AntoineH/BU_Stoch_pool
1
0
Fork 0
mirror of https://github.com/AntoineHX/BU_Stoch_pool.git synced 2025-05-06 02:30:46 +02:00
Code Issues Projects Releases Packages Wiki Activity

Benchmark NoCeil

Browse source
This commit is contained in:
Antoine Harlé 2020-06-24 01:57:32 -07:00
parent 115a8d80b5
commit 019310e60c
83 changed files with 15790 additions and 25 deletions
Download patch file Download diff file Expand all files Collapse all files

12
jobs/benchmark.sh
View file

@ -4,7 +4,7 @@
#SBATCH --mem=32000M #32000M # Memory proportional to CPUs: 32000 Cedar, 64000 Graham.
#SBATCH --account=def-mpederso
#SBATCH --time=10:00:00
#SBATCH --job-name=Benchmark-MyLeNetMatNormalNoceil
#SBATCH --job-name=Benchmark-MyLeNetMatStochBUNoceil
#SBATCH --output=log/%x-%j.out
#SBATCH --mail-user=harle.collette.antoine@gmail.com
#SBATCH --mail-type=END
@ -20,10 +20,10 @@ source ~/virtual_env/stoch_pool/bin/activate
cd ../
time python main.py \
-n MyLeNetMatNormalNoceil \
-ep 100 \
-n MyLeNetMatStochBUNoceil \
-ep 50 \
-sc cosine \
-lr 5e-2 \
-rf '../res/benchmark_NoCeil/' \
-k 1 \
-pf __k1_$SLURM_ARRAY_TASK_ID
-rf 'res/benchmark_NoCeil/' \
-k 4 \
-pf __k4_$SLURM_ARRAY_TASK_ID

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44511714.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k1_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.32 / 46.57
Loss : 1.71 / 1.47
Time: 45.96672709006816
---------
Epoch: 5
Acc : 70.95 / 70.58
Loss : 0.83 / 0.86
Time: 275.3688823627308
---------
Epoch: 10
Acc : 79.10 / 74.53
Loss : 0.60 / 0.74
Time: 504.9038645885885
---------
Epoch: 15
Acc : 83.93 / 80.50
Loss : 0.46 / 0.57
Time: 734.5577887063846
---------
Epoch: 20
Acc : 87.32 / 80.67
Loss : 0.37 / 0.55
Time: 964.6286195870489
---------
Epoch: 25
Acc : 90.39 / 82.75
Loss : 0.29 / 0.50
Time: 1194.849686741829
---------
Epoch: 30
Acc : 93.07 / 82.63
Loss : 0.22 / 0.51
Time: 1426.0590728316456
---------
Epoch: 35
Acc : 95.51 / 82.89
Loss : 0.15 / 0.52
Time: 1657.5508981496096
---------
Epoch: 40
Acc : 97.45 / 84.48
Loss : 0.11 / 0.49
Time: 1887.298833809793
---------
Epoch: 45
Acc : 98.55 / 84.93
Loss : 0.08 / 0.49
Time: 2116.886452781968
---------
Best Acc : 85.09
Training time (min): 38.34314283429024
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k1_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k1_3 " saved !
real 38m33.682s
user 53m21.131s
sys 12m1.798s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44511729.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k2_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.12 / 45.47
Loss : 1.72 / 1.50
Time: 123.19412856083363
---------
Epoch: 5
Acc : 71.71 / 68.77
Loss : 0.81 / 0.91
Time: 739.1145824743435
---------
Epoch: 10
Acc : 79.86 / 77.88
Loss : 0.58 / 0.64
Time: 1355.644656511955
---------
Epoch: 15
Acc : 84.55 / 77.81
Loss : 0.45 / 0.64
Time: 1971.8144962312654
---------
Epoch: 20
Acc : 87.95 / 81.82
Loss : 0.35 / 0.54
Time: 2587.859218424186
---------
Epoch: 25
Acc : 91.09 / 80.44
Loss : 0.26 / 0.57
Time: 3204.394698444754
---------
Epoch: 30
Acc : 93.85 / 82.45
Loss : 0.19 / 0.54
Time: 3820.647819074802
---------
Epoch: 35
Acc : 96.26 / 83.40
Loss : 0.13 / 0.51
Time: 4437.080120751634
---------
Epoch: 40
Acc : 98.12 / 85.19
Loss : 0.09 / 0.49
Time: 5053.203467056155
---------
Epoch: 45
Acc : 99.15 / 84.79
Loss : 0.06 / 0.50
Time: 5669.440995325334
---------
Best Acc : 85.19
Training time (min): 102.70833032748972
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k2_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k2_3 " saved !
real 102m58.647s
user 99m45.063s
sys 32m18.510s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44511735.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k3_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.95 / 45.98
Loss : 1.72 / 1.51
Time: 247.98328754398972
---------
Epoch: 5
Acc : 71.70 / 72.71
Loss : 0.81 / 0.78
Time: 1488.6181818954647
---------
Epoch: 10
Acc : 80.03 / 75.58
Loss : 0.57 / 0.72
Time: 2729.465384825133
---------
Epoch: 15
Acc : 84.79 / 80.80
Loss : 0.44 / 0.57
Time: 3970.437553227879
---------
Epoch: 20
Acc : 88.45 / 81.73
Loss : 0.34 / 0.54
Time: 5211.3001955430955
---------
Epoch: 25
Acc : 91.57 / 82.44
Loss : 0.26 / 0.53
Time: 6452.203316219151
---------
Epoch: 30
Acc : 94.42 / 83.35
Loss : 0.18 / 0.51
Time: 7693.365694181062
---------
Epoch: 35
Acc : 96.76 / 83.32
Loss : 0.12 / 0.53
Time: 8934.607322337106
---------
Epoch: 40
Acc : 98.65 / 84.05
Loss : 0.07 / 0.52
Time: 10175.596159897745
---------
Epoch: 45
Acc : 99.41 / 84.61
Loss : 0.05 / 0.52
Time: 11416.773865584284
---------
Best Acc : 84.74
Training time (min): 206.82600436269615
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k3_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k3_3 " saved !
real 207m3.639s
user 164m26.994s
sys 70m29.354s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44511739.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k4_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 37.25 / 46.81
Loss : 1.71 / 1.45
Time: 412.9022887162864
---------
Epoch: 5
Acc : 72.74 / 72.93
Loss : 0.79 / 0.79
Time: 2478.4713746113703
---------
Epoch: 10
Acc : 80.47 / 78.57
Loss : 0.56 / 0.63
Time: 4544.841249100864
---------
Epoch: 15
Acc : 84.95 / 80.01
Loss : 0.43 / 0.59
Time: 6611.430219493806
---------
Epoch: 20
Acc : 88.65 / 81.20
Loss : 0.33 / 0.55
Time: 8678.567663419992
---------
Epoch: 25
Acc : 91.80 / 82.36
Loss : 0.25 / 0.52
Time: 10747.367348232307
---------
Epoch: 30
Acc : 94.57 / 82.16
Loss : 0.18 / 0.56
Time: 12816.555869823322
---------
Epoch: 35
Acc : 97.08 / 83.29
Loss : 0.11 / 0.54
Time: 14885.712881685235
---------
Epoch: 40
Acc : 98.70 / 84.38
Loss : 0.07 / 0.51
Time: 16954.642080742866
---------
Epoch: 45
Acc : 99.55 / 84.41
Loss : 0.05 / 0.52
Time: 19023.663426892832
---------
Best Acc : 84.63
Training time (min): 344.65982003790947
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k4_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k4_3 " saved !
real 344m54.291s
user 251m47.869s
sys 119m40.516s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512423.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k1_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 35.43 / 48.21
Loss : 1.73 / 1.43
Time: 46.12261764425784
---------
Epoch: 5
Acc : 70.44 / 69.29
Loss : 0.85 / 0.87
Time: 276.4114668201655
---------
Epoch: 10
Acc : 78.80 / 76.67
Loss : 0.61 / 0.68
Time: 506.5132551435381
---------
Epoch: 15
Acc : 83.43 / 80.02
Loss : 0.48 / 0.58
Time: 736.7161982152611
---------
Epoch: 20
Acc : 86.87 / 80.73
Loss : 0.38 / 0.55
Time: 966.8520798627287
---------
Epoch: 25
Acc : 90.07 / 83.00
Loss : 0.30 / 0.51
Time: 1197.1624066643417
---------
Epoch: 30
Acc : 92.45 / 82.59
Loss : 0.23 / 0.52
Time: 1427.6633465271443
---------
Epoch: 35
Acc : 94.93 / 83.86
Loss : 0.17 / 0.48
Time: 1658.0701657654718
---------
Epoch: 40
Acc : 96.97 / 84.58
Loss : 0.12 / 0.48
Time: 1888.4067949829623
---------
Epoch: 45
Acc : 98.12 / 84.77
Loss : 0.09 / 0.49
Time: 2118.8804153930396
---------
Best Acc : 84.89
Training time (min): 38.38393786516972
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k1_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k1_1 " saved !
real 38m36.671s
user 52m22.204s
sys 12m0.649s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512424.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k1_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.34 / 46.99
Loss : 1.72 / 1.43
Time: 45.97680170927197
---------
Epoch: 5
Acc : 70.85 / 70.73
Loss : 0.83 / 0.83
Time: 275.01454721949995
---------
Epoch: 10
Acc : 78.84 / 75.75
Loss : 0.61 / 0.70
Time: 504.2062480049208
---------
Epoch: 15
Acc : 83.54 / 79.54
Loss : 0.47 / 0.60
Time: 733.3581873718649
---------
Epoch: 20
Acc : 86.82 / 81.49
Loss : 0.38 / 0.54
Time: 962.6645513242111
---------
Epoch: 25
Acc : 89.89 / 81.93
Loss : 0.29 / 0.54
Time: 1191.833936353214
---------
Epoch: 30
Acc : 92.88 / 83.44
Loss : 0.22 / 0.49
Time: 1421.0683761807159
---------
Epoch: 35
Acc : 95.38 / 83.74
Loss : 0.16 / 0.50
Time: 1650.2434674538672
---------
Epoch: 40
Acc : 97.26 / 84.41
Loss : 0.11 / 0.50
Time: 1879.4910161253065
---------
Epoch: 45
Acc : 98.51 / 84.84
Loss : 0.08 / 0.49
Time: 2108.9041818892583
---------
Best Acc : 84.94
Training time (min): 38.20426822579466
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k1_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k1_2 " saved !
real 38m24.983s
user 54m3.179s
sys 11m47.645s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512425.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k2_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.99 / 46.90
Loss : 1.71 / 1.47
Time: 123.3850756874308
---------
Epoch: 5
Acc : 71.92 / 70.74
Loss : 0.81 / 0.84
Time: 740.4402446337044
---------
Epoch: 10
Acc : 79.88 / 76.11
Loss : 0.58 / 0.70
Time: 1357.5990120908245
---------
Epoch: 15
Acc : 84.43 / 80.63
Loss : 0.45 / 0.56
Time: 1975.4050657227635
---------
Epoch: 20
Acc : 88.04 / 80.34
Loss : 0.35 / 0.57
Time: 2593.1981341233477
---------
Epoch: 25
Acc : 91.30 / 82.08
Loss : 0.26 / 0.53
Time: 3211.109001107514
---------
Epoch: 30
Acc : 94.06 / 82.95
Loss : 0.19 / 0.51
Time: 3828.7535809641704
---------
Epoch: 35
Acc : 96.45 / 83.79
Loss : 0.13 / 0.50
Time: 4446.439873588271
---------
Epoch: 40
Acc : 98.28 / 84.44
Loss : 0.08 / 0.51
Time: 5064.219255410135
---------
Epoch: 45
Acc : 99.23 / 84.78
Loss : 0.06 / 0.51
Time: 5681.4778324710205
---------
Best Acc : 84.92
Training time (min): 102.92114744419231
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k2_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k2_1 " saved !
real 103m8.605s
user 98m5.937s
sys 33m32.017s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512426.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k2_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.03 / 48.99
Loss : 1.73 / 1.44
Time: 123.21687715034932
---------
Epoch: 5
Acc : 71.80 / 72.91
Loss : 0.82 / 0.78
Time: 739.4670832967386
---------
Epoch: 10
Acc : 79.74 / 78.42
Loss : 0.58 / 0.62
Time: 1355.8709250381216
---------
Epoch: 15
Acc : 84.44 / 79.47
Loss : 0.45 / 0.60
Time: 1972.987782953307
---------
Epoch: 20
Acc : 87.94 / 82.06
Loss : 0.35 / 0.53
Time: 2590.0748241245747
---------
Epoch: 25
Acc : 91.02 / 82.41
Loss : 0.27 / 0.53
Time: 3207.198580466211
---------
Epoch: 30
Acc : 93.75 / 83.29
Loss : 0.20 / 0.51
Time: 3824.246278378181
---------
Epoch: 35
Acc : 96.25 / 84.13
Loss : 0.13 / 0.49
Time: 4441.094482817687
---------
Epoch: 40
Acc : 98.26 / 84.04
Loss : 0.09 / 0.53
Time: 5057.893284536898
---------
Epoch: 45
Acc : 99.16 / 84.72
Loss : 0.06 / 0.51
Time: 5674.5474897976965
---------
Best Acc : 84.99
Training time (min): 102.79756409196804
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k2_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k2_2 " saved !
real 103m1.616s
user 96m38.047s
sys 34m9.080s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512427.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k3_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.16 / 46.20
Loss : 1.72 / 1.48
Time: 247.73547955136746
---------
Epoch: 5
Acc : 72.15 / 72.22
Loss : 0.80 / 0.79
Time: 1486.8846261743456
---------
Epoch: 10
Acc : 80.06 / 77.68
Loss : 0.57 / 0.65
Time: 2726.5283604683354
---------
Epoch: 15
Acc : 84.93 / 78.28
Loss : 0.44 / 0.62
Time: 3966.002595074475
---------
Epoch: 20
Acc : 88.39 / 81.81
Loss : 0.34 / 0.53
Time: 5205.35461238306
---------
Epoch: 25
Acc : 91.51 / 82.47
Loss : 0.26 / 0.52
Time: 6445.20023989398
---------
Epoch: 30
Acc : 94.24 / 83.09
Loss : 0.18 / 0.51
Time: 7685.129119827412
---------
Epoch: 35
Acc : 96.49 / 84.29
Loss : 0.13 / 0.49
Time: 8925.508030240424
---------
Epoch: 40
Acc : 98.33 / 84.07
Loss : 0.08 / 0.51
Time: 10166.46677840408
---------
Epoch: 45
Acc : 99.31 / 84.71
Loss : 0.05 / 0.51
Time: 11407.179788754322
---------
Best Acc : 84.74
Training time (min): 206.67341564196784
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k3_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k3_1 " saved !
real 206m56.537s
user 164m29.383s
sys 69m32.644s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512428.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k3_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.50 / 46.49
Loss : 1.72 / 1.47
Time: 247.84990867320448
---------
Epoch: 5
Acc : 71.88 / 68.91
Loss : 0.80 / 0.89
Time: 1487.8860270297155
---------
Epoch: 10
Acc : 80.39 / 78.03
Loss : 0.57 / 0.64
Time: 2728.763779843226
---------
Epoch: 15
Acc : 85.00 / 80.67
Loss : 0.44 / 0.57
Time: 3969.208628097549
---------
Epoch: 20
Acc : 88.49 / 80.28
Loss : 0.34 / 0.57
Time: 5209.553945231251
---------
Epoch: 25
Acc : 91.64 / 82.31
Loss : 0.25 / 0.53
Time: 6450.321215154603
---------
Epoch: 30
Acc : 94.55 / 82.92
Loss : 0.18 / 0.50
Time: 7690.907434923574
---------
Epoch: 35
Acc : 96.83 / 83.45
Loss : 0.12 / 0.52
Time: 8931.464527050033
---------
Epoch: 40
Acc : 98.60 / 84.15
Loss : 0.07 / 0.51
Time: 10172.187801788561
---------
Epoch: 45
Acc : 99.43 / 84.93
Loss : 0.05 / 0.51
Time: 11412.973015635274
---------
Best Acc : 84.93
Training time (min): 206.75552665510526
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k3_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k3_2 " saved !
real 206m59.928s
user 165m3.232s
sys 69m9.800s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512429.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k4_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 35.99 / 47.12
Loss : 1.73 / 1.43
Time: 413.15663176588714
---------
Epoch: 5
Acc : 72.51 / 69.89
Loss : 0.78 / 0.87
Time: 2480.58620422706
---------
Epoch: 10
Acc : 80.56 / 76.70
Loss : 0.56 / 0.67
Time: 4549.522261521779
---------
Epoch: 15
Acc : 85.25 / 79.85
Loss : 0.43 / 0.59
Time: 6618.715358470567
---------
Epoch: 20
Acc : 88.85 / 80.16
Loss : 0.33 / 0.58
Time: 8688.104032479227
---------
Epoch: 25
Acc : 91.80 / 82.65
Loss : 0.25 / 0.51
Time: 10757.39982889127
---------
Epoch: 30
Acc : 94.70 / 83.66
Loss : 0.17 / 0.50
Time: 12826.700856998563
---------
Epoch: 35
Acc : 97.03 / 83.37
Loss : 0.11 / 0.54
Time: 14895.382152607664
---------
Epoch: 40
Acc : 98.66 / 84.33
Loss : 0.07 / 0.51
Time: 16965.023936793208
---------
Epoch: 45
Acc : 99.53 / 84.55
Loss : 0.05 / 0.52
Time: 19034.066288430244
---------
Best Acc : 84.77
Training time (min): 344.79890690402436
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k4_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k4_1 " saved !
real 345m2.785s
user 255m9.074s
sys 118m22.763s

73
jobs/log/Benchmark-MyLeNetMatNormalNoceil-44512430.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='__k4_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.29 / 46.76
Loss : 1.72 / 1.45
Time: 413.4330958928913
---------
Epoch: 5
Acc : 72.84 / 73.12
Loss : 0.78 / 0.77
Time: 2481.541083334014
---------
Epoch: 10
Acc : 80.54 / 77.11
Loss : 0.56 / 0.67
Time: 4550.451236174442
---------
Epoch: 15
Acc : 85.16 / 80.44
Loss : 0.43 / 0.57
Time: 6619.68952109199
---------
Epoch: 20
Acc : 88.79 / 81.69
Loss : 0.33 / 0.54
Time: 8689.100086900406
---------
Epoch: 25
Acc : 91.94 / 81.74
Loss : 0.25 / 0.58
Time: 10758.489818972535
---------
Epoch: 30
Acc : 94.65 / 82.18
Loss : 0.17 / 0.56
Time: 12827.756879460067
---------
Epoch: 35
Acc : 97.02 / 84.21
Loss : 0.11 / 0.51
Time: 14896.387511295266
---------
Epoch: 40
Acc : 98.82 / 84.45
Loss : 0.07 / 0.51
Time: 16964.967554398812
---------
Epoch: 45
Acc : 99.51 / 85.09
Loss : 0.05 / 0.51
Time: 19033.166706252843
---------
Best Acc : 85.09
Training time (min): 344.792250782003
Log :" res/benchmark_NoCeil/log/MyLeNetMatNormalNoceil-50epochs__k4_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatNormalNoceil-50epochs__k4_2 " saved !
real 345m2.760s
user 257m11.566s
sys 114m37.729s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44511791.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k1_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.71 / 22.38
Loss : 2.05 / 1.98
Time: 23.942733087576926
---------
Epoch: 5
Acc : 39.64 / 32.36
Loss : 1.65 / 1.86
Time: 144.20551797002554
---------
Epoch: 10
Acc : 45.79 / 33.74
Loss : 1.50 / 1.76
Time: 264.1239112522453
---------
Epoch: 15
Acc : 49.18 / 37.71
Loss : 1.40 / 1.66
Time: 382.70462130755186
---------
Epoch: 20
Acc : 52.63 / 41.70
Loss : 1.32 / 1.58
Time: 501.70173474773765
---------
Epoch: 25
Acc : 55.52 / 42.16
Loss : 1.24 / 1.55
Time: 620.2704349383712
---------
Epoch: 30
Acc : 57.60 / 42.07
Loss : 1.19 / 1.61
Time: 740.046815501526
---------
Epoch: 35
Acc : 59.83 / 40.80
Loss : 1.13 / 1.64
Time: 860.0331608653069
---------
Epoch: 40
Acc : 60.89 / 41.76
Loss : 1.09 / 1.63
Time: 980.0542129781097
---------
Epoch: 45
Acc : 62.60 / 42.83
Loss : 1.05 / 1.61
Time: 1099.9315661629662
---------
Best Acc : 47.46
Training time (min): 19.902773757139222
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k1_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k1_3 " saved !
real 20m6.443s
user 41m56.250s
sys 4m44.181s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44511796.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k2_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 23.07 / 27.72
Loss : 2.04 / 1.89
Time: 43.24283229280263
---------
Epoch: 5
Acc : 40.82 / 34.26
Loss : 1.62 / 1.78
Time: 260.49624280724674
---------
Epoch: 10
Acc : 46.79 / 32.58
Loss : 1.47 / 1.81
Time: 476.04811190254986
---------
Epoch: 15
Acc : 49.99 / 35.99
Loss : 1.39 / 1.74
Time: 692.3408845262602
---------
Epoch: 20
Acc : 53.41 / 35.56
Loss : 1.30 / 1.71
Time: 908.646314191632
---------
Epoch: 25
Acc : 55.78 / 40.94
Loss : 1.23 / 1.63
Time: 1123.2973771356046
---------
Epoch: 30
Acc : 58.41 / 44.97
Loss : 1.16 / 1.52
Time: 1339.4196774363518
---------
Epoch: 35
Acc : 60.69 / 43.89
Loss : 1.10 / 1.55
Time: 1554.9195355875418
---------
Epoch: 40
Acc : 62.63 / 46.52
Loss : 1.05 / 1.48
Time: 1770.575025911443
---------
Epoch: 45
Acc : 63.64 / 44.60
Loss : 1.03 / 1.56
Time: 1986.1898049293086
---------
Best Acc : 46.52
Training time (min): 35.97862558372629
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k2_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k2_3 " saved !
real 36m11.005s
user 53m35.069s
sys 9m56.397s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44511805.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k3_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.60 / 30.29
Loss : 2.07 / 1.83
Time: 73.45093404874206
---------
Epoch: 5
Acc : 41.07 / 36.65
Loss : 1.62 / 1.71
Time: 438.5278573250398
---------
Epoch: 10
Acc : 46.55 / 39.90
Loss : 1.48 / 1.61
Time: 804.3547679651529
---------
Epoch: 15
Acc : 50.45 / 38.47
Loss : 1.37 / 1.60
Time: 1169.9524947265163
---------
Epoch: 20
Acc : 54.17 / 39.78
Loss : 1.28 / 1.64
Time: 1535.7289634384215
---------
Epoch: 25
Acc : 56.16 / 42.75
Loss : 1.23 / 1.57
Time: 1901.3407105300575
---------
Epoch: 30
Acc : 58.65 / 46.69
Loss : 1.16 / 1.47
Time: 2266.849103649147
---------
Epoch: 35
Acc : 60.92 / 42.35
Loss : 1.10 / 1.66
Time: 2632.664029543288
---------
Epoch: 40
Acc : 63.07 / 41.75
Loss : 1.05 / 1.66
Time: 2998.279637400992
---------
Epoch: 45
Acc : 64.36 / 44.86
Loss : 1.01 / 1.58
Time: 3364.5860096393153
---------
Best Acc : 48.28
Training time (min): 60.951772359386084
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k3_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k3_3 " saved !
real 61m10.551s
user 70m16.339s
sys 18m43.649s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44511816.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k4_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.84 / 23.02
Loss : 2.06 / 2.02
Time: 109.14865664206445
---------
Epoch: 5
Acc : 41.31 / 34.38
Loss : 1.61 / 1.74
Time: 651.8169922940433
---------
Epoch: 10
Acc : 47.19 / 34.92
Loss : 1.46 / 1.75
Time: 1196.2202987512574
---------
Epoch: 15
Acc : 51.14 / 40.63
Loss : 1.35 / 1.61
Time: 1740.0807679975405
---------
Epoch: 20
Acc : 53.28 / 37.40
Loss : 1.30 / 1.66
Time: 2285.094068882987
---------
Epoch: 25
Acc : 56.24 / 44.48
Loss : 1.22 / 1.50
Time: 2828.706437432207
---------
Epoch: 30
Acc : 58.89 / 46.33
Loss : 1.15 / 1.50
Time: 3372.260067921132
---------
Epoch: 35
Acc : 61.42 / 43.65
Loss : 1.09 / 1.54
Time: 3917.1703928979114
---------
Epoch: 40
Acc : 63.16 / 45.52
Loss : 1.04 / 1.52
Time: 4460.238603447564
---------
Epoch: 45
Acc : 63.59 / 44.41
Loss : 1.03 / 1.61
Time: 5004.6891468744725
---------
Best Acc : 46.61
Training time (min): 90.65389749701134
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k4_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k4_3 " saved !
real 90m54.391s
user 89m26.614s
sys 28m46.091s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512439.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k1_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.40 / 33.33
Loss : 2.07 / 1.88
Time: 23.69926338084042
---------
Epoch: 5
Acc : 39.01 / 32.53
Loss : 1.66 / 1.82
Time: 142.7658422710374
---------
Epoch: 10
Acc : 45.48 / 38.35
Loss : 1.49 / 1.62
Time: 260.89411601331085
---------
Epoch: 15
Acc : 48.70 / 35.44
Loss : 1.41 / 1.69
Time: 379.99177754390985
---------
Epoch: 20
Acc : 52.10 / 38.83
Loss : 1.33 / 1.69
Time: 497.3336850358173
---------
Epoch: 25
Acc : 54.84 / 42.42
Loss : 1.26 / 1.55
Time: 616.1871792878956
---------
Epoch: 30
Acc : 57.42 / 40.18
Loss : 1.19 / 1.60
Time: 734.6347392667085
---------
Epoch: 35
Acc : 60.00 / 43.08
Loss : 1.12 / 1.55
Time: 852.3725217897445
---------
Epoch: 40
Acc : 61.69 / 43.45
Loss : 1.07 / 1.55
Time: 971.9417355190963
---------
Epoch: 45
Acc : 62.51 / 45.85
Loss : 1.05 / 1.53
Time: 1090.8980515552685
---------
Best Acc : 46.76
Training time (min): 19.73898232462816
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k1_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k1_1 " saved !
real 19m56.378s
user 41m16.163s
sys 4m49.607s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512440.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k1_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 21.82 / 26.99
Loss : 2.06 / 1.88
Time: 23.74811160005629
---------
Epoch: 5
Acc : 39.08 / 38.52
Loss : 1.66 / 1.71
Time: 144.02691594604403
---------
Epoch: 10
Acc : 45.10 / 35.87
Loss : 1.51 / 1.83
Time: 264.6037198500708
---------
Epoch: 15
Acc : 48.87 / 38.52
Loss : 1.41 / 1.64
Time: 384.93165539298207
---------
Epoch: 20
Acc : 52.06 / 38.93
Loss : 1.33 / 1.64
Time: 506.40114451851696
---------
Epoch: 25
Acc : 55.01 / 39.44
Loss : 1.25 / 1.66
Time: 628.3252009730786
---------
Epoch: 30
Acc : 57.14 / 37.17
Loss : 1.19 / 1.69
Time: 748.4351985407993
---------
Epoch: 35
Acc : 59.74 / 42.78
Loss : 1.13 / 1.62
Time: 869.0831191660836
---------
Epoch: 40
Acc : 61.07 / 43.44
Loss : 1.09 / 1.58
Time: 989.6546215675771
---------
Epoch: 45
Acc : 63.31 / 43.17
Loss : 1.04 / 1.63
Time: 1112.0879259789363
---------
Best Acc : 46.04
Training time (min): 20.136301828424134
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k1_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k1_2 " saved !
real 20m20.589s
user 42m38.377s
sys 4m32.509s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512441.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k2_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 23.02 / 28.97
Loss : 2.04 / 1.88
Time: 42.91845454927534
---------
Epoch: 5
Acc : 40.13 / 31.49
Loss : 1.63 / 1.85
Time: 257.2308983253315
---------
Epoch: 10
Acc : 47.02 / 41.68
Loss : 1.46 / 1.60
Time: 472.04093196988106
---------
Epoch: 15
Acc : 50.61 / 39.56
Loss : 1.37 / 1.65
Time: 686.8101648585871
---------
Epoch: 20
Acc : 53.33 / 38.19
Loss : 1.30 / 1.61
Time: 901.7532791886479
---------
Epoch: 25
Acc : 55.35 / 37.57
Loss : 1.24 / 1.67
Time: 1115.9920925060287
---------
Epoch: 30
Acc : 58.07 / 40.61
Loss : 1.17 / 1.58
Time: 1330.7566032465547
---------
Epoch: 35
Acc : 60.48 / 42.13
Loss : 1.11 / 1.58
Time: 1545.4550960762426
---------
Epoch: 40
Acc : 62.28 / 41.08
Loss : 1.06 / 1.68
Time: 1760.2057344606146
---------
Epoch: 45
Acc : 63.52 / 43.85
Loss : 1.02 / 1.59
Time: 1975.1974748587236
---------
Best Acc : 46.05
Training time (min): 35.78298388926002
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k2_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k2_1 " saved !
real 35m59.647s
user 52m30.243s
sys 10m6.934s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512442.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k2_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.17 / 31.05
Loss : 2.06 / 1.81
Time: 43.00148150417954
---------
Epoch: 5
Acc : 40.88 / 36.09
Loss : 1.61 / 1.71
Time: 257.9188695559278
---------
Epoch: 10
Acc : 46.77 / 41.97
Loss : 1.47 / 1.64
Time: 473.44908204767853
---------
Epoch: 15
Acc : 49.94 / 40.83
Loss : 1.39 / 1.63
Time: 688.832499277778
---------
Epoch: 20
Acc : 53.27 / 39.07
Loss : 1.30 / 1.57
Time: 903.5729479668662
---------
Epoch: 25
Acc : 56.13 / 38.69
Loss : 1.23 / 1.69
Time: 1118.620039199479
---------
Epoch: 30
Acc : 58.55 / 43.83
Loss : 1.16 / 1.56
Time: 1333.6320593841374
---------
Epoch: 35
Acc : 60.22 / 42.94
Loss : 1.11 / 1.55
Time: 1548.2262184657156
---------
Epoch: 40
Acc : 62.57 / 46.98
Loss : 1.05 / 1.49
Time: 1763.601978128776
---------
Epoch: 45
Acc : 63.72 / 45.46
Loss : 1.03 / 1.55
Time: 1978.3571852482855
---------
Best Acc : 47.21
Training time (min): 35.84222391460401
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k2_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k2_2 " saved !
real 36m2.860s
user 51m48.891s
sys 9m59.159s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512443.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k3_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.43 / 31.52
Loss : 2.07 / 1.80
Time: 73.24458603095263
---------
Epoch: 5
Acc : 40.97 / 37.28
Loss : 1.61 / 1.74
Time: 439.50911753159016
---------
Epoch: 10
Acc : 46.77 / 45.71
Loss : 1.47 / 1.57
Time: 805.9838715335354
---------
Epoch: 15
Acc : 51.05 / 37.82
Loss : 1.36 / 1.66
Time: 1172.2754459958524
---------
Epoch: 20
Acc : 53.40 / 37.39
Loss : 1.29 / 1.69
Time: 1538.9271471500397
---------
Epoch: 25
Acc : 56.16 / 47.24
Loss : 1.22 / 1.48
Time: 1904.104914849624
---------
Epoch: 30
Acc : 58.15 / 40.97
Loss : 1.17 / 1.67
Time: 2269.141853663139
---------
Epoch: 35
Acc : 61.22 / 42.66
Loss : 1.09 / 1.60
Time: 2635.261180281639
---------
Epoch: 40
Acc : 63.04 / 43.25
Loss : 1.04 / 1.60
Time: 3000.9633819106966
---------
Epoch: 45
Acc : 63.74 / 44.18
Loss : 1.02 / 1.63
Time: 3366.7220483692363
---------
Best Acc : 47.24
Training time (min): 60.98429485363886
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k3_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k3_1 " saved !
real 61m11.687s
user 70m18.044s
sys 18m52.141s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512444.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k3_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.47 / 28.87
Loss : 2.07 / 1.87
Time: 73.43172709643841
---------
Epoch: 5
Acc : 40.79 / 36.61
Loss : 1.62 / 1.77
Time: 440.08674016874284
---------
Epoch: 10
Acc : 46.63 / 31.72
Loss : 1.48 / 1.78
Time: 805.6908490704373
---------
Epoch: 15
Acc : 50.33 / 36.32
Loss : 1.38 / 1.65
Time: 1171.5075644087046
---------
Epoch: 20
Acc : 53.78 / 35.36
Loss : 1.28 / 1.70
Time: 1537.8829597318545
---------
Epoch: 25
Acc : 55.99 / 40.15
Loss : 1.22 / 1.60
Time: 1903.7469382034615
---------
Epoch: 30
Acc : 58.90 / 46.06
Loss : 1.15 / 1.45
Time: 2269.2510411664844
---------
Epoch: 35
Acc : 61.37 / 41.17
Loss : 1.09 / 1.70
Time: 2634.5049549862742
---------
Epoch: 40
Acc : 62.56 / 41.62
Loss : 1.05 / 1.69
Time: 2999.9693449428305
---------
Epoch: 45
Acc : 64.12 / 44.62
Loss : 1.01 / 1.56
Time: 3365.504422593862
---------
Best Acc : 47.18
Training time (min): 60.982219754951075
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k3_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k3_2 " saved !
real 61m11.800s
user 71m34.006s
sys 17m52.119s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512445.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k4_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.37 / 26.51
Loss : 2.07 / 1.96
Time: 108.77128831204027
---------
Epoch: 5
Acc : 40.45 / 37.46
Loss : 1.63 / 1.71
Time: 650.9957964411005
---------
Epoch: 10
Acc : 46.97 / 42.16
Loss : 1.46 / 1.63
Time: 1196.1481666192412
---------
Epoch: 15
Acc : 50.84 / 40.25
Loss : 1.37 / 1.64
Time: 1738.2220820374787
---------
Epoch: 20
Acc : 54.10 / 40.06
Loss : 1.28 / 1.63
Time: 2281.5323149422184
---------
Epoch: 25
Acc : 56.00 / 38.17
Loss : 1.23 / 1.71
Time: 2825.1198084335774
---------
Epoch: 30
Acc : 58.56 / 41.46
Loss : 1.16 / 1.54
Time: 3369.311863881536
---------
Epoch: 35
Acc : 61.36 / 41.71
Loss : 1.09 / 1.60
Time: 3912.360254479572
---------
Epoch: 40
Acc : 63.18 / 45.78
Loss : 1.04 / 1.53
Time: 4456.764227654785
---------
Epoch: 45
Acc : 63.95 / 43.66
Loss : 1.02 / 1.67
Time: 4999.635746097192
---------
Best Acc : 45.90
Training time (min): 90.55816794208562
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k4_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k4_1 " saved !
real 90m46.566s
user 91m1.517s
sys 28m30.486s

73
jobs/log/Benchmark-MyLeNetMatStochBUNoceil-44512446.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='__k4_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.48 / 27.01
Loss : 2.08 / 1.90
Time: 108.83058886602521
---------
Epoch: 5
Acc : 40.69 / 34.36
Loss : 1.63 / 1.72
Time: 652.1599301332608
---------
Epoch: 10
Acc : 46.78 / 39.97
Loss : 1.47 / 1.61
Time: 1195.5011438680813
---------
Epoch: 15
Acc : 50.49 / 34.57
Loss : 1.37 / 1.80
Time: 1737.6985349021852
---------
Epoch: 20
Acc : 53.32 / 39.84
Loss : 1.30 / 1.62
Time: 2280.7298378217965
---------
Epoch: 25
Acc : 56.46 / 40.06
Loss : 1.22 / 1.60
Time: 2822.9770746883005
---------
Epoch: 30
Acc : 59.26 / 40.83
Loss : 1.14 / 1.60
Time: 3365.840932168998
---------
Epoch: 35
Acc : 60.35 / 41.57
Loss : 1.11 / 1.60
Time: 3910.3378865057603
---------
Epoch: 40
Acc : 63.47 / 45.50
Loss : 1.03 / 1.51
Time: 4454.4913957370445
---------
Epoch: 45
Acc : 63.91 / 42.31
Loss : 1.02 / 1.67
Time: 4998.679032381624
---------
Best Acc : 47.46
Training time (min): 90.5504788563742
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochBUNoceil-50epochs__k4_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochBUNoceil-50epochs__k4_2 " saved !
real 90m46.666s
user 90m46.955s
sys 27m20.062s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44511750.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k4_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.19 / 31.17
Loss : 1.99 / 1.84
Time: 138.68438915722072
---------
Epoch: 5
Acc : 48.85 / 40.39
Loss : 1.42 / 1.68
Time: 832.1555920224637
---------
Epoch: 10
Acc : 56.04 / 34.76
Loss : 1.23 / 1.76
Time: 1525.8188216267154
---------
Epoch: 15
Acc : 60.87 / 38.62
Loss : 1.10 / 1.81
Time: 2219.536086712964
---------
Epoch: 20
Acc : 63.73 / 40.78
Loss : 1.02 / 1.74
Time: 2913.3058100305498
---------
Epoch: 25
Acc : 66.55 / 39.24
Loss : 0.95 / 1.68
Time: 3607.307444162667
---------
Epoch: 30
Acc : 69.17 / 47.70
Loss : 0.88 / 1.48
Time: 4300.989568569697
---------
Epoch: 35
Acc : 71.81 / 41.68
Loss : 0.80 / 1.62
Time: 4994.865097979084
---------
Epoch: 40
Acc : 73.30 / 45.42
Loss : 0.76 / 1.55
Time: 5688.842534529977
---------
Epoch: 45
Acc : 74.59 / 46.82
Loss : 0.72 / 1.52
Time: 6382.720301232301
---------
Best Acc : 47.70
Training time (min): 115.63096341982795
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k4_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k4_3 " saved !
real 115m50.643s
user 105m32.016s
sys 37m36.485s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44511756.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k3_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.51 / 29.79
Loss : 1.99 / 1.85
Time: 88.86880144290626
---------
Epoch: 5
Acc : 48.73 / 36.95
Loss : 1.41 / 1.75
Time: 533.375917987898
---------
Epoch: 10
Acc : 56.15 / 35.34
Loss : 1.23 / 1.73
Time: 978.0676415488124
---------
Epoch: 15
Acc : 60.08 / 38.06
Loss : 1.13 / 1.68
Time: 1422.6883844202384
---------
Epoch: 20
Acc : 64.37 / 39.67
Loss : 1.00 / 1.65
Time: 1867.38415880315
---------
Epoch: 25
Acc : 66.28 / 35.38
Loss : 0.95 / 1.96
Time: 2312.077256356366
---------
Epoch: 30
Acc : 68.93 / 41.81
Loss : 0.88 / 1.61
Time: 2756.5770405810326
---------
Epoch: 35
Acc : 71.02 / 44.29
Loss : 0.82 / 1.61
Time: 3200.8784932522103
---------
Epoch: 40
Acc : 73.21 / 45.35
Loss : 0.76 / 1.58
Time: 3645.313464949839
---------
Epoch: 45
Acc : 75.00 / 45.84
Loss : 0.71 / 1.56
Time: 4089.6597863975912
---------
Best Acc : 46.56
Training time (min): 74.08285954588403
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k3_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k3_3 " saved !
real 74m17.415s
user 79m1.446s
sys 24m21.474s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44511761.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k2_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.55 / 22.22
Loss : 1.98 / 1.94
Time: 48.120111235417426
---------
Epoch: 5
Acc : 49.35 / 35.96
Loss : 1.40 / 1.68
Time: 288.35652240365744
---------
Epoch: 10
Acc : 55.80 / 36.23
Loss : 1.25 / 1.73
Time: 528.6361616123468
---------
Epoch: 15
Acc : 60.81 / 40.07
Loss : 1.11 / 1.62
Time: 769.0124968513846
---------
Epoch: 20
Acc : 62.94 / 41.54
Loss : 1.05 / 1.64
Time: 1009.3565639145672
---------
Epoch: 25
Acc : 65.76 / 40.07
Loss : 0.97 / 1.59
Time: 1249.6207106374204
---------
Epoch: 30
Acc : 68.49 / 41.41
Loss : 0.89 / 1.68
Time: 1489.9397344766185
---------
Epoch: 35
Acc : 71.38 / 44.63
Loss : 0.81 / 1.55
Time: 1730.3411526689306
---------
Epoch: 40
Acc : 72.98 / 43.74
Loss : 0.77 / 1.64
Time: 1970.7049634410068
---------
Epoch: 45
Acc : 74.14 / 44.63
Loss : 0.74 / 1.59
Time: 2211.1649585356936
---------
Best Acc : 45.64
Training time (min): 40.060455758000415
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k2_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k2_3 " saved !
real 40m16.163s
user 55m29.844s
sys 13m1.891s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44511783.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k1_3', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.31 / 32.71
Loss : 1.97 / 1.80
Time: 23.04006605129689
---------
Epoch: 5
Acc : 47.91 / 37.56
Loss : 1.44 / 1.63
Time: 137.0359864262864
---------
Epoch: 10
Acc : 54.32 / 38.65
Loss : 1.28 / 1.77
Time: 250.97384090255946
---------
Epoch: 15
Acc : 59.07 / 35.07
Loss : 1.15 / 1.64
Time: 365.1462406516075
---------
Epoch: 20
Acc : 61.75 / 33.37
Loss : 1.07 / 1.84
Time: 479.1790603613481
---------
Epoch: 25
Acc : 64.43 / 40.65
Loss : 1.00 / 1.69
Time: 593.1814860291779
---------
Epoch: 30
Acc : 67.26 / 43.06
Loss : 0.92 / 1.60
Time: 707.3531251512468
---------
Epoch: 35
Acc : 69.03 / 42.53
Loss : 0.88 / 1.67
Time: 821.4921944644302
---------
Epoch: 40
Acc : 71.71 / 42.02
Loss : 0.81 / 1.68
Time: 935.5345202535391
---------
Epoch: 45
Acc : 72.13 / 45.46
Loss : 0.79 / 1.57
Time: 1049.5861575081944
---------
Best Acc : 46.35
Training time (min): 19.014818465150892
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k1_3.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k1_3 " saved !
real 19m13.464s
user 39m30.936s
sys 5m26.684s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512431.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k4_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 24.25 / 27.11
Loss : 2.01 / 1.89
Time: 138.53814427461475
---------
Epoch: 5
Acc : 48.78 / 28.38
Loss : 1.41 / 1.92
Time: 831.1760958814994
---------
Epoch: 10
Acc : 55.77 / 35.29
Loss : 1.24 / 1.85
Time: 1524.006879854016
---------
Epoch: 15
Acc : 60.26 / 34.95
Loss : 1.12 / 1.77
Time: 2216.9549421677366
---------
Epoch: 20
Acc : 63.12 / 37.17
Loss : 1.04 / 1.74
Time: 2910.0242391834036
---------
Epoch: 25
Acc : 66.15 / 40.66
Loss : 0.95 / 1.67
Time: 3602.922821316868
---------
Epoch: 30
Acc : 68.97 / 43.07
Loss : 0.89 / 1.63
Time: 4295.794840999879
---------
Epoch: 35
Acc : 71.39 / 43.45
Loss : 0.81 / 1.66
Time: 4988.862918260507
---------
Epoch: 40
Acc : 73.23 / 45.10
Loss : 0.76 / 1.61
Time: 5682.054097885266
---------
Epoch: 45
Acc : 74.62 / 44.73
Loss : 0.73 / 1.64
Time: 6375.051586847752
---------
Best Acc : 48.23
Training time (min): 115.4906927387075
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k4_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k4_1 " saved !
real 115m43.858s
user 105m46.985s
sys 37m41.831s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512432.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=4, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k4_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.53 / 31.27
Loss : 1.98 / 1.83
Time: 138.69305455219
---------
Epoch: 5
Acc : 49.02 / 37.01
Loss : 1.41 / 1.75
Time: 832.2641551829875
---------
Epoch: 10
Acc : 56.74 / 39.72
Loss : 1.22 / 1.72
Time: 1526.109748011455
---------
Epoch: 15
Acc : 60.87 / 42.71
Loss : 1.11 / 1.59
Time: 2219.8009980134666
---------
Epoch: 20
Acc : 63.56 / 39.88
Loss : 1.03 / 1.67
Time: 2913.56171762757
---------
Epoch: 25
Acc : 66.38 / 41.00
Loss : 0.95 / 1.56
Time: 3607.454514555633
---------
Epoch: 30
Acc : 68.79 / 42.87
Loss : 0.88 / 1.56
Time: 4301.280980648473
---------
Epoch: 35
Acc : 71.48 / 43.23
Loss : 0.81 / 1.58
Time: 4995.14941327367
---------
Epoch: 40
Acc : 73.54 / 46.15
Loss : 0.75 / 1.57
Time: 5689.165943298489
---------
Epoch: 45
Acc : 74.85 / 46.25
Loss : 0.72 / 1.57
Time: 6383.023234537803
---------
Best Acc : 47.23
Training time (min): 115.63643312873319
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k4_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k4_2 " saved !
real 115m51.013s
user 106m48.549s
sys 38m31.911s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512433.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k3_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.43 / 30.27
Loss : 1.98 / 1.86
Time: 88.86627784278244
---------
Epoch: 5
Acc : 48.91 / 32.66
Loss : 1.41 / 1.83
Time: 533.099832713604
---------
Epoch: 10
Acc : 56.05 / 37.94
Loss : 1.23 / 1.67
Time: 977.4172119349241
---------
Epoch: 15
Acc : 60.28 / 38.37
Loss : 1.11 / 1.70
Time: 1421.747462933883
---------
Epoch: 20
Acc : 63.74 / 37.67
Loss : 1.02 / 1.68
Time: 1866.0527866259217
---------
Epoch: 25
Acc : 66.02 / 38.42
Loss : 0.96 / 1.74
Time: 2310.4425071775913
---------
Epoch: 30
Acc : 69.11 / 41.55
Loss : 0.88 / 1.64
Time: 2755.050319542177
---------
Epoch: 35
Acc : 71.41 / 46.73
Loss : 0.82 / 1.52
Time: 3199.463610848412
---------
Epoch: 40
Acc : 73.18 / 43.02
Loss : 0.76 / 1.65
Time: 3643.856560504064
---------
Epoch: 45
Acc : 74.14 / 45.88
Loss : 0.74 / 1.58
Time: 4088.410287500359
---------
Best Acc : 47.19
Training time (min): 74.06591026883883
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k3_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k3_1 " saved !
real 74m18.154s
user 77m19.376s
sys 24m31.717s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512434.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=3, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k3_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 26.23 / 24.77
Loss : 1.96 / 1.95
Time: 88.89952981937677
---------
Epoch: 5
Acc : 49.19 / 37.63
Loss : 1.41 / 1.69
Time: 533.3709430936724
---------
Epoch: 10
Acc : 56.72 / 39.58
Loss : 1.21 / 1.61
Time: 978.0299946693704
---------
Epoch: 15
Acc : 60.20 / 37.36
Loss : 1.11 / 1.73
Time: 1422.6836077561602
---------
Epoch: 20
Acc : 63.17 / 37.28
Loss : 1.03 / 1.73
Time: 1867.3761521866545
---------
Epoch: 25
Acc : 66.56 / 41.14
Loss : 0.95 / 1.61
Time: 2311.9993200674653
---------
Epoch: 30
Acc : 68.86 / 40.12
Loss : 0.89 / 1.72
Time: 2756.85310437344
---------
Epoch: 35
Acc : 71.20 / 45.03
Loss : 0.82 / 1.60
Time: 3201.5983270090073
---------
Epoch: 40
Acc : 73.10 / 43.67
Loss : 0.77 / 1.62
Time: 3646.298190155998
---------
Epoch: 45
Acc : 75.00 / 45.30
Loss : 0.72 / 1.58
Time: 4091.1723202215508
---------
Best Acc : 48.57
Training time (min): 74.11512494976633
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k3_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k3_2 " saved !
real 74m20.018s
user 77m49.963s
sys 24m34.997s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512435.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k2_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.34 / 29.78
Loss : 1.98 / 1.85
Time: 47.96411813329905
---------
Epoch: 5
Acc : 48.65 / 42.37
Loss : 1.42 / 1.55
Time: 287.80597504321486
---------
Epoch: 10
Acc : 56.07 / 38.00
Loss : 1.23 / 1.70
Time: 527.6232785293832
---------
Epoch: 15
Acc : 59.94 / 36.03
Loss : 1.12 / 1.76
Time: 767.5103654088452
---------
Epoch: 20
Acc : 62.89 / 43.49
Loss : 1.04 / 1.52
Time: 1007.4215006893501
---------
Epoch: 25
Acc : 65.90 / 43.04
Loss : 0.96 / 1.55
Time: 1247.2988663418218
---------
Epoch: 30
Acc : 68.11 / 41.09
Loss : 0.91 / 1.59
Time: 1487.2040980625898
---------
Epoch: 35
Acc : 71.02 / 40.45
Loss : 0.82 / 1.66
Time: 1727.0953295668587
---------
Epoch: 40
Acc : 72.37 / 46.05
Loss : 0.78 / 1.51
Time: 1966.9677904183045
---------
Epoch: 45
Acc : 73.65 / 46.43
Loss : 0.75 / 1.54
Time: 2206.9109367653728
---------
Best Acc : 47.12
Training time (min): 39.980595599099374
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k2_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k2_1 " saved !
real 40m12.627s
user 55m24.324s
sys 12m21.461s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512436.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=2, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k2_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.09 / 28.73
Loss : 1.99 / 1.90
Time: 48.0987977925688
---------
Epoch: 5
Acc : 49.20 / 26.62
Loss : 1.40 / 1.83
Time: 288.2156368214637
---------
Epoch: 10
Acc : 55.36 / 35.66
Loss : 1.24 / 1.78
Time: 528.4270763546228
---------
Epoch: 15
Acc : 60.04 / 38.56
Loss : 1.12 / 1.69
Time: 768.740109086968
---------
Epoch: 20
Acc : 62.17 / 36.83
Loss : 1.06 / 1.72
Time: 1008.971360463649
---------
Epoch: 25
Acc : 66.36 / 39.98
Loss : 0.95 / 1.54
Time: 1249.2037892034277
---------
Epoch: 30
Acc : 68.35 / 42.86
Loss : 0.90 / 1.60
Time: 1489.428648560308
---------
Epoch: 35
Acc : 70.29 / 40.06
Loss : 0.84 / 1.69
Time: 1729.7566568944603
---------
Epoch: 40
Acc : 72.70 / 41.93
Loss : 0.78 / 1.69
Time: 1970.0739753739908
---------
Epoch: 45
Acc : 74.19 / 43.76
Loss : 0.74 / 1.63
Time: 2210.493873266503
---------
Best Acc : 47.52
Training time (min): 40.04782316264075
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k2_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k2_2 " saved !
real 40m15.453s
user 55m3.384s
sys 12m52.035s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512437.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k1_1', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.68 / 28.52
Loss : 1.97 / 1.88
Time: 22.80050491821021
---------
Epoch: 5
Acc : 47.17 / 33.95
Loss : 1.45 / 1.80
Time: 136.88718870654702
---------
Epoch: 10
Acc : 54.93 / 35.39
Loss : 1.26 / 1.72
Time: 250.9757927497849
---------
Epoch: 15
Acc : 58.73 / 41.05
Loss : 1.15 / 1.68
Time: 364.77219730988145
---------
Epoch: 20
Acc : 61.29 / 36.65
Loss : 1.09 / 1.75
Time: 478.69057740084827
---------
Epoch: 25
Acc : 64.54 / 38.35
Loss : 1.00 / 1.87
Time: 592.9431653441861
---------
Epoch: 30
Acc : 67.13 / 38.56
Loss : 0.93 / 1.71
Time: 707.10416641552
---------
Epoch: 35
Acc : 69.71 / 42.13
Loss : 0.86 / 1.68
Time: 821.3001165790483
---------
Epoch: 40
Acc : 71.33 / 42.99
Loss : 0.81 / 1.67
Time: 935.0861265612766
---------
Epoch: 45
Acc : 72.50 / 44.22
Loss : 0.78 / 1.64
Time: 1049.0620342679322
---------
Best Acc : 46.14
Training time (min): 18.99982411774496
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k1_1.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k1_1 " saved !
real 19m12.168s
user 40m45.219s
sys 5m20.754s

73
jobs/log/Benchmark-MyLeNetMatStochNoceil-44512438.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=50, k=1, lr=0.05, net='MyLeNetMatStochNoceil', postfix='__k1_2', res_folder='res/benchmark_NoCeil/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.65 / 27.44
Loss : 1.98 / 1.93
Time: 22.87828525621444
---------
Epoch: 5
Acc : 47.19 / 34.34
Loss : 1.45 / 1.73
Time: 136.54613589681685
---------
Epoch: 10
Acc : 54.83 / 34.87
Loss : 1.26 / 1.79
Time: 250.1823681294918
---------
Epoch: 15
Acc : 58.61 / 36.11
Loss : 1.16 / 1.67
Time: 364.1771142780781
---------
Epoch: 20
Acc : 61.51 / 41.69
Loss : 1.07 / 1.61
Time: 477.91743985097855
---------
Epoch: 25
Acc : 64.63 / 42.91
Loss : 1.00 / 1.60
Time: 591.6240369612351
---------
Epoch: 30
Acc : 66.87 / 42.92
Loss : 0.94 / 1.60
Time: 705.5203553121537
---------
Epoch: 35
Acc : 69.64 / 40.97
Loss : 0.86 / 1.71
Time: 819.437248964794
---------
Epoch: 40
Acc : 71.34 / 45.05
Loss : 0.82 / 1.57
Time: 933.206932798028
---------
Epoch: 45
Acc : 72.89 / 46.70
Loss : 0.78 / 1.52
Time: 1047.1215672474355
---------
Best Acc : 47.16
Training time (min): 18.968624996906147
Log :" res/benchmark_NoCeil/log/MyLeNetMatStochNoceil-50epochs__k1_2.json " saved !
Plot :" res/benchmark_NoCeil/MyLeNetMatStochNoceil-50epochs__k1_2 " saved !
real 19m10.751s
user 39m32.089s
sys 5m24.696s

17
jobs/log/MyLeNetMatNormalNoceil-44492595.out
View file

@ -1,17 +0,0 @@
Namespace(batch=128, dataset='CIFAR10', epochs=10, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='_noCrop', res_folder='res/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 36.21 / 49.24
Loss : 1.72 / 1.42
Time: 247.9577119993046
---------
Epoch: 1
Acc : 52.45 / 53.66
Loss : 1.32 / 1.29
Time: 496.28596297279
slurmstepd: error: *** JOB 44492595 ON cdr294 CANCELLED AT 2020-06-23T01:58:42 ***

73
jobs/log/MyLeNetMatNormalNoceil-44494271.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=10, lr=0.05, net='MyLeNetMatNormalNoceil', postfix='_noCrop_k3', res_folder='res/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 35.84 / 47.42
Loss : 1.73 / 1.45
Time: 247.71314711589366
---------
Epoch: 1
Acc : 53.34 / 60.40
Loss : 1.30 / 1.14
Time: 495.9195477189496
---------
Epoch: 2
Acc : 60.44 / 60.80
Loss : 1.11 / 1.08
Time: 743.7779585365206
---------
Epoch: 3
Acc : 65.48 / 65.42
Loss : 0.98 / 0.98
Time: 991.7510572513565
---------
Epoch: 4
Acc : 69.41 / 69.60
Loss : 0.87 / 0.86
Time: 1239.6623603589833
---------
Epoch: 5
Acc : 72.88 / 71.52
Loss : 0.78 / 0.80
Time: 1487.751311905682
---------
Epoch: 6
Acc : 75.24 / 73.46
Loss : 0.71 / 0.76
Time: 1735.6824928615242
---------
Epoch: 7
Acc : 77.73 / 74.89
Loss : 0.64 / 0.72
Time: 1983.8558715265244
---------
Epoch: 8
Acc : 79.20 / 77.12
Loss : 0.60 / 0.66
Time: 2231.74981981609
---------
Epoch: 9
Acc : 80.25 / 77.54
Loss : 0.57 / 0.65
Time: 2479.908313873224
---------
Best Acc : 77.54
Training time (min): 41.33180533062356
Log :" res/log/MyLeNetMatNormalNoceil-10epochs_noCrop_k3.json " saved !
Plot :" res/MyLeNetMatNormalNoceil-10epochs_noCrop_k3 " saved !
real 41m40.414s
user 32m50.590s
sys 14m15.666s

73
jobs/log/MyLeNetMatStochBUNoceil-44494341.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=10, lr=0.05, net='MyLeNetMatStochBUNoceil', postfix='_noCrop_k3', res_folder='res/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 22.73 / 30.49
Loss : 2.06 / 1.90
Time: 73.46249075420201
---------
Epoch: 1
Acc : 30.11 / 30.61
Loss : 1.87 / 1.79
Time: 146.35232928954065
---------
Epoch: 2
Acc : 34.14 / 33.30
Loss : 1.77 / 1.73
Time: 219.46285958588123
---------
Epoch: 3
Acc : 38.18 / 34.64
Loss : 1.68 / 1.74
Time: 292.7825370589271
---------
Epoch: 4
Acc : 40.50 / 37.73
Loss : 1.62 / 1.68
Time: 366.0361840762198
---------
Epoch: 5
Acc : 43.22 / 38.42
Loss : 1.54 / 1.61
Time: 439.2800999926403
---------
Epoch: 6
Acc : 46.57 / 39.54
Loss : 1.47 / 1.61
Time: 513.1152702141553
---------
Epoch: 7
Acc : 48.72 / 38.11
Loss : 1.41 / 1.69
Time: 586.3596808109432
---------
Epoch: 8
Acc : 50.07 / 37.34
Loss : 1.37 / 1.67
Time: 660.0147361075506
---------
Epoch: 9
Acc : 51.24 / 40.30
Loss : 1.35 / 1.60
Time: 733.1694793645293
---------
Best Acc : 40.30
Training time (min): 12.219491492429126
Log :" res/log/MyLeNetMatStochBUNoceil-10epochs_noCrop_k3.json " saved !
Plot :" res/MyLeNetMatStochBUNoceil-10epochs_noCrop_k3 " saved !
real 12m34.154s
user 14m9.534s
sys 3m43.740s

73
jobs/log/MyLeNetMatStochNoceil-44494296.out Normal file
View file

@ -0,0 +1,73 @@
Namespace(batch=128, dataset='CIFAR10', epochs=10, lr=0.05, net='MyLeNetMatStochNoceil', postfix='_noCrop_k3', res_folder='res/', resume=False, scheduler='cosine', stoch=False, warmup_ep=5, warmup_mul=0)
==> Preparing data..
==> Building model..
==> Training model..
---------
Epoch: 0
Acc : 25.40 / 29.56
Loss : 1.99 / 1.83
Time: 88.82820883858949
---------
Epoch: 1
Acc : 35.13 / 32.19
Loss : 1.74 / 1.77
Time: 177.5576635012403
---------
Epoch: 2
Acc : 40.73 / 35.05
Loss : 1.61 / 1.76
Time: 266.2516449401155
---------
Epoch: 3
Acc : 45.58 / 38.76
Loss : 1.49 / 1.71
Time: 355.2165524791926
---------
Epoch: 4
Acc : 49.07 / 35.94
Loss : 1.41 / 1.71
Time: 443.9428597986698
---------
Epoch: 5
Acc : 52.32 / 32.03
Loss : 1.32 / 1.86
Time: 532.6662928508595
---------
Epoch: 6
Acc : 54.68 / 40.80
Loss : 1.26 / 1.59
Time: 621.3865936575457
---------
Epoch: 7
Acc : 58.03 / 41.34
Loss : 1.17 / 1.62
Time: 710.462741760537
---------
Epoch: 8
Acc : 59.08 / 44.26
Loss : 1.15 / 1.53
Time: 799.1782552432269
---------
Epoch: 9
Acc : 60.38 / 42.60
Loss : 1.11 / 1.58
Time: 887.9221545867622
---------
Best Acc : 44.26
Training time (min): 14.798702753568069
Log :" res/log/MyLeNetMatStochNoceil-10epochs_noCrop_k3.json " saved !
Plot :" res/MyLeNetMatStochNoceil-10epochs_noCrop_k3 " saved !
real 15m9.606s
user 15m45.480s
sys 4m48.638s