Inference and Analysis of scATAC-seq Rhythmicity using GeneRhythm
This tutorial provides a step-by-step guide on performing inference, analysis, and visualization of scATAC-seq rhythmicity in a single dataset using GeneRhythm. We demonstrate the diverse capabilities of GeneRhythm by applying it to scATAC-seq data obtained from mouse tissue samples.
GeneRhythm leverages user-provided scATAC-seq data by integrating wavelet transformation with deep generative modeling. This approach enables the extraction of frequency-domain features that complement traditional time-domain analyses, allowing for a more comprehensive understanding of gene expression dynamics.
[1]:
import subprocess
from Build_graph import *
from Frequency_extract import *
from GCN_VAE import *
from Show_result import *
Part 1: Data loading, time and frequency information acquisition
Obtain time information with monocle3
GeneRhythm firstly utilized monocle3 to obtain the trajectory information of single-cell ATAC-seq data. trajectory_inference.R is the script to run monocle3. dataset indicated the name of dataset. mtx, barcode and peak indicate the position of the single-cell ATAC-seq data.Based on the trajectory information, GeneRhythm can derive time information (Peak expression chagnes on the trajectory pesodu-time path).
[2]:
dataset = 'mouse_atherosclerotic_plaque_immune_cells_Arsenic_ATAC'
mtx = './dataset/mouse_atherosclerotic_plaque_immune_cells_ATAC/Arsenic/matrix.mtx'
barcode = './dataset/mouse_atherosclerotic_plaque_immune_cells_ATAC/Arsenic/barcodes.csv'
peak = './dataset/mouse_atherosclerotic_plaque_immune_cells_ATAC/Arsenic/features.csv'
subprocess.run(["Rscript", "trajectory_inference.R", dataset, mtx， barcode, peak])
Frequency information generation
In this step, we harness the power of wavelet transformation to extract detailed frequency information from the scATAC-seq data. By leveraging the trajectory information obtained from Monocle3—which orders cells along a pseudotime axis—we can capture both the temporal progression and the underlying periodic patterns of peak expression.
Wavelet transformation decomposes the peak expression profiles into various frequency components, enabling us to detect subtle oscillations and rhythmic behaviors that are often not apparent in the time domain alone. This multi-scale analysis helps reveal hidden periodicities.
By integrating the frequency-domain features with the temporal trajectory and expression, we achieve a more nuanced analysis. The frequency information complements the time-domain data, enhancing peak clustering. Ultimately, this combined approach provides a comprehensive view of the dynamic changes in peak expression, paving the way for deeper insights into cellular functions and regulatory processes.
[3]:
trajectory_info = pd.read_csv("mouse_atherosclerotic_plaque_immune_cells_Arsenic_ATAC.csv")
mtx = sc.read_mtx(mtx)
mtx = mtx.X.T
barcode = pd.read_csv(barcode,sep=',',index_col=0)
gene = pd.read_csv(peak,header=0,index_col = 0)
adata = anndata.AnnData(mtx,barcode,peak)
frequency_extract(trajectory_info, adata, dataset)
Part 2: Model preparation and training
Modle training
In this stage, the model is trained using time, frequency, and expression data. This integration allows the model to learn a comprehensive latent embedding that encapsulates the intricate relationships among peaks. Once the model has been trained, we apply the Leiden algorithm to the latent space to identify peak clusters. These clusters represent groups of genes with similar expression dynamics and regulatory patterns, providing a valuable basis for further biological insights and downstream analysis.
[4]:
GeneRhythm_Model(input_data = 'mouse_atherosclerotic_plaque_immune_cells_Arsenic_ATAC.npy',sc_data = adata)
____________________________
Start Training VAE...
torch.Size([92807, 250])
Epochs: 1, AvgLoss: 33.0622
torch.Size([92807, 250])
Epochs: 2, AvgLoss: 32.9728
torch.Size([92807, 250])
Epochs: 3, AvgLoss: 32.8834
torch.Size([92807, 250])
Epochs: 4, AvgLoss: 32.7933
torch.Size([92807, 250])
Epochs: 5, AvgLoss: 32.7026
torch.Size([92807, 250])
Epochs: 6, AvgLoss: 32.6092
torch.Size([92807, 250])
Epochs: 7, AvgLoss: 32.5159
torch.Size([92807, 250])
Epochs: 8, AvgLoss: 32.4197
torch.Size([92807, 250])
Epochs: 9, AvgLoss: 32.3222
torch.Size([92807, 250])
Epochs: 10, AvgLoss: 32.2203
torch.Size([92807, 250])
Epochs: 11, AvgLoss: 32.1164
torch.Size([92807, 250])
Epochs: 12, AvgLoss: 32.0080
torch.Size([92807, 250])
Epochs: 13, AvgLoss: 31.8945
torch.Size([92807, 250])
Epochs: 14, AvgLoss: 31.7772
torch.Size([92807, 250])
Epochs: 15, AvgLoss: 31.6531
torch.Size([92807, 250])
Epochs: 16, AvgLoss: 31.5217
torch.Size([92807, 250])
Epochs: 17, AvgLoss: 31.3871
torch.Size([92807, 250])
Epochs: 18, AvgLoss: 31.2414
torch.Size([92807, 250])
Epochs: 19, AvgLoss: 31.0848
torch.Size([92807, 250])
Epochs: 20, AvgLoss: 30.9225
torch.Size([92807, 250])
Epochs: 21, AvgLoss: 30.7477
torch.Size([92807, 250])
Epochs: 22, AvgLoss: 30.5590
torch.Size([92807, 250])
Epochs: 23, AvgLoss: 30.3605
torch.Size([92807, 250])
Epochs: 24, AvgLoss: 30.1408
torch.Size([92807, 250])
Epochs: 25, AvgLoss: 29.9114
torch.Size([92807, 250])
Epochs: 26, AvgLoss: 29.6574
torch.Size([92807, 250])
Epochs: 27, AvgLoss: 29.3821
torch.Size([92807, 250])
Epochs: 28, AvgLoss: 29.0907
torch.Size([92807, 250])
Epochs: 29, AvgLoss: 28.7726
torch.Size([92807, 250])
Epochs: 30, AvgLoss: 28.4275
torch.Size([92807, 250])
Epochs: 31, AvgLoss: 28.0552
torch.Size([92807, 250])
Epochs: 32, AvgLoss: 27.6415
torch.Size([92807, 250])
Epochs: 33, AvgLoss: 27.1980
torch.Size([92807, 250])
Epochs: 34, AvgLoss: 26.7197
torch.Size([92807, 250])
Epochs: 35, AvgLoss: 26.1953
torch.Size([92807, 250])
Epochs: 36, AvgLoss: 25.6184
torch.Size([92807, 250])
Epochs: 37, AvgLoss: 25.0203
torch.Size([92807, 250])
Epochs: 38, AvgLoss: 24.3616
torch.Size([92807, 250])
Epochs: 39, AvgLoss: 23.6622
torch.Size([92807, 250])
Epochs: 40, AvgLoss: 22.9133
torch.Size([92807, 250])
Epochs: 41, AvgLoss: 22.1017
torch.Size([92807, 250])
Epochs: 42, AvgLoss: 21.2524
torch.Size([92807, 250])
Epochs: 43, AvgLoss: 20.3655
torch.Size([92807, 250])
Epochs: 44, AvgLoss: 19.4586
torch.Size([92807, 250])
Epochs: 45, AvgLoss: 18.5328
torch.Size([92807, 250])
Epochs: 46, AvgLoss: 17.5815
torch.Size([92807, 250])
Epochs: 47, AvgLoss: 16.6289
torch.Size([92807, 250])
Epochs: 48, AvgLoss: 15.6969
torch.Size([92807, 250])
Epochs: 49, AvgLoss: 14.8131
torch.Size([92807, 250])
Epochs: 50, AvgLoss: 13.9498
torch.Size([92807, 250])
Epochs: 51, AvgLoss: 13.1690
torch.Size([92807, 250])
Epochs: 52, AvgLoss: 12.4450
torch.Size([92807, 250])
Epochs: 53, AvgLoss: 11.8147
torch.Size([92807, 250])
Epochs: 54, AvgLoss: 11.2545
torch.Size([92807, 250])
Epochs: 55, AvgLoss: 10.7855
torch.Size([92807, 250])
Epochs: 56, AvgLoss: 10.3982
torch.Size([92807, 250])
Epochs: 57, AvgLoss: 10.0746
torch.Size([92807, 250])
Epochs: 58, AvgLoss: 9.8260
torch.Size([92807, 250])
Epochs: 59, AvgLoss: 9.6198
torch.Size([92807, 250])
Epochs: 60, AvgLoss: 9.4721
torch.Size([92807, 250])
Epochs: 61, AvgLoss: 9.3546
torch.Size([92807, 250])
Epochs: 62, AvgLoss: 9.2668
torch.Size([92807, 250])
Epochs: 63, AvgLoss: 9.1971
torch.Size([92807, 250])
Epochs: 64, AvgLoss: 9.1442
torch.Size([92807, 250])
Epochs: 65, AvgLoss: 9.1057
torch.Size([92807, 250])
Epochs: 66, AvgLoss: 9.0725
torch.Size([92807, 250])
Epochs: 67, AvgLoss: 9.0413
torch.Size([92807, 250])
Epochs: 68, AvgLoss: 9.0158
torch.Size([92807, 250])
Epochs: 69, AvgLoss: 8.9891
torch.Size([92807, 250])
Epochs: 70, AvgLoss: 8.9672
torch.Size([92807, 250])
Epochs: 71, AvgLoss: 8.9393
torch.Size([92807, 250])
Epochs: 72, AvgLoss: 8.9172
torch.Size([92807, 250])
Epochs: 73, AvgLoss: 8.8937
torch.Size([92807, 250])
Epochs: 74, AvgLoss: 8.8694
torch.Size([92807, 250])
Epochs: 75, AvgLoss: 8.8449
torch.Size([92807, 250])
Epochs: 76, AvgLoss: 8.8214
torch.Size([92807, 250])
Epochs: 77, AvgLoss: 8.8015
torch.Size([92807, 250])
Epochs: 78, AvgLoss: 8.7762
torch.Size([92807, 250])
Epochs: 79, AvgLoss: 8.7565
torch.Size([92807, 250])
Epochs: 80, AvgLoss: 8.7375
torch.Size([92807, 250])
Epochs: 81, AvgLoss: 8.7191
torch.Size([92807, 250])
Epochs: 82, AvgLoss: 8.7013
torch.Size([92807, 250])
Epochs: 83, AvgLoss: 8.6838
torch.Size([92807, 250])
Epochs: 84, AvgLoss: 8.6708
torch.Size([92807, 250])
Epochs: 85, AvgLoss: 8.6539
torch.Size([92807, 250])
Epochs: 86, AvgLoss: 8.6389
torch.Size([92807, 250])
Epochs: 87, AvgLoss: 8.6299
torch.Size([92807, 250])
Epochs: 88, AvgLoss: 8.6140
torch.Size([92807, 250])
Epochs: 89, AvgLoss: 8.6057
torch.Size([92807, 250])
Epochs: 90, AvgLoss: 8.5920
torch.Size([92807, 250])
Epochs: 91, AvgLoss: 8.5812
torch.Size([92807, 250])
Epochs: 92, AvgLoss: 8.5716
torch.Size([92807, 250])
Epochs: 93, AvgLoss: 8.5608
torch.Size([92807, 250])
Epochs: 94, AvgLoss: 8.5528
torch.Size([92807, 250])
Epochs: 95, AvgLoss: 8.5411
torch.Size([92807, 250])
Epochs: 96, AvgLoss: 8.5335
torch.Size([92807, 250])
Epochs: 97, AvgLoss: 8.5236
torch.Size([92807, 250])
Epochs: 98, AvgLoss: 8.5158
torch.Size([92807, 250])
Epochs: 99, AvgLoss: 8.5065
torch.Size([92807, 250])
Epochs: 100, AvgLoss: 8.4973
torch.Size([92807, 250])
Epochs: 101, AvgLoss: 8.4888
torch.Size([92807, 250])
Epochs: 102, AvgLoss: 8.4822
torch.Size([92807, 250])
Epochs: 103, AvgLoss: 8.4722
torch.Size([92807, 250])
Epochs: 104, AvgLoss: 8.4654
torch.Size([92807, 250])
Epochs: 105, AvgLoss: 8.4591
torch.Size([92807, 250])
Epochs: 106, AvgLoss: 8.4484
torch.Size([92807, 250])
Epochs: 107, AvgLoss: 8.4415
torch.Size([92807, 250])
Epochs: 108, AvgLoss: 8.4336
torch.Size([92807, 250])
Epochs: 109, AvgLoss: 8.4295
torch.Size([92807, 250])
Epochs: 110, AvgLoss: 8.4185
torch.Size([92807, 250])
Epochs: 111, AvgLoss: 8.4110
torch.Size([92807, 250])
Epochs: 112, AvgLoss: 8.4041
torch.Size([92807, 250])
Epochs: 113, AvgLoss: 8.3973
torch.Size([92807, 250])
Epochs: 114, AvgLoss: 8.3891
torch.Size([92807, 250])
Epochs: 115, AvgLoss: 8.3825
torch.Size([92807, 250])
Epochs: 116, AvgLoss: 8.3756
torch.Size([92807, 250])
Epochs: 117, AvgLoss: 8.3684
torch.Size([92807, 250])
Epochs: 118, AvgLoss: 8.3647
torch.Size([92807, 250])
Epochs: 119, AvgLoss: 8.3563
torch.Size([92807, 250])
Epochs: 120, AvgLoss: 8.3482
torch.Size([92807, 250])
Epochs: 121, AvgLoss: 8.3403
torch.Size([92807, 250])
Epochs: 122, AvgLoss: 8.3349
torch.Size([92807, 250])
Epochs: 123, AvgLoss: 8.3273
torch.Size([92807, 250])
Epochs: 124, AvgLoss: 8.3217
torch.Size([92807, 250])
Epochs: 125, AvgLoss: 8.3147
torch.Size([92807, 250])
Epochs: 126, AvgLoss: 8.3106
torch.Size([92807, 250])
Epochs: 127, AvgLoss: 8.3042
torch.Size([92807, 250])
Epochs: 128, AvgLoss: 8.2972
torch.Size([92807, 250])
Epochs: 129, AvgLoss: 8.2947
torch.Size([92807, 250])
Epochs: 130, AvgLoss: 8.2856
torch.Size([92807, 250])
Epochs: 131, AvgLoss: 8.2821
torch.Size([92807, 250])
Epochs: 132, AvgLoss: 8.2755
torch.Size([92807, 250])
Epochs: 133, AvgLoss: 8.2710
torch.Size([92807, 250])
Epochs: 134, AvgLoss: 8.2642
torch.Size([92807, 250])
Epochs: 135, AvgLoss: 8.2602
torch.Size([92807, 250])
Epochs: 136, AvgLoss: 8.2557
torch.Size([92807, 250])
Epochs: 137, AvgLoss: 8.2490
torch.Size([92807, 250])
Epochs: 138, AvgLoss: 8.2435
torch.Size([92807, 250])
Epochs: 139, AvgLoss: 8.2420
torch.Size([92807, 250])
Epochs: 140, AvgLoss: 8.2352
torch.Size([92807, 250])
Epochs: 141, AvgLoss: 8.2299
torch.Size([92807, 250])
Epochs: 142, AvgLoss: 8.2281
torch.Size([92807, 250])
Epochs: 143, AvgLoss: 8.2220
torch.Size([92807, 250])
Epochs: 144, AvgLoss: 8.2178
torch.Size([92807, 250])
Epochs: 145, AvgLoss: 8.2109
torch.Size([92807, 250])
Epochs: 146, AvgLoss: 8.2079
torch.Size([92807, 250])
Epochs: 147, AvgLoss: 8.2033
torch.Size([92807, 250])
Epochs: 148, AvgLoss: 8.2003
torch.Size([92807, 250])
Epochs: 149, AvgLoss: 8.1973
torch.Size([92807, 250])
Epochs: 150, AvgLoss: 8.1907
torch.Size([92807, 250])
Epochs: 151, AvgLoss: 8.1863
torch.Size([92807, 250])
Epochs: 152, AvgLoss: 8.1800
torch.Size([92807, 250])
Epochs: 153, AvgLoss: 8.1784
torch.Size([92807, 250])
Epochs: 154, AvgLoss: 8.1746
torch.Size([92807, 250])
Epochs: 155, AvgLoss: 8.1690
torch.Size([92807, 250])
Epochs: 156, AvgLoss: 8.1643
torch.Size([92807, 250])
Epochs: 157, AvgLoss: 8.1594
torch.Size([92807, 250])
Epochs: 158, AvgLoss: 8.1559
torch.Size([92807, 250])
Epochs: 159, AvgLoss: 8.1531
torch.Size([92807, 250])
Epochs: 160, AvgLoss: 8.1488
torch.Size([92807, 250])
Epochs: 161, AvgLoss: 8.1438
torch.Size([92807, 250])
Epochs: 162, AvgLoss: 8.1407
torch.Size([92807, 250])
Epochs: 163, AvgLoss: 8.1370
torch.Size([92807, 250])
Epochs: 164, AvgLoss: 8.1353
torch.Size([92807, 250])
Epochs: 165, AvgLoss: 8.1287
torch.Size([92807, 250])
Epochs: 166, AvgLoss: 8.1267
torch.Size([92807, 250])
Epochs: 167, AvgLoss: 8.1224
torch.Size([92807, 250])
Epochs: 168, AvgLoss: 8.1181
torch.Size([92807, 250])
Epochs: 169, AvgLoss: 8.1155
torch.Size([92807, 250])
Epochs: 170, AvgLoss: 8.1131
torch.Size([92807, 250])
Epochs: 171, AvgLoss: 8.1088
torch.Size([92807, 250])
Epochs: 172, AvgLoss: 8.1040
torch.Size([92807, 250])
Epochs: 173, AvgLoss: 8.1001
torch.Size([92807, 250])
Epochs: 174, AvgLoss: 8.0966
torch.Size([92807, 250])
Epochs: 175, AvgLoss: 8.0928
torch.Size([92807, 250])
Epochs: 176, AvgLoss: 8.0890
torch.Size([92807, 250])
Epochs: 177, AvgLoss: 8.0839
torch.Size([92807, 250])
Epochs: 178, AvgLoss: 8.0816
torch.Size([92807, 250])
Epochs: 179, AvgLoss: 8.0782
torch.Size([92807, 250])
Epochs: 180, AvgLoss: 8.0729
torch.Size([92807, 250])
Epochs: 181, AvgLoss: 8.0665
torch.Size([92807, 250])
Epochs: 182, AvgLoss: 8.0631
torch.Size([92807, 250])
Epochs: 183, AvgLoss: 8.0610
torch.Size([92807, 250])
Epochs: 184, AvgLoss: 8.0555
torch.Size([92807, 250])
Epochs: 185, AvgLoss: 8.0518
torch.Size([92807, 250])
Epochs: 186, AvgLoss: 8.0476
torch.Size([92807, 250])
Epochs: 187, AvgLoss: 8.0402
torch.Size([92807, 250])
Epochs: 188, AvgLoss: 8.0384
torch.Size([92807, 250])
Epochs: 189, AvgLoss: 8.0331
torch.Size([92807, 250])
Epochs: 190, AvgLoss: 8.0291
torch.Size([92807, 250])
Epochs: 191, AvgLoss: 8.0250
torch.Size([92807, 250])
Epochs: 192, AvgLoss: 8.0194
torch.Size([92807, 250])
Epochs: 193, AvgLoss: 8.0174
torch.Size([92807, 250])
Epochs: 194, AvgLoss: 8.0120
torch.Size([92807, 250])
Epochs: 195, AvgLoss: 8.0062
torch.Size([92807, 250])
Epochs: 196, AvgLoss: 8.0011
torch.Size([92807, 250])
Epochs: 197, AvgLoss: 7.9972
torch.Size([92807, 250])
Epochs: 198, AvgLoss: 7.9922
torch.Size([92807, 250])
Epochs: 199, AvgLoss: 7.9875
torch.Size([92807, 250])
Epochs: 200, AvgLoss: 7.9832
torch.Size([92807, 250])
Epochs: 201, AvgLoss: 7.9764
torch.Size([92807, 250])
Epochs: 202, AvgLoss: 7.9737
torch.Size([92807, 250])
Epochs: 203, AvgLoss: 7.9684
torch.Size([92807, 250])
Epochs: 204, AvgLoss: 7.9638
torch.Size([92807, 250])
Epochs: 205, AvgLoss: 7.9615
torch.Size([92807, 250])
Epochs: 206, AvgLoss: 7.9530
torch.Size([92807, 250])
Epochs: 207, AvgLoss: 7.9489
torch.Size([92807, 250])
Epochs: 208, AvgLoss: 7.9444
torch.Size([92807, 250])
Epochs: 209, AvgLoss: 7.9418
torch.Size([92807, 250])
Epochs: 210, AvgLoss: 7.9340
torch.Size([92807, 250])
Epochs: 211, AvgLoss: 7.9302
torch.Size([92807, 250])
Epochs: 212, AvgLoss: 7.9265
torch.Size([92807, 250])
Epochs: 213, AvgLoss: 7.9204
torch.Size([92807, 250])
Epochs: 214, AvgLoss: 7.9169
torch.Size([92807, 250])
Epochs: 215, AvgLoss: 7.9104
torch.Size([92807, 250])
Epochs: 216, AvgLoss: 7.9031
torch.Size([92807, 250])
Epochs: 217, AvgLoss: 7.8990
torch.Size([92807, 250])
Epochs: 218, AvgLoss: 7.8912
torch.Size([92807, 250])
Epochs: 219, AvgLoss: 7.8885
torch.Size([92807, 250])
Epochs: 220, AvgLoss: 7.8823
torch.Size([92807, 250])
Epochs: 221, AvgLoss: 7.8766
torch.Size([92807, 250])
Epochs: 222, AvgLoss: 7.8750
torch.Size([92807, 250])
Epochs: 223, AvgLoss: 7.8680
torch.Size([92807, 250])
Epochs: 224, AvgLoss: 7.8623
torch.Size([92807, 250])
Epochs: 225, AvgLoss: 7.8562
torch.Size([92807, 250])
Epochs: 226, AvgLoss: 7.8504
torch.Size([92807, 250])
Epochs: 227, AvgLoss: 7.8443
torch.Size([92807, 250])
Epochs: 228, AvgLoss: 7.8405
torch.Size([92807, 250])
Epochs: 229, AvgLoss: 7.8340
torch.Size([92807, 250])
Epochs: 230, AvgLoss: 7.8278
torch.Size([92807, 250])
Epochs: 231, AvgLoss: 7.8241
torch.Size([92807, 250])
Epochs: 232, AvgLoss: 7.8167
torch.Size([92807, 250])
Epochs: 233, AvgLoss: 7.8109
torch.Size([92807, 250])
Epochs: 234, AvgLoss: 7.8079
torch.Size([92807, 250])
Epochs: 235, AvgLoss: 7.8012
torch.Size([92807, 250])
Epochs: 236, AvgLoss: 7.7954
torch.Size([92807, 250])
Epochs: 237, AvgLoss: 7.7872
torch.Size([92807, 250])
Epochs: 238, AvgLoss: 7.7820
torch.Size([92807, 250])
Epochs: 239, AvgLoss: 7.7768
torch.Size([92807, 250])
Epochs: 240, AvgLoss: 7.7710
torch.Size([92807, 250])
Epochs: 241, AvgLoss: 7.7646
torch.Size([92807, 250])
Epochs: 242, AvgLoss: 7.7573
torch.Size([92807, 250])
Epochs: 243, AvgLoss: 7.7520
torch.Size([92807, 250])
Epochs: 244, AvgLoss: 7.7464
torch.Size([92807, 250])
Epochs: 245, AvgLoss: 7.7387
torch.Size([92807, 250])
Epochs: 246, AvgLoss: 7.7330
torch.Size([92807, 250])
Epochs: 247, AvgLoss: 7.7261
torch.Size([92807, 250])
Epochs: 248, AvgLoss: 7.7189
torch.Size([92807, 250])
Epochs: 249, AvgLoss: 7.7153
torch.Size([92807, 250])
Epochs: 250, AvgLoss: 7.7051
torch.Size([92807, 250])
Epochs: 251, AvgLoss: 7.7024
torch.Size([92807, 250])
Epochs: 252, AvgLoss: 7.6935
torch.Size([92807, 250])
Epochs: 253, AvgLoss: 7.6864
torch.Size([92807, 250])
Epochs: 254, AvgLoss: 7.6818
torch.Size([92807, 250])
Epochs: 255, AvgLoss: 7.6740
torch.Size([92807, 250])
Epochs: 256, AvgLoss: 7.6680
torch.Size([92807, 250])
Epochs: 257, AvgLoss: 7.6596
torch.Size([92807, 250])
Epochs: 258, AvgLoss: 7.6537
torch.Size([92807, 250])
Epochs: 259, AvgLoss: 7.6469
torch.Size([92807, 250])
Epochs: 260, AvgLoss: 7.6408
torch.Size([92807, 250])
Epochs: 261, AvgLoss: 7.6344
torch.Size([92807, 250])
Epochs: 262, AvgLoss: 7.6258
torch.Size([92807, 250])
Epochs: 263, AvgLoss: 7.6193
torch.Size([92807, 250])
Epochs: 264, AvgLoss: 7.6125
torch.Size([92807, 250])
Epochs: 265, AvgLoss: 7.6052
torch.Size([92807, 250])
Epochs: 266, AvgLoss: 7.5986
torch.Size([92807, 250])
Epochs: 267, AvgLoss: 7.5939
torch.Size([92807, 250])
Epochs: 268, AvgLoss: 7.5843
torch.Size([92807, 250])
Epochs: 269, AvgLoss: 7.5801
torch.Size([92807, 250])
Epochs: 270, AvgLoss: 7.5724
torch.Size([92807, 250])
Epochs: 271, AvgLoss: 7.5673
torch.Size([92807, 250])
Epochs: 272, AvgLoss: 7.5595
torch.Size([92807, 250])
Epochs: 273, AvgLoss: 7.5525
torch.Size([92807, 250])
Epochs: 274, AvgLoss: 7.5460
torch.Size([92807, 250])
Epochs: 275, AvgLoss: 7.5395
torch.Size([92807, 250])
Epochs: 276, AvgLoss: 7.5354
torch.Size([92807, 250])
Epochs: 277, AvgLoss: 7.5278
torch.Size([92807, 250])
Epochs: 278, AvgLoss: 7.5234
torch.Size([92807, 250])
Epochs: 279, AvgLoss: 7.5172
torch.Size([92807, 250])
Epochs: 280, AvgLoss: 7.5126
torch.Size([92807, 250])
Epochs: 281, AvgLoss: 7.5049
torch.Size([92807, 250])
Epochs: 282, AvgLoss: 7.5000
torch.Size([92807, 250])
Epochs: 283, AvgLoss: 7.4945
torch.Size([92807, 250])
Epochs: 284, AvgLoss: 7.4882
torch.Size([92807, 250])
Epochs: 285, AvgLoss: 7.4843
torch.Size([92807, 250])
Epochs: 286, AvgLoss: 7.4782
torch.Size([92807, 250])
Epochs: 287, AvgLoss: 7.4730
torch.Size([92807, 250])
Epochs: 288, AvgLoss: 7.4687
torch.Size([92807, 250])
Epochs: 289, AvgLoss: 7.4639
torch.Size([92807, 250])
Epochs: 290, AvgLoss: 7.4588
torch.Size([92807, 250])
Epochs: 291, AvgLoss: 7.4555
torch.Size([92807, 250])
Epochs: 292, AvgLoss: 7.4486
torch.Size([92807, 250])
Epochs: 293, AvgLoss: 7.4441
torch.Size([92807, 250])
Epochs: 294, AvgLoss: 7.4398
torch.Size([92807, 250])
Epochs: 295, AvgLoss: 7.4354
torch.Size([92807, 250])
Epochs: 296, AvgLoss: 7.4303
torch.Size([92807, 250])
Epochs: 297, AvgLoss: 7.4253
torch.Size([92807, 250])
Epochs: 298, AvgLoss: 7.4212
torch.Size([92807, 250])
Epochs: 299, AvgLoss: 7.4171
torch.Size([92807, 250])
Epochs: 300, AvgLoss: 7.4139
torch.Size([92807, 250])
Epochs: 301, AvgLoss: 7.4079
torch.Size([92807, 250])
Epochs: 302, AvgLoss: 7.4035
torch.Size([92807, 250])
Epochs: 303, AvgLoss: 7.3983
torch.Size([92807, 250])
Epochs: 304, AvgLoss: 7.3936
torch.Size([92807, 250])
Epochs: 305, AvgLoss: 7.3900
torch.Size([92807, 250])
Epochs: 306, AvgLoss: 7.3841
torch.Size([92807, 250])
Epochs: 307, AvgLoss: 7.3803
torch.Size([92807, 250])
Epochs: 308, AvgLoss: 7.3755
torch.Size([92807, 250])
Epochs: 309, AvgLoss: 7.3708
torch.Size([92807, 250])
Epochs: 310, AvgLoss: 7.3670
torch.Size([92807, 250])
Epochs: 311, AvgLoss: 7.3620
torch.Size([92807, 250])
Epochs: 312, AvgLoss: 7.3574
torch.Size([92807, 250])
Epochs: 313, AvgLoss: 7.3534
torch.Size([92807, 250])
Epochs: 314, AvgLoss: 7.3489
torch.Size([92807, 250])
Epochs: 315, AvgLoss: 7.3439
torch.Size([92807, 250])
Epochs: 316, AvgLoss: 7.3384
torch.Size([92807, 250])
Epochs: 317, AvgLoss: 7.3322
torch.Size([92807, 250])
Epochs: 318, AvgLoss: 7.3295
torch.Size([92807, 250])
Epochs: 319, AvgLoss: 7.3246
torch.Size([92807, 250])
Epochs: 320, AvgLoss: 7.3179
torch.Size([92807, 250])
Epochs: 321, AvgLoss: 7.3138
torch.Size([92807, 250])
Epochs: 322, AvgLoss: 7.3097
torch.Size([92807, 250])
Epochs: 323, AvgLoss: 7.3032
torch.Size([92807, 250])
Epochs: 324, AvgLoss: 7.2978
torch.Size([92807, 250])
Epochs: 325, AvgLoss: 7.2919
torch.Size([92807, 250])
Epochs: 326, AvgLoss: 7.2855
torch.Size([92807, 250])
Epochs: 327, AvgLoss: 7.2812
torch.Size([92807, 250])
Epochs: 328, AvgLoss: 7.2757
torch.Size([92807, 250])
Epochs: 329, AvgLoss: 7.2703
torch.Size([92807, 250])
Epochs: 330, AvgLoss: 7.2635
torch.Size([92807, 250])
Epochs: 331, AvgLoss: 7.2569
torch.Size([92807, 250])
Epochs: 332, AvgLoss: 7.2500
torch.Size([92807, 250])
Epochs: 333, AvgLoss: 7.2431
torch.Size([92807, 250])
Epochs: 334, AvgLoss: 7.2357
torch.Size([92807, 250])
Epochs: 335, AvgLoss: 7.2289
torch.Size([92807, 250])
Epochs: 336, AvgLoss: 7.2225
torch.Size([92807, 250])
Epochs: 337, AvgLoss: 7.2138
torch.Size([92807, 250])
Epochs: 338, AvgLoss: 7.2060
torch.Size([92807, 250])
Epochs: 339, AvgLoss: 7.1964
torch.Size([92807, 250])
Epochs: 340, AvgLoss: 7.1897
torch.Size([92807, 250])
Epochs: 341, AvgLoss: 7.1806
torch.Size([92807, 250])
Epochs: 342, AvgLoss: 7.1719
torch.Size([92807, 250])
Epochs: 343, AvgLoss: 7.1619
torch.Size([92807, 250])
Epochs: 344, AvgLoss: 7.1518
torch.Size([92807, 250])
Epochs: 345, AvgLoss: 7.1406
torch.Size([92807, 250])
Epochs: 346, AvgLoss: 7.1310
torch.Size([92807, 250])
Epochs: 347, AvgLoss: 7.1203
torch.Size([92807, 250])
Epochs: 348, AvgLoss: 7.1082
torch.Size([92807, 250])
Epochs: 349, AvgLoss: 7.0953
torch.Size([92807, 250])
Epochs: 350, AvgLoss: 7.0831
torch.Size([92807, 250])
Epochs: 351, AvgLoss: 7.0707
torch.Size([92807, 250])
Epochs: 352, AvgLoss: 7.0560
torch.Size([92807, 250])
Epochs: 353, AvgLoss: 7.0423
torch.Size([92807, 250])
Epochs: 354, AvgLoss: 7.0266
torch.Size([92807, 250])
Epochs: 355, AvgLoss: 7.0107
torch.Size([92807, 250])
Epochs: 356, AvgLoss: 6.9938
torch.Size([92807, 250])
Epochs: 357, AvgLoss: 6.9769
torch.Size([92807, 250])
Epochs: 358, AvgLoss: 6.9595
torch.Size([92807, 250])
Epochs: 359, AvgLoss: 6.9405
torch.Size([92807, 250])
Epochs: 360, AvgLoss: 6.9213
torch.Size([92807, 250])
Epochs: 361, AvgLoss: 6.9008
torch.Size([92807, 250])
Epochs: 362, AvgLoss: 6.8793
torch.Size([92807, 250])
Epochs: 363, AvgLoss: 6.8574
torch.Size([92807, 250])
Epochs: 364, AvgLoss: 6.8363
torch.Size([92807, 250])
Epochs: 365, AvgLoss: 6.8116
torch.Size([92807, 250])
Epochs: 366, AvgLoss: 6.7889
torch.Size([92807, 250])
Epochs: 367, AvgLoss: 6.7648
torch.Size([92807, 250])
Epochs: 368, AvgLoss: 6.7398
torch.Size([92807, 250])
Epochs: 369, AvgLoss: 6.7151
torch.Size([92807, 250])
Epochs: 370, AvgLoss: 6.6895
torch.Size([92807, 250])
Epochs: 371, AvgLoss: 6.6641
torch.Size([92807, 250])
Epochs: 372, AvgLoss: 6.6380
torch.Size([92807, 250])
Epochs: 373, AvgLoss: 6.6120
torch.Size([92807, 250])
Epochs: 374, AvgLoss: 6.5858
torch.Size([92807, 250])
Epochs: 375, AvgLoss: 6.5602
torch.Size([92807, 250])
Epochs: 376, AvgLoss: 6.5357
torch.Size([92807, 250])
Epochs: 377, AvgLoss: 6.5108
torch.Size([92807, 250])
Epochs: 378, AvgLoss: 6.4859
torch.Size([92807, 250])
Epochs: 379, AvgLoss: 6.4622
torch.Size([92807, 250])
Epochs: 380, AvgLoss: 6.4388
torch.Size([92807, 250])
Epochs: 381, AvgLoss: 6.4174
torch.Size([92807, 250])
Epochs: 382, AvgLoss: 6.3961
torch.Size([92807, 250])
Epochs: 383, AvgLoss: 6.3756
torch.Size([92807, 250])
Epochs: 384, AvgLoss: 6.3569
torch.Size([92807, 250])
Epochs: 385, AvgLoss: 6.3374
torch.Size([92807, 250])
Epochs: 386, AvgLoss: 6.3194
torch.Size([92807, 250])
Epochs: 387, AvgLoss: 6.3022
torch.Size([92807, 250])
Epochs: 388, AvgLoss: 6.2856
torch.Size([92807, 250])
Epochs: 389, AvgLoss: 6.2683
torch.Size([92807, 250])
Epochs: 390, AvgLoss: 6.2539
torch.Size([92807, 250])
Epochs: 391, AvgLoss: 6.2383
torch.Size([92807, 250])
Epochs: 392, AvgLoss: 6.2237
torch.Size([92807, 250])
Epochs: 393, AvgLoss: 6.2108
torch.Size([92807, 250])
Epochs: 394, AvgLoss: 6.1971
torch.Size([92807, 250])
Epochs: 395, AvgLoss: 6.1849
torch.Size([92807, 250])
Epochs: 396, AvgLoss: 6.1722
torch.Size([92807, 250])
Epochs: 397, AvgLoss: 6.1608
torch.Size([92807, 250])
Epochs: 398, AvgLoss: 6.1497
torch.Size([92807, 250])
Epochs: 399, AvgLoss: 6.1392
torch.Size([92807, 250])
Epochs: 400, AvgLoss: 6.1280
torch.Size([92807, 250])
Epochs: 401, AvgLoss: 6.1176
torch.Size([92807, 250])
Epochs: 402, AvgLoss: 6.1078
torch.Size([92807, 250])
Epochs: 403, AvgLoss: 6.0982
torch.Size([92807, 250])
Epochs: 404, AvgLoss: 6.0891
torch.Size([92807, 250])
Epochs: 405, AvgLoss: 6.0806
torch.Size([92807, 250])
Epochs: 406, AvgLoss: 6.0715
torch.Size([92807, 250])
Epochs: 407, AvgLoss: 6.0633
torch.Size([92807, 250])
Epochs: 408, AvgLoss: 6.0552
torch.Size([92807, 250])
Epochs: 409, AvgLoss: 6.0472
torch.Size([92807, 250])
Epochs: 410, AvgLoss: 6.0396
torch.Size([92807, 250])
Epochs: 411, AvgLoss: 6.0325
torch.Size([92807, 250])
Epochs: 412, AvgLoss: 6.0246
torch.Size([92807, 250])
Epochs: 413, AvgLoss: 6.0176
torch.Size([92807, 250])
Epochs: 414, AvgLoss: 6.0115
torch.Size([92807, 250])
Epochs: 415, AvgLoss: 6.0046
torch.Size([92807, 250])
Epochs: 416, AvgLoss: 5.9976
torch.Size([92807, 250])
Epochs: 417, AvgLoss: 5.9909
torch.Size([92807, 250])
Epochs: 418, AvgLoss: 5.9844
torch.Size([92807, 250])
Epochs: 419, AvgLoss: 5.9780
torch.Size([92807, 250])
Epochs: 420, AvgLoss: 5.9732
torch.Size([92807, 250])
Epochs: 421, AvgLoss: 5.9660
torch.Size([92807, 250])
Epochs: 422, AvgLoss: 5.9607
torch.Size([92807, 250])
Epochs: 423, AvgLoss: 5.9552
torch.Size([92807, 250])
Epochs: 424, AvgLoss: 5.9498
torch.Size([92807, 250])
Epochs: 425, AvgLoss: 5.9442
torch.Size([92807, 250])
Epochs: 426, AvgLoss: 5.9388
torch.Size([92807, 250])
Epochs: 427, AvgLoss: 5.9334
torch.Size([92807, 250])
Epochs: 428, AvgLoss: 5.9283
torch.Size([92807, 250])
Epochs: 429, AvgLoss: 5.9230
torch.Size([92807, 250])
Epochs: 430, AvgLoss: 5.9177
torch.Size([92807, 250])
Epochs: 431, AvgLoss: 5.9129
torch.Size([92807, 250])
Epochs: 432, AvgLoss: 5.9080
torch.Size([92807, 250])
Epochs: 433, AvgLoss: 5.9037
torch.Size([92807, 250])
Epochs: 434, AvgLoss: 5.8990
torch.Size([92807, 250])
Epochs: 435, AvgLoss: 5.8943
torch.Size([92807, 250])
Epochs: 436, AvgLoss: 5.8891
torch.Size([92807, 250])
Epochs: 437, AvgLoss: 5.8852
torch.Size([92807, 250])
Epochs: 438, AvgLoss: 5.8798
torch.Size([92807, 250])
Epochs: 439, AvgLoss: 5.8754
torch.Size([92807, 250])
Epochs: 440, AvgLoss: 5.8720
torch.Size([92807, 250])
Epochs: 441, AvgLoss: 5.8673
torch.Size([92807, 250])
Epochs: 442, AvgLoss: 5.8627
torch.Size([92807, 250])
Epochs: 443, AvgLoss: 5.8584
torch.Size([92807, 250])
Epochs: 444, AvgLoss: 5.8541
torch.Size([92807, 250])
Epochs: 445, AvgLoss: 5.8497
torch.Size([92807, 250])
Epochs: 446, AvgLoss: 5.8457
torch.Size([92807, 250])
Epochs: 447, AvgLoss: 5.8420
torch.Size([92807, 250])
Epochs: 448, AvgLoss: 5.8375
torch.Size([92807, 250])
Epochs: 449, AvgLoss: 5.8334
torch.Size([92807, 250])
Epochs: 450, AvgLoss: 5.8300
torch.Size([92807, 250])
Epochs: 451, AvgLoss: 5.8255
torch.Size([92807, 250])
Epochs: 452, AvgLoss: 5.8218
torch.Size([92807, 250])
Epochs: 453, AvgLoss: 5.8182
torch.Size([92807, 250])
Epochs: 454, AvgLoss: 5.8141
torch.Size([92807, 250])
Epochs: 455, AvgLoss: 5.8107
torch.Size([92807, 250])
Epochs: 456, AvgLoss: 5.8064
torch.Size([92807, 250])
Epochs: 457, AvgLoss: 5.8032
torch.Size([92807, 250])
Epochs: 458, AvgLoss: 5.7997
torch.Size([92807, 250])
Epochs: 459, AvgLoss: 5.7959
torch.Size([92807, 250])
Epochs: 460, AvgLoss: 5.7925
torch.Size([92807, 250])
Epochs: 461, AvgLoss: 5.7893
torch.Size([92807, 250])
Epochs: 462, AvgLoss: 5.7850
torch.Size([92807, 250])
Epochs: 463, AvgLoss: 5.7819
torch.Size([92807, 250])
Epochs: 464, AvgLoss: 5.7779
torch.Size([92807, 250])
Epochs: 465, AvgLoss: 5.7750
torch.Size([92807, 250])
Epochs: 466, AvgLoss: 5.7712
torch.Size([92807, 250])
Epochs: 467, AvgLoss: 5.7679
torch.Size([92807, 250])
Epochs: 468, AvgLoss: 5.7648
torch.Size([92807, 250])
Epochs: 469, AvgLoss: 5.7606
torch.Size([92807, 250])
Epochs: 470, AvgLoss: 5.7575
torch.Size([92807, 250])
Epochs: 471, AvgLoss: 5.7544
torch.Size([92807, 250])
Epochs: 472, AvgLoss: 5.7506
torch.Size([92807, 250])
Epochs: 473, AvgLoss: 5.7473
torch.Size([92807, 250])
Epochs: 474, AvgLoss: 5.7440
torch.Size([92807, 250])
Epochs: 475, AvgLoss: 5.7407
torch.Size([92807, 250])
Epochs: 476, AvgLoss: 5.7368
torch.Size([92807, 250])
Epochs: 477, AvgLoss: 5.7339
torch.Size([92807, 250])
Epochs: 478, AvgLoss: 5.7301
torch.Size([92807, 250])
Epochs: 479, AvgLoss: 5.7268
torch.Size([92807, 250])
Epochs: 480, AvgLoss: 5.7238
torch.Size([92807, 250])
Epochs: 481, AvgLoss: 5.7202
torch.Size([92807, 250])
Epochs: 482, AvgLoss: 5.7166
torch.Size([92807, 250])
Epochs: 483, AvgLoss: 5.7132
torch.Size([92807, 250])
Epochs: 484, AvgLoss: 5.7097
torch.Size([92807, 250])
Epochs: 485, AvgLoss: 5.7060
torch.Size([92807, 250])
Epochs: 486, AvgLoss: 5.7030
torch.Size([92807, 250])
Epochs: 487, AvgLoss: 5.6993
torch.Size([92807, 250])
Epochs: 488, AvgLoss: 5.6961
torch.Size([92807, 250])
Epochs: 489, AvgLoss: 5.6929
torch.Size([92807, 250])
Epochs: 490, AvgLoss: 5.6895
torch.Size([92807, 250])
Epochs: 491, AvgLoss: 5.6859
torch.Size([92807, 250])
Epochs: 492, AvgLoss: 5.6825
torch.Size([92807, 250])
Epochs: 493, AvgLoss: 5.6798
torch.Size([92807, 250])
Epochs: 494, AvgLoss: 5.6761
torch.Size([92807, 250])
Epochs: 495, AvgLoss: 5.6727
torch.Size([92807, 250])
Epochs: 496, AvgLoss: 5.6692
torch.Size([92807, 250])
Epochs: 497, AvgLoss: 5.6658
torch.Size([92807, 250])
Epochs: 498, AvgLoss: 5.6625
torch.Size([92807, 250])
Epochs: 499, AvgLoss: 5.6590
torch.Size([92807, 250])
Epochs: 500, AvgLoss: 5.6557
torch.Size([92807, 250])
Epochs: 501, AvgLoss: 5.6525
torch.Size([92807, 250])
Epochs: 502, AvgLoss: 5.6489
torch.Size([92807, 250])
Epochs: 503, AvgLoss: 5.6455
torch.Size([92807, 250])
Epochs: 504, AvgLoss: 5.6421
torch.Size([92807, 250])
Epochs: 505, AvgLoss: 5.6386
torch.Size([92807, 250])
Epochs: 506, AvgLoss: 5.6352
torch.Size([92807, 250])
Epochs: 507, AvgLoss: 5.6315
torch.Size([92807, 250])
Epochs: 508, AvgLoss: 5.6281
torch.Size([92807, 250])
Epochs: 509, AvgLoss: 5.6247
torch.Size([92807, 250])
Epochs: 510, AvgLoss: 5.6215
torch.Size([92807, 250])
Epochs: 511, AvgLoss: 5.6181
torch.Size([92807, 250])
Epochs: 512, AvgLoss: 5.6143
torch.Size([92807, 250])
Epochs: 513, AvgLoss: 5.6107
torch.Size([92807, 250])
Epochs: 514, AvgLoss: 5.6073
torch.Size([92807, 250])
Epochs: 515, AvgLoss: 5.6041
torch.Size([92807, 250])
Epochs: 516, AvgLoss: 5.6003
torch.Size([92807, 250])
Epochs: 517, AvgLoss: 5.5969
torch.Size([92807, 250])
Epochs: 518, AvgLoss: 5.5935
torch.Size([92807, 250])
Epochs: 519, AvgLoss: 5.5902
torch.Size([92807, 250])
Epochs: 520, AvgLoss: 5.5867
torch.Size([92807, 250])
Epochs: 521, AvgLoss: 5.5833
torch.Size([92807, 250])
Epochs: 522, AvgLoss: 5.5798
torch.Size([92807, 250])
Epochs: 523, AvgLoss: 5.5766
torch.Size([92807, 250])
Epochs: 524, AvgLoss: 5.5729
torch.Size([92807, 250])
Epochs: 525, AvgLoss: 5.5696
torch.Size([92807, 250])
Epochs: 526, AvgLoss: 5.5663
torch.Size([92807, 250])
Epochs: 527, AvgLoss: 5.5630
torch.Size([92807, 250])
Epochs: 528, AvgLoss: 5.5594
torch.Size([92807, 250])
Epochs: 529, AvgLoss: 5.5564
torch.Size([92807, 250])
Epochs: 530, AvgLoss: 5.5533
torch.Size([92807, 250])
Epochs: 531, AvgLoss: 5.5499
torch.Size([92807, 250])
Epochs: 532, AvgLoss: 5.5465
torch.Size([92807, 250])
Epochs: 533, AvgLoss: 5.5435
torch.Size([92807, 250])
Epochs: 534, AvgLoss: 5.5399
torch.Size([92807, 250])
Epochs: 535, AvgLoss: 5.5370
torch.Size([92807, 250])
Epochs: 536, AvgLoss: 5.5338
torch.Size([92807, 250])
Epochs: 537, AvgLoss: 5.5306
torch.Size([92807, 250])
Epochs: 538, AvgLoss: 5.5274
torch.Size([92807, 250])
Epochs: 539, AvgLoss: 5.5241
torch.Size([92807, 250])
Epochs: 540, AvgLoss: 5.5214
torch.Size([92807, 250])
Epochs: 541, AvgLoss: 5.5181
torch.Size([92807, 250])
Epochs: 542, AvgLoss: 5.5151
torch.Size([92807, 250])
Epochs: 543, AvgLoss: 5.5120
torch.Size([92807, 250])
Epochs: 544, AvgLoss: 5.5091
torch.Size([92807, 250])
Epochs: 545, AvgLoss: 5.5060
torch.Size([92807, 250])
Epochs: 546, AvgLoss: 5.5029
torch.Size([92807, 250])
Epochs: 547, AvgLoss: 5.4999
torch.Size([92807, 250])
Epochs: 548, AvgLoss: 5.4968
torch.Size([92807, 250])
Epochs: 549, AvgLoss: 5.4936
torch.Size([92807, 250])
Epochs: 550, AvgLoss: 5.4910
torch.Size([92807, 250])
Epochs: 551, AvgLoss: 5.4881
torch.Size([92807, 250])
Epochs: 552, AvgLoss: 5.4852
torch.Size([92807, 250])
Epochs: 553, AvgLoss: 5.4825
torch.Size([92807, 250])
Epochs: 554, AvgLoss: 5.4793
torch.Size([92807, 250])
Epochs: 555, AvgLoss: 5.4763
torch.Size([92807, 250])
Epochs: 556, AvgLoss: 5.4732
torch.Size([92807, 250])
Epochs: 557, AvgLoss: 5.4705
torch.Size([92807, 250])
Epochs: 558, AvgLoss: 5.4672
torch.Size([92807, 250])
Epochs: 559, AvgLoss: 5.4648
torch.Size([92807, 250])
Epochs: 560, AvgLoss: 5.4614
torch.Size([92807, 250])
Epochs: 561, AvgLoss: 5.4589
torch.Size([92807, 250])
Epochs: 562, AvgLoss: 5.4560
torch.Size([92807, 250])
Epochs: 563, AvgLoss: 5.4532
torch.Size([92807, 250])
Epochs: 564, AvgLoss: 5.4503
torch.Size([92807, 250])
Epochs: 565, AvgLoss: 5.4473
torch.Size([92807, 250])
Epochs: 566, AvgLoss: 5.4441
torch.Size([92807, 250])
Epochs: 567, AvgLoss: 5.4412
torch.Size([92807, 250])
Epochs: 568, AvgLoss: 5.4385
torch.Size([92807, 250])
Epochs: 569, AvgLoss: 5.4355
torch.Size([92807, 250])
Epochs: 570, AvgLoss: 5.4323
torch.Size([92807, 250])
Epochs: 571, AvgLoss: 5.4297
torch.Size([92807, 250])
Epochs: 572, AvgLoss: 5.4264
torch.Size([92807, 250])
Epochs: 573, AvgLoss: 5.4236
torch.Size([92807, 250])
Epochs: 574, AvgLoss: 5.4206
torch.Size([92807, 250])
Epochs: 575, AvgLoss: 5.4175
torch.Size([92807, 250])
Epochs: 576, AvgLoss: 5.4142
torch.Size([92807, 250])
Epochs: 577, AvgLoss: 5.4112
torch.Size([92807, 250])
Epochs: 578, AvgLoss: 5.4080
torch.Size([92807, 250])
Epochs: 579, AvgLoss: 5.4050
torch.Size([92807, 250])
Epochs: 580, AvgLoss: 5.4018
torch.Size([92807, 250])
Epochs: 581, AvgLoss: 5.3989
torch.Size([92807, 250])
Epochs: 582, AvgLoss: 5.3957
torch.Size([92807, 250])
Epochs: 583, AvgLoss: 5.3925
torch.Size([92807, 250])
Epochs: 584, AvgLoss: 5.3892
torch.Size([92807, 250])
Epochs: 585, AvgLoss: 5.3864
torch.Size([92807, 250])
Epochs: 586, AvgLoss: 5.3830
torch.Size([92807, 250])
Epochs: 587, AvgLoss: 5.3797
torch.Size([92807, 250])
Epochs: 588, AvgLoss: 5.3766
torch.Size([92807, 250])
Epochs: 589, AvgLoss: 5.3732
torch.Size([92807, 250])
Epochs: 590, AvgLoss: 5.3698
torch.Size([92807, 250])
Epochs: 591, AvgLoss: 5.3664
torch.Size([92807, 250])
Epochs: 592, AvgLoss: 5.3629
torch.Size([92807, 250])
Epochs: 593, AvgLoss: 5.3595
torch.Size([92807, 250])
Epochs: 594, AvgLoss: 5.3562
torch.Size([92807, 250])
Epochs: 595, AvgLoss: 5.3526
torch.Size([92807, 250])
Epochs: 596, AvgLoss: 5.3490
torch.Size([92807, 250])
Epochs: 597, AvgLoss: 5.3454
torch.Size([92807, 250])
Epochs: 598, AvgLoss: 5.3416
torch.Size([92807, 250])
Epochs: 599, AvgLoss: 5.3378
torch.Size([92807, 250])
Epochs: 600, AvgLoss: 5.3341
torch.Size([92807, 250])
Epochs: 601, AvgLoss: 5.3302
torch.Size([92807, 250])
Epochs: 602, AvgLoss: 5.3262
torch.Size([92807, 250])
Epochs: 603, AvgLoss: 5.3219
torch.Size([92807, 250])
Epochs: 604, AvgLoss: 5.3179
torch.Size([92807, 250])
Epochs: 605, AvgLoss: 5.3138
torch.Size([92807, 250])
Epochs: 606, AvgLoss: 5.3095
torch.Size([92807, 250])
Epochs: 607, AvgLoss: 5.3046
torch.Size([92807, 250])
Epochs: 608, AvgLoss: 5.3004
torch.Size([92807, 250])
Epochs: 609, AvgLoss: 5.2956
torch.Size([92807, 250])
Epochs: 610, AvgLoss: 5.2904
torch.Size([92807, 250])
Epochs: 611, AvgLoss: 5.2855
torch.Size([92807, 250])
Epochs: 612, AvgLoss: 5.2802
torch.Size([92807, 250])
Epochs: 613, AvgLoss: 5.2745
torch.Size([92807, 250])
Epochs: 614, AvgLoss: 5.2687
torch.Size([92807, 250])
Epochs: 615, AvgLoss: 5.2629
torch.Size([92807, 250])
Epochs: 616, AvgLoss: 5.2563
torch.Size([92807, 250])
Epochs: 617, AvgLoss: 5.2496
torch.Size([92807, 250])
Epochs: 618, AvgLoss: 5.2430
torch.Size([92807, 250])
Epochs: 619, AvgLoss: 5.2358
torch.Size([92807, 250])
Epochs: 620, AvgLoss: 5.2285
torch.Size([92807, 250])
Epochs: 621, AvgLoss: 5.2214
torch.Size([92807, 250])
Epochs: 622, AvgLoss: 5.2139
torch.Size([92807, 250])
Epochs: 623, AvgLoss: 5.2072
torch.Size([92807, 250])
Epochs: 624, AvgLoss: 5.2001
torch.Size([92807, 250])
Epochs: 625, AvgLoss: 5.1937
torch.Size([92807, 250])
Epochs: 626, AvgLoss: 5.1873
torch.Size([92807, 250])
Epochs: 627, AvgLoss: 5.1811
torch.Size([92807, 250])
Epochs: 628, AvgLoss: 5.1751
torch.Size([92807, 250])
Epochs: 629, AvgLoss: 5.1689
torch.Size([92807, 250])
Epochs: 630, AvgLoss: 5.1623
torch.Size([92807, 250])
Epochs: 631, AvgLoss: 5.1564
torch.Size([92807, 250])
Epochs: 632, AvgLoss: 5.1504
torch.Size([92807, 250])
Epochs: 633, AvgLoss: 5.1448
torch.Size([92807, 250])
Epochs: 634, AvgLoss: 5.1397
torch.Size([92807, 250])
Epochs: 635, AvgLoss: 5.1347
torch.Size([92807, 250])
Epochs: 636, AvgLoss: 5.1301
torch.Size([92807, 250])
Epochs: 637, AvgLoss: 5.1258
torch.Size([92807, 250])
Epochs: 638, AvgLoss: 5.1215
torch.Size([92807, 250])
Epochs: 639, AvgLoss: 5.1174
torch.Size([92807, 250])
Epochs: 640, AvgLoss: 5.1134
torch.Size([92807, 250])
Epochs: 641, AvgLoss: 5.1097
torch.Size([92807, 250])
Epochs: 642, AvgLoss: 5.1059
torch.Size([92807, 250])
Epochs: 643, AvgLoss: 5.1022
torch.Size([92807, 250])
Epochs: 644, AvgLoss: 5.0988
torch.Size([92807, 250])
Epochs: 645, AvgLoss: 5.0953
torch.Size([92807, 250])
Epochs: 646, AvgLoss: 5.0922
torch.Size([92807, 250])
Epochs: 647, AvgLoss: 5.0896
torch.Size([92807, 250])
Epochs: 648, AvgLoss: 5.0865
torch.Size([92807, 250])
Epochs: 649, AvgLoss: 5.0838
torch.Size([92807, 250])
Epochs: 650, AvgLoss: 5.0809
torch.Size([92807, 250])
Epochs: 651, AvgLoss: 5.0782
torch.Size([92807, 250])
Epochs: 652, AvgLoss: 5.0754
torch.Size([92807, 250])
Epochs: 653, AvgLoss: 5.0728
torch.Size([92807, 250])
Epochs: 654, AvgLoss: 5.0702
torch.Size([92807, 250])
Epochs: 655, AvgLoss: 5.0675
torch.Size([92807, 250])
Epochs: 656, AvgLoss: 5.0653
torch.Size([92807, 250])
Epochs: 657, AvgLoss: 5.0627
torch.Size([92807, 250])
Epochs: 658, AvgLoss: 5.0604
torch.Size([92807, 250])
Epochs: 659, AvgLoss: 5.0578
torch.Size([92807, 250])
Epochs: 660, AvgLoss: 5.0554
torch.Size([92807, 250])
Epochs: 661, AvgLoss: 5.0532
torch.Size([92807, 250])
Epochs: 662, AvgLoss: 5.0511
torch.Size([92807, 250])
Epochs: 663, AvgLoss: 5.0490
torch.Size([92807, 250])
Epochs: 664, AvgLoss: 5.0466
torch.Size([92807, 250])
Epochs: 665, AvgLoss: 5.0445
torch.Size([92807, 250])
Epochs: 666, AvgLoss: 5.0422
torch.Size([92807, 250])
Epochs: 667, AvgLoss: 5.0402
torch.Size([92807, 250])
Epochs: 668, AvgLoss: 5.0379
torch.Size([92807, 250])
Epochs: 669, AvgLoss: 5.0358
torch.Size([92807, 250])
Epochs: 670, AvgLoss: 5.0339
torch.Size([92807, 250])
Epochs: 671, AvgLoss: 5.0320
torch.Size([92807, 250])
Epochs: 672, AvgLoss: 5.0299
torch.Size([92807, 250])
Epochs: 673, AvgLoss: 5.0278
torch.Size([92807, 250])
Epochs: 674, AvgLoss: 5.0260
torch.Size([92807, 250])
Epochs: 675, AvgLoss: 5.0241
torch.Size([92807, 250])
Epochs: 676, AvgLoss: 5.0222
torch.Size([92807, 250])
Epochs: 677, AvgLoss: 5.0201
torch.Size([92807, 250])
Epochs: 678, AvgLoss: 5.0181
torch.Size([92807, 250])
Epochs: 679, AvgLoss: 5.0163
torch.Size([92807, 250])
Epochs: 680, AvgLoss: 5.0145
torch.Size([92807, 250])
Epochs: 681, AvgLoss: 5.0126
torch.Size([92807, 250])
Epochs: 682, AvgLoss: 5.0107
torch.Size([92807, 250])
Epochs: 683, AvgLoss: 5.0089
torch.Size([92807, 250])
Epochs: 684, AvgLoss: 5.0070
torch.Size([92807, 250])
Epochs: 685, AvgLoss: 5.0051
torch.Size([92807, 250])
Epochs: 686, AvgLoss: 5.0035
torch.Size([92807, 250])
Epochs: 687, AvgLoss: 5.0016
torch.Size([92807, 250])
Epochs: 688, AvgLoss: 4.9997
torch.Size([92807, 250])
Epochs: 689, AvgLoss: 4.9979
torch.Size([92807, 250])
Epochs: 690, AvgLoss: 4.9961
torch.Size([92807, 250])
Epochs: 691, AvgLoss: 4.9945
torch.Size([92807, 250])
Epochs: 692, AvgLoss: 4.9926
torch.Size([92807, 250])
Epochs: 693, AvgLoss: 4.9909
torch.Size([92807, 250])
Epochs: 694, AvgLoss: 4.9891
torch.Size([92807, 250])
Epochs: 695, AvgLoss: 4.9874
torch.Size([92807, 250])
Epochs: 696, AvgLoss: 4.9855
torch.Size([92807, 250])
Epochs: 697, AvgLoss: 4.9840
torch.Size([92807, 250])
Epochs: 698, AvgLoss: 4.9822
torch.Size([92807, 250])
Epochs: 699, AvgLoss: 4.9804
torch.Size([92807, 250])
Epochs: 700, AvgLoss: 4.9787
torch.Size([92807, 250])
Epochs: 701, AvgLoss: 4.9770
torch.Size([92807, 250])
Epochs: 702, AvgLoss: 4.9752
torch.Size([92807, 250])
Epochs: 703, AvgLoss: 4.9738
torch.Size([92807, 250])
Epochs: 704, AvgLoss: 4.9718
torch.Size([92807, 250])
Epochs: 705, AvgLoss: 4.9702
torch.Size([92807, 250])
Epochs: 706, AvgLoss: 4.9685
torch.Size([92807, 250])
Epochs: 707, AvgLoss: 4.9669
torch.Size([92807, 250])
Epochs: 708, AvgLoss: 4.9652
torch.Size([92807, 250])
Epochs: 709, AvgLoss: 4.9634
torch.Size([92807, 250])
Epochs: 710, AvgLoss: 4.9617
torch.Size([92807, 250])
Epochs: 711, AvgLoss: 4.9600
torch.Size([92807, 250])
Epochs: 712, AvgLoss: 4.9583
torch.Size([92807, 250])
Epochs: 713, AvgLoss: 4.9568
torch.Size([92807, 250])
Epochs: 714, AvgLoss: 4.9551
torch.Size([92807, 250])
Epochs: 715, AvgLoss: 4.9533
torch.Size([92807, 250])
Epochs: 716, AvgLoss: 4.9518
torch.Size([92807, 250])
Epochs: 717, AvgLoss: 4.9501
torch.Size([92807, 250])
Epochs: 718, AvgLoss: 4.9485
torch.Size([92807, 250])
Epochs: 719, AvgLoss: 4.9469
torch.Size([92807, 250])
Epochs: 720, AvgLoss: 4.9450
torch.Size([92807, 250])
Epochs: 721, AvgLoss: 4.9435
torch.Size([92807, 250])
Epochs: 722, AvgLoss: 4.9420
torch.Size([92807, 250])
Epochs: 723, AvgLoss: 4.9401
torch.Size([92807, 250])
Epochs: 724, AvgLoss: 4.9384
torch.Size([92807, 250])
Epochs: 725, AvgLoss: 4.9369
torch.Size([92807, 250])
Epochs: 726, AvgLoss: 4.9351
torch.Size([92807, 250])
Epochs: 727, AvgLoss: 4.9337
torch.Size([92807, 250])
Epochs: 728, AvgLoss: 4.9320
torch.Size([92807, 250])
Epochs: 729, AvgLoss: 4.9303
torch.Size([92807, 250])
Epochs: 730, AvgLoss: 4.9286
torch.Size([92807, 250])
Epochs: 731, AvgLoss: 4.9270
torch.Size([92807, 250])
Epochs: 732, AvgLoss: 4.9253
torch.Size([92807, 250])
Epochs: 733, AvgLoss: 4.9236
torch.Size([92807, 250])
Epochs: 734, AvgLoss: 4.9220
torch.Size([92807, 250])
Epochs: 735, AvgLoss: 4.9204
torch.Size([92807, 250])
Epochs: 736, AvgLoss: 4.9187
torch.Size([92807, 250])
Epochs: 737, AvgLoss: 4.9169
torch.Size([92807, 250])
Epochs: 738, AvgLoss: 4.9152
torch.Size([92807, 250])
Epochs: 739, AvgLoss: 4.9134
torch.Size([92807, 250])
Epochs: 740, AvgLoss: 4.9118
torch.Size([92807, 250])
Epochs: 741, AvgLoss: 4.9101
torch.Size([92807, 250])
Epochs: 742, AvgLoss: 4.9083
torch.Size([92807, 250])
Epochs: 743, AvgLoss: 4.9065
torch.Size([92807, 250])
Epochs: 744, AvgLoss: 4.9048
torch.Size([92807, 250])
Epochs: 745, AvgLoss: 4.9029
torch.Size([92807, 250])
Epochs: 746, AvgLoss: 4.9012
torch.Size([92807, 250])
Epochs: 747, AvgLoss: 4.8997
torch.Size([92807, 250])
Epochs: 748, AvgLoss: 4.8979
torch.Size([92807, 250])
Epochs: 749, AvgLoss: 4.8960
torch.Size([92807, 250])
Epochs: 750, AvgLoss: 4.8943
torch.Size([92807, 250])
Epochs: 751, AvgLoss: 4.8927
torch.Size([92807, 250])
Epochs: 752, AvgLoss: 4.8909
torch.Size([92807, 250])
Epochs: 753, AvgLoss: 4.8893
torch.Size([92807, 250])
Epochs: 754, AvgLoss: 4.8876
torch.Size([92807, 250])
Epochs: 755, AvgLoss: 4.8861
torch.Size([92807, 250])
Epochs: 756, AvgLoss: 4.8844
torch.Size([92807, 250])
Epochs: 757, AvgLoss: 4.8829
torch.Size([92807, 250])
Epochs: 758, AvgLoss: 4.8811
torch.Size([92807, 250])
Epochs: 759, AvgLoss: 4.8796
torch.Size([92807, 250])
Epochs: 760, AvgLoss: 4.8779
torch.Size([92807, 250])
Epochs: 761, AvgLoss: 4.8765
torch.Size([92807, 250])
Epochs: 762, AvgLoss: 4.8749
torch.Size([92807, 250])
Epochs: 763, AvgLoss: 4.8735
torch.Size([92807, 250])
Epochs: 764, AvgLoss: 4.8719
torch.Size([92807, 250])
Epochs: 765, AvgLoss: 4.8702
torch.Size([92807, 250])
Epochs: 766, AvgLoss: 4.8686
torch.Size([92807, 250])
Epochs: 767, AvgLoss: 4.8672
torch.Size([92807, 250])
Epochs: 768, AvgLoss: 4.8659
torch.Size([92807, 250])
Epochs: 769, AvgLoss: 4.8642
torch.Size([92807, 250])
Epochs: 770, AvgLoss: 4.8625
torch.Size([92807, 250])
Epochs: 771, AvgLoss: 4.8611
torch.Size([92807, 250])
Epochs: 772, AvgLoss: 4.8597
torch.Size([92807, 250])
Epochs: 773, AvgLoss: 4.8583
torch.Size([92807, 250])
Epochs: 774, AvgLoss: 4.8567
torch.Size([92807, 250])
Epochs: 775, AvgLoss: 4.8552
torch.Size([92807, 250])
Epochs: 776, AvgLoss: 4.8537
torch.Size([92807, 250])
Epochs: 777, AvgLoss: 4.8523
torch.Size([92807, 250])
Epochs: 778, AvgLoss: 4.8508
torch.Size([92807, 250])
Epochs: 779, AvgLoss: 4.8494
torch.Size([92807, 250])
Epochs: 780, AvgLoss: 4.8480
torch.Size([92807, 250])
Epochs: 781, AvgLoss: 4.8463
torch.Size([92807, 250])
Epochs: 782, AvgLoss: 4.8450
torch.Size([92807, 250])
Epochs: 783, AvgLoss: 4.8435
torch.Size([92807, 250])
Epochs: 784, AvgLoss: 4.8420
torch.Size([92807, 250])
Epochs: 785, AvgLoss: 4.8406
torch.Size([92807, 250])
Epochs: 786, AvgLoss: 4.8392
torch.Size([92807, 250])
Epochs: 787, AvgLoss: 4.8377
torch.Size([92807, 250])
Epochs: 788, AvgLoss: 4.8363
torch.Size([92807, 250])
Epochs: 789, AvgLoss: 4.8349
torch.Size([92807, 250])
Epochs: 790, AvgLoss: 4.8333
torch.Size([92807, 250])
Epochs: 791, AvgLoss: 4.8322
torch.Size([92807, 250])
Epochs: 792, AvgLoss: 4.8307
torch.Size([92807, 250])
Epochs: 793, AvgLoss: 4.8293
torch.Size([92807, 250])
Epochs: 794, AvgLoss: 4.8277
torch.Size([92807, 250])
Epochs: 795, AvgLoss: 4.8264
torch.Size([92807, 250])
Epochs: 796, AvgLoss: 4.8250
torch.Size([92807, 250])
Epochs: 797, AvgLoss: 4.8236
torch.Size([92807, 250])
Epochs: 798, AvgLoss: 4.8221
torch.Size([92807, 250])
Epochs: 799, AvgLoss: 4.8207
torch.Size([92807, 250])
Epochs: 800, AvgLoss: 4.8191
torch.Size([92807, 250])
Epochs: 801, AvgLoss: 4.8178
torch.Size([92807, 250])
Epochs: 802, AvgLoss: 4.8166
torch.Size([92807, 250])
Epochs: 803, AvgLoss: 4.8152
torch.Size([92807, 250])
Epochs: 804, AvgLoss: 4.8138
torch.Size([92807, 250])
Epochs: 805, AvgLoss: 4.8124
torch.Size([92807, 250])
Epochs: 806, AvgLoss: 4.8111
torch.Size([92807, 250])
Epochs: 807, AvgLoss: 4.8097
torch.Size([92807, 250])
Epochs: 808, AvgLoss: 4.8082
torch.Size([92807, 250])
Epochs: 809, AvgLoss: 4.8067
torch.Size([92807, 250])
Epochs: 810, AvgLoss: 4.8054
torch.Size([92807, 250])
Epochs: 811, AvgLoss: 4.8041
torch.Size([92807, 250])
Epochs: 812, AvgLoss: 4.8027
torch.Size([92807, 250])
Epochs: 813, AvgLoss: 4.8015
torch.Size([92807, 250])
Epochs: 814, AvgLoss: 4.8000
torch.Size([92807, 250])
Epochs: 815, AvgLoss: 4.7984
torch.Size([92807, 250])
Epochs: 816, AvgLoss: 4.7971
torch.Size([92807, 250])
Epochs: 817, AvgLoss: 4.7957
torch.Size([92807, 250])
Epochs: 818, AvgLoss: 4.7946
torch.Size([92807, 250])
Epochs: 819, AvgLoss: 4.7930
torch.Size([92807, 250])
Epochs: 820, AvgLoss: 4.7915
torch.Size([92807, 250])
Epochs: 821, AvgLoss: 4.7903
torch.Size([92807, 250])
Epochs: 822, AvgLoss: 4.7889
torch.Size([92807, 250])
Epochs: 823, AvgLoss: 4.7874
torch.Size([92807, 250])
Epochs: 824, AvgLoss: 4.7861
torch.Size([92807, 250])
Epochs: 825, AvgLoss: 4.7847
torch.Size([92807, 250])
Epochs: 826, AvgLoss: 4.7835
torch.Size([92807, 250])
Epochs: 827, AvgLoss: 4.7822
torch.Size([92807, 250])
Epochs: 828, AvgLoss: 4.7804
torch.Size([92807, 250])
Epochs: 829, AvgLoss: 4.7793
torch.Size([92807, 250])
Epochs: 830, AvgLoss: 4.7777
torch.Size([92807, 250])
Epochs: 831, AvgLoss: 4.7764
torch.Size([92807, 250])
Epochs: 832, AvgLoss: 4.7752
torch.Size([92807, 250])
Epochs: 833, AvgLoss: 4.7737
torch.Size([92807, 250])
Epochs: 834, AvgLoss: 4.7724
torch.Size([92807, 250])
Epochs: 835, AvgLoss: 4.7710
torch.Size([92807, 250])
Epochs: 836, AvgLoss: 4.7700
torch.Size([92807, 250])
Epochs: 837, AvgLoss: 4.7683
torch.Size([92807, 250])
Epochs: 838, AvgLoss: 4.7670
torch.Size([92807, 250])
Epochs: 839, AvgLoss: 4.7655
torch.Size([92807, 250])
Epochs: 840, AvgLoss: 4.7642
torch.Size([92807, 250])
Epochs: 841, AvgLoss: 4.7629
torch.Size([92807, 250])
Epochs: 842, AvgLoss: 4.7616
torch.Size([92807, 250])
Epochs: 843, AvgLoss: 4.7600
torch.Size([92807, 250])
Epochs: 844, AvgLoss: 4.7587
torch.Size([92807, 250])
Epochs: 845, AvgLoss: 4.7574
torch.Size([92807, 250])
Epochs: 846, AvgLoss: 4.7561
torch.Size([92807, 250])
Epochs: 847, AvgLoss: 4.7547
torch.Size([92807, 250])
Epochs: 848, AvgLoss: 4.7534
torch.Size([92807, 250])
Epochs: 849, AvgLoss: 4.7520
torch.Size([92807, 250])
Epochs: 850, AvgLoss: 4.7507
torch.Size([92807, 250])
Epochs: 851, AvgLoss: 4.7491
torch.Size([92807, 250])
Epochs: 852, AvgLoss: 4.7478
torch.Size([92807, 250])
Epochs: 853, AvgLoss: 4.7465
torch.Size([92807, 250])
Epochs: 854, AvgLoss: 4.7452
torch.Size([92807, 250])
Epochs: 855, AvgLoss: 4.7437
torch.Size([92807, 250])
Epochs: 856, AvgLoss: 4.7424
torch.Size([92807, 250])
Epochs: 857, AvgLoss: 4.7410
torch.Size([92807, 250])
Epochs: 858, AvgLoss: 4.7395
torch.Size([92807, 250])
Epochs: 859, AvgLoss: 4.7383
torch.Size([92807, 250])
Epochs: 860, AvgLoss: 4.7367
torch.Size([92807, 250])
Epochs: 861, AvgLoss: 4.7355
torch.Size([92807, 250])
Epochs: 862, AvgLoss: 4.7342
torch.Size([92807, 250])
Epochs: 863, AvgLoss: 4.7328
torch.Size([92807, 250])
Epochs: 864, AvgLoss: 4.7313
torch.Size([92807, 250])
Epochs: 865, AvgLoss: 4.7299
torch.Size([92807, 250])
Epochs: 866, AvgLoss: 4.7286
torch.Size([92807, 250])
Epochs: 867, AvgLoss: 4.7272
torch.Size([92807, 250])
Epochs: 868, AvgLoss: 4.7256
torch.Size([92807, 250])
Epochs: 869, AvgLoss: 4.7243
torch.Size([92807, 250])
Epochs: 870, AvgLoss: 4.7229
torch.Size([92807, 250])
Epochs: 871, AvgLoss: 4.7215
torch.Size([92807, 250])
Epochs: 872, AvgLoss: 4.7201
torch.Size([92807, 250])
Epochs: 873, AvgLoss: 4.7185
torch.Size([92807, 250])
Epochs: 874, AvgLoss: 4.7174
torch.Size([92807, 250])
Epochs: 875, AvgLoss: 4.7157
torch.Size([92807, 250])
Epochs: 876, AvgLoss: 4.7143
torch.Size([92807, 250])
Epochs: 877, AvgLoss: 4.7130
torch.Size([92807, 250])
Epochs: 878, AvgLoss: 4.7115
torch.Size([92807, 250])
Epochs: 879, AvgLoss: 4.7101
torch.Size([92807, 250])
Epochs: 880, AvgLoss: 4.7088
torch.Size([92807, 250])
Epochs: 881, AvgLoss: 4.7072
torch.Size([92807, 250])
Epochs: 882, AvgLoss: 4.7059
torch.Size([92807, 250])
Epochs: 883, AvgLoss: 4.7045
torch.Size([92807, 250])
Epochs: 884, AvgLoss: 4.7030
torch.Size([92807, 250])
Epochs: 885, AvgLoss: 4.7017
torch.Size([92807, 250])
Epochs: 886, AvgLoss: 4.7001
torch.Size([92807, 250])
Epochs: 887, AvgLoss: 4.6987
torch.Size([92807, 250])
Epochs: 888, AvgLoss: 4.6971
torch.Size([92807, 250])
Epochs: 889, AvgLoss: 4.6957
torch.Size([92807, 250])
Epochs: 890, AvgLoss: 4.6942
torch.Size([92807, 250])
Epochs: 891, AvgLoss: 4.6927
torch.Size([92807, 250])
Epochs: 892, AvgLoss: 4.6913
torch.Size([92807, 250])
Epochs: 893, AvgLoss: 4.6900
torch.Size([92807, 250])
Epochs: 894, AvgLoss: 4.6885
torch.Size([92807, 250])
Epochs: 895, AvgLoss: 4.6870
torch.Size([92807, 250])
Epochs: 896, AvgLoss: 4.6857
torch.Size([92807, 250])
Epochs: 897, AvgLoss: 4.6841
torch.Size([92807, 250])
Epochs: 898, AvgLoss: 4.6827
torch.Size([92807, 250])
Epochs: 899, AvgLoss: 4.6811
torch.Size([92807, 250])
Epochs: 900, AvgLoss: 4.6795
torch.Size([92807, 250])
Epochs: 901, AvgLoss: 4.6783
torch.Size([92807, 250])
Epochs: 902, AvgLoss: 4.6768
torch.Size([92807, 250])
Epochs: 903, AvgLoss: 4.6753
torch.Size([92807, 250])
Epochs: 904, AvgLoss: 4.6739
torch.Size([92807, 250])
Epochs: 905, AvgLoss: 4.6724
torch.Size([92807, 250])
Epochs: 906, AvgLoss: 4.6709
torch.Size([92807, 250])
Epochs: 907, AvgLoss: 4.6697
torch.Size([92807, 250])
Epochs: 908, AvgLoss: 4.6680
torch.Size([92807, 250])
Epochs: 909, AvgLoss: 4.6667
torch.Size([92807, 250])
Epochs: 910, AvgLoss: 4.6653
torch.Size([92807, 250])
Epochs: 911, AvgLoss: 4.6640
torch.Size([92807, 250])
Epochs: 912, AvgLoss: 4.6624
torch.Size([92807, 250])
Epochs: 913, AvgLoss: 4.6610
torch.Size([92807, 250])
Epochs: 914, AvgLoss: 4.6596
torch.Size([92807, 250])
Epochs: 915, AvgLoss: 4.6579
torch.Size([92807, 250])
Epochs: 916, AvgLoss: 4.6566
torch.Size([92807, 250])
Epochs: 917, AvgLoss: 4.6553
torch.Size([92807, 250])
Epochs: 918, AvgLoss: 4.6538
torch.Size([92807, 250])
Epochs: 919, AvgLoss: 4.6524
torch.Size([92807, 250])
Epochs: 920, AvgLoss: 4.6510
torch.Size([92807, 250])
Epochs: 921, AvgLoss: 4.6495
torch.Size([92807, 250])
Epochs: 922, AvgLoss: 4.6483
torch.Size([92807, 250])
Epochs: 923, AvgLoss: 4.6468
torch.Size([92807, 250])
Epochs: 924, AvgLoss: 4.6454
torch.Size([92807, 250])
Epochs: 925, AvgLoss: 4.6441
torch.Size([92807, 250])
Epochs: 926, AvgLoss: 4.6425
torch.Size([92807, 250])
Epochs: 927, AvgLoss: 4.6412
torch.Size([92807, 250])
Epochs: 928, AvgLoss: 4.6398
torch.Size([92807, 250])
Epochs: 929, AvgLoss: 4.6386
torch.Size([92807, 250])
Epochs: 930, AvgLoss: 4.6372
torch.Size([92807, 250])
Epochs: 931, AvgLoss: 4.6357
torch.Size([92807, 250])
Epochs: 932, AvgLoss: 4.6345
torch.Size([92807, 250])
Epochs: 933, AvgLoss: 4.6329
torch.Size([92807, 250])
Epochs: 934, AvgLoss: 4.6317
torch.Size([92807, 250])
Epochs: 935, AvgLoss: 4.6304
torch.Size([92807, 250])
Epochs: 936, AvgLoss: 4.6288
torch.Size([92807, 250])
Epochs: 937, AvgLoss: 4.6275
torch.Size([92807, 250])
Epochs: 938, AvgLoss: 4.6261
torch.Size([92807, 250])
Epochs: 939, AvgLoss: 4.6247
torch.Size([92807, 250])
Epochs: 940, AvgLoss: 4.6233
torch.Size([92807, 250])
Epochs: 941, AvgLoss: 4.6218
torch.Size([92807, 250])
Epochs: 942, AvgLoss: 4.6205
torch.Size([92807, 250])
Epochs: 943, AvgLoss: 4.6191
torch.Size([92807, 250])
Epochs: 944, AvgLoss: 4.6178
torch.Size([92807, 250])
Epochs: 945, AvgLoss: 4.6163
torch.Size([92807, 250])
Epochs: 946, AvgLoss: 4.6151
torch.Size([92807, 250])
Epochs: 947, AvgLoss: 4.6137
torch.Size([92807, 250])
Epochs: 948, AvgLoss: 4.6124
torch.Size([92807, 250])
Epochs: 949, AvgLoss: 4.6112
torch.Size([92807, 250])
Epochs: 950, AvgLoss: 4.6098
torch.Size([92807, 250])
Epochs: 951, AvgLoss: 4.6082
torch.Size([92807, 250])
Epochs: 952, AvgLoss: 4.6070
torch.Size([92807, 250])
Epochs: 953, AvgLoss: 4.6057
torch.Size([92807, 250])
Epochs: 954, AvgLoss: 4.6043
torch.Size([92807, 250])
Epochs: 955, AvgLoss: 4.6030
torch.Size([92807, 250])
Epochs: 956, AvgLoss: 4.6016
torch.Size([92807, 250])
Epochs: 957, AvgLoss: 4.6002
torch.Size([92807, 250])
Epochs: 958, AvgLoss: 4.5987
torch.Size([92807, 250])
Epochs: 959, AvgLoss: 4.5976
torch.Size([92807, 250])
Epochs: 960, AvgLoss: 4.5961
torch.Size([92807, 250])
Epochs: 961, AvgLoss: 4.5948
torch.Size([92807, 250])
Epochs: 962, AvgLoss: 4.5934
torch.Size([92807, 250])
Epochs: 963, AvgLoss: 4.5921
torch.Size([92807, 250])
Epochs: 964, AvgLoss: 4.5906
torch.Size([92807, 250])
Epochs: 965, AvgLoss: 4.5892
torch.Size([92807, 250])
Epochs: 966, AvgLoss: 4.5880
torch.Size([92807, 250])
Epochs: 967, AvgLoss: 4.5865
torch.Size([92807, 250])
Epochs: 968, AvgLoss: 4.5849
torch.Size([92807, 250])
Epochs: 969, AvgLoss: 4.5837
torch.Size([92807, 250])
Epochs: 970, AvgLoss: 4.5823
torch.Size([92807, 250])
Epochs: 971, AvgLoss: 4.5809
torch.Size([92807, 250])
Epochs: 972, AvgLoss: 4.5796
torch.Size([92807, 250])
Epochs: 973, AvgLoss: 4.5781
torch.Size([92807, 250])
Epochs: 974, AvgLoss: 4.5767
torch.Size([92807, 250])
Epochs: 975, AvgLoss: 4.5752
torch.Size([92807, 250])
Epochs: 976, AvgLoss: 4.5738
torch.Size([92807, 250])
Epochs: 977, AvgLoss: 4.5724
torch.Size([92807, 250])
Epochs: 978, AvgLoss: 4.5710
torch.Size([92807, 250])
Epochs: 979, AvgLoss: 4.5695
torch.Size([92807, 250])
Epochs: 980, AvgLoss: 4.5682
torch.Size([92807, 250])
Epochs: 981, AvgLoss: 4.5666
torch.Size([92807, 250])
Epochs: 982, AvgLoss: 4.5652
torch.Size([92807, 250])
Epochs: 983, AvgLoss: 4.5637
torch.Size([92807, 250])
Epochs: 984, AvgLoss: 4.5623
torch.Size([92807, 250])
Epochs: 985, AvgLoss: 4.5609
torch.Size([92807, 250])
Epochs: 986, AvgLoss: 4.5595
torch.Size([92807, 250])
Epochs: 987, AvgLoss: 4.5580
torch.Size([92807, 250])
Epochs: 988, AvgLoss: 4.5567
torch.Size([92807, 250])
Epochs: 989, AvgLoss: 4.5552
torch.Size([92807, 250])
Epochs: 990, AvgLoss: 4.5538
torch.Size([92807, 250])
Epochs: 991, AvgLoss: 4.5523
torch.Size([92807, 250])
Epochs: 992, AvgLoss: 4.5508
torch.Size([92807, 250])
Epochs: 993, AvgLoss: 4.5494
torch.Size([92807, 250])
Epochs: 994, AvgLoss: 4.5481
torch.Size([92807, 250])
Epochs: 995, AvgLoss: 4.5466
torch.Size([92807, 250])
Epochs: 996, AvgLoss: 4.5452
torch.Size([92807, 250])
Epochs: 997, AvgLoss: 4.5438
torch.Size([92807, 250])
Epochs: 998, AvgLoss: 4.5423
torch.Size([92807, 250])
Epochs: 999, AvgLoss: 4.5409
torch.Size([92807, 250])
Epochs: 1000, AvgLoss: 4.5395
Training for VAE has been done.
Part 3: Showing result
In this phase, we visualize the outcomes of our analysis by displaying both the time and frequency information for each identified peak cluster. These visualizations help in interpreting the complex dynamics captured by the model and provide insights into the temporal progression and rhythmicity of peak expression within each cluster.
[5]:
show_result(peak,trajectory_info)
Only the first eligible cluster’s plots are ploted out. Other clusters’s plots are saved to pdf files.
Part 4: Differential frequency peaks
In this step, we perform a comparative analysis to identify peaks that show significant differences in frequency between arsenic and control. By statistically evaluating these differences, we can uncover frequency-specific regulatory elements or patterns that are condition-dependent. This analysis provides deeper biological insights into how rhythmicity may vary across different peaks, thereby highlighting potential functional elements involved in dynamic processes.
Firstly, we need to obtain the frequency information of control dataset.
[6]:
dataset = 'mouse_atherosclerotic_plaque_immune_cells_Control_ATAC'
mtx = './dataset/mouse_atherosclerotic_plaque_immune_cells_ATAC/Control/matrix.mtx'
barcode = './dataset/mouse_atherosclerotic_plaque_immune_cells_ATAC/Control/barcodes.csv'
peak = './dataset/mouse_atherosclerotic_plaque_immune_cells_ATAC/Control/features.csv'
subprocess.run(["Rscript", "trajectory_inference.R", dataset, mtx， barcode, peak])
trajectory_info = pd.read_csv("mouse_atherosclerotic_plaque_immune_cells_Control_ATAC.csv")
Then, run the differential frequency to obtain the differential frequency peaks.
[7]:
differential_frequency('mouse_atherosclerotic_plaque_immune_cells_Arsenic_ATAC.csv','mouse_atherosclerotic_plaque_immune_cells_Control_ATAC.csv','Arsenic','Control',peak)
