diff --git a/fedformer/FEDformer.ipynb b/fedformer/FEDformer.ipynb
new file mode 100644
index 0000000..3e21068
--- /dev/null
+++ b/fedformer/FEDformer.ipynb
@@ -0,0 +1 @@
+{"cells":[{"cell_type":"code","execution_count":32,"id":"420561b7","metadata":{"id":"420561b7","executionInfo":{"status":"ok","timestamp":1657091681882,"user_tz":240,"elapsed":314,"user":{"displayName":"Hena Ghonia","userId":"03246241722682988409"}}},"outputs":[],"source":["%matplotlib inline\n","from matplotlib import pyplot as plt\n","import matplotlib.dates as mdates\n","\n","from itertools import islice"]},{"cell_type":"code","execution_count":33,"id":"b10c3dd3","metadata":{"scrolled":true,"id":"b10c3dd3","executionInfo":{"status":"ok","timestamp":1657091683978,"user_tz":240,"elapsed":248,"user":{"displayName":"Hena Ghonia","userId":"03246241722682988409"}}},"outputs":[],"source":["from gluonts.evaluation import make_evaluation_predictions, Evaluator\n","from gluonts.dataset.repository.datasets import get_dataset\n","\n","from estimator import FEDformerEstimator"]},{"cell_type":"code","execution_count":34,"id":"d018b7fb","metadata":{"id":"d018b7fb","executionInfo":{"status":"ok","timestamp":1657091684462,"user_tz":240,"elapsed":3,"user":{"displayName":"Hena Ghonia","userId":"03246241722682988409"}}},"outputs":[],"source":["dataset = get_dataset(\"electricity\")"]},{"cell_type":"code","execution_count":35,"id":"50025289","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"50025289","executionInfo":{"status":"ok","timestamp":1657091686062,"user_tz":240,"elapsed":4,"user":{"displayName":"Hena Ghonia","userId":"03246241722682988409"}},"outputId":"8e5ad35b-caca-49fc-bf54-397e8af05b53"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["TrainDatasets(metadata=MetaData(freq='1H', target=None, feat_static_cat=[CategoricalFeatureInfo(name='feat_static_cat_0', cardinality='321')], feat_static_real=[], feat_dynamic_real=[], feat_dynamic_cat=[], prediction_length=24), train=DatasetCollection(datasets=[Map(data=JsonLinesFile(path=PosixPath('/root/.mxnet/gluon-ts/datasets/electricity/train/data.json.gz')))], interleave=False), test=DatasetCollection(datasets=[Map(data=JsonLinesFile(path=PosixPath('/root/.mxnet/gluon-ts/datasets/electricity/test/data.json.gz')))], interleave=False))"]},"metadata":{},"execution_count":35}],"source":["dataset"]},{"cell_type":"code","execution_count":36,"id":"e772234f","metadata":{"id":"e772234f","executionInfo":{"status":"ok","timestamp":1657091687089,"user_tz":240,"elapsed":3,"user":{"displayName":"Hena Ghonia","userId":"03246241722682988409"}}},"outputs":[],"source":["estimator = FEDformerEstimator(\n","    freq='h',\n","    prediction_length=dataset.metadata.prediction_length,\n","    context_length=dataset.metadata.prediction_length*7,\n","    dim_feedforward=16,\n","    num_feat_static_cat=1,\n","    cardinality=[321],\n","    embedding_dimension=[3],\n","    # attention hyper-params\n","    num_encoder_layers=2,\n","    num_decoder_layers=1,\n","    nhead=1,\n","    activation=\"relu\",\n","    moving_avg=[24],\n","    # training params\n","    batch_size=128,\n","    num_batches_per_epoch=50,\n","    trainer_kwargs=dict(max_epochs=1, accelerator='gpu', gpus=1),\n",")"]},{"cell_type":"code","execution_count":37,"id":"22d804e4","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000,"referenced_widgets":["15f43598310a460d98c2f5fd7b4ed68a","f93223e98409437da542e499168b5d2e","487fa08c566d42849d02f5447eb7fd51","14d504febb0c40719bef98e2a67ab670","9781b8be1681499988883461a84e6c2f","65828c15c48547dc8a4167f2efc09198","a0b2398815ba4515b4d079c629cf3a06","71f31567159f4551bcaf6f2586daf29b","d7af616078794d5087a0f713ff9b1d88","2031ecb3c32245dd95e73c2eb86189b1","aedfd2f7bf80486c9ed3602c6b1762e6"]},"id":"22d804e4","executionInfo":{"status":"error","timestamp":1657091693249,"user_tz":240,"elapsed":4710,"user":{"displayName":"Hena Ghonia","userId":"03246241722682988409"}},"outputId":"6966479c-f461-423e-d9b8-3d2b04c7c131"},"outputs":[{"output_type":"stream","name":"stdout","text":["fourier enhanced block used!\n","modes=64, index=[0, 1, 2, 3, 6, 7, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 28, 29, 30, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 50, 53, 54, 55, 57, 59, 60, 61, 63, 65, 68, 70, 72, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83]\n","fourier enhanced block used!\n","modes=64, index=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53]\n"," fourier enhanced cross attention used!\n","modes_q=54, index_q=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53]\n","modes_kv=64, index_kv=[0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 19, 20, 21, 22, 25, 26, 27, 28, 30, 32, 33, 36, 37, 39, 40, 42, 43, 44, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 59, 60, 62, 64, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 80, 81, 83]\n","dim_feedforward 16\n","enc_modes: 64, dec_modes: 54\n","encoder_self_att FourierBlock()\n"]},{"output_type":"stream","name":"stderr","text":["/usr/local/lib/python3.7/dist-packages/pytorch_lightning/utilities/parsing.py:262: UserWarning: Attribute 'model' is an instance of `nn.Module` and is already saved during checkpointing. It is recommended to ignore them using `self.save_hyperparameters(ignore=['model'])`.\n","  f\"Attribute {k!r} is an instance of `nn.Module` and is already saved during checkpointing.\"\n","GPU available: True, used: True\n","TPU available: False, using: 0 TPU cores\n","IPU available: False, using: 0 IPUs\n","HPU available: False, using: 0 HPUs\n","/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/configuration_validator.py:133: UserWarning: You defined a `validation_step` but have no `val_dataloader`. Skipping val loop.\n","  rank_zero_warn(\"You defined a `validation_step` but have no `val_dataloader`. Skipping val loop.\")\n","LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n","\n","  | Name  | Type           | Params\n","-----------------------------------------\n","0 | model | FEDformerModel | 97.9 K\n","-----------------------------------------\n","97.9 K    Trainable params\n","0         Non-trainable params\n","97.9 K    Total params\n","0.391     Total estimated model params size (MB)\n"]},{"output_type":"display_data","data":{"text/plain":["Training: 0it [00:00, ?it/s]"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"15f43598310a460d98c2f5fd7b4ed68a"}},"metadata":{}},{"output_type":"stream","name":"stdout","text":["transformer_inputs torch.Size([128, 192, 50])\n","context_length 168\n","enc_input: torch.Size([128, 168, 50])\n","torch.Size([128, 168, 50])\n","query proj torch.Size([128, 168, 50])\n"]},{"output_type":"error","ename":"RuntimeError","evalue":"ignored","traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)","\u001b[0;32m<ipython-input-37-ecb55c4d2def>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m predictor = estimator.train(\n\u001b[1;32m      2\u001b[0m     \u001b[0mtraining_data\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m     \u001b[0mnum_workers\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m     \u001b[0;31m# shuffle_buffer_length=1024\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m )\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/gluonts/torch/model/estimator.py\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(self, training_data, validation_data, num_workers, shuffle_buffer_length, cache_data, ckpt_path, **kwargs)\u001b[0m\n\u001b[1;32m    235\u001b[0m             \u001b[0mcache_data\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcache_data\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    236\u001b[0m             \u001b[0mckpt_path\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mckpt_path\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 237\u001b[0;31m             \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    238\u001b[0m         ).predictor\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/gluonts/torch/model/estimator.py\u001b[0m in \u001b[0;36mtrain_model\u001b[0;34m(self, training_data, validation_data, num_workers, shuffle_buffer_length, cache_data, ckpt_path, **kwargs)\u001b[0m\n\u001b[1;32m    199\u001b[0m             \u001b[0mtrain_dataloaders\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtraining_data_loader\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    200\u001b[0m             \u001b[0mval_dataloaders\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidation_data_loader\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 201\u001b[0;31m             \u001b[0mckpt_path\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mckpt_path\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    202\u001b[0m         )\n\u001b[1;32m    203\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)\u001b[0m\n\u001b[1;32m    769\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    770\u001b[0m         self._call_and_handle_interrupt(\n\u001b[0;32m--> 771\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_fit_impl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_dataloaders\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_dataloaders\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdatamodule\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mckpt_path\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    772\u001b[0m         )\n\u001b[1;32m    773\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_call_and_handle_interrupt\u001b[0;34m(self, trainer_fn, *args, **kwargs)\u001b[0m\n\u001b[1;32m    721\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlauncher\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlaunch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrainer_fn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrainer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    722\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 723\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mtrainer_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    724\u001b[0m         \u001b[0;31m# TODO: treat KeyboardInterrupt as BaseException (delete the code below) in v1.7\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    725\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mexception\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_fit_impl\u001b[0;34m(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)\u001b[0m\n\u001b[1;32m    809\u001b[0m             \u001b[0mckpt_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel_provided\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel_connected\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    810\u001b[0m         )\n\u001b[0;32m--> 811\u001b[0;31m         \u001b[0mresults\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mckpt_path\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mckpt_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    812\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    813\u001b[0m         \u001b[0;32massert\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstopped\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_run\u001b[0;34m(self, model, ckpt_path)\u001b[0m\n\u001b[1;32m   1234\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_checkpoint_connector\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mresume_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1235\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1236\u001b[0;31m         \u001b[0mresults\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_stage\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1237\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1238\u001b[0m         \u001b[0mlog\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdetail\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"{self.__class__.__name__}: trainer tearing down\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_run_stage\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m   1321\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredicting\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1322\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_predict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1323\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_train\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1324\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1325\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_pre_training_routine\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_run_train\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m   1351\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1352\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_detect_anomaly\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_detect_anomaly\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1353\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1354\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1355\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_run_evaluate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0m_EVALUATE_OUTPUT\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/base.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    202\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    203\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 204\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    205\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    206\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/fit_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    267\u001b[0m         )\n\u001b[1;32m    268\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"run_training_epoch\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 269\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_outputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mepoch_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_data_fetcher\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    270\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    271\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/base.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    202\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    203\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 204\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    205\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    206\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/epoch/training_epoch_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self, data_fetcher)\u001b[0m\n\u001b[1;32m    206\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    207\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"run_training_batch\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 208\u001b[0;31m                 \u001b[0mbatch_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbatch_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    209\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    210\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbatch_progress\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mincrement_processed\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/base.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    202\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    203\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 204\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    205\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    206\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/batch/training_batch_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self, batch, batch_idx)\u001b[0m\n\u001b[1;32m     86\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautomatic_optimization\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     87\u001b[0m             \u001b[0moptimizers\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_get_active_optimizers\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_frequencies\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 88\u001b[0;31m             \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msplit_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     89\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     90\u001b[0m             \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmanual_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msplit_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/base.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    202\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    203\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 204\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    205\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    206\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self, batch, *args, **kwargs)\u001b[0m\n\u001b[1;32m    205\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_batch_idx\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    206\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizers\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptim_progress\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_position\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 207\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    208\u001b[0m         )\n\u001b[1;32m    209\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloss\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m_run_optimization\u001b[0;34m(self, split_batch, batch_idx, optimizer, opt_idx)\u001b[0m\n\u001b[1;32m    254\u001b[0m         \u001b[0;31m# gradient update with accumulated gradients\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    255\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 256\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizer_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopt_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    257\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    258\u001b[0m         \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconsume_result\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m_optimizer_step\u001b[0;34m(self, optimizer, opt_idx, batch_idx, train_step_and_backward_closure)\u001b[0m\n\u001b[1;32m    376\u001b[0m             \u001b[0mon_tpu\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0maccelerator\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mTPUAccelerator\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    377\u001b[0m             \u001b[0musing_native_amp\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mamp_backend\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mAMPType\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mNATIVE\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 378\u001b[0;31m             \u001b[0musing_lbfgs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mis_lbfgs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    379\u001b[0m         )\n\u001b[1;32m    380\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_call_lightning_module_hook\u001b[0;34m(self, hook_name, pl_module, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1593\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1594\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"[LightningModule]{pl_module.__class__.__name__}.{hook_name}\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1595\u001b[0;31m             \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1596\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1597\u001b[0m         \u001b[0;31m# restore current_fx when nested context\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/core/lightning.py\u001b[0m in \u001b[0;36moptimizer_step\u001b[0;34m(self, epoch, batch_idx, optimizer, optimizer_idx, optimizer_closure, on_tpu, using_native_amp, using_lbfgs)\u001b[0m\n\u001b[1;32m   1644\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1645\u001b[0m         \"\"\"\n\u001b[0;32m-> 1646\u001b[0;31m         \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0moptimizer_closure\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1647\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1648\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0moptimizer_zero_grad\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepoch\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mOptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/core/optimizer.py\u001b[0m in \u001b[0;36mstep\u001b[0;34m(self, closure, **kwargs)\u001b[0m\n\u001b[1;32m    166\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    167\u001b[0m         \u001b[0;32massert\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_strategy\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 168\u001b[0;31m         \u001b[0mstep_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_strategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    169\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    170\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_on_after_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/strategies/strategy.py\u001b[0m in \u001b[0;36moptimizer_step\u001b[0;34m(self, optimizer, opt_idx, closure, model, **kwargs)\u001b[0m\n\u001b[1;32m    191\u001b[0m         \"\"\"\n\u001b[1;32m    192\u001b[0m         \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 193\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprecision_plugin\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopt_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    194\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    195\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_setup_model_and_optimizers\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mModule\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizers\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mList\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mOptimizer\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTuple\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mModule\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mList\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mOptimizer\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/plugins/precision/precision_plugin.py\u001b[0m in \u001b[0;36moptimizer_step\u001b[0;34m(self, model, optimizer, optimizer_idx, closure, **kwargs)\u001b[0m\n\u001b[1;32m    153\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpl\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mLightningModule\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    154\u001b[0m             \u001b[0mclosure\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpartial\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_wrap_closure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 155\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    156\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    157\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_track_grad_norm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrainer\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;34m\"pl.Trainer\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/optimizer.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     86\u001b[0m                 \u001b[0mprofile_name\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"Optimizer.step#{}.step\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__class__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     87\u001b[0m                 \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecord_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mprofile_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 88\u001b[0;31m                     \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     89\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mwrapper\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     90\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/autograd/grad_mode.py\u001b[0m in \u001b[0;36mdecorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     25\u001b[0m         \u001b[0;32mdef\u001b[0m \u001b[0mdecorate_context\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     26\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclone\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 27\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     28\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mcast\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mF\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdecorate_context\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     29\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/adam.py\u001b[0m in \u001b[0;36mstep\u001b[0;34m(self, closure)\u001b[0m\n\u001b[1;32m     98\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mclosure\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     99\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0menable_grad\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 100\u001b[0;31m                 \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    101\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    102\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mgroup\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparam_groups\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/plugins/precision/precision_plugin.py\u001b[0m in \u001b[0;36m_wrap_closure\u001b[0;34m(self, model, optimizer, optimizer_idx, closure)\u001b[0m\n\u001b[1;32m    138\u001b[0m         \u001b[0mconsistent\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mthe\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0mPrecisionPlugin\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m \u001b[0msubclasses\u001b[0m \u001b[0mthat\u001b[0m \u001b[0mcannot\u001b[0m \u001b[0;32mpass\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m \u001b[0mdirectly\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    139\u001b[0m         \"\"\"\n\u001b[0;32m--> 140\u001b[0;31m         \u001b[0mclosure_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    141\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_after_closure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    142\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mclosure_result\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    146\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    147\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mAny\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mAny\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mOptional\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mTensor\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 148\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    149\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_result\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    150\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36mclosure\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    132\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    133\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mAny\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mAny\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mClosureResult\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 134\u001b[0;31m         \u001b[0mstep_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_step_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    135\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    136\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mstep_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclosure_loss\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m_training_step\u001b[0;34m(self, split_batch, batch_idx, opt_idx)\u001b[0m\n\u001b[1;32m    425\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    426\u001b[0m         \u001b[0;31m# manually capture logged metrics\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 427\u001b[0;31m         \u001b[0mtraining_step_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call_strategy_hook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"training_step\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mstep_kwargs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    428\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpost_training_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    429\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_call_strategy_hook\u001b[0;34m(self, hook_name, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1763\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1764\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"[Strategy]{self.strategy.__class__.__name__}.{hook_name}\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1765\u001b[0;31m             \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1766\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1767\u001b[0m         \u001b[0;31m# restore current_fx when nested context\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/pytorch_lightning/strategies/strategy.py\u001b[0m in \u001b[0;36mtraining_step\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    331\u001b[0m         \"\"\"\n\u001b[1;32m    332\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprecision_plugin\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_step_context\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 333\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtraining_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    334\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    335\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpost_training_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/lightning_module.py\u001b[0m in \u001b[0;36mtraining_step\u001b[0;34m(self, batch, batch_idx)\u001b[0m\n\u001b[1;32m     24\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mtraining_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     25\u001b[0m         \u001b[0;34m\"\"\"Execute training step\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 26\u001b[0;31m         \u001b[0mtrain_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     27\u001b[0m         self.log(\n\u001b[1;32m     28\u001b[0m             \u001b[0;34m\"train_loss\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1108\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1109\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1110\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1111\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1112\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/lightning_module.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, batch)\u001b[0m\n\u001b[1;32m     69\u001b[0m         )\n\u001b[1;32m     70\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'transformer_inputs'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtransformer_inputs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 71\u001b[0;31m         \u001b[0mparams\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutput_params\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtransformer_inputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     72\u001b[0m         \u001b[0mdistr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutput_distribution\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mscale\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     73\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/module.py\u001b[0m in \u001b[0;36moutput_params\u001b[0;34m(self, transformer_inputs)\u001b[0m\n\u001b[1;32m   1622\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0moutput_params\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtransformer_inputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1623\u001b[0m         \u001b[0menc_input\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtransformer_inputs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcontext_length\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m...\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1624\u001b[0;31m         \u001b[0mdec_input\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtransformer_inputs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcontext_length\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m...\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1625\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1626\u001b[0m         \u001b[0menc_out\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mencoder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0menc_input\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1108\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1109\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1110\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1111\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1112\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/module.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x, attn_mask)\u001b[0m\n\u001b[1;32m    483\u001b[0m             \u001b[0mattns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mattn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    484\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 485\u001b[0;31m             \u001b[0;32mfor\u001b[0m \u001b[0mattn_layer\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mattn_layers\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    486\u001b[0m                 \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mattn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mattn_layer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mattn_mask\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mattn_mask\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    487\u001b[0m                 \u001b[0mattns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mattn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1108\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1109\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1110\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1111\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1112\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/module.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x, attn_mask)\u001b[0m\n\u001b[1;32m    451\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mattn_mask\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    452\u001b[0m         new_x, attn = self.attention(\n\u001b[0;32m--> 453\u001b[0;31m             \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    454\u001b[0m             \u001b[0mattn_mask\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mattn_mask\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    455\u001b[0m         )\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1108\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1109\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1110\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1111\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1112\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/module.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, queries, keys, values, attn_mask)\u001b[0m\n\u001b[1;32m    345\u001b[0m             \u001b[0mqueries\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    346\u001b[0m             \u001b[0mkeys\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 347\u001b[0;31m             \u001b[0mvalues\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    348\u001b[0m             \u001b[0mattn_mask\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    349\u001b[0m         )\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1108\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1109\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1110\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1111\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1112\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/module.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, q, k, v, mask)\u001b[0m\n\u001b[1;32m    753\u001b[0m         \u001b[0mx_ft\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfft\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrfft\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    754\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_ft\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 755\u001b[0;31m         \u001b[0;31m# Perform Fourier neural operations\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    756\u001b[0m         \u001b[0mout_ft\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mzeros\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mB\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mH\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mE\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mL\u001b[0m \u001b[0;34m//\u001b[0m \u001b[0;36m2\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdevice\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcfloat\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    757\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mwi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/MyDrive/Udem/Sem2/Representation_Learning/FEDformer/module.py\u001b[0m in \u001b[0;36mcompl_mul1d\u001b[0;34m(self, input, weights)\u001b[0m\n\u001b[1;32m    742\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mcompl_mul1d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweights\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    743\u001b[0m         \u001b[0;31m# (batch, in_channel, x ), (in_channel, out_channel, x) -> (batch, out_channel, x)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 744\u001b[0;31m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'input fft'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    745\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'weight fft'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweights\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    746\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meinsum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"bhi,hio->bho\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweights\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;31m# hio->bho\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/functional.py\u001b[0m in \u001b[0;36meinsum\u001b[0;34m(*args)\u001b[0m\n\u001b[1;32m    328\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0meinsum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mequation\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0m_operands\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    329\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 330\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0m_VF\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meinsum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mequation\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moperands\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# type: ignore[attr-defined]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    331\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    332\u001b[0m \u001b[0;31m# Wrapper around _histogramdd and _histogramdd_bin_edges needed due to (Tensor, Tensor[]) return type.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mRuntimeError\u001b[0m: einsum(): operands do not broadcast with remapped shapes [original->remapped]: [128, 1, 50]->[128, 1, 1, 50] [8, 6, 6]->[1, 8, 6, 6]"]}],"source":["predictor = estimator.train(\n","    training_data=dataset.train,\n","    num_workers=1,\n","    # shuffle_buffer_length=1024\n",")"]},{"cell_type":"code","execution_count":null,"id":"11a47d5a","metadata":{"id":"11a47d5a"},"outputs":[],"source":["forecast_it, ts_it = make_evaluation_predictions(\n","    dataset=dataset.test, \n","    predictor=predictor\n",")"]},{"cell_type":"code","execution_count":null,"id":"e4e94932","metadata":{"id":"e4e94932"},"outputs":[],"source":["forecasts = list(forecast_it)"]},{"cell_type":"code","execution_count":null,"id":"17c5e570","metadata":{"id":"17c5e570"},"outputs":[],"source":["tss = list(ts_it)"]},{"cell_type":"code","execution_count":null,"id":"9985be71","metadata":{"id":"9985be71"},"outputs":[],"source":["evaluator = Evaluator()"]},{"cell_type":"code","execution_count":null,"id":"cca60b1e","metadata":{"id":"cca60b1e"},"outputs":[],"source":["agg_metrics, ts_metrics = evaluator(iter(tss), iter(forecasts))"]},{"cell_type":"code","execution_count":null,"id":"92389256","metadata":{"id":"92389256"},"outputs":[],"source":["agg_metrics"]},{"cell_type":"code","execution_count":null,"id":"23878611","metadata":{"id":"23878611"},"outputs":[],"source":["plt.figure(figsize=(20, 15))\n","date_formater = mdates.DateFormatter('%b, %d')\n","plt.rcParams.update({'font.size': 15})\n","\n","for idx, (forecast, ts) in islice(enumerate(zip(forecasts, tss)), 9):\n","    ax = plt.subplot(3, 3, idx+1)\n","\n","    plt.plot(ts[-4 * dataset.metadata.prediction_length:], label=\"target\", )\n","    forecast.plot( color='g')\n","    plt.xticks(rotation=60)\n","    plt.title(forecast.item_id)\n","    ax.xaxis.set_major_formatter(date_formater)\n","\n","plt.gcf().tight_layout()\n","plt.legend()\n","plt.show()"]},{"cell_type":"code","execution_count":null,"id":"f34e7aa7","metadata":{"id":"f34e7aa7"},"outputs":[],"source":[""]}],"metadata":{"kernelspec":{"display_name":"Python 3 (ipykernel)","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.8.10"},"colab":{"name":"FEDformer.ipynb","provenance":[],"collapsed_sections":[],"machine_shape":"hm"},"accelerator":"GPU","gpuClass":"standard","widgets":{"application/vnd.jupyter.widget-state+json":{"15f43598310a460d98c2f5fd7b4ed68a":{"model_module":"@jupyter-widgets/controls","model_name":"HBoxModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_f93223e98409437da542e499168b5d2e","IPY_MODEL_487fa08c566d42849d02f5447eb7fd51","IPY_MODEL_14d504febb0c40719bef98e2a67ab670"],"layout":"IPY_MODEL_9781b8be1681499988883461a84e6c2f"}},"f93223e98409437da542e499168b5d2e":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_65828c15c48547dc8a4167f2efc09198","placeholder":"​","style":"IPY_MODEL_a0b2398815ba4515b4d079c629cf3a06","value":"Epoch 0: "}},"487fa08c566d42849d02f5447eb7fd51":{"model_module":"@jupyter-widgets/controls","model_name":"FloatProgressModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"","description":"","description_tooltip":null,"layout":"IPY_MODEL_71f31567159f4551bcaf6f2586daf29b","max":1,"min":0,"orientation":"horizontal","style":"IPY_MODEL_d7af616078794d5087a0f713ff9b1d88","value":0}},"14d504febb0c40719bef98e2a67ab670":{"model_module":"@jupyter-widgets/controls","model_name":"HTMLModel","model_module_version":"1.5.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_2031ecb3c32245dd95e73c2eb86189b1","placeholder":"​","style":"IPY_MODEL_aedfd2f7bf80486c9ed3602c6b1762e6","value":" 0/? [00:00&lt;?, ?it/s]"}},"9781b8be1681499988883461a84e6c2f":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":"inline-flex","flex":null,"flex_flow":"row wrap","grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":"100%"}},"65828c15c48547dc8a4167f2efc09198":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a0b2398815ba4515b4d079c629cf3a06":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"71f31567159f4551bcaf6f2586daf29b":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":"2","flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"d7af616078794d5087a0f713ff9b1d88":{"model_module":"@jupyter-widgets/controls","model_name":"ProgressStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"2031ecb3c32245dd95e73c2eb86189b1":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"aedfd2f7bf80486c9ed3602c6b1762e6":{"model_module":"@jupyter-widgets/controls","model_name":"DescriptionStyleModel","model_module_version":"1.5.0","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}}}}},"nbformat":4,"nbformat_minor":5}
\ No newline at end of file
diff --git a/fedformer/estimator.py b/fedformer/estimator.py
new file mode 100644
index 0000000..a4f0e28
--- /dev/null
+++ b/fedformer/estimator.py
@@ -0,0 +1,356 @@
+from typing import Any, Dict, Iterable, List, Optional
+
+import torch
+from gluonts.core.component import validated
+from gluonts.dataset.common import Dataset
+from gluonts.dataset.field_names import FieldName
+from gluonts.itertools import Cyclic, IterableSlice, PseudoShuffled
+from gluonts.time_feature import TimeFeature, time_features_from_frequency_str
+from gluonts.torch.model.estimator import PyTorchLightningEstimator
+from gluonts.torch.model.predictor import PyTorchPredictor
+from gluonts.torch.distributions import DistributionOutput, StudentTOutput
+from gluonts.torch.modules.loss import DistributionLoss, NegativeLogLikelihood
+from gluonts.torch.util import IterableDataset
+from gluonts.transform import (
+    AddAgeFeature,
+    AddObservedValuesIndicator,
+    AddTimeFeatures,
+    AsNumpyArray,
+    Chain,
+    ExpectedNumInstanceSampler,
+    InstanceSplitter,
+    RemoveFields,
+    SelectFields,
+    SetField,
+    TestSplitSampler,
+    Transformation,
+    ValidationSplitSampler,
+    VstackFeatures,
+)
+from gluonts.transform.sampler import InstanceSampler
+from torch.utils.data import DataLoader
+
+from lightning_module import FEDformerLightningModule
+from module import FEDformerModel
+import random
+PREDICTION_INPUT_NAMES = [
+    "feat_static_cat",
+    "feat_static_real",
+    "past_time_feat",
+    "past_target",
+    "past_observed_values",
+    "future_time_feat",
+]
+
+TRAINING_INPUT_NAMES = PREDICTION_INPUT_NAMES + [
+    "future_target",
+    "future_observed_values",
+]
+
+
+class FEDformerEstimator(PyTorchLightningEstimator):
+    @validated()
+    def __init__(
+        self,
+        freq: str,
+        prediction_length: int,
+        nhead: int,
+        num_encoder_layers: int, 
+        num_decoder_layers: int, 
+        dim_feedforward: int = 2048, #dimension of fcn
+        version:str = 'Fourier',#Fourier, Wavelets
+        features: str = 'M',#options:[M, S, MS]; M:multivariate predict multivariate, 'S':univariate predict univariate, MS:multivariate predict univariate' 
+        target: str = '0T',#'target feature in S or MS task'
+        checkpoints:str = './checkpoints/', #location of model checkpoints
+        modes:int = 64,
+        mode_select:str = 'random',#options: [random, low]
+        base:str = 'legendre',
+        cross_activation:str = 'tanh',
+        L: int = 3,
+        #forecasting task
+        context_length: Optional[int] = None,#seq_len : input sequence length
+        label_length: Optional[int] = 48,#start token length
+        # model argument
+        input_size: int = 1, #encoder input size
+        dec_in: int = 7, #decoder input size
+        c_out: int = 7, #output size
+        moving_avg: List[int] = [24],
+        factor:int = 1,
+        activation: str = "gelu",
+        dropout: float = 0.05,
+        embed: str = "timeF",#options:[timeF, fixed, learned]
+        output_attention: bool = True,#whether to output attention in encoder
+        distil: bool = True,
+        num_feat_dynamic_real: int = 0,
+        num_feat_static_cat: int = 0,
+        num_feat_static_real: int = 0,
+        cardinality: Optional[List[int]] = None,
+        embedding_dimension: Optional[List[int]] = None,
+        distr_output: DistributionOutput = StudentTOutput(),
+        loss: DistributionLoss = NegativeLogLikelihood(),
+        scaling: bool = True,
+        lags_seq: Optional[List[int]] = None,
+        time_features: Optional[List[TimeFeature]] = None,
+        num_parallel_samples: int = 100,
+        batch_size: int = 32,
+        num_batches_per_epoch: int = 512,
+        trainer_kwargs: Optional[Dict[str, Any]] = dict(),
+        train_sampler: Optional[InstanceSampler] = None,
+        validation_sampler: Optional[InstanceSampler] = None,
+    ) -> None:
+        trainer_kwargs = {
+            "max_epochs": 100,
+            **trainer_kwargs,
+        }
+        super().__init__(trainer_kwargs=trainer_kwargs)
+        self.version = version
+        self.modes = modes
+        self.mode_select = mode_select
+        self.base = base
+        self.cross_activation = cross_activation
+        self.L = L
+        self.freq = freq
+        self.context_length = (
+            context_length if context_length is not None else prediction_length
+        )
+        self.label_length = label_length
+        self.prediction_length = prediction_length
+        self.distr_output = distr_output
+        self.loss = loss
+        
+        self.input_size = input_size
+        # self.dec_in = dec_in
+        self.c_out = c_out
+        self.nhead = nhead
+        self.num_encoder_layers = num_encoder_layers
+        self.num_decoder_layers = num_decoder_layers
+        self.activation = activation
+        self.dim_feedforward = dim_feedforward
+        self.dropout = dropout
+        self.output_attention = output_attention
+        # self.distil = distil
+        self.factor = factor
+        self.moving_avg = moving_avg
+        self.embed = embed
+        
+        self.num_feat_dynamic_real = num_feat_dynamic_real
+        self.num_feat_static_cat = num_feat_static_cat
+        self.num_feat_static_real = num_feat_static_real
+        self.cardinality = (
+            cardinality if cardinality and num_feat_static_cat > 0 else [1]
+        )
+        self.embedding_dimension = embedding_dimension
+        self.scaling = scaling
+        self.lags_seq = lags_seq
+        self.time_features = (
+            time_features
+            if time_features is not None
+            else time_features_from_frequency_str(self.freq)
+        )
+
+        self.num_parallel_samples = num_parallel_samples
+        self.batch_size = batch_size
+        self.num_batches_per_epoch = num_batches_per_epoch
+
+        self.train_sampler = train_sampler or ExpectedNumInstanceSampler(
+            num_instances=1.0, min_future=prediction_length
+        )
+        self.validation_sampler = validation_sampler or ValidationSplitSampler(
+            min_future=prediction_length
+        )
+
+    def create_transformation(self) -> Transformation:
+        remove_field_names = []
+        if self.num_feat_static_real == 0:
+            remove_field_names.append(FieldName.FEAT_STATIC_REAL)
+        if self.num_feat_dynamic_real == 0:
+            remove_field_names.append(FieldName.FEAT_DYNAMIC_REAL)
+
+        return Chain(
+            [RemoveFields(field_names=remove_field_names)]
+            + (
+                [SetField(output_field=FieldName.FEAT_STATIC_CAT, value=[0])]
+                if not self.num_feat_static_cat > 0
+                else []
+            )
+            + (
+                [SetField(output_field=FieldName.FEAT_STATIC_REAL, value=[0.0])]
+                if not self.num_feat_static_real > 0
+                else []
+            )
+            + [
+                AsNumpyArray(
+                    field=FieldName.FEAT_STATIC_CAT,
+                    expected_ndim=1,
+                    dtype=int,
+                ),
+                AsNumpyArray(
+                    field=FieldName.FEAT_STATIC_REAL,
+                    expected_ndim=1,
+                ),
+                AsNumpyArray(
+                    field=FieldName.TARGET,
+                    # in the following line, we add 1 for the time dimension
+                    expected_ndim=1 + len(self.distr_output.event_shape),
+                ),
+                AddObservedValuesIndicator(
+                    target_field=FieldName.TARGET,
+                    output_field=FieldName.OBSERVED_VALUES,
+                ),
+                AddTimeFeatures(
+                    start_field=FieldName.START,
+                    target_field=FieldName.TARGET,
+                    output_field=FieldName.FEAT_TIME,
+                    time_features=self.time_features,
+                    pred_length=self.prediction_length,
+                ),
+                AddAgeFeature(
+                    target_field=FieldName.TARGET,
+                    output_field=FieldName.FEAT_AGE,
+                    pred_length=self.prediction_length,
+                    log_scale=True,
+                ),
+                VstackFeatures(
+                    output_field=FieldName.FEAT_TIME,
+                    input_fields=[FieldName.FEAT_TIME, FieldName.FEAT_AGE]
+                    + (
+                        [FieldName.FEAT_DYNAMIC_REAL]
+                        if self.num_feat_dynamic_real > 0
+                        else []
+                    ),
+                ),
+            ]
+        )
+
+    def _create_instance_splitter(self, module: FEDformerLightningModule, mode: str):
+        assert mode in ["training", "validation", "test"]
+
+        instance_sampler = {
+            "training": self.train_sampler,
+            "validation": self.validation_sampler,
+            "test": TestSplitSampler(),
+        }[mode]
+
+        return InstanceSplitter(
+            target_field=FieldName.TARGET,
+            is_pad_field=FieldName.IS_PAD,
+            start_field=FieldName.START,
+            forecast_start_field=FieldName.FORECAST_START,
+            instance_sampler=instance_sampler,
+            past_length=module.model._past_length,
+            future_length=self.prediction_length,
+            time_series_fields=[
+                FieldName.FEAT_TIME,
+                FieldName.OBSERVED_VALUES,
+            ],
+            dummy_value=self.distr_output.value_in_support,
+        )
+
+    def create_training_data_loader(
+        self,
+        data: Dataset,
+        module: FEDformerLightningModule,
+        shuffle_buffer_length: Optional[int] = None,
+        **kwargs,
+    ) -> Iterable:
+        transformation = self._create_instance_splitter(
+            module, "training"
+        ) + SelectFields(TRAINING_INPUT_NAMES)
+
+        training_instances = transformation.apply(
+            Cyclic(data)
+            if shuffle_buffer_length is None
+            else PseudoShuffled(
+                Cyclic(data), shuffle_buffer_length=shuffle_buffer_length
+            )
+        )
+
+        return IterableSlice(
+            iter(
+                DataLoader(
+                    IterableDataset(training_instances),
+                    batch_size=self.batch_size,
+                    **kwargs,
+                )
+            ),
+            self.num_batches_per_epoch,
+        )
+
+    def create_validation_data_loader(
+        self,
+        data: Dataset,
+        module: FEDformerLightningModule,
+        **kwargs,
+    ) -> Iterable:
+        transformation = self._create_instance_splitter(
+            module, "validation"
+        ) + SelectFields(TRAINING_INPUT_NAMES)
+
+        validation_instances = transformation.apply(data)
+
+        return DataLoader(
+            IterableDataset(validation_instances),
+            batch_size=self.batch_size,
+            **kwargs,
+        )
+
+    def create_predictor(
+        self,
+        transformation: Transformation,
+        module: FEDformerLightningModule,
+    ) -> PyTorchPredictor:
+        prediction_splitter = self._create_instance_splitter(module, "test")
+
+        return PyTorchPredictor(
+            input_transform=transformation + prediction_splitter,
+            input_names=PREDICTION_INPUT_NAMES,
+            prediction_net=module.model,
+            batch_size=self.batch_size,
+            prediction_length=self.prediction_length,
+            device=torch.device("cuda" if torch.cuda.is_available() else "cpu"),
+        )
+
+    def create_lightning_module(self) -> FEDformerLightningModule:
+        model = FEDformerModel(
+        version = self.version,#Fourier, Wavelets
+        freq = self.freq,
+        modes = self.modes,
+        mode_select = self.mode_select,
+        base = self.base,
+        cross_activation = self.cross_activation,
+        L = self.L,
+        
+        #forecasting task
+        prediction_length = self.prediction_length,
+        context_length = self.context_length,#seq_len : input sequence length
+        label_length = self.label_length,#start token length
+        
+        # model arguments
+        nhead = self.nhead,
+        num_encoder_layers = self.num_encoder_layers, 
+        num_decoder_layers = self.num_decoder_layers, 
+        dim_feedforward = self.dim_feedforward, #dimension of fcn
+        input_size = self.input_size, #encoder input size
+        # dec_in = self.dec_in, #decoder input size
+        c_out =self.c_out, #output size
+        moving_avg = self.moving_avg,
+        factor = self.factor,
+        scaling = self.scaling,
+        activation = self.activation,
+        dropout = self.dropout,
+        embed = self.embed,#options:[timeF, fixed, learned]
+        output_attention = self.output_attention,#whether to output attention in encoder
+        num_feat_dynamic_real=1
+        + self.num_feat_dynamic_real
+        + len(self.time_features),
+        num_feat_static_real=max(1, self.num_feat_static_real),
+        num_feat_static_cat=max(1, self.num_feat_static_cat),
+        cardinality=self.cardinality,
+        embedding_dimension=self.embedding_dimension,
+        # univariate input
+        distr_output=self.distr_output,
+        lags_seq=self.lags_seq,
+        num_parallel_samples=self.num_parallel_samples,
+        )
+
+        return FEDformerLightningModule(model=model, loss=self.loss)
diff --git a/fedformer/lightning_module.py b/fedformer/lightning_module.py
new file mode 100644
index 0000000..56448fb
--- /dev/null
+++ b/fedformer/lightning_module.py
@@ -0,0 +1,81 @@
+import pytorch_lightning as pl
+import torch
+from gluonts.torch.modules.loss import DistributionLoss, NegativeLogLikelihood
+from gluonts.torch.util import weighted_average
+
+from module import FEDformerModel
+
+
+class FEDformerLightningModule(pl.LightningModule):
+    def __init__(
+        self,
+        model: FEDformerModel,
+        loss: DistributionLoss = NegativeLogLikelihood(),
+        lr: float = 1e-3,
+        weight_decay: float = 1e-8,
+    ) -> None:
+        super().__init__()
+        self.save_hyperparameters()
+        self.model = model
+        self.loss = loss
+        self.lr = lr
+        self.weight_decay = weight_decay
+
+    def training_step(self, batch, batch_idx: int):
+        """Execute training step"""
+        train_loss = self(batch)
+        self.log(
+            "train_loss",
+            train_loss,
+            on_epoch=True,
+            on_step=False,
+            prog_bar=True,
+        )
+        return train_loss
+
+    def validation_step(self, batch, batch_idx: int):
+        """Execute validation step"""
+        with torch.inference_mode():
+            val_loss = self(batch)
+        self.log("val_loss", val_loss, on_epoch=True, on_step=False, prog_bar=True)
+        return val_loss
+
+    def configure_optimizers(self):
+        """Returns the optimizer to use"""
+        return torch.optim.Adam(
+            self.model.parameters(),
+            lr=self.lr,
+            weight_decay=self.weight_decay,
+        )
+
+    def forward(self, batch):
+        feat_static_cat = batch["feat_static_cat"]
+        feat_static_real = batch["feat_static_real"]
+        past_time_feat = batch["past_time_feat"]
+        past_target = batch["past_target"]
+        future_time_feat = batch["future_time_feat"]
+        future_target = batch["future_target"]
+        past_observed_values = batch["past_observed_values"]
+        future_observed_values = batch["future_observed_values"]
+
+        transformer_inputs, scale, _ = self.model.create_network_inputs(
+            feat_static_cat,
+            feat_static_real,
+            past_time_feat,
+            past_target,
+            past_observed_values,
+            future_time_feat,
+            future_target,
+        )
+        print('transformer_inputs', transformer_inputs.size())
+        params = self.model.output_params(transformer_inputs)
+        distr = self.model.output_distribution(params, scale)
+
+        loss_values = self.loss(distr, future_target)
+
+        if len(self.model.target_shape) == 0:
+            loss_weights = future_observed_values
+        else:
+            loss_weights = future_observed_values.min(dim=-1, keepdim=False)
+
+        return weighted_average(loss_values, weights=loss_weights)
diff --git a/fedformer/module.py b/fedformer/module.py
new file mode 100644
index 0000000..22c4d5f
--- /dev/null
+++ b/fedformer/module.py
@@ -0,0 +1,1722 @@
+from math import sqrt
+import math
+from typing import List, Optional
+
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from gluonts.core.component import validated
+from gluonts.time_feature import get_lags_for_frequency
+from gluonts.torch.distributions import DistributionOutput, StudentTOutput
+from gluonts.torch.modules.feature import FeatureEmbedder
+from gluonts.torch.modules.scaler import MeanScaler, NOPScaler
+import random
+from matplotlib import pyplot as plt
+from torch.nn.functional import interpolate
+
+class TriangularCausalMask:
+    def __init__(self, B, L, device="cpu"):
+        mask_shape = [B, 1, L, L]
+        with torch.no_grad():
+            self._mask = torch.triu(
+                torch.ones(mask_shape, dtype=torch.bool), diagonal=1
+            ).to(device)
+
+    @property
+    def mask(self):
+        return self._mask
+
+
+class ProbMask:
+    def __init__(self, B, H, L, index, scores, device="cpu"):
+        _mask = torch.ones(L, scores.shape[-1], dtype=torch.bool).to(device).triu(1)
+        _mask_ex = _mask[None, None, :].expand(B, H, L, scores.shape[-1])
+        indicator = _mask_ex[
+            torch.arange(B)[:, None, None], torch.arange(H)[None, :, None], index, :
+        ].to(device)
+        self._mask = indicator.view(scores.shape).to(device)
+
+    @property
+    def mask(self):
+        return self._mask
+
+class LocalMask():
+    def __init__(self, B, L,S,device="cpu"):
+        mask_shape = [B, 1, L, S]
+        with torch.no_grad():
+            self.len = math.ceil(np.log2(L))
+            self._mask1 = torch.triu(torch.ones(mask_shape, dtype=torch.bool), diagonal=1).to(device)
+            self._mask2 = ~torch.triu(torch.ones(mask_shape,dtype=torch.bool),diagonal=-self.len).to(device)
+            self._mask = self._mask1+self._mask2
+    @property
+    def mask(self):
+        return self._mask
+        
+def adjust_learning_rate(optimizer, epoch, args):
+    # lr = args.learning_rate * (0.2 ** (epoch // 2))
+    if args.lradj == 'type1':
+        lr_adjust = {epoch: args.learning_rate * (0.5 ** ((epoch - 1) // 1))}
+    elif args.lradj == 'type2':
+        lr_adjust = {
+            2: 5e-5, 4: 1e-5, 6: 5e-6, 8: 1e-6,
+            10: 5e-7, 15: 1e-7, 20: 5e-8
+        }
+    elif args.lradj =='type3':
+        lr_adjust = {epoch: args.learning_rate}
+    elif args.lradj == 'type4':
+        lr_adjust = {epoch: args.learning_rate * (0.9 ** ((epoch - 1) // 1))}
+    if epoch in lr_adjust.keys():
+        lr = lr_adjust[epoch]
+        for param_group in optimizer.param_groups:
+            param_group['lr'] = lr
+        print('Updating learning rate to {}'.format(lr))
+
+
+class EarlyStopping:
+    def __init__(self, patience=7, verbose=False, delta=0):
+        self.patience = patience
+        self.verbose = verbose
+        self.counter = 0
+        self.best_score = None
+        self.early_stop = False
+        self.val_loss_min = np.Inf
+        self.delta = delta
+
+    def __call__(self, val_loss, model, path):
+        score = -val_loss
+        if self.best_score is None:
+            self.best_score = score
+            self.save_checkpoint(val_loss, model, path)
+        elif score < self.best_score + self.delta:
+            self.counter += 1
+            print(f'EarlyStopping counter: {self.counter} out of {self.patience}')
+            if self.counter >= self.patience:
+                self.early_stop = True
+        else:
+            self.best_score = score
+            self.save_checkpoint(val_loss, model, path)
+            self.counter = 0
+
+    def save_checkpoint(self, val_loss, model, path):
+        if self.verbose:
+            print(f'Validation loss decreased ({self.val_loss_min:.6f} --> {val_loss:.6f}).  Saving model ...')
+        torch.save(model.state_dict(), path + '/' + 'checkpoint.pth')
+        self.val_loss_min = val_loss
+
+
+class dotdict(dict):
+    """dot.notation access to dictionary attributes"""
+    __getattr__ = dict.get
+    __setattr__ = dict.__setitem__
+    __delattr__ = dict.__delitem__
+
+
+class StandardScaler():
+    def __init__(self, mean, std):
+        self.mean = mean
+        self.std = std
+
+    def transform(self, data):
+        return (data - self.mean) / self.std
+
+    def inverse_transform(self, data):
+        return (data * self.std) + self.mean
+
+
+def visual(true, preds=None, name='./pic/test.pdf'):
+    """
+    Results visualization
+    """
+    plt.figure()
+    plt.plot(true, label='GroundTruth', linewidth=2)
+    if preds is not None:
+        plt.plot(preds, label='Prediction', linewidth=2)
+    plt.legend()
+    plt.savefig(name, bbox_inches='tight')
+        
+def decor_time(func):
+    def func2(*args, **kw):
+        now = time.time()
+        y = func(*args, **kw)
+        t = time.time() - now
+        print('call <{}>, time={}'.format(func.__name__, t))
+        return y
+    return func2
+
+
+class AutoCorrelation(nn.Module):
+    """
+    AutoCorrelation Mechanism with the following two phases:
+    (1) period-based dependencies discovery
+    (2) time delay aggregation
+    This block can replace the self-attention family mechanism seamlessly.
+    """
+    def __init__(self, mask_flag=True, factor=1, scale=None, attention_dropout=0.1, output_attention=False, wavelet=False):
+        super(AutoCorrelation, self).__init__()
+        print('Autocorrelation used !')
+        self.factor = factor
+        self.scale = scale
+        self.mask_flag = mask_flag
+        self.output_attention = output_attention
+        self.dropout = nn.Dropout(attention_dropout)
+        self.agg = None
+        self.use_wavelet = wavelet
+
+    # @decor_time
+    def time_delay_agg_training(self, values, corr):
+        """
+        SpeedUp version of Autocorrelation (a batch-normalization style design)
+        This is for the training phase.
+        """
+        head = values.shape[1]
+        channel = values.shape[2]
+        length = values.shape[3]
+        # find top k
+        top_k = int(self.factor * math.log(length))
+        mean_value = torch.mean(torch.mean(corr, dim=1), dim=1)
+        index = torch.topk(torch.mean(mean_value, dim=0), top_k, dim=-1)[1]
+        weights = torch.stack([mean_value[:, index[i]] for i in range(top_k)], dim=-1)
+        # update corr
+        tmp_corr = torch.softmax(weights, dim=-1)
+        # aggregation
+        tmp_values = values
+        delays_agg = torch.zeros_like(values).float()
+        for i in range(top_k):
+            pattern = torch.roll(tmp_values, -int(index[i]), -1)
+            delays_agg = delays_agg + pattern * \
+                         (tmp_corr[:, i].unsqueeze(1).unsqueeze(1).unsqueeze(1).repeat(1, head, channel, length))
+        return delays_agg  # size=[B, H, d, S]
+
+    def time_delay_agg_inference(self, values, corr):
+        """
+        SpeedUp version of Autocorrelation (a batch-normalization style design)
+        This is for the inference phase.
+        """
+        batch = values.shape[0]
+        head = values.shape[1]
+        channel = values.shape[2]
+        length = values.shape[3]
+        # index init
+        init_index = torch.arange(length).unsqueeze(0).unsqueeze(0).unsqueeze(0).repeat(batch, head, channel, 1).cuda()
+        # find top k
+        top_k = int(self.factor * math.log(length))
+        mean_value = torch.mean(torch.mean(corr, dim=1), dim=1)
+        weights = torch.topk(mean_value, top_k, dim=-1)[0]
+        delay = torch.topk(mean_value, top_k, dim=-1)[1]
+        # update corr
+        tmp_corr = torch.softmax(weights, dim=-1)
+        # aggregation
+        tmp_values = values.repeat(1, 1, 1, 2)
+        delays_agg = torch.zeros_like(values).float()
+        for i in range(top_k):
+            tmp_delay = init_index + delay[:, i].unsqueeze(1).unsqueeze(1).unsqueeze(1).repeat(1, head, channel, length)
+            pattern = torch.gather(tmp_values, dim=-1, index=tmp_delay)
+            delays_agg = delays_agg + pattern * \
+                         (tmp_corr[:, i].unsqueeze(1).unsqueeze(1).unsqueeze(1).repeat(1, head, channel, length))
+        return delays_agg
+
+    def time_delay_agg_full(self, values, corr):
+        """
+        Standard version of Autocorrelation
+        """
+        batch = values.shape[0]
+        head = values.shape[1]
+        channel = values.shape[2]
+        length = values.shape[3]
+        # index init
+        init_index = torch.arange(length).unsqueeze(0).unsqueeze(0).unsqueeze(0).repeat(batch, head, channel, 1).cuda()
+        # find top k
+        top_k = int(self.factor * math.log(length))
+        weights = torch.topk(corr, top_k, dim=-1)[0]
+        delay = torch.topk(corr, top_k, dim=-1)[1]
+        # update corr
+        tmp_corr = torch.softmax(weights, dim=-1)
+        # aggregation
+        tmp_values = values.repeat(1, 1, 1, 2)
+        delays_agg = torch.zeros_like(values).float()
+        for i in range(top_k):
+            tmp_delay = init_index + delay[..., i].unsqueeze(-1)
+            pattern = torch.gather(tmp_values, dim=-1, index=tmp_delay)
+            delays_agg = delays_agg + pattern * (tmp_corr[..., i].unsqueeze(-1))
+        return delays_agg
+
+    def forward(self, queries, keys, values, attn_mask):
+        B, L, H, E = queries.shape
+        _, S, _, D = values.shape
+        if L > S:
+            zeros = torch.zeros_like(queries[:, :(L - S), :]).float()
+            values = torch.cat([values, zeros], dim=1)
+            keys = torch.cat([keys, zeros], dim=1)
+        else:
+            values = values[:, :L, :, :]
+            keys = keys[:, :L, :, :]
+
+        # period-based dependencies
+        if self.use_wavelet != 2:
+            if self.use_wavelet == 1:
+                j_list = self.j_list
+                queries = queries.reshape([B, L, -1])
+                keys = keys.reshape([B, L, -1])
+                Ql, Qh_list = self.dwt1d(queries.transpose(1, 2))  # [B, H*D, L]
+                Kl, Kh_list = self.dwt1d(keys.transpose(1, 2))
+                qs = [queries.transpose(1, 2)] + Qh_list + [Ql]  # [B, H*D, L]
+                ks = [keys.transpose(1, 2)] + Kh_list + [Kl]
+                q_list = []
+                k_list = []
+                for q, k, j in zip(qs, ks, j_list):
+                    q_list += [interpolate(q, scale_factor=j, mode='linear')[:, :, -L:]]
+                    k_list += [interpolate(k, scale_factor=j, mode='linear')[:, :, -L:]]
+                queries = torch.stack([i.reshape([B, H, E, L]) for i in q_list], dim=3).reshape([B, H, -1, L]).permute(0, 3, 1, 2)
+                keys = torch.stack([i.reshape([B, H, E, L]) for i in k_list], dim=3).reshape([B, H, -1, L]).permute(0, 3, 1, 2)
+            else:
+                pass
+            q_fft = torch.fft.rfft(queries.permute(0, 2, 3, 1).contiguous(), dim=-1)  # size=[B, H, E, L]
+            k_fft = torch.fft.rfft(keys.permute(0, 2, 3, 1).contiguous(), dim=-1)
+            res = q_fft * torch.conj(k_fft)
+            corr = torch.fft.irfft(res, dim=-1) # size=[B, H, E, L]
+
+            # time delay agg
+            if self.training:
+                V = self.time_delay_agg_training(values.permute(0, 2, 3, 1).contiguous(), corr).permute(0, 3, 1, 2)  # [B, L, H, E], [B, H, E, L] -> [B, L, H, E]
+            else:
+                V = self.time_delay_agg_inference(values.permute(0, 2, 3, 1).contiguous(), corr).permute(0, 3, 1, 2)
+        else:
+            V_list = []
+            queries = queries.reshape([B, L, -1])
+            keys = keys.reshape([B, L, -1])
+            values = values.reshape([B, L, -1])
+            Ql, Qh_list = self.dwt1d(queries.transpose(1, 2))  # [B, H*D, L]
+            Kl, Kh_list = self.dwt1d(keys.transpose(1, 2))
+            Vl, Vh_list = self.dwt1d(values.transpose(1, 2))
+            qs = Qh_list + [Ql]  # [B, H*D, L]
+            ks = Kh_list + [Kl]
+            vs = Vh_list + [Vl]
+            for q, k, v in zip(qs, ks, vs):
+                q = q.reshape([B, H, E, -1])
+                k = k.reshape([B, H, E, -1])
+                v = v.reshape([B, H, E, -1]).permute(0, 3, 1, 2)
+                q_fft = torch.fft.rfft(q.contiguous(), dim=-1)
+                k_fft = torch.fft.rfft(k.contiguous(), dim=-1)
+                res = q_fft * torch.conj(k_fft)
+                corr = torch.fft.irfft(res, dim=-1)  # [B, H, E, L]
+                if self.training:
+                    V = self.time_delay_agg_training(v.permute(0, 2, 3, 1).contiguous(), corr).permute(0, 3, 1, 2)
+                else:
+                    V = self.time_delay_agg_inference(v.permute(0, 2, 3, 1).contiguous(), corr).permute(0, 3, 1, 2)
+                V_list += [V]
+            Vl = V_list[-1].reshape([B, -1, H*E]).transpose(1, 2)
+            Vh_list = [i.reshape([B, -1, H*E]).transpose(1, 2) for i in V_list[:-1]]
+            V = self.dwt1div((Vl, Vh_list)).reshape([B, H, E, -1]).permute(0, 3, 1, 2)
+            # corr = self.dwt1div((V_list[-1], V_list[:-1]))
+
+        if self.output_attention:
+            return (V.contiguous(), corr.permute(0, 3, 1, 2))  # size = [B, L, H, E]
+        else:
+            return (V.contiguous(), None)
+
+
+class AutoCorrelationLayer(nn.Module):
+    def __init__(self, correlation, d_model, n_heads, d_keys=None,
+                 d_values=None):
+        super(AutoCorrelationLayer, self).__init__()
+
+        d_keys = d_keys or (d_model // n_heads)
+        d_values = d_values or (d_model // n_heads)
+
+        self.inner_correlation = correlation
+        self.query_projection = nn.Linear(d_model, d_keys * n_heads)
+        self.key_projection = nn.Linear(d_model, d_keys * n_heads)
+        self.value_projection = nn.Linear(d_model, d_values * n_heads)
+        self.out_projection = nn.Linear(d_values * n_heads, d_model)
+        self.n_heads = n_heads
+
+    def forward(self, queries, keys, values, attn_mask):
+        B, L, _ = queries.shape
+        _, S, _ = keys.shape
+        H = self.n_heads
+        print(queries.size())
+        print('query proj',self.query_projection(queries).size())
+        queries = self.query_projection(queries).view(B, L, H, -1)
+        keys = self.key_projection(keys).view(B, S, H, -1)
+        values = self.value_projection(values).view(B, S, H, -1)
+
+        out, attn = self.inner_correlation(
+            queries,
+            keys,
+            values,
+            attn_mask
+        )
+
+        out = out.view(B, L, -1)
+        return self.out_projection(out), attn
+        
+class my_Layernorm(nn.Module):
+    """
+    Special designed layernorm for the seasonal part
+    """
+    def __init__(self, channels):
+        super(my_Layernorm, self).__init__()
+        self.layernorm = nn.LayerNorm(channels)
+
+    def forward(self, x):
+        x_hat = self.layernorm(x)
+        bias = torch.mean(x_hat, dim=1).unsqueeze(1).repeat(1, x.shape[1], 1)
+        return x_hat - bias
+
+
+class moving_avg(nn.Module):
+    """
+    Moving average block to highlight the trend of time series
+    """
+    def __init__(self, kernel_size, stride):
+        super(moving_avg, self).__init__()
+        self.kernel_size = kernel_size
+        self.avg = nn.AvgPool1d(kernel_size=kernel_size, stride=stride, padding=0)
+
+    def forward(self, x):
+        # padding on the both ends of time series
+        front = x[:, 0:1, :].repeat(1, self.kernel_size - 1-math.floor((self.kernel_size - 1) // 2), 1)
+        end = x[:, -1:, :].repeat(1, math.floor((self.kernel_size - 1) // 2), 1)
+        x = torch.cat([front, x, end], dim=1)
+        x = self.avg(x.permute(0, 2, 1))
+        x = x.permute(0, 2, 1)
+        return x
+
+
+class series_decomp(nn.Module):
+    """
+    Series decomposition block
+    """
+    def __init__(self, kernel_size):
+        super(series_decomp, self).__init__()
+        self.moving_avg = moving_avg(kernel_size, stride=1)
+
+    def forward(self, x):
+        moving_mean = self.moving_avg(x)
+        res = x - moving_mean
+        return res, moving_mean
+
+
+class series_decomp_multi(nn.Module):
+    """
+    Series decomposition block
+    """
+    def __init__(self, kernel_size):
+        super(series_decomp_multi, self).__init__()
+        self.moving_avg = [moving_avg(kernel, stride=1) for kernel in kernel_size]
+        self.layer = torch.nn.Linear(1, len(kernel_size))
+
+    def forward(self, x):
+        moving_mean=[]
+        for func in self.moving_avg:
+            moving_avg = func(x)
+            moving_mean.append(moving_avg.unsqueeze(-1))
+        moving_mean=torch.cat(moving_mean,dim=-1)
+        moving_mean = torch.sum(moving_mean*nn.Softmax(-1)(self.layer(x.unsqueeze(-1))),dim=-1)
+        res = x - moving_mean
+        return res, moving_mean 
+
+
+class FourierDecomp(nn.Module):
+    def __init__(self):
+        super(FourierDecomp, self).__init__()
+        pass
+
+    def forward(self, x):
+        x_ft = torch.fft.rfft(x, dim=-1)
+
+
+class EncoderLayer(nn.Module):
+    """
+    Autoformer encoder layer with the progressive decomposition architecture
+    """
+    def __init__(self, attention, d_model, d_ff=None, moving_avg=25, dropout=0.1, activation="relu"):
+        super(EncoderLayer, self).__init__()
+        d_ff = d_ff or 4 * d_model
+        self.attention = attention
+        self.conv1 = nn.Conv1d(in_channels=d_model, out_channels=d_ff, kernel_size=1, bias=False)
+        self.conv2 = nn.Conv1d(in_channels=d_ff, out_channels=d_model, kernel_size=1, bias=False)
+
+        if isinstance(moving_avg, list):
+            self.decomp1 = series_decomp_multi(moving_avg)
+            self.decomp2 = series_decomp_multi(moving_avg)
+        else:
+            self.decomp1 = series_decomp(moving_avg)
+            self.decomp2 = series_decomp(moving_avg)
+
+        self.dropout = nn.Dropout(dropout)
+        self.activation = F.relu if activation == "relu" else F.gelu
+
+    def forward(self, x, attn_mask=None):
+        new_x, attn = self.attention(
+            x, x, x,
+            attn_mask=attn_mask
+        )
+        x = x + self.dropout(new_x)
+        x, _ = self.decomp1(x)
+        y = x
+        y = self.dropout(self.activation(self.conv1(y.transpose(-1, 1))))
+        y = self.dropout(self.conv2(y).transpose(-1, 1))
+        res, _ = self.decomp2(x + y)
+        return res, attn
+
+
+class Encoder(nn.Module):
+    """
+    Autoformer encoder
+    """
+    def __init__(self, attn_layers, conv_layers=None, norm_layer=None):
+        super(Encoder, self).__init__()
+        self.attn_layers = nn.ModuleList(attn_layers)
+        self.conv_layers = nn.ModuleList(conv_layers) if conv_layers is not None else None
+        self.norm = norm_layer
+
+    def forward(self, x, attn_mask=None):
+        attns = []
+        if self.conv_layers is not None:
+            for attn_layer, conv_layer in zip(self.attn_layers, self.conv_layers):
+                x, attn = attn_layer(x, attn_mask=attn_mask)
+                x = conv_layer(x)
+                attns.append(attn)
+            x, attn = self.attn_layers[-1](x)
+            attns.append(attn)
+        else:
+            for attn_layer in self.attn_layers:
+                x, attn = attn_layer(x, attn_mask=attn_mask)
+                attns.append(attn)
+
+        if self.norm is not None:
+            x = self.norm(x)
+
+        return x, attns
+
+class DecoderLayer(nn.Module):
+    """
+    Autoformer decoder layer with the progressive decomposition architecture
+    """
+    def __init__(self, self_attention, cross_attention, d_model, c_out, d_ff=None,
+                 moving_avg=25, dropout=0.1, activation="relu"):
+        super(DecoderLayer, self).__init__()
+        d_ff = d_ff or 4 * d_model
+        self.self_attention = self_attention
+        self.cross_attention = cross_attention
+        self.conv1 = nn.Conv1d(in_channels=d_model, out_channels=d_ff, kernel_size=1, bias=False)
+        self.conv2 = nn.Conv1d(in_channels=d_ff, out_channels=d_model, kernel_size=1, bias=False)
+
+        if isinstance(moving_avg, list):
+            self.decomp1 = series_decomp_multi(moving_avg)
+            self.decomp2 = series_decomp_multi(moving_avg)
+            self.decomp3 = series_decomp_multi(moving_avg)
+        else:
+            self.decomp1 = series_decomp(moving_avg)
+            self.decomp2 = series_decomp(moving_avg)
+            self.decomp3 = series_decomp(moving_avg)
+
+        self.dropout = nn.Dropout(dropout)
+        self.projection = nn.Conv1d(in_channels=d_model, out_channels=c_out, kernel_size=3, stride=1, padding=1,
+                                    padding_mode='circular', bias=False)
+        self.activation = F.relu if activation == "relu" else F.gelu
+
+    def forward(self, x, cross, x_mask=None, cross_mask=None):
+        x = x + self.dropout(self.self_attention(
+            x, x, x,
+            attn_mask=x_mask
+        )[0])
+
+        x, trend1 = self.decomp1(x)
+        x = x + self.dropout(self.cross_attention(
+            x, cross, cross,
+            attn_mask=cross_mask
+        )[0])
+
+        x, trend2 = self.decomp2(x)
+        y = x
+        y = self.dropout(self.activation(self.conv1(y.transpose(-1, 1))))
+        y = self.dropout(self.conv2(y).transpose(-1, 1))
+        x, trend3 = self.decomp3(x + y)
+
+        residual_trend = trend1 + trend2 + trend3
+        residual_trend = self.projection(residual_trend.permute(0, 2, 1)).transpose(1, 2)
+        return x, residual_trend
+
+
+class Decoder(nn.Module):
+    """
+    Autoformer encoder
+    """
+    def __init__(self, layers, norm_layer=None, projection=None):
+        super(Decoder, self).__init__()
+        self.layers = nn.ModuleList(layers)
+        self.norm = norm_layer
+        self.projection = projection
+
+    def forward(self, x, cross, x_mask=None, cross_mask=None, trend=None):
+        for layer in self.layers:
+            x, residual_trend = layer(x, cross, x_mask=x_mask, cross_mask=cross_mask)
+            trend = trend + residual_trend
+
+        if self.norm is not None:
+            x = self.norm(x)
+
+        if self.projection is not None:
+            x = self.projection(x)
+        return x, trend
+        
+class PositionalEmbedding(nn.Module):
+    def __init__(self, d_model, max_len=5000):
+        super(PositionalEmbedding, self).__init__()
+        # Compute the positional encodings once in log space.
+        pe = torch.zeros(max_len, d_model).float()
+        pe.require_grad = False
+
+        position = torch.arange(0, max_len).float().unsqueeze(1)
+        div_term = (torch.arange(0, d_model, 2).float() * -(math.log(10000.0) / d_model)).exp()
+
+        pe[:, 0::2] = torch.sin(position * div_term)
+        pe[:, 1::2] = torch.cos(position * div_term)
+
+        pe = pe.unsqueeze(0)
+        self.register_buffer('pe', pe)
+
+    def forward(self, x):
+        return self.pe[:, :x.size(1)]
+
+
+class TokenEmbedding(nn.Module):
+    def __init__(self, c_in, d_model):
+        super(TokenEmbedding, self).__init__()
+        padding = 1 if torch.__version__ >= '1.5.0' else 2
+        self.tokenConv = nn.Conv1d(in_channels=c_in, out_channels=d_model,
+                                   kernel_size=3, padding=padding, padding_mode='circular', bias=False)
+        for m in self.modules():
+            if isinstance(m, nn.Conv1d):
+                nn.init.kaiming_normal_(m.weight, mode='fan_in', nonlinearity='leaky_relu')
+
+    def forward(self, x):
+        x = self.tokenConv(x.permute(0, 2, 1)).transpose(1, 2)
+        return x
+
+
+class FixedEmbedding(nn.Module):
+    def __init__(self, c_in, d_model):
+        super(FixedEmbedding, self).__init__()
+
+        w = torch.zeros(c_in, d_model).float()
+        w.require_grad = False
+
+        position = torch.arange(0, c_in).float().unsqueeze(1)
+        div_term = (torch.arange(0, d_model, 2).float() * -(math.log(10000.0) / d_model)).exp()
+
+        w[:, 0::2] = torch.sin(position * div_term)
+        w[:, 1::2] = torch.cos(position * div_term)
+
+        self.emb = nn.Embedding(c_in, d_model)
+        self.emb.weight = nn.Parameter(w, requires_grad=False)
+
+    def forward(self, x):
+        return self.emb(x).detach()
+
+
+class TemporalEmbedding(nn.Module):
+    def __init__(self, d_model, embed_type='fixed', freq='h'):
+        super(TemporalEmbedding, self).__init__()
+
+        minute_size = 4
+        hour_size = 24
+        weekday_size = 7
+        day_size = 32
+        month_size = 13
+
+        Embed = FixedEmbedding if embed_type == 'fixed' else nn.Embedding
+        if freq == 't':
+            self.minute_embed = Embed(minute_size, d_model)
+        self.hour_embed = Embed(hour_size, d_model)
+        self.weekday_embed = Embed(weekday_size, d_model)
+        self.day_embed = Embed(day_size, d_model)
+        self.month_embed = Embed(month_size, d_model)
+
+    def forward(self, x):
+        x = x.long()
+
+        minute_x = self.minute_embed(x[:, :, 4]) if hasattr(self, 'minute_embed') else 0.
+        hour_x = self.hour_embed(x[:, :, 3])
+        weekday_x = self.weekday_embed(x[:, :, 2])
+        day_x = self.day_embed(x[:, :, 1])
+        month_x = self.month_embed(x[:, :, 0])
+
+        return hour_x + weekday_x + day_x + month_x + minute_x
+
+
+class TimeFeatureEmbedding(nn.Module):
+    def __init__(self, d_model, embed_type='timeF', freq='h'):
+        super(TimeFeatureEmbedding, self).__init__()
+
+        freq_map = {'h': 4, 't': 5, 's': 6, 'm': 1, 'a': 1, 'w': 2, 'd': 3, 'b': 3}
+        d_inp = freq_map[freq]
+        self.embed = nn.Linear(d_inp, d_model, bias=False)
+
+    def forward(self, x):
+        return self.embed(x)
+
+
+class DataEmbedding(nn.Module):
+    def __init__(self, c_in, d_model, embed_type='fixed', freq='h', dropout=0.1):
+        super(DataEmbedding, self).__init__()
+
+        self.value_embedding = TokenEmbedding(c_in=c_in, d_model=d_model)
+        self.position_embedding = PositionalEmbedding(d_model=d_model)
+        self.temporal_embedding = TemporalEmbedding(d_model=d_model, embed_type=embed_type,
+                                                    freq=freq) if embed_type != 'timeF' else TimeFeatureEmbedding(
+            d_model=d_model, embed_type=embed_type, freq=freq)
+        self.dropout = nn.Dropout(p=dropout)
+
+    def forward(self, x, x_mark):
+        x = self.value_embedding(x) + self.temporal_embedding(x_mark) + self.position_embedding(x)
+        return self.dropout(x)
+
+class DataEmbedding_onlypos(nn.Module):
+    def __init__(self, c_in, d_model, embed_type='fixed', freq='h', dropout=0.1):
+        super(DataEmbedding_onlypos, self).__init__()
+
+        self.value_embedding = TokenEmbedding(c_in=c_in, d_model=d_model)
+        self.position_embedding = PositionalEmbedding(d_model=d_model)
+        self.dropout = nn.Dropout(p=dropout)
+
+    def forward(self, x, x_mark):
+        x = self.value_embedding(x) + self.position_embedding(x)
+        return self.dropout(x)
+    
+class DataEmbedding_wo_pos(nn.Module):
+    def __init__(self, c_in, d_model, embed_type='fixed', freq='h', dropout=0.1):
+        super(DataEmbedding_wo_pos, self).__init__()
+
+        self.value_embedding = TokenEmbedding(c_in=c_in, d_model=d_model)
+        self.position_embedding = PositionalEmbedding(d_model=d_model)
+        self.temporal_embedding = TemporalEmbedding(d_model=d_model, embed_type=embed_type,
+                                                    freq=freq) if embed_type != 'timeF' else TimeFeatureEmbedding(
+            d_model=d_model, embed_type=embed_type, freq=freq)
+        self.dropout = nn.Dropout(p=dropout)
+
+    def forward(self, x, x_mark):
+        # try:
+        x = self.value_embedding(x) + self.temporal_embedding(x_mark)
+        # except:
+        #     a = 1
+        return self.dropout(x)
+        
+def get_frequency_modes(seq_len, modes=64, mode_select_method='random'):
+    """
+    get modes on frequency domain:
+    'random' means sampling randomly;
+    'else' means sampling the lowest modes;
+    """
+    modes = min(modes, seq_len//2)
+    if mode_select_method == 'random':
+        index = list(range(0, seq_len // 2))
+        np.random.shuffle(index)
+        index = index[:modes]
+    else:
+        index = list(range(0, modes))
+    index.sort()
+    return index
+
+
+# ########## fourier layer #############
+class FourierBlock(nn.Module):
+    def __init__(self, in_channels, out_channels, seq_len, modes=0, mode_select_method='random'):
+        super(FourierBlock, self).__init__()
+        print('fourier enhanced block used!')
+        """
+        1D Fourier block. It performs representation learning on frequency domain, 
+        it does FFT, linear transform, and Inverse FFT.    
+        """
+        # get modes on frequency domain
+        self.index = get_frequency_modes(seq_len, modes=modes, mode_select_method=mode_select_method)
+        print('modes={}, index={}'.format(modes, self.index))
+
+        self.scale = (1 / (in_channels * out_channels))
+        self.weights1 = nn.Parameter(self.scale * torch.rand(8, in_channels // 8, out_channels // 8, len(self.index), dtype=torch.cfloat))
+
+    # Complex multiplication
+    def compl_mul1d(self, input, weights):
+        # (batch, in_channel, x ), (in_channel, out_channel, x) -> (batch, out_channel, x)
+        print('input fft', input.size())
+        print('weight fft', weights.size())
+        return torch.einsum("bhi,hio->bho", input, weights)# hio->bho
+
+    def forward(self, q, k, v, mask):
+        # size = [B, L, H, E]
+        B, L, H, E = q.shape
+        x = q.permute(0, 2, 3, 1)
+        # Compute Fourier coefficients
+        x_ft = torch.fft.rfft(x, dim=-1)
+        print(x_ft.size())
+        # Perform Fourier neural operations
+        out_ft = torch.zeros(B, H, E, L // 2 + 1, device=x.device, dtype=torch.cfloat)
+        for wi, i in enumerate(self.index):
+            out_ft[:, :, :, wi] = self.compl_mul1d(x_ft[:, :, :, i], self.weights1[:, :, :, wi])
+        # Return to time domain
+        x = torch.fft.irfft(out_ft, n=x.size(-1))
+        return (x, None)
+
+
+# ########## Fourier Cross Former ####################
+class FourierCrossAttention(nn.Module):
+    def __init__(self, in_channels, out_channels, seq_len_q, seq_len_kv, modes=64, mode_select_method='random',
+                 activation='tanh', policy=0):
+        super(FourierCrossAttention, self).__init__()
+        print(' fourier enhanced cross attention used!')
+        """
+        1D Fourier Cross Attention layer. It does FFT, linear transform, attention mechanism and Inverse FFT.    
+        """
+        self.activation = activation
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        # get modes for queries and keys (& values) on frequency domain
+        self.index_q = get_frequency_modes(seq_len_q, modes=modes, mode_select_method=mode_select_method)
+        self.index_kv = get_frequency_modes(seq_len_kv, modes=modes, mode_select_method=mode_select_method)
+
+        print('modes_q={}, index_q={}'.format(len(self.index_q), self.index_q))
+        print('modes_kv={}, index_kv={}'.format(len(self.index_kv), self.index_kv))
+
+        self.scale = (1 / (in_channels * out_channels))
+        self.weights1 = nn.Parameter(
+            self.scale * torch.rand(8, in_channels // 8, out_channels // 8, len(self.index_q), dtype=torch.cfloat))
+
+    # Complex multiplication
+    def compl_mul1d(self, input, weights):
+        # (batch, in_channel, x ), (in_channel, out_channel, x) -> (batch, out_channel, x)
+        return torch.einsum("bhi,hoi->boi", input, weights)#bhi,hio->bho"
+
+    def forward(self, q, k, v, mask):
+        # size = [B, L, H, E]
+        B, L, H, E = q.shape
+        xq = q.permute(0, 2, 3, 1)  # size = [B, H, E, L]
+        xk = k.permute(0, 2, 3, 1)
+        xv = v.permute(0, 2, 3, 1)
+
+        # Compute Fourier coefficients
+        xq_ft_ = torch.zeros(B, H, E, len(self.index_q), device=xq.device, dtype=torch.cfloat)
+        xq_ft = torch.fft.rfft(xq, dim=-1)
+        for i, j in enumerate(self.index_q):
+            xq_ft_[:, :, :, i] = xq_ft[:, :, :, j]
+        xk_ft_ = torch.zeros(B, H, E, len(self.index_kv), device=xq.device, dtype=torch.cfloat)
+        xk_ft = torch.fft.rfft(xk, dim=-1)
+        for i, j in enumerate(self.index_kv):
+            xk_ft_[:, :, :, i] = xk_ft[:, :, :, j]
+
+        # perform attention mechanism on frequency domain
+        xqk_ft = (torch.einsum("bhex,bhey->bhxy", xq_ft_, xk_ft_))
+        if self.activation == 'tanh':
+            xqk_ft = xqk_ft.tanh()
+        elif self.activation == 'softmax':
+            xqk_ft = torch.softmax(abs(xqk_ft), dim=-1)
+            xqk_ft = torch.complex(xqk_ft, torch.zeros_like(xqk_ft))
+        else:
+            raise Exception('{} actiation function is not implemented'.format(self.activation))
+        xqkv_ft = torch.einsum("bhxy,bhey->bhex", xqk_ft, xk_ft_)
+        xqkvw = torch.einsum("bhex,heox->bhox", xqkv_ft, self.weights1)
+        out_ft = torch.zeros(B, H, E, L // 2 + 1, device=xq.device, dtype=torch.cfloat)
+        for i, j in enumerate(self.index_q):
+            out_ft[:, :, :, j] = xqkvw[:, :, :, i]
+        # Return to time domain
+        out = torch.fft.irfft(out_ft / self.in_channels / self.out_channels, n=xq.size(-1))
+        return (out, None)   
+        
+class MultiWaveletTransform(nn.Module):
+    """
+    1D multiwavelet block.
+    """
+
+    def __init__(self, ich=1, k=8, alpha=16, c=128,
+                 nCZ=1, L=0, base='legendre', attention_dropout=0.1):
+        super(MultiWaveletTransform, self).__init__()
+        print('base', base)
+        self.k = k
+        self.c = c
+        self.L = L
+        self.nCZ = nCZ
+        self.Lk0 = nn.Linear(ich, c * k)
+        self.Lk1 = nn.Linear(c * k, ich)
+        self.ich = ich
+        self.MWT_CZ = nn.ModuleList(MWT_CZ1d(k, alpha, L, c, base) for i in range(nCZ))
+
+    def forward(self, queries, keys, values, attn_mask):
+        B, L, H, E = queries.shape
+        _, S, _, D = values.shape
+        if L > S:
+            zeros = torch.zeros_like(queries[:, :(L - S), :]).float()
+            values = torch.cat([values, zeros], dim=1)
+            keys = torch.cat([keys, zeros], dim=1)
+        else:
+            values = values[:, :L, :, :]
+            keys = keys[:, :L, :, :]
+        values = values.view(B, L, -1)
+
+        V = self.Lk0(values).view(B, L, self.c, -1)
+        for i in range(self.nCZ):
+            V = self.MWT_CZ[i](V)
+            if i < self.nCZ - 1:
+                V = F.relu(V)
+
+        V = self.Lk1(V.view(B, L, -1))
+        V = V.view(B, L, -1, D)
+        return (V.contiguous(), None)
+
+
+class MultiWaveletCross(nn.Module):
+    """
+    1D Multiwavelet Cross Attention layer.
+    """
+
+    def __init__(self, in_channels, out_channels, seq_len_q, seq_len_kv, modes, c=64,
+                 k=8, ich=512,
+                 L=0,
+                 base='legendre',
+                 mode_select_method='random',
+                 initializer=None, activation='tanh',
+                 **kwargs):
+        super(MultiWaveletCross, self).__init__()
+        print('base', base)
+
+        self.c = c
+        self.k = k
+        self.L = L
+        H0, H1, G0, G1, PHI0, PHI1 = get_filter(base, k)
+        H0r = H0 @ PHI0
+        G0r = G0 @ PHI0
+        H1r = H1 @ PHI1
+        G1r = G1 @ PHI1
+        
+        H0r[np.abs(H0r) < 1e-8] = 0
+        H1r[np.abs(H1r) < 1e-8] = 0
+        G0r[np.abs(G0r) < 1e-8] = 0
+        G1r[np.abs(G1r) < 1e-8] = 0
+        self.max_item = 3
+
+        self.attn1 = FourierCrossAttentionW(in_channels=in_channels, out_channels=out_channels, seq_len_q=seq_len_q,
+                                            seq_len_kv=seq_len_kv, modes=modes, activation=activation,
+                                            mode_select_method=mode_select_method)
+        self.attn2 = FourierCrossAttentionW(in_channels=in_channels, out_channels=out_channels, seq_len_q=seq_len_q,
+                                            seq_len_kv=seq_len_kv, modes=modes, activation=activation,
+                                            mode_select_method=mode_select_method)
+        self.attn3 = FourierCrossAttentionW(in_channels=in_channels, out_channels=out_channels, seq_len_q=seq_len_q,
+                                            seq_len_kv=seq_len_kv, modes=modes, activation=activation,
+                                            mode_select_method=mode_select_method)
+        self.attn4 = FourierCrossAttentionW(in_channels=in_channels, out_channels=out_channels, seq_len_q=seq_len_q,
+                                            seq_len_kv=seq_len_kv, modes=modes, activation=activation,
+                                            mode_select_method=mode_select_method)
+        self.T0 = nn.Linear(k, k)
+        self.register_buffer('ec_s', torch.Tensor(
+            np.concatenate((H0.T, H1.T), axis=0)))
+        self.register_buffer('ec_d', torch.Tensor(
+            np.concatenate((G0.T, G1.T), axis=0)))
+
+        self.register_buffer('rc_e', torch.Tensor(
+            np.concatenate((H0r, G0r), axis=0)))
+        self.register_buffer('rc_o', torch.Tensor(
+            np.concatenate((H1r, G1r), axis=0)))
+
+        self.Lk = nn.Linear(ich, c * k)
+        self.Lq = nn.Linear(ich, c * k)
+        self.Lv = nn.Linear(ich, c * k)
+        self.out = nn.Linear(c * k, ich)
+        self.modes1 = modes
+
+    def forward(self, q, k, v, mask=None):
+        B, N, H, E = q.shape  # (B, N, H, E) torch.Size([3, 768, 8, 2])
+        _, S, _, _ = k.shape  # (B, S, H, E) torch.Size([3, 96, 8, 2])
+
+        q = q.view(q.shape[0], q.shape[1], -1)
+        k = k.view(k.shape[0], k.shape[1], -1)
+        v = v.view(v.shape[0], v.shape[1], -1)
+        q = self.Lq(q)
+        q = q.view(q.shape[0], q.shape[1], self.c, self.k)
+        k = self.Lk(k)
+        k = k.view(k.shape[0], k.shape[1], self.c, self.k)
+        v = self.Lv(v)
+        v = v.view(v.shape[0], v.shape[1], self.c, self.k)
+
+        if N > S:
+            zeros = torch.zeros_like(q[:, :(N - S), :]).float()
+            v = torch.cat([v, zeros], dim=1)
+            k = torch.cat([k, zeros], dim=1)
+        else:
+            v = v[:, :N, :, :]
+            k = k[:, :N, :, :]
+
+        ns = math.floor(np.log2(N))
+        nl = pow(2, math.ceil(np.log2(N)))
+        extra_q = q[:, 0:nl - N, :, :]
+        extra_k = k[:, 0:nl - N, :, :]
+        extra_v = v[:, 0:nl - N, :, :]
+        q = torch.cat([q, extra_q], 1)
+        k = torch.cat([k, extra_k], 1)
+        v = torch.cat([v, extra_v], 1)
+
+        Ud_q = torch.jit.annotate(List[Tuple[Tensor]], [])
+        Ud_k = torch.jit.annotate(List[Tuple[Tensor]], [])
+        Ud_v = torch.jit.annotate(List[Tuple[Tensor]], [])
+
+        Us_q = torch.jit.annotate(List[Tensor], [])
+        Us_k = torch.jit.annotate(List[Tensor], [])
+        Us_v = torch.jit.annotate(List[Tensor], [])
+
+        Ud = torch.jit.annotate(List[Tensor], [])
+        Us = torch.jit.annotate(List[Tensor], [])
+
+        # decompose
+        for i in range(ns - self.L):
+            # print('q shape',q.shape)
+            d, q = self.wavelet_transform(q)
+            Ud_q += [tuple([d, q])]
+            Us_q += [d]
+        for i in range(ns - self.L):
+            d, k = self.wavelet_transform(k)
+            Ud_k += [tuple([d, k])]
+            Us_k += [d]
+        for i in range(ns - self.L):
+            d, v = self.wavelet_transform(v)
+            Ud_v += [tuple([d, v])]
+            Us_v += [d]
+        for i in range(ns - self.L):
+            dk, sk = Ud_k[i], Us_k[i]
+            dq, sq = Ud_q[i], Us_q[i]
+            dv, sv = Ud_v[i], Us_v[i]
+            Ud += [self.attn1(dq[0], dk[0], dv[0], mask)[0] + self.attn2(dq[1], dk[1], dv[1], mask)[0]]
+            Us += [self.attn3(sq, sk, sv, mask)[0]]
+        v = self.attn4(q, k, v, mask)[0]
+
+        # reconstruct
+        for i in range(ns - 1 - self.L, -1, -1):
+            v = v + Us[i]
+            v = torch.cat((v, Ud[i]), -1)
+            v = self.evenOdd(v)
+        v = self.out(v[:, :N, :, :].contiguous().view(B, N, -1))
+        return (v.contiguous(), None)
+
+    def wavelet_transform(self, x):
+        xa = torch.cat([x[:, ::2, :, :],
+                        x[:, 1::2, :, :],
+                        ], -1)
+        d = torch.matmul(xa, self.ec_d)
+        s = torch.matmul(xa, self.ec_s)
+        return d, s
+
+    def evenOdd(self, x):
+        B, N, c, ich = x.shape  # (B, N, c, k)
+        assert ich == 2 * self.k
+        x_e = torch.matmul(x, self.rc_e)
+        x_o = torch.matmul(x, self.rc_o)
+
+        x = torch.zeros(B, N * 2, c, self.k,
+                        device=x.device)
+        x[..., ::2, :, :] = x_e
+        x[..., 1::2, :, :] = x_o
+        return x
+
+
+class FourierCrossAttentionW(nn.Module):
+    def __init__(self, in_channels, out_channels, seq_len_q, seq_len_kv, modes=16, activation='tanh',
+                 mode_select_method='random'):
+        super(FourierCrossAttentionW, self).__init__()
+        print('corss fourier correlation used!')
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.modes1 = modes
+        self.activation = activation
+
+    def forward(self, q, k, v, mask):
+        B, L, E, H = q.shape
+
+        xq = q.permute(0, 3, 2, 1)  # size = [B, H, E, L] torch.Size([3, 8, 64, 512])
+        xk = k.permute(0, 3, 2, 1)
+        xv = v.permute(0, 3, 2, 1)
+        self.index_q = list(range(0, min(int(L // 2), self.modes1)))
+        self.index_k_v = list(range(0, min(int(xv.shape[3] // 2), self.modes1)))
+
+        # Compute Fourier coefficients
+        xq_ft_ = torch.zeros(B, H, E, len(self.index_q), device=xq.device, dtype=torch.cfloat)
+        xq_ft = torch.fft.rfft(xq, dim=-1)
+        for i, j in enumerate(self.index_q):
+            xq_ft_[:, :, :, i] = xq_ft[:, :, :, j]
+
+        xk_ft_ = torch.zeros(B, H, E, len(self.index_k_v), device=xq.device, dtype=torch.cfloat)
+        xk_ft = torch.fft.rfft(xk, dim=-1)
+        for i, j in enumerate(self.index_k_v):
+            xk_ft_[:, :, :, i] = xk_ft[:, :, :, j]
+        xqk_ft = (torch.einsum("bhex,bhey->bhxy", xq_ft_, xk_ft_))
+        if self.activation == 'tanh':
+            xqk_ft = xqk_ft.tanh()
+        elif self.activation == 'softmax':
+            xqk_ft = torch.softmax(abs(xqk_ft), dim=-1)
+            xqk_ft = torch.complex(xqk_ft, torch.zeros_like(xqk_ft))
+        else:
+            raise Exception('{} actiation function is not implemented'.format(self.activation))
+        xqkv_ft = torch.einsum("bhxy,bhey->bhex", xqk_ft, xk_ft_)
+
+        xqkvw = xqkv_ft
+        out_ft = torch.zeros(B, H, E, L // 2 + 1, device=xq.device, dtype=torch.cfloat)
+        for i, j in enumerate(self.index_q):
+            out_ft[:, :, :, j] = xqkvw[:, :, :, i]
+
+        out = torch.fft.irfft(out_ft / self.in_channels / self.out_channels, n=xq.size(-1)).permute(0, 3, 2, 1)
+        # size = [B, L, H, E]
+        return (out, None)
+
+
+class sparseKernelFT1d(nn.Module):
+    def __init__(self,
+                 k, alpha, c=1,
+                 nl=1,
+                 initializer=None,
+                 **kwargs):
+        super(sparseKernelFT1d, self).__init__()
+
+        self.modes1 = alpha
+        self.scale = (1 / (c * k * c * k))
+        self.weights1 = nn.Parameter(self.scale * torch.rand(c * k, c * k, self.modes1, dtype=torch.cfloat))
+        self.weights1.requires_grad = True
+        self.k = k
+
+    def compl_mul1d(self, x, weights):
+        # (batch, in_channel, x ), (in_channel, out_channel, x) -> (batch, out_channel, x)
+        return torch.einsum("bix,iox->box", x, weights)
+
+    def forward(self, x):
+        B, N, c, k = x.shape  # (B, N, c, k)
+
+        x = x.view(B, N, -1)
+        x = x.permute(0, 2, 1)
+        x_fft = torch.fft.rfft(x)
+        # Multiply relevant Fourier modes
+        l = min(self.modes1, N // 2 + 1)
+        # l = N//2+1
+        out_ft = torch.zeros(B, c * k, N // 2 + 1, device=x.device, dtype=torch.cfloat)
+        out_ft[:, :, :l] = self.compl_mul1d(x_fft[:, :, :l], self.weights1[:, :, :l])
+        x = torch.fft.irfft(out_ft, n=N)
+        x = x.permute(0, 2, 1).view(B, N, c, k)
+        return x
+
+
+# ##
+class MWT_CZ1d(nn.Module):
+    def __init__(self,
+                 k=3, alpha=64,
+                 L=0, c=1,
+                 base='legendre',
+                 initializer=None,
+                 **kwargs):
+        super(MWT_CZ1d, self).__init__()
+
+        self.k = k
+        self.L = L
+        H0, H1, G0, G1, PHI0, PHI1 = get_filter(base, k)
+        H0r = H0 @ PHI0
+        G0r = G0 @ PHI0
+        H1r = H1 @ PHI1
+        G1r = G1 @ PHI1
+
+        H0r[np.abs(H0r) < 1e-8] = 0
+        H1r[np.abs(H1r) < 1e-8] = 0
+        G0r[np.abs(G0r) < 1e-8] = 0
+        G1r[np.abs(G1r) < 1e-8] = 0
+        self.max_item = 3
+
+        self.A = sparseKernelFT1d(k, alpha, c)
+        self.B = sparseKernelFT1d(k, alpha, c)
+        self.C = sparseKernelFT1d(k, alpha, c)
+
+        self.T0 = nn.Linear(k, k)
+
+        self.register_buffer('ec_s', torch.Tensor(
+            np.concatenate((H0.T, H1.T), axis=0)))
+        self.register_buffer('ec_d', torch.Tensor(
+            np.concatenate((G0.T, G1.T), axis=0)))
+
+        self.register_buffer('rc_e', torch.Tensor(
+            np.concatenate((H0r, G0r), axis=0)))
+        self.register_buffer('rc_o', torch.Tensor(
+            np.concatenate((H1r, G1r), axis=0)))
+
+    def forward(self, x):
+        B, N, c, k = x.shape  # (B, N, k)
+        ns = math.floor(np.log2(N))
+        nl = pow(2, math.ceil(np.log2(N)))
+        extra_x = x[:, 0:nl - N, :, :]
+        x = torch.cat([x, extra_x], 1)
+        Ud = torch.jit.annotate(List[Tensor], [])
+        Us = torch.jit.annotate(List[Tensor], [])
+        #         decompose
+        for i in range(ns - self.L):
+            # print('x shape',x.shape)
+            d, x = self.wavelet_transform(x)
+            Ud += [self.A(d) + self.B(x)]
+            Us += [self.C(d)]
+        x = self.T0(x)  # coarsest scale transform
+
+        #        reconstruct
+        for i in range(ns - 1 - self.L, -1, -1):
+            x = x + Us[i]
+            x = torch.cat((x, Ud[i]), -1)
+            x = self.evenOdd(x)
+        x = x[:, :N, :, :]
+
+        return x
+
+    def wavelet_transform(self, x):
+        xa = torch.cat([x[:, ::2, :, :],
+                        x[:, 1::2, :, :],
+                        ], -1)
+        d = torch.matmul(xa, self.ec_d)
+        s = torch.matmul(xa, self.ec_s)
+        return d, s
+
+    def evenOdd(self, x):
+
+        B, N, c, ich = x.shape  # (B, N, c, k)
+        assert ich == 2 * self.k
+        x_e = torch.matmul(x, self.rc_e)
+        x_o = torch.matmul(x, self.rc_o)
+
+        x = torch.zeros(B, N * 2, c, self.k,
+                        device=x.device)
+        x[..., ::2, :, :] = x_e
+        x[..., 1::2, :, :] = x_o
+        return x
+        
+class FullAttention(nn.Module):
+    def __init__(self, mask_flag=True, factor=5, scale=None, attention_dropout=0.1, output_attention=False):
+        super(FullAttention, self).__init__()
+        self.scale = scale
+        self.mask_flag = mask_flag
+        self.output_attention = output_attention
+        self.dropout = nn.Dropout(attention_dropout)
+
+    def forward(self, queries, keys, values, attn_mask):
+        B, L, H, E = queries.shape
+        _, S, _, D = values.shape
+        scale = self.scale or 1. / sqrt(E)
+
+        scores = torch.einsum("blhe,bshe->bhls", queries, keys)
+
+        if self.mask_flag:
+            if attn_mask is None:
+                attn_mask = TriangularCausalMask(B, L, device=queries.device)
+
+            scores.masked_fill_(attn_mask.mask, -np.inf)
+
+        A = self.dropout(torch.softmax(scale * scores, dim=-1))
+        V = torch.einsum("bhls,bshd->blhd", A, values)
+
+        if self.output_attention:
+            return (V.contiguous(), A)
+        else:
+            return (V.contiguous(), None)
+
+
+class ProbAttention(nn.Module):
+    def __init__(self, mask_flag=True, factor=5, scale=None, attention_dropout=0.1, output_attention=False):
+        super(ProbAttention, self).__init__()
+        self.factor = factor
+        self.scale = scale
+        self.mask_flag = mask_flag
+        self.output_attention = output_attention
+        self.dropout = nn.Dropout(attention_dropout)
+
+    def _prob_QK(self, Q, K, sample_k, n_top):  # n_top: c*ln(L_q)
+        # Q [B, H, L, D]
+        B, H, L_K, E = K.shape
+        _, _, L_Q, _ = Q.shape
+
+        # calculate the sampled Q_K
+        K_expand = K.unsqueeze(-3).expand(B, H, L_Q, L_K, E)
+        index_sample = torch.randint(L_K, (L_Q, sample_k))  # real U = U_part(factor*ln(L_k))*L_q
+        K_sample = K_expand[:, :, torch.arange(L_Q).unsqueeze(1), index_sample, :]
+        Q_K_sample = torch.matmul(Q.unsqueeze(-2), K_sample.transpose(-2, -1)).squeeze()
+
+        # find the Top_k query with sparisty measurement
+        M = Q_K_sample.max(-1)[0] - torch.div(Q_K_sample.sum(-1), L_K)
+        M_top = M.topk(n_top, sorted=False)[1]
+
+        # use the reduced Q to calculate Q_K
+        Q_reduce = Q[torch.arange(B)[:, None, None],
+                   torch.arange(H)[None, :, None],
+                   M_top, :]  # factor*ln(L_q)
+        Q_K = torch.matmul(Q_reduce, K.transpose(-2, -1))  # factor*ln(L_q)*L_k
+
+        return Q_K, M_top
+
+    def _get_initial_context(self, V, L_Q):
+        B, H, L_V, D = V.shape
+        if not self.mask_flag:
+            # V_sum = V.sum(dim=-2)
+            V_sum = V.mean(dim=-2)
+            contex = V_sum.unsqueeze(-2).expand(B, H, L_Q, V_sum.shape[-1]).clone()
+        else:  # use mask
+            assert (L_Q == L_V)  # requires that L_Q == L_V, i.e. for self-attention only
+            contex = V.cumsum(dim=-2)
+        return contex
+
+    def _update_context(self, context_in, V, scores, index, L_Q, attn_mask):
+        B, H, L_V, D = V.shape
+
+        if self.mask_flag:
+            attn_mask = ProbMask(B, H, L_Q, index, scores, device=V.device)
+            scores.masked_fill_(attn_mask.mask, -np.inf)
+
+        attn = torch.softmax(scores, dim=-1)  # nn.Softmax(dim=-1)(scores)
+
+        context_in[torch.arange(B)[:, None, None],
+        torch.arange(H)[None, :, None],
+        index, :] = torch.matmul(attn, V).type_as(context_in)
+        if self.output_attention:
+            attns = (torch.ones([B, H, L_V, L_V]) / L_V).type_as(attn).to(attn.device)
+            attns[torch.arange(B)[:, None, None], torch.arange(H)[None, :, None], index, :] = attn
+            return (context_in, attns)
+        else:
+            return (context_in, None)
+
+    def forward(self, queries, keys, values, attn_mask):
+        B, L_Q, H, D = queries.shape
+        _, L_K, _, _ = keys.shape
+
+        queries = queries.transpose(2, 1)
+        keys = keys.transpose(2, 1)
+        values = values.transpose(2, 1)
+
+        U_part = self.factor * np.ceil(np.log(L_K)).astype('int').item()  # c*ln(L_k)
+        u = self.factor * np.ceil(np.log(L_Q)).astype('int').item()  # c*ln(L_q)
+
+        U_part = U_part if U_part < L_K else L_K
+        u = u if u < L_Q else L_Q
+
+        scores_top, index = self._prob_QK(queries, keys, sample_k=U_part, n_top=u)
+
+        # add scale factor
+        scale = self.scale or 1. / sqrt(D)
+        if scale is not None:
+            scores_top = scores_top * scale
+        # get the context
+        context = self._get_initial_context(values, L_Q)
+        # update the context with selected top_k queries
+        context, attn = self._update_context(context, values, scores_top, index, L_Q, attn_mask)
+
+        return context.contiguous(), attn
+
+
+class FEDformerModel(nn.Module):
+    @validated()
+    def __init__(
+        self,
+        freq: str,
+        prediction_length: int,
+        nhead: int,
+        num_encoder_layers: int, 
+        num_decoder_layers: int, 
+        dim_feedforward: int = 16, #dimension of fcn
+        version:str = 'Fourier',#Fourier, Wavelets
+        features: str = 'M',#options:[M, S, MS]; M:multivariate predict multivariate, 'S':univariate predict univariate, MS:multivariate predict univariate' 
+        modes:int = 64,
+        mode_select:str = 'random',
+        base:str = 'legendre',
+        cross_activation:str = 'tanh',
+        L: int = 3,
+        #forecasting task
+        context_length: Optional[int] = None,#seq_len : input sequence length
+        label_length: Optional[int] = 48,#start token length
+        # model argument
+        input_size: int = 1, #encoder input size
+        # dec_in: int = 7, #decoder input size
+        c_out: int = 7, #output size
+        moving_avg: Optional[List[int]] = None,
+        factor:int = 1,
+        scaling:bool = True,
+        activation: str = "gelu",
+        dropout: float = 0.05,
+        embed: str = "timeF",#options:[timeF, fixed, learned]
+        output_attention: bool = True,#whether to output attention in encoder
+        num_feat_dynamic_real: int = 0,
+        num_feat_static_cat: int = 0,
+        num_feat_static_real: int = 0,
+        cardinality: Optional[List[int]] = None,
+        embedding_dimension: Optional[List[int]] = None,
+        distr_output: DistributionOutput = StudentTOutput(),
+        lags_seq: Optional[List[int]] = None,
+        num_parallel_samples: int = 100,
+    ) -> None:
+        super().__init__()
+
+        self.input_size = input_size
+
+        self.target_shape = distr_output.event_shape
+        self.num_feat_dynamic_real = num_feat_dynamic_real
+        self.num_feat_static_cat = num_feat_static_cat
+        self.num_feat_static_real = num_feat_static_real
+        self.embedding_dimension = (
+            embedding_dimension
+            if embedding_dimension is not None or cardinality is None
+            else [min(50, (cat + 1) // 2) for cat in cardinality]
+        )
+        self.lags_seq = lags_seq or get_lags_for_frequency(freq_str=freq)
+        self.num_parallel_samples = num_parallel_samples
+        self.history_length = context_length + max(self.lags_seq)
+        self.embedder = FeatureEmbedder(
+            cardinalities=cardinality,
+            embedding_dims=self.embedding_dimension,
+        )
+        if scaling:
+            self.scaler = MeanScaler(dim=1, keepdim=True)
+        else:
+            self.scaler = NOPScaler(dim=1, keepdim=True)
+
+        # total feature size
+        d_model = self.input_size * len(self.lags_seq) + self._number_of_features
+
+        self.context_length = context_length
+        self.prediction_length = prediction_length
+        self.distr_output = distr_output
+        self.param_proj = distr_output.get_args_proj(d_model)
+        self.moving_avg = moving_avg
+        self.version = version
+        self.mode_select = mode_select
+        self.modes = modes
+        self.label_length = label_length
+        self.output_attention = output_attention
+        self.dim_feedforward = dim_feedforward
+        # Decomp
+        kernel_size = self.moving_avg
+        if isinstance(kernel_size, list):
+            self.decomp = series_decomp_multi(kernel_size)
+        else:
+            self.decomp = series_decomp(kernel_size)
+
+        # Embedding
+        # The series-wise connection inherently contains the sequential information.
+        # Thus, we can discard the position embedding of transformers.
+        # self.enc_embedding = DataEmbedding_wo_pos(enc_in, d_model, embed, freq,
+        #                                           dropout)
+        # self.dec_embedding = DataEmbedding_wo_pos(dec_in, d_model, embed, freq,
+        #                                           dropout)
+        
+        if self.version == 'Wavelets':
+            encoder_self_att = MultiWaveletTransform(ich=d_model, L=L, base=base)
+            decoder_self_att = MultiWaveletTransform(ich=d_model, L=L, base=base)
+            decoder_cross_att = MultiWaveletCross(in_channels=d_model,
+                                                  out_channels=d_model,
+                                                  seq_len_q=self.context_length // 2 + self.prediction_length,
+                                                  seq_len_kv=self.context_length,
+                                                  modes=modes,
+                                                  ich=d_model,
+                                                  base=base,
+                                                  activation=cross_activation)
+        else:
+            encoder_self_att = FourierBlock(in_channels=d_model,
+                                            out_channels=d_model,
+                                            seq_len=self.context_length,
+                                            modes=modes,
+                                            mode_select_method=mode_select)
+            decoder_self_att = FourierBlock(in_channels=d_model,
+                                            out_channels=d_model,
+                                            seq_len=self.context_length//2+self.prediction_length,
+                                            modes=modes,
+                                            mode_select_method=mode_select)
+            decoder_cross_att = FourierCrossAttention(in_channels=d_model,
+                                                      out_channels=d_model,
+                                                      seq_len_q=self.context_length//2+self.prediction_length,
+                                                      seq_len_kv=self.context_length,
+                                                      modes=modes,
+                                                      mode_select_method=mode_select)
+        # Encoder
+        print('dim_feedforward',self.dim_feedforward)
+        enc_modes = int(min(modes, context_length//2))
+        dec_modes = int(min(modes, (context_length//2+self.prediction_length)//2))
+        print('enc_modes: {}, dec_modes: {}'.format(enc_modes, dec_modes))
+        print('encoder_self_att',encoder_self_att)
+        self.encoder = Encoder(
+            [
+                EncoderLayer(
+                    AutoCorrelationLayer(
+                        encoder_self_att,
+                        d_model, n_heads = nhead),
+                        
+                    d_model,
+                    d_ff = self.dim_feedforward,
+                    moving_avg=moving_avg,
+                    dropout=dropout,
+                    activation=activation,
+                ) for l in range(num_encoder_layers)
+            ],
+            norm_layer=my_Layernorm(d_model)
+        )
+        # Decoder
+        
+        self.decoder = Decoder(
+            [
+                DecoderLayer(
+                    AutoCorrelationLayer(
+                        decoder_self_att,
+                        d_model, n_heads = nhead),
+                    AutoCorrelationLayer(
+                        decoder_cross_att,
+                        d_model, n_heads = nhead),
+
+                    d_model,
+                    c_out = c_out,
+                    d_ff = self.dim_feedforward,
+                    moving_avg=self.moving_avg,
+                    dropout=dropout,
+                    activation=activation,
+                )
+                for l in range(num_decoder_layers)
+            ],
+            norm_layer=my_Layernorm(d_model),
+            projection=nn.Linear(d_model, c_out, bias=True)
+        )
+        
+
+        
+    @property
+    def _number_of_features(self) -> int:
+        return (
+            sum(self.embedding_dimension)
+            + self.num_feat_dynamic_real
+            + self.num_feat_static_real
+            + 1  # the log(scale)
+        )
+
+    @property
+    def _past_length(self) -> int:
+        return self.context_length + max(self.lags_seq)
+
+    def get_lagged_subsequences(
+        self, sequence: torch.Tensor, subsequences_length: int, shift: int = 0
+    ) -> torch.Tensor:
+        """
+        Returns lagged subsequences of a given sequence.
+        Parameters
+        ----------
+        sequence : Tensor
+            the sequence from which lagged subsequences should be extracted.
+            Shape: (N, T, C).
+        subsequences_length : int
+            length of the subsequences to be extracted.
+        shift: int
+            shift the lags by this amount back.
+        Returns
+        --------
+        lagged : Tensor
+            a tensor of shape (N, S, C, I), where S = subsequences_length and
+            I = len(indices), containing lagged subsequences. Specifically,
+            lagged[i, j, :, k] = sequence[i, -indices[k]-S+j, :].
+        """
+        sequence_length = sequence.shape[1]
+        indices = [lag - shift for lag in self.lags_seq]
+
+        assert max(indices) + subsequences_length <= sequence_length, (
+            f"lags cannot go further than history length, found lag {max(indices)} "
+            f"while history length is only {sequence_length}"
+        )
+
+        lagged_values = []
+        for lag_index in indices:
+            begin_index = -lag_index - subsequences_length
+            end_index = -lag_index if lag_index > 0 else None
+            lagged_values.append(sequence[:, begin_index:end_index, ...])
+        return torch.stack(lagged_values, dim=-1)
+
+    def _check_shapes(
+        self,
+        prior_input: torch.Tensor,
+        inputs: torch.Tensor,
+        features: Optional[torch.Tensor],
+    ) -> None:
+        assert len(prior_input.shape) == len(inputs.shape)
+        assert (
+            len(prior_input.shape) == 2 and self.input_size == 1
+        ) or prior_input.shape[2] == self.input_size
+        assert (len(inputs.shape) == 2 and self.input_size == 1) or inputs.shape[
+            -1
+        ] == self.input_size
+        assert (
+            features is None or features.shape[2] == self._number_of_features
+        ), f"{features.shape[2]}, expected {self._number_of_features}"
+
+    def create_network_inputs(
+        self,
+        feat_static_cat: torch.Tensor,
+        feat_static_real: torch.Tensor,
+        past_time_feat: torch.Tensor,
+        past_target: torch.Tensor,
+        past_observed_values: torch.Tensor,
+        future_time_feat: Optional[torch.Tensor] = None,
+        future_target: Optional[torch.Tensor] = None,
+    ):
+        # time feature
+        time_feat = (
+            torch.cat(
+                (
+                    past_time_feat[:, self._past_length - self.context_length :, ...],
+                    future_time_feat,
+                ),
+                dim=1,
+            )
+            if future_target is not None
+            else past_time_feat[:, self._past_length - self.context_length :, ...]
+        )
+
+        # target
+        context = past_target[:, -self.context_length :]
+        observed_context = past_observed_values[:, -self.context_length :]
+        _, scale = self.scaler(context, observed_context)
+
+        inputs = (
+            torch.cat((past_target, future_target), dim=1) / scale
+            if future_target is not None
+            else past_target / scale
+        )
+
+        inputs_length = (
+            self._past_length + self.prediction_length
+            if future_target is not None
+            else self._past_length
+        )
+        assert inputs.shape[1] == inputs_length
+
+        subsequences_length = (
+            self.context_length + self.prediction_length
+            if future_target is not None
+            else self.context_length
+        )
+
+        # embeddings
+        embedded_cat = self.embedder(feat_static_cat)
+        static_feat = torch.cat(
+            (embedded_cat, feat_static_real, scale.log()),
+            dim=1,
+        )
+        expanded_static_feat = static_feat.unsqueeze(1).expand(
+            -1, time_feat.shape[1], -1
+        )
+
+        features = torch.cat((expanded_static_feat, time_feat), dim=-1)
+
+        # self._check_shapes(prior_input, inputs, features)
+
+        # sequence = torch.cat((prior_input, inputs), dim=1)
+        lagged_sequence = self.get_lagged_subsequences(
+            sequence=inputs,
+            subsequences_length=subsequences_length,
+        )
+
+        lags_shape = lagged_sequence.shape
+        reshaped_lagged_sequence = lagged_sequence.reshape(
+            lags_shape[0], lags_shape[1], -1
+        )
+
+        transformer_inputs = torch.cat((reshaped_lagged_sequence, features), dim=-1)
+
+        return transformer_inputs, scale, static_feat
+
+    def output_params(self, transformer_inputs):
+        enc_input = transformer_inputs[:, : self.context_length, ...]
+        dec_input = transformer_inputs[:, self.context_length :, ...]
+
+        enc_out, _ = self.encoder(enc_input)
+        dec_output = self.decoder(dec_input, enc_out)
+
+        return self.param_proj(dec_output)
+
+    @torch.jit.ignore
+    def output_distribution(
+        self, params, scale=None, trailing_n=None
+    ) -> torch.distributions.Distribution:
+        sliced_params = params
+        if trailing_n is not None:
+            sliced_params = [p[:, -trailing_n:] for p in params]
+        return self.distr_output.distribution(sliced_params, scale=scale)
+
+    # for prediction
+    def forward(
+        self,
+        feat_static_cat: torch.Tensor,
+        feat_static_real: torch.Tensor,
+        past_time_feat: torch.Tensor,
+        past_target: torch.Tensor,
+        past_observed_values: torch.Tensor,
+        future_time_feat: torch.Tensor,
+        num_parallel_samples: Optional[int] = None,
+    ) -> torch.Tensor:
+
+        if num_parallel_samples is None:
+            num_parallel_samples = self.num_parallel_samples
+
+        encoder_inputs, scale, static_feat = self.create_network_inputs(
+            feat_static_cat,
+            feat_static_real,
+            past_time_feat,
+            past_target,
+            past_observed_values,
+        )
+
+        enc_out, _ = self.encoder(encoder_inputs)
+
+        repeated_scale = scale.repeat_interleave(
+            repeats=self.num_parallel_samples, dim=0
+        )
+
+        repeated_past_target = (
+            past_target.repeat_interleave(repeats=self.num_parallel_samples, dim=0)
+            / repeated_scale
+        )
+
+        expanded_static_feat = static_feat.unsqueeze(1).expand(
+            -1, future_time_feat.shape[1], -1
+        )
+        features = torch.cat((expanded_static_feat, future_time_feat), dim=-1)
+        repeated_features = features.repeat_interleave(
+            repeats=self.num_parallel_samples, dim=0
+        )
+
+        repeated_enc_out = enc_out.repeat_interleave(
+            repeats=self.num_parallel_samples, dim=0
+        )
+
+        future_samples = []
+
+        # greedy decoding
+        for k in range(self.prediction_length):
+            # self._check_shapes(repeated_past_target, next_sample, next_features)
+            # sequence = torch.cat((repeated_past_target, next_sample), dim=1)
+
+            lagged_sequence = self.get_lagged_subsequences(
+                sequence=repeated_past_target,
+                subsequences_length=1 + k,
+                shift=1,
+            )
+
+            lags_shape = lagged_sequence.shape
+            reshaped_lagged_sequence = lagged_sequence.reshape(
+                lags_shape[0], lags_shape[1], -1
+            )
+
+            decoder_input = torch.cat(
+                (reshaped_lagged_sequence, repeated_features[:, : k + 1]), dim=-1
+            )
+
+            output = self.decoder(decoder_input, repeated_enc_out)
+
+            params = self.param_proj(output[:, -1:])
+            distr = self.output_distribution(params, scale=repeated_scale)
+            next_sample = distr.sample()
+
+            repeated_past_target = torch.cat(
+                (repeated_past_target, next_sample / repeated_scale), dim=1
+            )
+            future_samples.append(next_sample)
+
+        concat_future_samples = torch.cat(future_samples, dim=1)
+        return concat_future_samples.reshape(
+            (-1, self.num_parallel_samples, self.prediction_length) + self.target_shape,
+        )
\ No newline at end of file