This page was generated from
docs/tutorial/python/imputation_2modalities.ipynb.
Interactive online version:
.
.
Imputation for two modalities#
[1]:
import os
import sys
os.environ["OMP_NUM_THREADS"] = "11"
os.environ["OPENBLAS_NUM_THREADS"] = "8" # export OPENBLAS_NUM_THREADS=4
os.environ["MKL_NUM_THREADS"] = "11" # export MKL_NUM_THREADS=6
os.environ["VECLIB_MAXIMUM_THREADS"] = "8" # export VECLIB_MAXIMUM_THREADS=4
os.environ["NUMEXPR_NUM_THREADS"] = "11" # export NUMEXPR_NUM_THREADS=6
os.environ["NUMBA_CACHE_DIR"]='/tmp/numba_cache'
import numpy as np
import pandas as pd
import scipy as sp
import scipy.sparse
import h5py
import tensorflow as tf
import matplotlib.pyplot as plt
import scanpy as sc
physical_devices = tf.config.list_physical_devices('GPU')
try:
tf.config.experimental.set_memory_growth(physical_devices[0], True)
except:
# Invalid device or cannot modify virtual devices once initialized.
pass
2025-06-07 03:44:20.607049: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2025-06-07 03:44:20.627913: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2025-06-07 03:44:20.627927: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2025-06-07 03:44:20.628482: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2025-06-07 03:44:20.631565: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: SSE4.1 SSE4.2 AVX AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
2025-06-07 03:44:34.386157: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:34.420339: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:34.420462: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
Load data#
[2]:
with h5py.File('dogma_cite_asap.h5', 'r') as f:
print(f.keys())
ADT_names = np.array(f['ADT_names'], dtype='S32').astype(str)
gene_names = np.array(f['gene_names'], dtype='S32').astype(str)
# Count matrices of three modalities: X for genes, Y for ADT, and Z for peaks.
X = sp.sparse.csc_matrix(
(np.array(f['RNA.data'], dtype=np.float32),
np.array(f['RNA.indices'], dtype=np.int32),
np.array(f['RNA.indptr'], dtype=np.int32)
),
shape = np.array(f['RNA.shape'], dtype=np.int32)).tocsc().astype(np.float32)
Y = np.array(f['ADT'], dtype=np.float32)
cell_types = np.array(f['cell_types'], dtype='S32').astype(str)
# The covariate matrix, the last column is the ids of datasets
batches = np.array(f['batches'], dtype=np.float32)
# The indices of features that are observed in each dataset.
id_X_cite = np.array(f['id_X_cite'], dtype=np.int32)
id_Y_cite = np.array(f['id_Y_cite'], dtype=np.int32)
id_Y_asap = np.array(f['id_Y_asap'], dtype=np.int32)
# The sample size
sample_sizes = np.array(f['sample_sizes'], dtype=np.int32)
# Preprocess the data
X = X.toarray()
X[batches[:,-1]!=2,:] = np.log(X[batches[:,-1]!=2,:]/np.sum(X[batches[:,-1]!=2,:], axis=1, keepdims=True)*1e4+1.)
Y = np.log(Y/np.sum(Y, axis=1, keepdims=True)*1e4+1.)
data = np.c_[X, Y]
# Subset the data, only use CITE-seq and ASAP-seq datasets for illustration.
batches = batches[sample_sizes[0]:]
data = data[sample_sizes[0]:]
sample_sizes = sample_sizes[1:]
gene_names = gene_names[id_X_cite] # remove genes not measured in the two datasets
id_Y_cite_val = np.random.choice(id_Y_cite, int(0.1*len(id_Y_cite)), replace=False)
id_Y_cite = np.setdiff1d(id_Y_cite, id_Y_cite_val)
dim_input_arr = np.array([len(gene_names),len(ADT_names)])
data = data[:, np.r_[id_X_cite, id_Y_asap+dim_input_arr[0]]]
data_raw = data.copy()
# The masks indicate missing pattern for the whole datasets.
masks_raw = - np.ones_like(data, dtype=np.float32)
masks_raw[:sample_sizes[0], :] = 0.
masks_raw[sample_sizes[0]:, dim_input_arr[0]:] = 0.
# randomly remove 20% of entries to be missing
# Each row corresponds to a cell, and each column corresponds to a feature.
# If the dataset is large, one can also provide a mask matrix with shape (num_datasets, num_features)
# along with a variable `id_datasets` to reduce memory usages.
# -1 means missing and 0 means observed.
masks = masks_raw.copy()
masks[np.random.rand(*masks.shape) < 0.2] = -1
data = data * (masks != -1)
masks = tf.convert_to_tensor(masks, dtype=tf.float32)
print('Sample sizes: {}. Feature sizes: {}.'.format(sample_sizes, dim_input_arr))
<KeysViewHDF5 ['ADT', 'ADT_names', 'RNA.data', 'RNA.indices', 'RNA.indptr', 'RNA.shape', 'batches', 'cell_ids', 'cell_types', 'gene_names', 'id_X_cite', 'id_X_dogma', 'id_Y_asap', 'id_Y_cite', 'id_Y_dogma', 'id_Z_asap', 'id_Z_dogma', 'peak_names', 'peaks.data', 'peaks.indices', 'peaks.indptr', 'peaks.shape', 'sample_sizes']>
Sample sizes: [8689 8535]. Feature sizes: [2374 227].
2025-06-07 03:44:36.886871: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:36.887006: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:36.887059: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:36.938079: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:36.938178: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:36.938254: I external/local_xla/xla/stream_executor/cuda/cuda_executor.cc:901] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
2025-06-07 03:44:36.938304: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1929] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 22282 MB memory: -> device: 0, name: NVIDIA GeForce RTX 4090, pci bus id: 0000:01:00.0, compute capability: 8.9
Set up config for model training#
[ ]:
import sys; sys.path.insert(0, '../../../../') # add parent folder path if not installed via PyPI
import scVAEIT
print(scVAEIT.__version__)
from scVAEIT import VAEIT
path_root = 'result/2/'
config = {
# Dimension of input features for [RNA, ADT, peaks]
'dim_input_arr': dim_input_arr,
# Blocks for [RNA, ADT]
'dim_block': np.array([len(gene_names),len(ADT_names)]), # input dimension of blocks
'dist_block':['NB','NB'], # distributions of blocks
'dim_block_enc':np.array([128, 64]), # dimension of first layer of the encoder
'dim_block_dec':np.array([128, 64]), # dimension of first layer of the decoder
'dim_block_embed':np.array([128, 64]), # mask embedding dimension
# Internal network structure
'dimensions':[32], # dimension of latent layers of encoder; the reversed is used for decoder
'dim_latent':8, # the latent dimension between the encoder and decoder
# Weights
'beta_unobs':.9, # weight for masked out observation; weight for observed values will be 1-beta_unobs.
'beta_modal':np.array([0.05,0.95]), # weights for 2 modalities, which can be adjusted based on loss in the first few epochs.
# Masking probability
"p_feat" : 0.5, # probability of randomly masking out an entry
}
batches_cate = batches
batches_cont = None
model = VAEIT(config, data, masks, None, batches_cate, batches_cont)
1.1.0
Missingness (overall): 0.5618
Missingness (per row): mean=0.5618, min=0.1696, max=0.9389
Missingness (per column): mean=0.5618, min=0.1923, max=0.6039
namespace(beta_kl=2.0, beta_unobs=0.9, beta_reverse=0.0, beta_modal=array([0.05, 0.95]), p_modal=None, p_feat=0.5, uni_block_names=array(['M-0', 'M-1'], dtype='<U23'), block_names=array(['M-0', 'M-1'], dtype='<U23'), dist_block=array(['NB', 'NB'], dtype='<U2'), dim_block=array([2374, 227]), dim_block_enc=array([128, 64], dtype=int32), dim_block_dec=array([128, 64], dtype=int32), skip_conn=False, mean_vals=<tf.Tensor: shape=(2601,), dtype=float32, numpy=array([0., 0., 0., ..., 0., 0., 0.], dtype=float32)>, min_vals=<tf.Tensor: shape=(2601,), dtype=float32, numpy=array([0., 0., 0., ..., 0., 0., 0.], dtype=float32)>, max_vals=<tf.Tensor: shape=(2601,), dtype=float32, numpy=
array([8.865503 , 8.865503 , 8.865503 , ..., 6.5426416, 6.5426416,
6.5426416], dtype=float32)>, max_disp=6.0, max_zi_prob=None, gamma=0.0, dim_input_arr=array([2374, 227]), dim_block_embed=array([128, 64]), dimensions=array([32], dtype=int32), dim_latent=8) (17224, 2601) (17224, 2601) (17224, 4)
[4]:
# Here we only train for 300 epochs for illustration purpose
hist = model.train(
num_epoch=300, batch_size=256, save_every_epoch=50,
verbose=True, checkpoint_dir=path_root+'checkpoint/')
Deleting old log directory at example/result/2/checkpoint/
Train - Start of epoch 1
2025-06-07 03:44:41.904367: I external/local_xla/xla/service/service.cc:168] XLA service 0x615982b752c0 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2025-06-07 03:44:41.904383: I external/local_xla/xla/service/service.cc:176] StreamExecutor device (0): NVIDIA GeForce RTX 4090, Compute Capability 8.9
2025-06-07 03:44:41.906431: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:269] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.
2025-06-07 03:44:41.914470: I external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:454] Loaded cuDNN version 8907
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1749267881.957988 3938238 device_compiler.h:186] Compiled cluster using XLA! This line is logged at most once for the lifetime of the process.
WARNING:tensorflow:5 out of the last 5 calls to <function _BaseOptimizer._update_step_xla at 0x75d57915f280> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.
WARNING:tensorflow:6 out of the last 6 calls to <function _BaseOptimizer._update_step_xla at 0x75d57915f280> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.
67/67 [==============================] - 6s 95ms/step - Reconstructed Loss: 178.8202
Epoch 1, Time elapsed: 0.11 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 187.82, 8.56, 27.83, 71.62, 68.92, 10.88, 0.00
Train - Start of epoch 2
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 104.4566
Epoch 2, Time elapsed: 0.14 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 106.74, 4.84, 22.03, 42.03, 35.02, 2.82, 0.00
Train - Start of epoch 3
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 71.8717
Epoch 3, Time elapsed: 0.18 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 70.97, 3.95, 14.95, 33.36, 17.21, 1.50, 0.00
Train - Start of epoch 4
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 61.5277
Epoch 4, Time elapsed: 0.21 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 61.66, 3.79, 9.38, 31.73, 15.69, 1.07, 0.00
Train - Start of epoch 5
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 60.5139
Epoch 5, Time elapsed: 0.25 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 57.61, 3.74, 6.78, 31.20, 15.04, 0.85, 0.00
Train - Start of epoch 6
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 53.1962
Epoch 6, Time elapsed: 0.28 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 55.60, 3.71, 5.78, 30.89, 14.51, 0.71, 0.00
Train - Start of epoch 7
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 56.6386
Epoch 7, Time elapsed: 0.32 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 54.65, 3.70, 5.38, 30.93, 14.04, 0.61, 0.00
Train - Start of epoch 8
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 57.3299
Epoch 8, Time elapsed: 0.35 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 54.33, 3.69, 5.19, 30.82, 14.10, 0.52, 0.00
Train - Start of epoch 9
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 54.3415
Epoch 9, Time elapsed: 0.39 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 54.22, 3.69, 5.09, 30.82, 14.15, 0.47, 0.00
Train - Start of epoch 10
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.5923
Epoch 10, Time elapsed: 0.42 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 53.85, 3.67, 5.02, 30.82, 13.91, 0.42, 0.00
Train - Start of epoch 11
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 52.7781
Epoch 11, Time elapsed: 0.46 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 53.54, 3.67, 4.97, 30.78, 13.74, 0.38, 0.00
Train - Start of epoch 12
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 54.8336
Epoch 12, Time elapsed: 0.49 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 53.45, 3.66, 4.94, 30.58, 13.92, 0.35, 0.00
Train - Start of epoch 13
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 53.1341
Epoch 13, Time elapsed: 0.53 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 53.17, 3.65, 4.92, 30.49, 13.78, 0.33, 0.00
Train - Start of epoch 14
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 53.4900
Epoch 14, Time elapsed: 0.57 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 53.06, 3.64, 4.90, 30.56, 13.64, 0.32, 0.00
Train - Start of epoch 15
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 53.0251
Epoch 15, Time elapsed: 0.6 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.88, 3.63, 4.88, 30.36, 13.69, 0.32, 0.00
Train - Start of epoch 16
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 53.0253
Epoch 16, Time elapsed: 0.63 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.64, 3.61, 4.85, 30.17, 13.71, 0.31, 0.00
Train - Start of epoch 17
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 53.0556
Epoch 17, Time elapsed: 0.67 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.69, 3.61, 4.85, 30.14, 13.78, 0.31, 0.00
Train - Start of epoch 18
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 52.4851
Epoch 18, Time elapsed: 0.7 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.32, 3.60, 4.84, 30.02, 13.55, 0.31, 0.00
Train - Start of epoch 19
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.9084
Epoch 19, Time elapsed: 0.74 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.33, 3.59, 4.82, 30.05, 13.57, 0.31, 0.00
Train - Start of epoch 20
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 52.5811
Epoch 20, Time elapsed: 0.78 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.07, 3.57, 4.80, 29.92, 13.48, 0.30, 0.00
Train - Start of epoch 21
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 52.4618
Epoch 21, Time elapsed: 0.81 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.11, 3.57, 4.79, 29.86, 13.60, 0.30, 0.00
Train - Start of epoch 22
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 53.7763
Epoch 22, Time elapsed: 0.85 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 52.14, 3.55, 4.77, 29.92, 13.60, 0.30, 0.00
Train - Start of epoch 23
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 51.4684
Epoch 23, Time elapsed: 0.88 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.93, 3.54, 4.75, 29.73, 13.63, 0.29, 0.00
Train - Start of epoch 24
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 53.0989
Epoch 24, Time elapsed: 0.92 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.99, 3.54, 4.75, 29.74, 13.67, 0.29, 0.00
Train - Start of epoch 25
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.6392
Epoch 25, Time elapsed: 0.95 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.57, 3.53, 4.72, 29.70, 13.35, 0.27, 0.00
Train - Start of epoch 26
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.8983
Epoch 26, Time elapsed: 0.99 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.44, 3.52, 4.71, 29.62, 13.31, 0.27, 0.00
Train - Start of epoch 27
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 51.1043
Epoch 27, Time elapsed: 1.02 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.38, 3.52, 4.70, 29.46, 13.45, 0.26, 0.00
Train - Start of epoch 28
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.0919
Epoch 28, Time elapsed: 1.06 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.35, 3.51, 4.68, 29.53, 13.36, 0.28, 0.00
Train - Start of epoch 29
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.6006
Epoch 29, Time elapsed: 1.09 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.43, 3.51, 4.67, 29.43, 13.55, 0.27, 0.00
Train - Start of epoch 30
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 52.0373
Epoch 30, Time elapsed: 1.13 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.19, 3.50, 4.66, 29.40, 13.36, 0.27, 0.00
Train - Start of epoch 31
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 51.9021
Epoch 31, Time elapsed: 1.16 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.23, 3.49, 4.64, 29.51, 13.32, 0.26, 0.00
Train - Start of epoch 32
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 50.9636
Epoch 32, Time elapsed: 1.2 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 51.10, 3.49, 4.63, 29.40, 13.32, 0.26, 0.00
Train - Start of epoch 33
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.0028
Epoch 33, Time elapsed: 1.24 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.98, 3.49, 4.62, 29.24, 13.38, 0.26, 0.00
Train - Start of epoch 34
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.2746
Epoch 34, Time elapsed: 1.27 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.91, 3.48, 4.61, 29.45, 13.11, 0.25, 0.00
Train - Start of epoch 35
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 50.3563
Epoch 35, Time elapsed: 1.31 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.99, 3.48, 4.61, 29.22, 13.43, 0.25, 0.00
Train - Start of epoch 36
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 49.2785
Epoch 36, Time elapsed: 1.34 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.85, 3.47, 4.60, 29.29, 13.24, 0.25, 0.00
Train - Start of epoch 37
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 51.8010
Epoch 37, Time elapsed: 1.38 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.62, 3.47, 4.58, 29.30, 13.02, 0.24, 0.00
Train - Start of epoch 38
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.2551
Epoch 38, Time elapsed: 1.42 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.53, 3.47, 4.58, 29.25, 12.98, 0.24, 0.00
Train - Start of epoch 39
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.4442
Epoch 39, Time elapsed: 1.45 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.66, 3.46, 4.57, 29.19, 13.19, 0.25, 0.00
Train - Start of epoch 40
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 52.0885
Epoch 40, Time elapsed: 1.48 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.49, 3.46, 4.56, 29.19, 13.03, 0.25, 0.00
Train - Start of epoch 41
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 51.9672
Epoch 41, Time elapsed: 1.52 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.58, 3.45, 4.55, 29.21, 13.14, 0.23, 0.00
Train - Start of epoch 42
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.1397
Epoch 42, Time elapsed: 1.56 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.62, 3.45, 4.54, 29.30, 13.09, 0.25, 0.00
Train - Start of epoch 43
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 51.9501
Epoch 43, Time elapsed: 1.59 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.54, 3.44, 4.54, 29.13, 13.18, 0.24, 0.00
Train - Start of epoch 44
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.8729
Epoch 44, Time elapsed: 1.63 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.74, 3.44, 4.53, 28.93, 13.60, 0.25, 0.00
Train - Start of epoch 45
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.9803
Epoch 45, Time elapsed: 1.66 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.35, 3.44, 4.53, 29.27, 12.86, 0.24, 0.00
Train - Start of epoch 46
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 51.4578
Epoch 46, Time elapsed: 1.7 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.35, 3.43, 4.52, 28.97, 13.19, 0.23, 0.00
Train - Start of epoch 47
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.5369
Epoch 47, Time elapsed: 1.73 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.36, 3.43, 4.51, 29.06, 13.12, 0.24, 0.00
Train - Start of epoch 48
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 52.9974
Epoch 48, Time elapsed: 1.77 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.11, 3.42, 4.50, 28.95, 13.00, 0.23, 0.00
Train - Start of epoch 49
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 50.2199
Epoch 49, Time elapsed: 1.8 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.10, 3.43, 4.50, 29.13, 12.81, 0.24, 0.00
Train - Start of epoch 50
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 51.4671
Epoch 50, Time elapsed: 1.84 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.08, 3.42, 4.49, 28.90, 13.04, 0.24, 0.00
<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=50>
Saved checkpoint for epoch 50: example/result/2/checkpoint/ckpt-1
Train - Start of epoch 51
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.6946
Epoch 51, Time elapsed: 1.9 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.98, 3.41, 4.49, 29.02, 12.82, 0.24, 0.00
Train - Start of epoch 52
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.2791
Epoch 52, Time elapsed: 1.94 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.13, 3.41, 4.48, 29.02, 12.97, 0.24, 0.00
Train - Start of epoch 53
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.4089
Epoch 53, Time elapsed: 1.97 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.05, 3.41, 4.47, 28.77, 13.17, 0.23, 0.00
Train - Start of epoch 54
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.5628
Epoch 54, Time elapsed: 2.01 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.20, 3.40, 4.46, 28.78, 13.31, 0.24, 0.00
Train - Start of epoch 55
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.9578
Epoch 55, Time elapsed: 2.04 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 50.06, 3.40, 4.46, 28.92, 13.05, 0.25, 0.00
Train - Start of epoch 56
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 50.4085
Epoch 56, Time elapsed: 2.08 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.91, 3.40, 4.45, 28.80, 13.02, 0.24, 0.00
Train - Start of epoch 57
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.1142
Epoch 57, Time elapsed: 2.11 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.80, 3.39, 4.44, 28.89, 12.83, 0.25, 0.00
Train - Start of epoch 58
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.5353
Epoch 58, Time elapsed: 2.15 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.66, 3.38, 4.43, 28.75, 12.83, 0.27, 0.00
Train - Start of epoch 59
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.8776
Epoch 59, Time elapsed: 2.18 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.59, 3.38, 4.42, 28.81, 12.71, 0.28, 0.00
Train - Start of epoch 60
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 52.1461
Epoch 60, Time elapsed: 2.22 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.80, 3.38, 4.42, 28.63, 13.09, 0.28, 0.00
Train - Start of epoch 61
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 51.0950
Epoch 61, Time elapsed: 2.25 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.62, 3.37, 4.42, 28.64, 12.90, 0.29, 0.00
Train - Start of epoch 62
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.6159
Epoch 62, Time elapsed: 2.29 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.50, 3.37, 4.40, 28.61, 12.83, 0.29, 0.00
Train - Start of epoch 63
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 52.1180
Epoch 63, Time elapsed: 2.32 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.52, 3.37, 4.40, 28.61, 12.84, 0.30, 0.00
Train - Start of epoch 64
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.3880
Epoch 64, Time elapsed: 2.36 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.72, 3.37, 4.40, 28.59, 13.06, 0.31, 0.00
Train - Start of epoch 65
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.8183
Epoch 65, Time elapsed: 2.4 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.26, 3.36, 4.39, 28.57, 12.64, 0.30, 0.00
Train - Start of epoch 66
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.3091
Epoch 66, Time elapsed: 2.43 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.46, 3.36, 4.38, 28.53, 12.88, 0.31, 0.00
Train - Start of epoch 67
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.2856
Epoch 67, Time elapsed: 2.47 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.50, 3.35, 4.38, 28.52, 12.93, 0.32, 0.00
Train - Start of epoch 68
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.6449
Epoch 68, Time elapsed: 2.5 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.25, 3.35, 4.37, 28.61, 12.60, 0.32, 0.00
Train - Start of epoch 69
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 50.3985
Epoch 69, Time elapsed: 2.53 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.34, 3.35, 4.37, 28.33, 12.99, 0.30, 0.00
Train - Start of epoch 70
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.2307
Epoch 70, Time elapsed: 2.57 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.31, 3.34, 4.36, 28.42, 12.88, 0.32, 0.00
Train - Start of epoch 71
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 51.8508
Epoch 71, Time elapsed: 2.6 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.25, 3.34, 4.36, 28.46, 12.77, 0.32, 0.00
Train - Start of epoch 72
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.9979
Epoch 72, Time elapsed: 2.64 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.09, 3.34, 4.35, 28.47, 12.61, 0.31, 0.00
Train - Start of epoch 73
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.2728
Epoch 73, Time elapsed: 2.68 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.17, 3.33, 4.34, 28.35, 12.83, 0.32, 0.00
Train - Start of epoch 74
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.1932
Epoch 74, Time elapsed: 2.71 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.19, 3.33, 4.34, 28.35, 12.86, 0.31, 0.00
Train - Start of epoch 75
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 49.4524
Epoch 75, Time elapsed: 2.75 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.01, 3.33, 4.34, 28.51, 12.52, 0.32, 0.00
Train - Start of epoch 76
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 50.0903
Epoch 76, Time elapsed: 2.78 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.15, 3.33, 4.33, 28.25, 12.92, 0.32, 0.00
Train - Start of epoch 77
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 50.1194
Epoch 77, Time elapsed: 2.82 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.05, 3.33, 4.33, 28.31, 12.78, 0.32, 0.00
Train - Start of epoch 78
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.4825
Epoch 78, Time elapsed: 2.85 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.76, 3.32, 4.32, 28.25, 12.56, 0.31, 0.00
Train - Start of epoch 79
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.6021
Epoch 79, Time elapsed: 2.89 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.90, 3.32, 4.31, 28.34, 12.62, 0.31, 0.00
Train - Start of epoch 80
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.0783
Epoch 80, Time elapsed: 2.92 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.78, 3.32, 4.31, 28.35, 12.48, 0.32, 0.00
Train - Start of epoch 81
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.1283
Epoch 81, Time elapsed: 2.96 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.97, 3.32, 4.31, 28.15, 12.88, 0.31, 0.00
Train - Start of epoch 82
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.7587
Epoch 82, Time elapsed: 2.99 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 49.00, 3.31, 4.30, 28.19, 12.88, 0.31, 0.00
Train - Start of epoch 83
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 50.0961
Epoch 83, Time elapsed: 3.02 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.73, 3.31, 4.29, 28.14, 12.67, 0.32, 0.00
Train - Start of epoch 84
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.1248
Epoch 84, Time elapsed: 3.06 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.77, 3.31, 4.29, 28.13, 12.71, 0.32, 0.00
Train - Start of epoch 85
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.9891
Epoch 85, Time elapsed: 3.09 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.74, 3.31, 4.29, 28.11, 12.73, 0.31, 0.00
Train - Start of epoch 86
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 48.5547
Epoch 86, Time elapsed: 3.13 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.49, 3.31, 4.29, 28.27, 12.33, 0.31, 0.00
Train - Start of epoch 87
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.8921
Epoch 87, Time elapsed: 3.16 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.72, 3.31, 4.28, 28.08, 12.75, 0.31, 0.00
Train - Start of epoch 88
67/67 [==============================] - 2s 34ms/step - Reconstructed Loss: 49.6782
Epoch 88, Time elapsed: 3.2 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.76, 3.30, 4.28, 28.06, 12.81, 0.32, 0.00
Train - Start of epoch 89
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.6908
Epoch 89, Time elapsed: 3.24 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.56, 3.30, 4.27, 28.19, 12.49, 0.31, 0.00
Train - Start of epoch 90
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.5490
Epoch 90, Time elapsed: 3.27 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.78, 3.30, 4.27, 27.99, 12.91, 0.31, 0.00
Train - Start of epoch 91
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.0337
Epoch 91, Time elapsed: 3.3 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.56, 3.30, 4.27, 28.13, 12.55, 0.31, 0.00
Train - Start of epoch 92
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.4025
Epoch 92, Time elapsed: 3.34 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.36, 3.29, 4.26, 28.07, 12.42, 0.31, 0.00
Train - Start of epoch 93
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.6372
Epoch 93, Time elapsed: 3.37 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.52, 3.29, 4.26, 27.98, 12.69, 0.31, 0.00
Train - Start of epoch 94
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.6416
Epoch 94, Time elapsed: 3.41 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.40, 3.29, 4.26, 28.25, 12.29, 0.31, 0.00
Train - Start of epoch 95
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 50.7436
Epoch 95, Time elapsed: 3.44 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.60, 3.29, 4.25, 28.04, 12.72, 0.30, 0.00
Train - Start of epoch 96
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.4002
Epoch 96, Time elapsed: 3.48 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.35, 3.29, 4.24, 28.13, 12.38, 0.31, 0.00
Train - Start of epoch 97
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.6671
Epoch 97, Time elapsed: 3.51 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.37, 3.29, 4.24, 27.95, 12.60, 0.30, 0.00
Train - Start of epoch 98
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 46.8496
Epoch 98, Time elapsed: 3.55 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.59, 3.28, 4.24, 27.96, 12.80, 0.31, 0.00
Train - Start of epoch 99
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.0241
Epoch 99, Time elapsed: 3.58 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.49, 3.28, 4.23, 28.01, 12.65, 0.31, 0.00
Train - Start of epoch 100
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.6947
Epoch 100, Time elapsed: 3.62 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.38, 3.28, 4.24, 28.11, 12.44, 0.31, 0.00
<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=100>
Saved checkpoint for epoch 100: example/result/2/checkpoint/ckpt-2
Train - Start of epoch 101
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.6655
Epoch 101, Time elapsed: 3.68 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.14, 3.28, 4.23, 28.06, 12.28, 0.30, 0.00
Train - Start of epoch 102
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.8981
Epoch 102, Time elapsed: 3.71 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.34, 3.28, 4.22, 28.06, 12.47, 0.31, 0.00
Train - Start of epoch 103
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.2580
Epoch 103, Time elapsed: 3.74 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.41, 3.28, 4.22, 28.08, 12.52, 0.30, 0.00
Train - Start of epoch 104
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.4097
Epoch 104, Time elapsed: 3.78 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.32, 3.27, 4.21, 27.99, 12.54, 0.30, 0.00
Train - Start of epoch 105
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.0353
Epoch 105, Time elapsed: 3.81 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.17, 3.27, 4.21, 28.03, 12.36, 0.30, 0.00
Train - Start of epoch 106
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.8740
Epoch 106, Time elapsed: 3.85 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.20, 3.27, 4.21, 27.93, 12.47, 0.31, 0.00
Train - Start of epoch 107
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.7308
Epoch 107, Time elapsed: 3.88 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.27, 3.27, 4.21, 27.86, 12.63, 0.30, 0.00
Train - Start of epoch 108
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.3254
Epoch 108, Time elapsed: 3.92 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.30, 3.27, 4.20, 28.11, 12.41, 0.31, 0.00
Train - Start of epoch 109
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.7671
Epoch 109, Time elapsed: 3.95 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.04, 3.27, 4.20, 27.84, 12.44, 0.30, 0.00
Train - Start of epoch 110
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.5436
Epoch 110, Time elapsed: 3.99 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.18, 3.26, 4.20, 27.98, 12.45, 0.29, 0.00
Train - Start of epoch 111
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.4635
Epoch 111, Time elapsed: 4.02 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.11, 3.26, 4.19, 27.94, 12.41, 0.30, 0.00
Train - Start of epoch 112
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.3480
Epoch 112, Time elapsed: 4.06 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.28, 3.27, 4.19, 27.93, 12.59, 0.30, 0.00
Train - Start of epoch 113
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.4992
Epoch 113, Time elapsed: 4.09 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.00, 3.26, 4.19, 27.95, 12.30, 0.30, 0.00
Train - Start of epoch 114
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.6776
Epoch 114, Time elapsed: 4.13 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.97, 3.26, 4.18, 27.90, 12.34, 0.29, 0.00
Train - Start of epoch 115
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 49.9045
Epoch 115, Time elapsed: 4.16 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.70, 3.26, 4.18, 27.93, 12.04, 0.30, 0.00
Train - Start of epoch 116
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.1948
Epoch 116, Time elapsed: 4.2 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.00, 3.26, 4.17, 27.94, 12.33, 0.30, 0.00
Train - Start of epoch 117
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.9947
Epoch 117, Time elapsed: 4.23 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.12, 3.25, 4.17, 27.87, 12.53, 0.30, 0.00
Train - Start of epoch 118
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.4674
Epoch 118, Time elapsed: 4.27 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.99, 3.25, 4.17, 27.85, 12.42, 0.30, 0.00
Train - Start of epoch 119
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.6954
Epoch 119, Time elapsed: 4.3 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.22, 3.25, 4.17, 27.91, 12.59, 0.30, 0.00
Train - Start of epoch 120
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.7153
Epoch 120, Time elapsed: 4.34 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.19, 3.25, 4.16, 27.68, 12.81, 0.29, 0.00
Train - Start of epoch 121
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.4231
Epoch 121, Time elapsed: 4.37 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.96, 3.25, 4.16, 27.84, 12.41, 0.29, 0.00
Train - Start of epoch 122
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.9184
Epoch 122, Time elapsed: 4.41 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.93, 3.25, 4.16, 27.91, 12.32, 0.29, 0.00
Train - Start of epoch 123
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 46.8689
Epoch 123, Time elapsed: 4.44 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.75, 3.24, 4.15, 27.78, 12.29, 0.29, 0.00
Train - Start of epoch 124
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.9576
Epoch 124, Time elapsed: 4.47 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.05, 3.25, 4.15, 27.79, 12.57, 0.30, 0.00
Train - Start of epoch 125
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.4398
Epoch 125, Time elapsed: 4.51 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.09, 3.25, 4.15, 27.69, 12.73, 0.28, 0.00
Train - Start of epoch 126
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.2286
Epoch 126, Time elapsed: 4.54 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.91, 3.24, 4.14, 27.93, 12.29, 0.30, 0.00
Train - Start of epoch 127
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.3623
Epoch 127, Time elapsed: 4.58 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.07, 3.24, 4.14, 27.71, 12.67, 0.30, 0.00
Train - Start of epoch 128
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.1527
Epoch 128, Time elapsed: 4.61 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.70, 3.24, 4.14, 27.82, 12.22, 0.29, 0.00
Train - Start of epoch 129
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.4416
Epoch 129, Time elapsed: 4.65 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.60, 3.24, 4.13, 27.74, 12.20, 0.29, 0.00
Train - Start of epoch 130
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.8017
Epoch 130, Time elapsed: 4.68 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.75, 3.24, 4.13, 27.69, 12.40, 0.29, 0.00
Train - Start of epoch 131
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.2396
Epoch 131, Time elapsed: 4.72 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.83, 3.24, 4.13, 27.86, 12.31, 0.29, 0.00
Train - Start of epoch 132
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.9627
Epoch 132, Time elapsed: 4.75 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.86, 3.24, 4.12, 27.91, 12.29, 0.30, 0.00
Train - Start of epoch 133
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 47.6593
Epoch 133, Time elapsed: 4.78 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 48.01, 3.24, 4.12, 27.78, 12.57, 0.29, 0.00
Train - Start of epoch 134
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.1454
Epoch 134, Time elapsed: 4.82 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.71, 3.24, 4.12, 27.79, 12.27, 0.29, 0.00
Train - Start of epoch 135
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 43.7758
Epoch 135, Time elapsed: 4.85 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.63, 3.23, 4.12, 27.81, 12.19, 0.29, 0.00
Train - Start of epoch 136
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 44.3605
Epoch 136, Time elapsed: 4.89 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.84, 3.24, 4.12, 27.88, 12.32, 0.29, 0.00
Train - Start of epoch 137
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.8829
Epoch 137, Time elapsed: 4.92 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.82, 3.23, 4.11, 27.83, 12.35, 0.29, 0.00
Train - Start of epoch 138
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 46.4798
Epoch 138, Time elapsed: 4.96 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.80, 3.23, 4.11, 27.74, 12.43, 0.29, 0.00
Train - Start of epoch 139
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.1866
Epoch 139, Time elapsed: 4.99 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.63, 3.24, 4.11, 27.74, 12.27, 0.28, 0.00
Train - Start of epoch 140
67/67 [==============================] - 2s 34ms/step - Reconstructed Loss: 46.5700
Epoch 140, Time elapsed: 5.03 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.75, 3.23, 4.10, 27.56, 12.58, 0.28, 0.00
Train - Start of epoch 141
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 50.0354
Epoch 141, Time elapsed: 5.07 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.72, 3.23, 4.10, 27.74, 12.35, 0.29, 0.00
Train - Start of epoch 142
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.5489
Epoch 142, Time elapsed: 5.1 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.42, 3.23, 4.09, 27.59, 12.23, 0.28, 0.00
Train - Start of epoch 143
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.8068
Epoch 143, Time elapsed: 5.14 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.68, 3.23, 4.09, 27.64, 12.43, 0.29, 0.00
Train - Start of epoch 144
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.2125
Epoch 144, Time elapsed: 5.17 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.71, 3.23, 4.09, 27.77, 12.33, 0.29, 0.00
Train - Start of epoch 145
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.5708
Epoch 145, Time elapsed: 5.2 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.58, 3.22, 4.08, 27.59, 12.40, 0.29, 0.00
Train - Start of epoch 146
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.8233
Epoch 146, Time elapsed: 5.24 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.51, 3.23, 4.08, 27.70, 12.22, 0.28, 0.00
Train - Start of epoch 147
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 49.4051
Epoch 147, Time elapsed: 5.27 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.53, 3.22, 4.08, 27.69, 12.26, 0.28, 0.00
Train - Start of epoch 148
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.0874
Epoch 148, Time elapsed: 5.31 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.52, 3.23, 4.08, 27.76, 12.17, 0.28, 0.00
Train - Start of epoch 149
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.7397
Epoch 149, Time elapsed: 5.34 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.59, 3.22, 4.07, 27.76, 12.25, 0.29, 0.00
Train - Start of epoch 150
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.1386
Epoch 150, Time elapsed: 5.38 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.62, 3.23, 4.07, 27.73, 12.31, 0.28, 0.00
<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=150>
Saved checkpoint for epoch 150: example/result/2/checkpoint/ckpt-3
Train - Start of epoch 151
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.7671
Epoch 151, Time elapsed: 5.43 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.41, 3.22, 4.06, 27.57, 12.29, 0.28, 0.00
Train - Start of epoch 152
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.0204
Epoch 152, Time elapsed: 5.47 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.44, 3.22, 4.06, 27.75, 12.11, 0.29, 0.00
Train - Start of epoch 153
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.8241
Epoch 153, Time elapsed: 5.5 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.52, 3.22, 4.06, 27.51, 12.46, 0.28, 0.00
Train - Start of epoch 154
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.1676
Epoch 154, Time elapsed: 5.54 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.62, 3.22, 4.06, 27.56, 12.52, 0.27, 0.00
Train - Start of epoch 155
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.5329
Epoch 155, Time elapsed: 5.57 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.52, 3.22, 4.06, 27.55, 12.40, 0.29, 0.00
Train - Start of epoch 156
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.6145
Epoch 156, Time elapsed: 5.61 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.34, 3.22, 4.06, 27.57, 12.21, 0.29, 0.00
Train - Start of epoch 157
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.2908
Epoch 157, Time elapsed: 5.64 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.59, 3.22, 4.05, 27.59, 12.45, 0.28, 0.00
Train - Start of epoch 158
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.6414
Epoch 158, Time elapsed: 5.68 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.43, 3.22, 4.05, 27.57, 12.32, 0.27, 0.00
Train - Start of epoch 159
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.9110
Epoch 159, Time elapsed: 5.71 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.34, 3.22, 4.05, 27.65, 12.14, 0.28, 0.00
Train - Start of epoch 160
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.2603
Epoch 160, Time elapsed: 5.75 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.40, 3.22, 4.05, 27.62, 12.23, 0.28, 0.00
Train - Start of epoch 161
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.8356
Epoch 161, Time elapsed: 5.78 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.33, 3.22, 4.04, 27.65, 12.14, 0.29, 0.00
Train - Start of epoch 162
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.5790
Epoch 162, Time elapsed: 5.81 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.36, 3.21, 4.03, 27.64, 12.19, 0.28, 0.00
Train - Start of epoch 163
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.5473
Epoch 163, Time elapsed: 5.85 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.56, 3.22, 4.04, 27.55, 12.47, 0.28, 0.00
Train - Start of epoch 164
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.2901
Epoch 164, Time elapsed: 5.88 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.36, 3.21, 4.03, 27.51, 12.33, 0.28, 0.00
Train - Start of epoch 165
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.2613
Epoch 165, Time elapsed: 5.91 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.23, 3.21, 4.03, 27.74, 11.97, 0.28, 0.00
Train - Start of epoch 166
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 45.9268
Epoch 166, Time elapsed: 5.95 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.16, 3.21, 4.02, 27.61, 12.04, 0.27, 0.00
Train - Start of epoch 167
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 46.7482
Epoch 167, Time elapsed: 5.98 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.37, 3.21, 4.02, 27.60, 12.25, 0.28, 0.00
Train - Start of epoch 168
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 44.8281
Epoch 168, Time elapsed: 6.01 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.38, 3.21, 4.02, 27.36, 12.52, 0.28, 0.00
Train - Start of epoch 169
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.7751
Epoch 169, Time elapsed: 6.05 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.49, 3.21, 4.02, 27.63, 12.33, 0.30, 0.00
Train - Start of epoch 170
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.1508
Epoch 170, Time elapsed: 6.08 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.48, 3.21, 4.01, 27.48, 12.51, 0.27, 0.00
Train - Start of epoch 171
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.0169
Epoch 171, Time elapsed: 6.12 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.35, 3.21, 4.01, 27.62, 12.22, 0.29, 0.00
Train - Start of epoch 172
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 47.2971
Epoch 172, Time elapsed: 6.15 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.53, 3.21, 4.02, 27.60, 12.41, 0.29, 0.00
Train - Start of epoch 173
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.9680
Epoch 173, Time elapsed: 6.19 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.42, 3.21, 4.01, 27.58, 12.35, 0.28, 0.00
Train - Start of epoch 174
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.9403
Epoch 174, Time elapsed: 6.22 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.25, 3.21, 4.01, 27.56, 12.21, 0.27, 0.00
Train - Start of epoch 175
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.5225
Epoch 175, Time elapsed: 6.26 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.37, 3.21, 4.00, 27.44, 12.43, 0.28, 0.00
Train - Start of epoch 176
67/67 [==============================] - 2s 33ms/step - Reconstructed Loss: 47.2550
Epoch 176, Time elapsed: 6.29 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.15, 3.21, 4.00, 27.47, 12.20, 0.27, 0.00
Train - Start of epoch 177
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.9100
Epoch 177, Time elapsed: 6.33 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.32, 3.21, 4.00, 27.57, 12.26, 0.28, 0.00
Train - Start of epoch 178
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 45.3915
Epoch 178, Time elapsed: 6.36 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.24, 3.21, 4.00, 27.58, 12.17, 0.28, 0.00
Train - Start of epoch 179
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.8790
Epoch 179, Time elapsed: 6.4 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.24, 3.20, 3.99, 27.52, 12.25, 0.28, 0.00
Train - Start of epoch 180
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.7349
Epoch 180, Time elapsed: 6.43 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.39, 3.20, 3.99, 27.45, 12.48, 0.27, 0.00
Train - Start of epoch 181
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.1739
Epoch 181, Time elapsed: 6.47 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.22, 3.21, 3.99, 27.60, 12.13, 0.29, 0.00
Train - Start of epoch 182
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.7884
Epoch 182, Time elapsed: 6.5 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.37, 3.20, 3.99, 27.47, 12.43, 0.29, 0.00
Train - Start of epoch 183
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.6262
Epoch 183, Time elapsed: 6.53 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.45, 3.20, 3.98, 27.49, 12.50, 0.28, 0.00
Train - Start of epoch 184
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.9248
Epoch 184, Time elapsed: 6.57 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.60, 3.20, 3.98, 27.45, 12.67, 0.29, 0.00
Train - Start of epoch 185
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.2149
Epoch 185, Time elapsed: 6.6 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.18, 3.20, 3.98, 27.58, 12.14, 0.27, 0.00
Train - Start of epoch 186
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.6707
Epoch 186, Time elapsed: 6.64 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.05, 3.20, 3.98, 27.49, 12.10, 0.28, 0.00
Train - Start of epoch 187
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.6844
Epoch 187, Time elapsed: 6.67 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.27, 3.20, 3.97, 27.46, 12.37, 0.27, 0.00
Train - Start of epoch 188
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.5835
Epoch 188, Time elapsed: 6.71 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.13, 3.20, 3.98, 27.47, 12.21, 0.28, 0.00
Train - Start of epoch 189
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.5384
Epoch 189, Time elapsed: 6.74 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.28, 3.20, 3.97, 27.41, 12.41, 0.29, 0.00
Train - Start of epoch 190
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.2642
Epoch 190, Time elapsed: 6.78 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.23, 3.20, 3.97, 27.49, 12.28, 0.28, 0.00
Train - Start of epoch 191
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.1097
Epoch 191, Time elapsed: 6.81 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.11, 3.20, 3.97, 27.50, 12.17, 0.27, 0.00
Train - Start of epoch 192
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.7131
Epoch 192, Time elapsed: 6.84 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.20, 3.20, 3.96, 27.53, 12.23, 0.29, 0.00
Train - Start of epoch 193
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.3144
Epoch 193, Time elapsed: 6.88 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.24, 3.20, 3.96, 27.49, 12.31, 0.27, 0.00
Train - Start of epoch 194
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.6667
Epoch 194, Time elapsed: 6.91 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.28, 3.20, 3.96, 27.44, 12.42, 0.27, 0.00
Train - Start of epoch 195
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.9985
Epoch 195, Time elapsed: 6.95 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.93, 3.19, 3.95, 27.30, 12.21, 0.27, 0.00
Train - Start of epoch 196
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.8408
Epoch 196, Time elapsed: 6.98 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.08, 3.20, 3.96, 27.56, 12.09, 0.28, 0.00
Train - Start of epoch 197
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.1835
Epoch 197, Time elapsed: 7.01 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.22, 3.20, 3.96, 27.47, 12.32, 0.28, 0.00
Train - Start of epoch 198
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.7089
Epoch 198, Time elapsed: 7.05 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.90, 3.20, 3.95, 27.57, 11.92, 0.27, 0.00
Train - Start of epoch 199
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.6964
Epoch 199, Time elapsed: 7.08 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.26, 3.20, 3.95, 27.52, 12.32, 0.28, 0.00
Train - Start of epoch 200
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.4922
Epoch 200, Time elapsed: 7.12 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.29, 3.20, 3.95, 27.51, 12.36, 0.28, 0.00
<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=200>
Saved checkpoint for epoch 200: example/result/2/checkpoint/ckpt-4
Train - Start of epoch 201
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 46.2172
Epoch 201, Time elapsed: 7.18 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.21, 3.19, 3.94, 27.39, 12.42, 0.28, 0.00
Train - Start of epoch 202
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.9010
Epoch 202, Time elapsed: 7.21 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.21, 3.20, 3.94, 27.52, 12.28, 0.28, 0.00
Train - Start of epoch 203
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.0207
Epoch 203, Time elapsed: 7.24 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.10, 3.20, 3.94, 27.43, 12.25, 0.28, 0.00
Train - Start of epoch 204
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.7954
Epoch 204, Time elapsed: 7.28 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.05, 3.19, 3.94, 27.50, 12.14, 0.29, 0.00
Train - Start of epoch 205
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 48.0454
Epoch 205, Time elapsed: 7.31 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.26, 3.19, 3.93, 27.46, 12.39, 0.28, 0.00
Train - Start of epoch 206
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.6789
Epoch 206, Time elapsed: 7.35 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.29, 3.19, 3.93, 27.50, 12.39, 0.28, 0.00
Train - Start of epoch 207
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.0162
Epoch 207, Time elapsed: 7.38 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.03, 3.20, 3.94, 27.47, 12.16, 0.27, 0.00
Train - Start of epoch 208
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.9440
Epoch 208, Time elapsed: 7.42 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.08, 3.20, 3.93, 27.53, 12.13, 0.29, 0.00
Train - Start of epoch 209
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.2347
Epoch 209, Time elapsed: 7.45 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.94, 3.20, 3.94, 27.42, 12.11, 0.27, 0.00
Train - Start of epoch 210
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.7746
Epoch 210, Time elapsed: 7.49 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.24, 3.19, 3.93, 27.37, 12.47, 0.28, 0.00
Train - Start of epoch 211
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 47.0703
Epoch 211, Time elapsed: 7.52 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.99, 3.19, 3.93, 27.44, 12.17, 0.27, 0.00
Train - Start of epoch 212
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.7168
Epoch 212, Time elapsed: 7.55 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.02, 3.19, 3.92, 27.43, 12.20, 0.27, 0.00
Train - Start of epoch 213
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.0151
Epoch 213, Time elapsed: 7.58 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.28, 3.20, 3.92, 27.38, 12.49, 0.29, 0.00
Train - Start of epoch 214
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 46.8750
Epoch 214, Time elapsed: 7.62 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.04, 3.19, 3.92, 27.41, 12.25, 0.27, 0.00
Train - Start of epoch 215
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.7992
Epoch 215, Time elapsed: 7.65 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.99, 3.19, 3.91, 27.44, 12.17, 0.27, 0.00
Train - Start of epoch 216
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.5146
Epoch 216, Time elapsed: 7.69 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.98, 3.19, 3.91, 27.42, 12.18, 0.28, 0.00
Train - Start of epoch 217
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.8657
Epoch 217, Time elapsed: 7.72 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.87, 3.19, 3.91, 27.44, 12.05, 0.27, 0.00
Train - Start of epoch 218
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.0412
Epoch 218, Time elapsed: 7.76 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.89, 3.19, 3.91, 27.30, 12.23, 0.27, 0.00
Train - Start of epoch 219
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 45.7186
Epoch 219, Time elapsed: 7.79 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.98, 3.19, 3.91, 27.41, 12.20, 0.27, 0.00
Train - Start of epoch 220
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.4213
Epoch 220, Time elapsed: 7.82 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.86, 3.19, 3.90, 27.47, 12.03, 0.26, 0.00
Train - Start of epoch 221
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.1619
Epoch 221, Time elapsed: 7.86 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.94, 3.19, 3.90, 27.46, 12.11, 0.28, 0.00
Train - Start of epoch 222
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 45.4866
Epoch 222, Time elapsed: 7.89 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.98, 3.19, 3.91, 27.56, 12.04, 0.28, 0.00
Train - Start of epoch 223
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.3440
Epoch 223, Time elapsed: 7.93 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.02, 3.19, 3.91, 27.41, 12.23, 0.28, 0.00
Train - Start of epoch 224
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.1865
Epoch 224, Time elapsed: 7.96 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.85, 3.19, 3.90, 27.51, 11.99, 0.27, 0.00
Train - Start of epoch 225
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.6991
Epoch 225, Time elapsed: 7.99 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.01, 3.19, 3.90, 27.46, 12.18, 0.28, 0.00
Train - Start of epoch 226
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.0580
Epoch 226, Time elapsed: 8.03 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.08, 3.19, 3.90, 27.52, 12.20, 0.27, 0.00
Train - Start of epoch 227
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.4047
Epoch 227, Time elapsed: 8.06 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.13, 3.19, 3.89, 27.40, 12.37, 0.28, 0.00
Train - Start of epoch 228
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.9976
Epoch 228, Time elapsed: 8.1 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.93, 3.19, 3.89, 27.43, 12.15, 0.27, 0.00
Train - Start of epoch 229
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 45.9257
Epoch 229, Time elapsed: 8.13 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.95, 3.19, 3.89, 27.48, 12.12, 0.28, 0.00
Train - Start of epoch 230
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 46.9692
Epoch 230, Time elapsed: 8.17 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.97, 3.19, 3.89, 27.38, 12.24, 0.27, 0.00
Train - Start of epoch 231
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.2643
Epoch 231, Time elapsed: 8.2 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.13, 3.19, 3.89, 27.31, 12.47, 0.28, 0.00
Train - Start of epoch 232
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.3282
Epoch 232, Time elapsed: 8.24 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.79, 3.19, 3.89, 27.35, 12.09, 0.26, 0.00
Train - Start of epoch 233
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.5958
Epoch 233, Time elapsed: 8.27 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.93, 3.19, 3.88, 27.32, 12.27, 0.28, 0.00
Train - Start of epoch 234
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.0760
Epoch 234, Time elapsed: 8.3 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.98, 3.18, 3.88, 27.38, 12.25, 0.28, 0.00
Train - Start of epoch 235
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.2090
Epoch 235, Time elapsed: 8.34 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.77, 3.19, 3.88, 27.46, 11.98, 0.27, 0.00
Train - Start of epoch 236
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.1357
Epoch 236, Time elapsed: 8.37 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.98, 3.19, 3.88, 27.31, 12.33, 0.27, 0.00
Train - Start of epoch 237
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.0905
Epoch 237, Time elapsed: 8.41 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.94, 3.19, 3.87, 27.40, 12.20, 0.28, 0.00
Train - Start of epoch 238
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.4859
Epoch 238, Time elapsed: 8.44 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.01, 3.19, 3.88, 27.49, 12.20, 0.27, 0.00
Train - Start of epoch 239
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.8739
Epoch 239, Time elapsed: 8.48 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.07, 3.18, 3.87, 27.49, 12.25, 0.28, 0.00
Train - Start of epoch 240
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.3052
Epoch 240, Time elapsed: 8.51 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.01, 3.19, 3.87, 27.42, 12.25, 0.27, 0.00
Train - Start of epoch 241
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.4480
Epoch 241, Time elapsed: 8.54 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.93, 3.18, 3.87, 27.48, 12.12, 0.28, 0.00
Train - Start of epoch 242
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 48.0134
Epoch 242, Time elapsed: 8.58 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.86, 3.18, 3.87, 27.43, 12.10, 0.28, 0.00
Train - Start of epoch 243
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.2994
Epoch 243, Time elapsed: 8.62 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.99, 3.18, 3.87, 27.39, 12.26, 0.28, 0.00
Train - Start of epoch 244
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.2075
Epoch 244, Time elapsed: 8.65 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.09, 3.18, 3.86, 27.43, 12.33, 0.28, 0.00
Train - Start of epoch 245
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 45.0724
Epoch 245, Time elapsed: 8.69 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.12, 3.18, 3.86, 27.50, 12.31, 0.27, 0.00
Train - Start of epoch 246
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 46.8852
Epoch 246, Time elapsed: 8.72 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.96, 3.18, 3.86, 27.30, 12.34, 0.27, 0.00
Train - Start of epoch 247
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.5600
Epoch 247, Time elapsed: 8.75 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.09, 3.18, 3.86, 27.31, 12.45, 0.28, 0.00
Train - Start of epoch 248
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 49.3318
Epoch 248, Time elapsed: 8.79 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.04, 3.18, 3.86, 27.31, 12.42, 0.27, 0.00
Train - Start of epoch 249
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.5495
Epoch 249, Time elapsed: 8.82 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.84, 3.18, 3.86, 27.33, 12.21, 0.26, 0.00
Train - Start of epoch 250
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.1971
Epoch 250, Time elapsed: 8.85 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.91, 3.18, 3.85, 27.46, 12.13, 0.28, 0.00
<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=250>
Saved checkpoint for epoch 250: example/result/2/checkpoint/ckpt-5
Train - Start of epoch 251
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.1060
Epoch 251, Time elapsed: 8.91 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.84, 3.18, 3.85, 27.41, 12.11, 0.28, 0.00
Train - Start of epoch 252
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.9170
Epoch 252, Time elapsed: 8.95 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.82, 3.18, 3.85, 27.40, 12.11, 0.27, 0.00
Train - Start of epoch 253
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 45.1682
Epoch 253, Time elapsed: 8.98 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.84, 3.18, 3.85, 27.44, 12.10, 0.27, 0.00
Train - Start of epoch 254
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.3222
Epoch 254, Time elapsed: 9.01 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.00, 3.18, 3.85, 27.35, 12.34, 0.28, 0.00
Train - Start of epoch 255
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.4963
Epoch 255, Time elapsed: 9.05 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.82, 3.18, 3.85, 27.37, 12.15, 0.27, 0.00
Train - Start of epoch 256
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.8660
Epoch 256, Time elapsed: 9.08 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.73, 3.18, 3.84, 27.25, 12.20, 0.27, 0.00
Train - Start of epoch 257
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.7552
Epoch 257, Time elapsed: 9.12 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.62, 3.18, 3.84, 27.35, 11.97, 0.27, 0.00
Train - Start of epoch 258
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.8873
Epoch 258, Time elapsed: 9.15 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.86, 3.18, 3.84, 27.47, 12.09, 0.28, 0.00
Train - Start of epoch 259
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.5836
Epoch 259, Time elapsed: 9.18 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.88, 3.18, 3.84, 27.40, 12.19, 0.27, 0.00
Train - Start of epoch 260
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.0684
Epoch 260, Time elapsed: 9.22 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.80, 3.19, 3.84, 27.53, 11.97, 0.27, 0.00
Train - Start of epoch 261
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.2586
Epoch 261, Time elapsed: 9.25 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.52, 3.18, 3.84, 27.49, 11.76, 0.26, 0.00
Train - Start of epoch 262
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.2602
Epoch 262, Time elapsed: 9.29 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.75, 3.18, 3.83, 27.37, 12.10, 0.27, 0.00
Train - Start of epoch 263
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 47.2382
Epoch 263, Time elapsed: 9.32 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.70, 3.18, 3.83, 27.19, 12.22, 0.27, 0.00
Train - Start of epoch 264
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.4126
Epoch 264, Time elapsed: 9.36 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.73, 3.18, 3.83, 27.43, 12.02, 0.27, 0.00
Train - Start of epoch 265
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.5876
Epoch 265, Time elapsed: 9.39 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.13, 3.18, 3.83, 27.30, 12.54, 0.28, 0.00
Train - Start of epoch 266
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 48.4587
Epoch 266, Time elapsed: 9.42 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 47.06, 3.18, 3.83, 27.46, 12.31, 0.28, 0.00
Train - Start of epoch 267
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.2227
Epoch 267, Time elapsed: 9.46 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.87, 3.18, 3.83, 27.35, 12.23, 0.28, 0.00
Train - Start of epoch 268
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 44.8841
Epoch 268, Time elapsed: 9.49 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.94, 3.18, 3.83, 27.36, 12.30, 0.27, 0.00
Train - Start of epoch 269
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.4972
Epoch 269, Time elapsed: 9.53 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.88, 3.18, 3.82, 27.38, 12.23, 0.27, 0.00
Train - Start of epoch 270
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.7286
Epoch 270, Time elapsed: 9.56 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.82, 3.18, 3.82, 27.41, 12.13, 0.27, 0.00
Train - Start of epoch 271
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 49.6774
Epoch 271, Time elapsed: 9.59 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.92, 3.18, 3.82, 27.36, 12.29, 0.28, 0.00
Train - Start of epoch 272
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 47.5476
Epoch 272, Time elapsed: 9.63 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.81, 3.18, 3.82, 27.43, 12.10, 0.27, 0.00
Train - Start of epoch 273
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.3370
Epoch 273, Time elapsed: 9.66 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.70, 3.18, 3.82, 27.29, 12.14, 0.27, 0.00
Train - Start of epoch 274
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 47.3766
Epoch 274, Time elapsed: 9.69 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.70, 3.18, 3.81, 27.27, 12.17, 0.27, 0.00
Train - Start of epoch 275
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 46.6907
Epoch 275, Time elapsed: 9.73 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.76, 3.18, 3.81, 27.35, 12.15, 0.28, 0.00
Train - Start of epoch 276
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.2845
Epoch 276, Time elapsed: 9.76 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.91, 3.18, 3.81, 27.21, 12.43, 0.27, 0.00
Train - Start of epoch 277
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.2970
Epoch 277, Time elapsed: 9.79 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.68, 3.18, 3.81, 27.47, 11.95, 0.27, 0.00
Train - Start of epoch 278
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.8481
Epoch 278, Time elapsed: 9.83 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.89, 3.18, 3.81, 27.38, 12.25, 0.28, 0.00
Train - Start of epoch 279
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.1377
Epoch 279, Time elapsed: 9.86 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.80, 3.18, 3.81, 27.31, 12.24, 0.27, 0.00
Train - Start of epoch 280
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 44.9855
Epoch 280, Time elapsed: 9.9 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.59, 3.18, 3.81, 27.34, 12.00, 0.27, 0.00
Train - Start of epoch 281
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.0509
Epoch 281, Time elapsed: 9.93 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.82, 3.18, 3.81, 27.18, 12.38, 0.28, 0.00
Train - Start of epoch 282
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.0602
Epoch 282, Time elapsed: 9.97 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.79, 3.18, 3.81, 27.31, 12.22, 0.27, 0.00
Train - Start of epoch 283
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.2836
Epoch 283, Time elapsed: 10.0 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.55, 3.18, 3.80, 27.36, 11.95, 0.27, 0.00
Train - Start of epoch 284
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 46.9941
Epoch 284, Time elapsed: 10.03 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.77, 3.18, 3.80, 27.32, 12.20, 0.27, 0.00
Train - Start of epoch 285
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 45.9119
Epoch 285, Time elapsed: 10.07 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.68, 3.18, 3.80, 27.27, 12.16, 0.27, 0.00
Train - Start of epoch 286
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.7913
Epoch 286, Time elapsed: 10.1 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.68, 3.18, 3.80, 27.33, 12.12, 0.26, 0.00
Train - Start of epoch 287
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.2617
Epoch 287, Time elapsed: 10.13 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.85, 3.18, 3.79, 27.32, 12.29, 0.27, 0.00
Train - Start of epoch 288
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.7005
Epoch 288, Time elapsed: 10.17 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.57, 3.18, 3.79, 27.37, 11.97, 0.26, 0.00
Train - Start of epoch 289
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 47.1884
Epoch 289, Time elapsed: 10.2 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.57, 3.18, 3.79, 27.37, 11.96, 0.27, 0.00
Train - Start of epoch 290
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 45.1959
Epoch 290, Time elapsed: 10.24 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.72, 3.18, 3.79, 27.34, 12.14, 0.27, 0.00
Train - Start of epoch 291
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.1662
Epoch 291, Time elapsed: 10.27 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.71, 3.18, 3.79, 27.39, 12.08, 0.27, 0.00
Train - Start of epoch 292
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 45.2051
Epoch 292, Time elapsed: 10.31 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.68, 3.18, 3.79, 27.25, 12.18, 0.28, 0.00
Train - Start of epoch 293
67/67 [==============================] - 2s 29ms/step - Reconstructed Loss: 47.4425
Epoch 293, Time elapsed: 10.34 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.75, 3.18, 3.79, 27.39, 12.11, 0.28, 0.00
Train - Start of epoch 294
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 45.8794
Epoch 294, Time elapsed: 10.37 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.64, 3.18, 3.79, 27.29, 12.10, 0.28, 0.00
Train - Start of epoch 295
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.1978
Epoch 295, Time elapsed: 10.41 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.86, 3.18, 3.79, 27.24, 12.38, 0.27, 0.00
Train - Start of epoch 296
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 46.4298
Epoch 296, Time elapsed: 10.44 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.61, 3.18, 3.79, 27.49, 11.88, 0.27, 0.00
Train - Start of epoch 297
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 48.5156
Epoch 297, Time elapsed: 10.47 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.83, 3.18, 3.78, 27.48, 12.12, 0.27, 0.00
Train - Start of epoch 298
67/67 [==============================] - 2s 30ms/step - Reconstructed Loss: 47.5198
Epoch 298, Time elapsed: 10.51 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.68, 3.18, 3.79, 27.34, 12.10, 0.27, 0.00
Train - Start of epoch 299
67/67 [==============================] - 2s 32ms/step - Reconstructed Loss: 46.8687
Epoch 299, Time elapsed: 10.54 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.70, 3.18, 3.78, 27.27, 12.19, 0.28, 0.00
Train - Start of epoch 300
67/67 [==============================] - 2s 31ms/step - Reconstructed Loss: 46.2144
Epoch 300, Time elapsed: 10.58 minutes
total, obs_M-0, obs_M-1, unobs_M-0, unobs_M-1, kl, mmd
train : 46.68, 3.18, 3.78, 27.50, 11.95, 0.27, 0.00
<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=300>
Saved checkpoint for epoch 300: example/result/2/checkpoint/ckpt-6
Train Done.
We can visualize the training loss. It fluctuates since we argument the data with random masking, which also provides regularization against overfitting. One may continue training until the loss gets stabilized.
[5]:
plt.plot(hist['train']['total'])
plt.xlabel('Epoch')
plt.ylabel('Trianing loss')
[5]:
Text(0, 0.5, 'Trianing loss')
For later usage, we can load the trained model through the load_model method without the need to retrain:
model.load_model(path_root+'checkpoint/')
Inspect imputation results#
[6]:
denoised_data = model.get_denoised_data()
[7]:
fig, axes = plt.subplots(1,2,figsize=(10,4))
axes[0].scatter(
data_raw[masks==0],
denoised_data[masks==0], s=0.1, alpha=0.1
)
axes[0].set_xlabel('Observed')
axes[0].set_ylabel('Denoised')
axes[1].scatter(
data_raw[(masks_raw==0)&(masks==-1)],
denoised_data[(masks_raw==0)&(masks==-1)], s=0.1, alpha=0.1
)
axes[1].set_xlabel('Observed')
axes[1].set_ylabel('Imputed')
plt.show()
