DeepAR Training gets stuck at some random epoch

PyTorch-Forecasting version: # v0.10.2 Multivariate networks (23/05/2022)
PyTorch version: pytorch 2.0.0 py3.9_cuda11.7_cudnn8.5.0_0 pytorch
Python version: python 3.9.16 h7a1cb2a_2
Operating System: Linux, Ubuntu 22.04.2 LTS
GPU: RTX4090

Expected behavior

I executed the below code in order to get optimal hyperparameter values by training DeepAR and validating after every epoch and using Optuna. I expected the training to run smoothly with 1 GPU. I only have 1 GPU.

Actual behavior

However, DeepAR Training gets stuck at some random epoch.

I could run the same code in Google colab error-free.

Code to reproduce the problem

#from ctypes import FormatError
import numpy as np

import warnings
warnings.filterwarnings("ignore")

import os,sys

# sys.path.append(os.path.abspath(os.path.join('C:\Work\WORK_PACKAGE\Demand_forecasting\github\DeepAR-pytorch\My_model\\2_freq_nbinom_LSTM')))

# sys.path.append(os.path.abspath(os.path.join('C:\Work\WORK_PACKAGE\Demand_forecasting\github\DeepAR-pytorch\My_model\\2_freq_nbinom_LSTM\\1_cluster_demand_prediction\data\weather_data')))
# sys.path.append(os.path.abspath(os.path.join('C:\Work\WORK_PACKAGE\Demand_forecasting\github\DeepAR-pytorch\My_model\2_freq_nbinom_LSTM\1_cluster_demand_prediction\data\demand_data')))

import torch
torch.use_deterministic_algorithms(True)

from pytorch_forecasting.data.encoders import TorchNormalizer
from pytorch_forecasting.metrics import SMAPE, RMSE
from torchmetrics import R2Score, SymmetricMeanAbsolutePercentageError, MeanSquaredError

import matplotlib.pyplot as plt
import pandas as pd
from pytorch_forecasting.data import TimeSeriesDataSet
from pytorch_forecasting.data.encoders import NaNLabelEncoder
from pytorch_lightning.callbacks import EarlyStopping, LearningRateMonitor
import pytorch_lightning as pl
import torch
from pytorch_forecasting.data.encoders import TorchNormalizer
import os,sys
import numpy as np
from statsmodels.tsa.stattools import adfuller
from statsmodels.tsa.seasonal import seasonal_decompose
from statsmodels.graphics.tsaplots import plot_acf
from statsmodels.graphics.tsaplots import plot_pacf
from statsmodels.tsa.stattools import acf,pacf
from scipy.signal import find_peaks
import operator
import statsmodels.api as sm
from itertools import combinations
import pickle
from pytorch_forecasting import Baseline
import random
from pytorch_forecasting import DeepAR,NegativeBinomialDistributionLoss
from itertools import product
from sklearn.metrics import mean_absolute_error, mean_squared_error
import optuna
from optuna.trial import TrialState
import plotly

random.seed(0)
torch.manual_seed(0)
np.random.seed(0)
## additional seeding to ensure reproduciblility.
pl.seed_everything(0)

Loss=NegativeBinomialDistributionLoss()

cov_lag_len= 0 

"""
Import pre-processed Data

response and target are the same thing
"""
tampines_all_clstr_train_dem_data = pd.read_csv('tampines_all_clstr_train_dem_data.csv')
tampines_all_clstr_val_dem_data = pd.read_csv('tampines_all_clstr_val_dem_data.csv')
tampines_all_clstr_test_dem_data = pd.read_csv('tampines_all_clstr_test_dem_data.csv')

tampines_all_clstr_train_dem_data = tampines_all_clstr_train_dem_data.drop(['Unnamed: 0'],axis=1)
tampines_all_clstr_val_dem_data = tampines_all_clstr_val_dem_data.drop(['Unnamed: 0'],axis=1)
tampines_all_clstr_test_dem_data = tampines_all_clstr_test_dem_data.drop(['Unnamed: 0'],axis=1)

train_data = tampines_all_clstr_train_dem_data
val_data = tampines_all_clstr_val_dem_data
test_data = tampines_all_clstr_test_dem_data

cov_lag_len = 1

#################### add date information ts ####################
#2021 oct 17 20:00:00
day=17
hour=(20+cov_lag_len)
if (20+cov_lag_len)>23:
  day = 18
  hour = hour%24

train_data["date"] = pd.Timestamp(year=2021, month=10, day=day, hour=hour ) + pd.to_timedelta(train_data.time_idx, "H")
train_data['_hour_of_day'] = train_data["date"].dt.hour.astype(str)
train_data['_day_of_week'] = train_data["date"].dt.dayofweek.astype(str)
train_data['_day_of_month'] = train_data["date"].dt.day.astype(str)
train_data['_day_of_year'] = train_data["date"].dt.dayofyear.astype(str)
train_data['_week_of_year'] = train_data["date"].dt.weekofyear.astype(str)
train_data['_month_of_year'] = train_data["date"].dt.month.astype(str)
train_data['_year'] = train_data["date"].dt.year.astype(str)
#################### add date information ts ####################

#################### add date information ts ####################
# val starts at 3/12/2021 09:00
day=3
hour=(9+cov_lag_len)
if (9+cov_lag_len)>23:
  day = 4
  hour = hour%24

val_data["date"] = pd.Timestamp(year=2021, month=12, day=day, hour=hour ) + pd.to_timedelta(val_data.time_idx, "H")
val_data['_hour_of_day'] = val_data["date"].dt.hour.astype(str)
val_data['_day_of_week'] = val_data["date"].dt.dayofweek.astype(str)
val_data['_day_of_month'] = val_data["date"].dt.day.astype(str)
val_data['_day_of_year'] = val_data["date"].dt.dayofyear.astype(str)
val_data['_week_of_year'] = val_data["date"].dt.weekofyear.astype(str)
val_data['_month_of_year'] = val_data["date"].dt.month.astype(str)
val_data['_year'] = val_data["date"].dt.year.astype(str)
#################### add date information ts ####################

#################### add date information ts ####################
# test starts at 16/12/2021 16:00
day=16
hour=(16+cov_lag_len)
if (16+cov_lag_len)>23:
  day = 17
  hour = hour%24

test_data["date"] = pd.Timestamp(year=2021, month=12, day=day, hour=hour ) + pd.to_timedelta(test_data.time_idx, "H")
test_data['_hour_of_day'] = test_data["date"].dt.hour.astype(str)
test_data['_day_of_week'] = test_data["date"].dt.dayofweek.astype(str)
test_data['_day_of_month'] = test_data["date"].dt.day.astype(str)
test_data['_day_of_year'] = test_data["date"].dt.dayofyear.astype(str)
test_data['_week_of_year'] = test_data["date"].dt.weekofyear.astype(str)
test_data['_month_of_year'] = test_data["date"].dt.month.astype(str)
test_data['_year'] = test_data["date"].dt.year.astype(str)
#################### add date information ts ####################

Target = 'target'

"""
Full Training Routine 
with bayesisan hyperparmeter search

Load data into TimeSeriesDataSet object

for fast development run
uncomment fast_dev_run = fdv_steps

"""

#early_stop_callback = EarlyStopping(monitor="val_loss", min_delta=1e-4, patience=p, verbose=False, mode="min")
lr_logger = LearningRateMonitor()

RMSE_list = [] # FIND minimum RMSE case
hyperparams_list = [] # FIND minimum RMSE case

# best_val_comb_idx=[17,21,51,52,53,54,61,62,63,82,83,84,109,110,111,143,144,145,195,218,219,220,232,233,234,236,237,238,280,338,339,340,344,345,346,386]
# best_val_train_epochs = [50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50, 50]

# best_val_comb_idx=[234]
# best_val_train_epochs = [50]

param_comb_cnt=-1
#for neu,lay,bat,lr,enc_len,pred_len,drop,cov_pair,num_ep in product(*[x for x in hparams_grid.values()]):

def objective(trial):  

  neu = trial.suggest_int(name="neu",low=500,high=700,step=10,log=False)
  lay = trial.suggest_int(name="lay",low=1,high=3,step=1,log=False)
  bat = trial.suggest_int(name="bat",low=4,high=8,step=4,log=False)
  lr = trial.suggest_float(name="lr",low=0.00001,high=0.0008,step=0.00001,log=False)
  enc_len = 24
  pred_len = 1
  drop = trial.suggest_float(name="dropout",low=0,high=0.2,step=0.2,log=False)
#   cov_pair = trial.suggest_categorical("cov_pair", cov_pairs_list)

  for num_ep in range(20,35,2):

    num_cols_list = []  

    cat_dict = {"_hour_of_day": NaNLabelEncoder(add_nan=True).fit(train_data._hour_of_day), \
    "_day_of_week": NaNLabelEncoder(add_nan=True).fit(train_data._day_of_week), "_day_of_month" : NaNLabelEncoder(add_nan=True).fit(train_data._day_of_month), "_day_of_year" : NaNLabelEncoder(add_nan=True).fit(train_data._day_of_year), \
        "_week_of_year": NaNLabelEncoder(add_nan=True).fit(train_data._week_of_year), "_month_of_year": NaNLabelEncoder(add_nan=True).fit(train_data._month_of_year) ,"_year": NaNLabelEncoder(add_nan=True).fit(train_data._year) }
    cat_list = ["_hour_of_day","_day_of_week","_day_of_month","_day_of_year","_week_of_year","_month_of_year","_year"]  

    num_cols_list.append('dem_lag_168') 
    num_cols_list.append('dem_lag_336')
    num_cols_list.append('inflow')
    # num_cols_list.append('inf_lag_168')  
    # num_cols_list.append('inf_lag_336')

    train_dataset = TimeSeriesDataSet(
        train_data,
        time_idx="time_idx",
        target=Target,
        categorical_encoders=cat_dict,
        group_ids=["group"],
        min_encoder_length=enc_len,
        max_encoder_length=enc_len,
        min_prediction_length=pred_len,
        max_prediction_length=pred_len,
        time_varying_unknown_reals=[Target],
        time_varying_known_reals=num_cols_list,
        time_varying_known_categoricals=cat_list,
        add_relative_time_idx=False,
        randomize_length=False,
        scalers={},
        target_normalizer=TorchNormalizer(method="identity",center=False,transformation=None )

    )

    val_dataset = TimeSeriesDataSet.from_dataset(train_dataset,val_data, stop_randomization=True, predict=False)
    test_dataset = TimeSeriesDataSet.from_dataset(train_dataset,test_data, stop_randomization=True)

    train_dataloader = train_dataset.to_dataloader(train=True, batch_size=bat)
    val_dataloader = val_dataset.to_dataloader(train=False, batch_size=bat)
    test_dataloader = test_dataset.to_dataloader(train=False, batch_size=bat)
    ######### Load DATA #############

    """
    Machine Learning predictions START
    1) DeepAR 

    """
    trainer = pl.Trainer(
        max_epochs=num_ep,
        gpus=-1, #-1
        auto_lr_find=False,
        gradient_clip_val=0.1,
        limit_train_batches=1.0,
        limit_val_batches=1.0,
        #fast_dev_run=fdv_steps,
        logger=True,
        #log_every_n_steps=10,
        # profiler=True,
        callbacks=[lr_logger]#, early_stop_callback],
        #enable_checkpointing=True,
        #default_root_dir="C:\Work\WORK_PACKAGE\Demand_forecasting\github\DeepAR-pytorch\My_model\2_freq_nbinom_LSTM\1_cluster_demand_prediction\logs"
    )

    #print(f"training routing:\n \n {trainer}")
    deepar = DeepAR.from_dataset(
        train_dataset,
        learning_rate=lr,
        hidden_size=neu,
        rnn_layers=lay,
        dropout=drop,
        loss=Loss,
        log_interval=20,
        log_val_interval=6,
        log_gradient_flow=False,
        # reduce_on_plateau_patience=3,
    )

    #print(f"Number of parameters in network: {deepar.size()/1e3:.1f}k")
    # print(f"Model :\n \n {deepar}")
    torch.set_num_threads(10)
    trainer.fit(
        deepar,
        train_dataloaders=train_dataloader,
        val_dataloaders=val_dataloader,
    )

    val_rmse = trainer.callback_metrics['val_RMSE'].item()
    trial.report(val_rmse, num_ep)

    # Handle pruning based on the intermediate value.
    if trial.should_prune():
        raise optuna.exceptions.TrialPruned()

  return val_rmse

########## optuna results #####################
if __name__ == "__main__":

  study = optuna.create_study(direction="minimize")
  study.optimize(objective, timeout=6000)

  pruned_trials = study.get_trials(deepcopy=False, states=[TrialState.PRUNED])
  complete_trials = study.get_trials(deepcopy=False, states=[TrialState.COMPLETE])

  print("Study statistics: ")
  print("  Number of finished trials: ", len(study.trials))
  print("  Number of pruned trials: ", len(pruned_trials))
  print("  Number of complete trials: ", len(complete_trials))

  print("Best trial:")
  trial = study.best_trial

  print("  Value: ", trial.value)

  print("  Params: ")
  for key, value in trial.params.items():
      print("    {}: {}".format(key, value))

  fig = optuna.visualization.plot_parallel_coordinate(study)
  fig.show()

  fig = optuna.visualization.plot_optimization_history(study)
  fig.show()

  fig = optuna.visualization.plot_slice(study)
  fig.show()

  fig = optuna.visualization.plot_param_importances(study)
  fig.show()
########## optuna results #####################

I was training the code in a jupyter notebook (xxx.ipynb). Should train using python scripts (xxx.py) I also updated all my packages to latest versions as below. THis solved the problem.

name: test-env
channels:
  - anaconda
  - pytorch
  - nvidia
  - conda-forge
  - defaults
dependencies:
  - _libgcc_mutex=0.1=conda_forge
  - _openmp_mutex=4.5=2_gnu
  - alembic=1.10.2=pyhd8ed1ab_0
  - asttokens=2.2.1=pyhd8ed1ab_0
  - backcall=0.2.0=pyh9f0ad1d_0
  - backports=1.0=pyhd8ed1ab_3
  - backports.functools_lru_cache=1.6.4=pyhd8ed1ab_0
  - blas=1.0=mkl
  - bottleneck=1.3.7=py310h0a54255_0
  - brotli=1.0.9=h166bdaf_8
  - brotli-bin=1.0.9=h166bdaf_8
  - brotlipy=0.7.0=py310h5764c6d_1005
  - bzip2=1.0.8=h7f98852_4
  - ca-certificates=2023.01.10=h06a4308_0
  - certifi=2022.12.7=py310h06a4308_0
  - cffi=1.15.1=py310h255011f_3
  - charset-normalizer=2.1.1=pyhd8ed1ab_0
  - cmaes=0.9.1=pyhd8ed1ab_0
  - colorama=0.4.6=pyhd8ed1ab_0
  - colorlog=6.7.0=py310hff52083_1
  - comm=0.1.3=pyhd8ed1ab_0
  - contourpy=1.0.7=py310hdf3cbec_0
  - cryptography=38.0.4=py310h597c629_0
  - cuda-cudart=11.7.99=0
  - cuda-cupti=11.7.101=0
  - cuda-libraries=11.7.1=0
  - cuda-nvrtc=11.7.99=0
  - cuda-nvtx=11.7.91=0
  - cuda-runtime=11.7.1=0
  - cycler=0.11.0=pyhd8ed1ab_0
  - dbus=1.13.18=hb2f20db_0
  - debugpy=1.6.6=py310heca2aa9_0
  - decorator=5.1.1=pyhd8ed1ab_0
  - executing=1.2.0=pyhd8ed1ab_0
  - expat=2.5.0=h27087fc_0
  - ffmpeg=4.3=hf484d3e_0
  - fftw=3.3.10=nompi_hf0379b8_106
  - filelock=3.10.7=pyhd8ed1ab_0
  - fontconfig=2.14.1=h52c9d5c_1
  - fonttools=4.39.3=py310h1fa729e_0
  - freetype=2.12.1=hca18f0e_1
  - fsspec=2023.3.0=pyhd8ed1ab_1
  - giflib=5.2.1=h0b41bf4_3
  - glib=2.69.1=he621ea3_2
  - gmp=6.2.1=h58526e2_0
  - gnutls=3.6.13=h85f3911_1
  - greenlet=2.0.2=py310heca2aa9_0
  - gst-plugins-base=1.14.1=h6a678d5_1
  - gstreamer=1.14.1=h5eee18b_1
  - icu=58.2=hf484d3e_1000
  - idna=3.4=pyhd8ed1ab_0
  - importlib-metadata=6.1.0=pyha770c72_0
  - importlib_metadata=6.1.0=hd8ed1ab_0
  - importlib_resources=5.12.0=pyhd8ed1ab_0
  - intel-openmp=2021.4.0=h06a4308_3561
  - ipykernel=6.22.0=pyh210e3f2_0
  - ipython=8.11.0=pyh41d4057_0
  - jedi=0.18.2=pyhd8ed1ab_0
  - jinja2=3.1.2=pyhd8ed1ab_1
  - joblib=1.2.0=pyhd8ed1ab_0
  - jpeg=9e=h0b41bf4_3
  - jupyter_client=8.1.0=pyhd8ed1ab_0
  - jupyter_core=5.3.0=py310hff52083_0
  - keyutils=1.6.1=h166bdaf_0
  - kiwisolver=1.4.4=py310hbf28c38_1
  - krb5=1.19.3=h3790be6_0
  - lame=3.100=h166bdaf_1003
  - lcms2=2.15=hfd0df8a_0
  - ld_impl_linux-64=2.40=h41732ed_0
  - lerc=4.0.0=h27087fc_0
  - libbrotlicommon=1.0.9=h166bdaf_8
  - libbrotlidec=1.0.9=h166bdaf_8
  - libbrotlienc=1.0.9=h166bdaf_8
  - libclang=10.0.1=default_hb85057a_2
  - libcublas=11.10.3.66=0
  - libcufft=10.7.2.124=h4fbf590_0
  - libcufile=1.6.0.25=0
  - libcurand=10.3.2.56=0
  - libcusolver=11.4.0.1=0
  - libcusparse=11.7.4.91=0
  - libdeflate=1.17=h0b41bf4_0
  - libedit=3.1.20191231=he28a2e2_2
  - libevent=2.1.12=h8f2d780_0
  - libffi=3.4.2=h7f98852_5
  - libgcc-ng=12.2.0=h65d4601_19
  - libgfortran-ng=12.2.0=h69a702a_19
  - libgfortran5=12.2.0=h337968e_19
  - libgomp=12.2.0=h65d4601_19
  - libiconv=1.17=h166bdaf_0
  - libllvm10=10.0.1=he513fc3_3
  - libnpp=11.7.4.75=0
  - libnvjpeg=11.8.0.2=0
  - libpng=1.6.39=h753d276_0
  - libpq=12.9=h16c4e8d_3
  - libsodium=1.0.18=h36c2ea0_1
  - libsqlite=3.40.0=h753d276_0
  - libstdcxx-ng=12.2.0=h46fd767_19
  - libtiff=4.5.0=h6adf6a1_2
  - libuuid=1.41.5=h5eee18b_0
  - libwebp=1.2.4=h1daa5a0_1
  - libwebp-base=1.2.4=h5eee18b_1
  - libxcb=1.15=h7f8727e_0
  - libxkbcommon=1.0.1=hfa300c1_0
  - libxml2=2.9.14=h74e7548_0
  - libxslt=1.1.35=h4e12654_0
  - libzlib=1.2.13=h166bdaf_4
  - lightning-utilities=0.8.0=pyhd8ed1ab_0
  - llvm-openmp=16.0.0=h417c0b6_0
  - mako=1.2.4=pyhd8ed1ab_0
  - markupsafe=2.1.2=py310h1fa729e_0
  - matplotlib=3.5.3=py310h06a4308_0
  - matplotlib-base=3.5.3=py310hf590b9c_0
  - matplotlib-inline=0.1.6=pyhd8ed1ab_0
  - mkl=2021.4.0=h06a4308_640
  - mkl-service=2.4.0=py310ha2c4b55_0
  - mkl_fft=1.3.1=py310hd6ae3a3_0
  - mkl_random=1.2.2=py310h00e6091_0
  - mpmath=1.3.0=pyhd8ed1ab_0
  - munkres=1.1.4=pyh9f0ad1d_0
  - ncurses=6.3=h27087fc_1
  - nest-asyncio=1.5.6=pyhd8ed1ab_0
  - nettle=3.6=he412f7d_0
  - networkx=3.0=pyhd8ed1ab_0
  - nspr=4.35=h27087fc_0
  - nss=3.89=he45b914_0
  - numexpr=2.8.4=py310h8879344_0
  - numpy=1.23.5=py310hd5efca6_0
  - numpy-base=1.23.5=py310h8e6c178_0
  - openh264=2.1.1=h4ff587b_0
  - openssl=1.1.1t=h7f8727e_0
  - optuna=3.1.0=pyhd8ed1ab_0
  - packaging=23.0=pyhd8ed1ab_0
  - pandas=1.5.3=py310h1128e8f_0
  - parso=0.8.3=pyhd8ed1ab_0
  - patsy=0.5.3=pyhd8ed1ab_0
  - pcre=8.45=h9c3ff4c_0
  - pexpect=4.8.0=pyh1a96a4e_2
  - pickleshare=0.7.5=py_1003
  - pillow=9.4.0=py310h6a678d5_0
  - pip=23.0.1=pyhd8ed1ab_0
  - platformdirs=3.2.0=pyhd8ed1ab_0
  - plotly=5.9.0=py310h06a4308_0
  - ply=3.11=py_1
  - pooch=1.7.0=pyha770c72_3
  - prompt-toolkit=3.0.38=pyha770c72_0
  - prompt_toolkit=3.0.38=hd8ed1ab_0
  - psutil=5.9.4=py310h5764c6d_0
  - ptyprocess=0.7.0=pyhd3deb0d_0
  - pure_eval=0.2.2=pyhd8ed1ab_0
  - pycparser=2.21=pyhd8ed1ab_0
  - pygments=2.14.0=pyhd8ed1ab_0
  - pyopenssl=23.1.1=pyhd8ed1ab_0
  - pyparsing=3.0.9=pyhd8ed1ab_0
  - pyqt=5.15.7=py310h6a678d5_1
  - pysocks=1.7.1=pyha2e5f31_6
  - python=3.10.9=h7a1cb2a_0
  - python-dateutil=2.8.2=pyhd8ed1ab_0
  - python_abi=3.10=2_cp310
  - pytorch=2.0.0=py3.10_cuda11.7_cudnn8.5.0_0
  - pytorch-cuda=11.7=h778d358_3
  - pytorch-lightning=1.9.0=pyhd8ed1ab_1
  - pytorch-mutex=1.0=cuda
  - pytz=2023.3=pyhd8ed1ab_0
  - pyyaml=6.0=py310h5764c6d_5
  - pyzmq=25.0.2=py310h059b190_0
  - qt-main=5.15.2=h327a75a_7
  - qt-webengine=5.15.9=hd2b0992_4
  - qtwebkit=5.212=h4eab89a_4
  - readline=8.2=h8228510_1
  - requests=2.28.2=pyhd8ed1ab_0
  - scikit-learn=1.2.0=py310h6a678d5_0
  - scipy=1.10.0=py310hd5efca6_0
  - setuptools=67.6.1=pyhd8ed1ab_0
  - sip=6.6.2=py310hd8f1fbe_0
  - six=1.16.0=pyh6c4a22f_0
  - sqlalchemy=2.0.7=py310h1fa729e_0
  - sqlite=3.40.1=h5082296_0
  - stack_data=0.6.2=pyhd8ed1ab_0
  - statsmodels=0.13.5=py310ha9d4c09_1
  - sympy=1.11.1=pyh04b8f61_3
  - tenacity=8.2.2=pyhd8ed1ab_0
  - threadpoolctl=3.1.0=pyh8a188c0_0
  - tk=8.6.12=h27826a3_0
  - toml=0.10.2=pyhd8ed1ab_0
  - torchaudio=2.0.0=py310_cu117
  - torchmetrics=0.11.4=pyhd8ed1ab_0
  - torchtriton=2.0.0=py310
  - torchvision=0.15.0=py310_cu117
  - tornado=6.2=py310h5764c6d_1
  - tqdm=4.65.0=pyhd8ed1ab_1
  - traitlets=5.9.0=pyhd8ed1ab_0
  - typing-extensions=4.5.0=hd8ed1ab_0
  - typing_extensions=4.5.0=pyha770c72_0
  - tzdata=2023c=h71feb2d_0
  - unicodedata2=15.0.0=py310h5764c6d_0
  - urllib3=1.26.15=pyhd8ed1ab_0
  - wcwidth=0.2.6=pyhd8ed1ab_0
  - wheel=0.40.0=pyhd8ed1ab_0
  - xz=5.2.10=h5eee18b_1
  - yaml=0.2.5=h7f98852_2
  - zeromq=4.3.4=h9c3ff4c_1
  - zipp=3.15.0=pyhd8ed1ab_0
  - zlib=1.2.13=h166bdaf_4
  - zstd=1.5.2=h3eb15da_6
  - pip:
      - absl-py==1.4.0
      - cachetools==5.3.0
      - google-auth==2.17.0
      - google-auth-oauthlib==0.4.6
      - grpcio==1.53.0
      - markdown==3.4.3
      - oauthlib==3.2.2
      - protobuf==4.22.1
      - pyasn1==0.4.8
      - pyasn1-modules==0.2.8
      - pyqt5-sip==12.11.0
      - requests-oauthlib==1.3.1
      - rsa==4.9
      - tensorboard==2.12.0
      - tensorboard-data-server==0.7.0
      - tensorboard-plugin-wit==1.8.1
      - werkzeug==2.2.3
prefix: /home/optimusprime/miniconda3/envs/test-env

sktime / pytorch-forecasting