention matrix
+    EarthSpecificBias = reshape(EarthSpecificBias, target_shape=(self.window_size[0]*self.window_size[1]*self.window_size[2], self.window_size[0]*self.window_size[1]*self.window_size[2], self.type_of_windows, self.head_number))
+    EarthSpecificBias = TransposeDimensions(EarthSpecificBias, (2, 3, 0, 1))
+    EarthSpecificBias = reshape(EarthSpecificBias, target_shape = [1]+EarthSpecificBias.shape)
+
+    # Add the Earth-Specific bias to the attention matrix
+    attention = attention + EarthSpecificBias
+
+    # Mask the attention between non-adjacent pixels, e.g., simply add -100 to the masked element.
+    attention = self.mask_attention(attention, mask)
+    attention = self.softmax(attention)
+    attention = self.dropout(attention)
+
+    # Calculated the tensor after spatial mixing.
+    x = attention @ value.T # @ denote matrix multiplication
+
+    # Reshape tensor to the original shape
+    x = TransposeDimensions(x, (0, 2, 1))
+    x = reshape(x, target_shape = original_shape)
+
+    # Linear layer to post-process operated tensor
+    x = self.linear2(x)
+    x = self.dropout(x)
+    return x
+  
+class Mlp:
+  def __init__(self, dim, dropout_rate):
+    '''MLP layers, same as most vision transformer architectures.'''
+    self.linear1 = Linear(dim, dim * 4)
+    self.linear2 = Linear(dim * 4, dim)
+    self.activation = GeLU()
+    self.drop = DropOut(drop_rate=dropout_rate)
+    
+  def forward(self, x):
+    x = self.linear(x)
+    x = self.activation(x)
+    x = self.drop(x)
+    x = self.linear(x)
+    x = self.drop(x)
+    return x
+
+def PerlinNoise():
+  '''Generate random Perlin noise: we follow https://github.com/pvigier/perlin-numpy/ to calculate the perlin noise.'''
+  # Define number of noise
+  octaves = 3
+  # Define the scaling factor of noise
+  noise_scale = 0.2
+  # Define the number of periods of noise along the axis
+  period_number = 12
+  # The size of an input slice
+  H, W = 721, 1440
+  # Scaling factor between two octaves
+  persistence = 0.5
+  # see https://github.com/pvigier/perlin-numpy/ for the implementation of GenerateFractalNoise (e.g., from perlin_numpy import generate_fractal_noise_3d)
+  perlin_noise = noise_scale*GenerateFractalNoise((H, W), (period_number, period_number), octaves, persistence)
+  return perlin_noise

requirements_cpu.txt

@@ -0,0 +1,3 @@
+numpy
+onnx==1.13.1
+onnxruntime==1.14.0

requirements_gpu.txt

@@ -0,0 +1,3 @@
+numpy
+onnx==1.12.0
+onnxruntime-gpu==1.14.0