Hugging Face's logo

YuqianFu
/
CausalStyleAdv

TensorBoard
Model card Files
xet
Metrics Community
YuqianFu commited on
Commit
197d4ca
·
verified ·
1 Parent(s): b5aa0c4

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitignore +81 -0
  2. 52715.error +309 -0
  3. 52715.log +114 -0
  4. 52729.error +309 -0
  5. 52729.log +114 -0
  6. Meta-causal/code-stage1-pipeline/56451.error +297 -0
  7. Meta-causal/code-stage1-pipeline/56451.log +0 -0
  8. Meta-causal/code-stage1-pipeline/56452.error +302 -0
  9. Meta-causal/code-stage1-pipeline/56452.log +0 -0
  10. Meta-causal/code-stage1-pipeline/56454.error +3 -0
  11. Meta-causal/code-stage1-pipeline/56454.log +0 -0
  12. Meta-causal/code-stage1-pipeline/56455.error +4 -0
  13. Meta-causal/code-stage1-pipeline/56455.log +0 -0
  14. Meta-causal/code-stage1-pipeline/56456.error +3 -0
  15. Meta-causal/code-stage1-pipeline/56456.log +0 -0
  16. Meta-causal/code-stage1-pipeline/56457.error +4 -0
  17. Meta-causal/code-stage1-pipeline/56457.log +0 -0
  18. Meta-causal/code-stage1-pipeline/56458.error +3 -0
  19. Meta-causal/code-stage1-pipeline/56458.log +0 -0
  20. Meta-causal/code-stage1-pipeline/56526.error +31 -0
  21. Meta-causal/code-stage1-pipeline/56526.log +4 -0
  22. Meta-causal/code-stage1-pipeline/56527.error +31 -0
  23. Meta-causal/code-stage1-pipeline/56527.log +3 -0
  24. Meta-causal/code-stage1-pipeline/56528.error +3 -0
  25. Meta-causal/code-stage1-pipeline/56528.log +0 -0
  26. Meta-causal/code-stage1-pipeline/56529.error +1 -0
  27. Meta-causal/code-stage1-pipeline/56529.log +0 -0
  28. Meta-causal/code-stage1-pipeline/56540.error +4 -0
  29. Meta-causal/code-stage1-pipeline/56540.log +151 -0
  30. Meta-causal/code-stage1-pipeline/56541.error +0 -0
  31. Meta-causal/code-stage1-pipeline/56541.log +432 -0
  32. Meta-causal/code-stage1-pipeline/AllEpochs_test_digit_v13.py +101 -0
  33. Meta-causal/code-stage1-pipeline/AllEpochs_test_pacs_v13.py +103 -0
  34. Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py +861 -0
  35. Meta-causal/code-stage1-pipeline/env.yaml +119 -0
  36. Meta-causal/code-stage1-pipeline/main_my_joint_v13_auto.py +279 -0
  37. Meta-causal/code-stage1-pipeline/main_test_digit_v13.py +85 -0
  38. Meta-causal/code-stage1-pipeline/main_test_pacs_v13.py +89 -0
  39. Meta-causal/code-stage1-pipeline/network/adaptor_v2.py +63 -0
  40. Meta-causal/code-stage1-pipeline/network/mnist_net_my.py +104 -0
  41. Meta-causal/code-stage1-pipeline/network/resnet.py +101 -0
  42. Meta-causal/code-stage1-pipeline/network/wideresnet.py +86 -0
  43. Meta-causal/code-stage1-pipeline/run_PACS/run_my_joint_v13_test.sh +35 -0
  44. Meta-causal/code-stage1-pipeline/run_digits/run_my_joint_test.sh +34 -0
  45. Meta-causal/code-stage1-pipeline/saved-PACS/art_painting/CA_multiple_16fa_v2_ep70_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5/events.out.tfevents.1719926752.hala +3 -0
  46. Meta-causal/code-stage1-pipeline/saved-PACS/art_painting/CA_multiple_16fa_v2_ep70_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5/log.log +1 -0
  47. Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/events.out.tfevents.1719925086.hala +3 -0
  48. Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/events.out.tfevents.1719925314.hala +3 -0
  49. Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/events.out.tfevents.1719925652.hala +3 -0
  50. Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/log.log +1 -0
.gitignore ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # no IntelliJ files
2
+ .idea
3
+
4
+ # don't upload macOS folder info
5
+ *.DS_Store
6
+
7
+ # don't upload node_modules from npm test
8
+ node_modules/*
9
+ flow-typed/*
10
+
11
+ # potential files generated by golang
12
+ bin/
13
+
14
+ # don't upload webpack bundle file
15
+ app/dist/
16
+
17
+ # potential integration testing data directory
18
+ # test_data/
19
+ /data
20
+
21
+ #python
22
+ *.pyc
23
+ __pycache__/
24
+
25
+ # pytype
26
+ .pytype
27
+
28
+ # vscode sftp settings
29
+ .vscode/sftp.json
30
+
31
+ # vscode launch settings
32
+ .vscode/launch.json
33
+
34
+ # redis
35
+ *.rdb
36
+
37
+ # mypy
38
+ .mypy_cache
39
+
40
+ # jest coverage cache
41
+ coverage/
42
+
43
+ # downloaded repos and models
44
+ scalabel/bot/experimental/*
45
+
46
+
47
+ # python virtual environment
48
+ env/
49
+
50
+ # vscode workspace configuration
51
+ *.code-workspace
52
+
53
+ # sphinx build folder
54
+ _build/
55
+
56
+ # media files are not in this repo
57
+ doc/media
58
+
59
+ # ignore rope db cache
60
+ .vscode/.ropeproject
61
+
62
+ # python build
63
+ build/
64
+ dist/
65
+
66
+ # coverage
67
+ .coverage*
68
+
69
+ # package default workspace
70
+ /output
71
+
72
+ *.tmp
73
+ *.zip
74
+
75
+ # local test logs and scripts
76
+ log/
77
+ /*.sh
78
+ wandb/
79
+
80
+ # No lightning logs
81
+ lightning_logs/
52715.error ADDED
@@ -0,0 +1,309 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Solving dependencies
2
+ Installing conda packages
3
+ Empty environment created at prefix: /scratch/yuqian_fu/micromamba/envs/auto-zcubaqpyrbpe
4
+ error libmamba Could not lock non-existing path '/scratch/yuqian_fu/micromamba/pkgs'
5
+ Transaction
6
+
7
+ Prefix: /scratch/yuqian_fu/micromamba/envs/auto-zcubaqpyrbpe
8
+
9
+
10
+
11
+ No specs added or removed.
12
+
13
+ Package Version Build Channel Size
14
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
15
+ Install:
16
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
17
+
18
+ + _libgcc_mutex 0.1 conda_forge conda-forge
19
+ + _openmp_mutex 4.5 2_kmp_llvm conda-forge
20
+ + blas 2.116 mkl conda-forge
21
+ + blas-devel 3.9.0 16_linux64_mkl conda-forge
22
+ + brotli-python 1.1.0 py311hb755f60_1 conda-forge
23
+ + bzip2 1.0.8 hd590300_5 conda-forge
24
+ + ca-certificates 2024.6.2 hbcca054_0 conda-forge
25
+ + certifi 2024.6.2 pyhd8ed1ab_0 conda-forge
26
+ + cffi 1.16.0 py311hb3a22ac_0 conda-forge
27
+ + charset-normalizer 3.3.2 pyhd8ed1ab_0 conda-forge
28
+ + cuda-cudart 12.1.105 0 nvidia
29
+ + cuda-cupti 12.1.105 0 /work/conda/cache/nvidia
30
+ + cuda-libraries 12.1.0 0 nvidia
31
+ + cuda-nvrtc 12.1.105 0 /work/conda/cache/nvidia
32
+ + cuda-nvtx 12.1.105 0 nvidia
33
+ + cuda-opencl 12.5.39 0 nvidia
34
+ + cuda-runtime 12.1.0 0 nvidia
35
+ + cuda-version 12.5 3 nvidia
36
+ + ffmpeg 4.3 hf484d3e_0 /work/conda/cache/pytorch
37
+ + filelock 3.15.4 pyhd8ed1ab_0 conda-forge
38
+ + freetype 2.12.1 h267a509_2 conda-forge
39
+ + gmp 6.3.0 hac33072_2 conda-forge
40
+ + gmpy2 2.1.5 py311hc4f1f91_1 conda-forge
41
+ + gnutls 3.6.13 h85f3911_1 /work/conda/cache/conda-forge
42
+ + h2 4.1.0 pyhd8ed1ab_0 conda-forge
43
+ + hpack 4.0.0 pyh9f0ad1d_0 conda-forge
44
+ + hyperframe 6.0.1 pyhd8ed1ab_0 conda-forge
45
+ + icu 73.2 h59595ed_0 /work/conda/cache/conda-forge
46
+ + idna 3.7 pyhd8ed1ab_0 conda-forge
47
+ + jinja2 3.1.4 pyhd8ed1ab_0 conda-forge
48
+ + jpeg 9e h166bdaf_2 conda-forge
49
+ + lame 3.100 h166bdaf_1003 conda-forge
50
+ + lcms2 2.15 hfd0df8a_0 conda-forge
51
+ + ld_impl_linux-64 2.40 hf3520f5_7 conda-forge
52
+ + lerc 4.0.0 h27087fc_0 conda-forge
53
+ + libblas 3.9.0 16_linux64_mkl conda-forge
54
+ + libcblas 3.9.0 16_linux64_mkl conda-forge
55
+ + libcublas 12.1.0.26 0 /work/conda/cache/nvidia
56
+ + libcufft 11.0.2.4 0 /work/conda/cache/nvidia
57
+ + libcufile 1.10.0.4 0 nvidia
58
+ + libcurand 10.3.6.39 0 nvidia
59
+ + libcusolver 11.4.4.55 0 /work/conda/cache/nvidia
60
+ + libcusparse 12.0.2.55 0 /work/conda/cache/nvidia
61
+ + libdeflate 1.17 h0b41bf4_0 conda-forge
62
+ + libexpat 2.6.2 h59595ed_0 conda-forge
63
+ + libffi 3.4.2 h7f98852_5 conda-forge
64
+ + libgcc-ng 13.2.0 h77fa898_13 conda-forge
65
+ + libgfortran-ng 13.2.0 h69a702a_13 conda-forge
66
+ + libgfortran5 13.2.0 h3d2ce59_13 conda-forge
67
+ + libhwloc 2.10.0 default_h5622ce7_1001 conda-forge
68
+ + libiconv 1.17 hd590300_2 conda-forge
69
+ + libjpeg-turbo 2.0.0 h9bf148f_0 pytorch
70
+ + liblapack 3.9.0 16_linux64_mkl conda-forge
71
+ + liblapacke 3.9.0 16_linux64_mkl conda-forge
72
+ + libnpp 12.0.2.50 0 /work/conda/cache/nvidia
73
+ + libnsl 2.0.1 hd590300_0 conda-forge
74
+ + libnvjitlink 12.1.105 0 /work/conda/cache/nvidia
75
+ + libnvjpeg 12.1.1.14 0 /work/conda/cache/nvidia
76
+ + libpng 1.6.43 h2797004_0 conda-forge
77
+ + libsqlite 3.46.0 hde9e2c9_0 conda-forge
78
+ + libstdcxx-ng 13.2.0 hc0a3c3a_13 conda-forge
79
+ + libtiff 4.5.0 h6adf6a1_2 conda-forge
80
+ + libuuid 2.38.1 h0b41bf4_0 conda-forge
81
+ + libwebp-base 1.4.0 hd590300_0 conda-forge
82
+ + libxcb 1.13 h7f98852_1004 conda-forge
83
+ + libxcrypt 4.4.36 hd590300_1 conda-forge
84
+ + libxml2 2.12.7 hc051c1a_1 conda-forge
85
+ + libzlib 1.2.13 h4ab18f5_6 conda-forge
86
+ + llvm-openmp 15.0.7 h0cdce71_0 /work/conda/cache/conda-forge
87
+ + markupsafe 2.1.5 py311h459d7ec_0 conda-forge
88
+ + mkl 2022.1.0 h84fe81f_915 /work/conda/cache/conda-forge
89
+ + mkl-devel 2022.1.0 ha770c72_916 conda-forge
90
+ + mkl-include 2022.1.0 h84fe81f_915 conda-forge
91
+ + mpc 1.3.1 hfe3b2da_0 conda-forge
92
+ + mpfr 4.2.1 h9458935_1 conda-forge
93
+ + mpmath 1.3.0 pyhd8ed1ab_0 conda-forge
94
+ + ncurses 6.5 h59595ed_0 conda-forge
95
+ + nettle 3.6 he412f7d_0 /work/conda/cache/conda-forge
96
+ + networkx 3.3 pyhd8ed1ab_1 /work/conda/cache/conda-forge
97
+ + numpy 2.0.0 py311h1461c94_0 conda-forge
98
+ + openh264 2.1.1 h780b84a_0 /work/conda/cache/conda-forge
99
+ + openjpeg 2.5.0 hfec8fc6_2 conda-forge
100
+ + openssl 3.3.1 h4ab18f5_1 conda-forge
101
+ + pandas 2.2.2 py311h14de704_1 conda-forge
102
+ + pillow 9.4.0 py311h50def17_1 conda-forge
103
+ + pip 24.0 pyhd8ed1ab_0 /work/conda/cache/conda-forge
104
+ + pthread-stubs 0.4 h36c2ea0_1001 conda-forge
105
+ + pycparser 2.22 pyhd8ed1ab_0 conda-forge
106
+ + pysocks 1.7.1 pyha2e5f31_6 conda-forge
107
+ + python 3.11.9 hb806964_0_cpython /work/conda/cache/conda-forge
108
+ + python-dateutil 2.9.0 pyhd8ed1ab_0 conda-forge
109
+ + python-tzdata 2024.1 pyhd8ed1ab_0 conda-forge
110
+ + python_abi 3.11 4_cp311 conda-forge
111
+ + pytorch 2.3.1 py3.11_cuda12.1_cudnn8.9.2_0 pytorch
112
+ + pytorch-cuda 12.1 ha16c6d3_5 pytorch
113
+ + pytorch-mutex 1.0 cuda pytorch
114
+ + pytz 2024.1 pyhd8ed1ab_0 conda-forge
115
+ + pyyaml 6.0.1 py311h459d7ec_1 conda-forge
116
+ + readline 8.2 h8228510_1 conda-forge
117
+ + requests 2.32.3 pyhd8ed1ab_0 conda-forge
118
+ + setuptools 70.1.1 pyhd8ed1ab_0 conda-forge
119
+ + six 1.16.0 pyh6c4a22f_0 conda-forge
120
+ + sympy 1.12.1 pypyh2585a3b_103 conda-forge
121
+ + tbb 2021.12.0 h297d8ca_1 conda-forge
122
+ + tk 8.6.13 noxft_h4845f30_101 /work/conda/cache/conda-forge
123
+ + torchtriton 2.3.1 py311 pytorch
124
+ + torchvision 0.18.1 py311_cu121 pytorch
125
+ + typing_extensions 4.12.2 pyha770c72_0 conda-forge
126
+ + tzdata 2024a h0c530f3_0 conda-forge
127
+ + urllib3 2.2.2 pyhd8ed1ab_1 conda-forge
128
+ + wheel 0.43.0 pyhd8ed1ab_1 conda-forge
129
+ + xorg-libxau 1.0.11 hd590300_0 conda-forge
130
+ + xorg-libxdmcp 1.1.3 h516909a_0 conda-forge
131
+ + xz 5.2.6 h166bdaf_0 conda-forge
132
+ + yaml 0.2.5 h7f98852_2 conda-forge
133
+ + zlib 1.2.13 h4ab18f5_6 conda-forge
134
+ + zstandard 0.22.0 py311hb6f056b_1 conda-forge
135
+ + zstd 1.5.6 ha6fb4c9_0 conda-forge
136
+
137
+ Summary:
138
+
139
+ Install: 118 packages
140
+
141
+ Total download: 0 B
142
+
143
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
144
+
145
+
146
+
147
+ Transaction starting
148
+ Linking libcublas-12.1.0.26-0
149
+ Linking libcufft-11.0.2.4-0
150
+ Linking libcusolver-11.4.4.55-0
151
+ Linking libcusparse-12.0.2.55-0
152
+ Linking libnpp-12.0.2.50-0
153
+ Linking libnvjitlink-12.1.105-0
154
+ Linking cuda-cudart-12.1.105-0
155
+ Linking cuda-nvrtc-12.1.105-0
156
+ Linking libnvjpeg-12.1.1.14-0
157
+ Linking cuda-cupti-12.1.105-0
158
+ Linking cuda-nvtx-12.1.105-0
159
+ Linking pytorch-mutex-1.0-cuda
160
+ Linking _libgcc_mutex-0.1-conda_forge
161
+ Linking mkl-include-2022.1.0-h84fe81f_915
162
+ Linking python_abi-3.11-4_cp311
163
+ Linking ld_impl_linux-64-2.40-hf3520f5_7
164
+ Linking ca-certificates-2024.6.2-hbcca054_0
165
+ Linking libgcc-ng-13.2.0-h77fa898_13
166
+ Linking libzlib-1.2.13-h4ab18f5_6
167
+ Linking llvm-openmp-15.0.7-h0cdce71_0
168
+ Linking _openmp_mutex-4.5-2_kmp_llvm
169
+ Linking xorg-libxdmcp-1.1.3-h516909a_0
170
+ Linking pthread-stubs-0.4-h36c2ea0_1001
171
+ Linking xorg-libxau-1.0.11-hd590300_0
172
+ Linking libwebp-base-1.4.0-hd590300_0
173
+ Linking libdeflate-1.17-h0b41bf4_0
174
+ Linking jpeg-9e-h166bdaf_2
175
+ Linking libffi-3.4.2-h7f98852_5
176
+ Linking tk-8.6.13-noxft_h4845f30_101
177
+ Linking openssl-3.3.1-h4ab18f5_1
178
+ Linking libxcrypt-4.4.36-hd590300_1
179
+ Linking libsqlite-3.46.0-hde9e2c9_0
180
+ Linking yaml-0.2.5-h7f98852_2
181
+ Linking ncurses-6.5-h59595ed_0
182
+ Linking libgfortran5-13.2.0-h3d2ce59_13
183
+ Linking lame-3.100-h166bdaf_1003
184
+ Linking nettle-3.6-he412f7d_0
185
+ Linking zlib-1.2.13-h4ab18f5_6
186
+ Linking libstdcxx-ng-13.2.0-hc0a3c3a_13
187
+ Linking libiconv-1.17-hd590300_2
188
+ Linking bzip2-1.0.8-hd590300_5
189
+ Linking libpng-1.6.43-h2797004_0
190
+ Linking xz-5.2.6-h166bdaf_0
191
+ Linking libuuid-2.38.1-h0b41bf4_0
192
+ Linking libnsl-2.0.1-hd590300_0
193
+ Linking libexpat-2.6.2-h59595ed_0
194
+ Linking libxcb-1.13-h7f98852_1004
195
+ Linking readline-8.2-h8228510_1
196
+ Linking libgfortran-ng-13.2.0-h69a702a_13
197
+ Linking icu-73.2-h59595ed_0
198
+ Linking zstd-1.5.6-ha6fb4c9_0
199
+ Linking lerc-4.0.0-h27087fc_0
200
+ Linking openh264-2.1.1-h780b84a_0
201
+ Linking gnutls-3.6.13-h85f3911_1
202
+ Linking gmp-6.3.0-hac33072_2
203
+ Linking freetype-2.12.1-h267a509_2
204
+ Linking libxml2-2.12.7-hc051c1a_1
205
+ Linking libtiff-4.5.0-h6adf6a1_2
206
+ Linking mpfr-4.2.1-h9458935_1
207
+ Linking libhwloc-2.10.0-default_h5622ce7_1001
208
+ Linking openjpeg-2.5.0-hfec8fc6_2
209
+ Linking lcms2-2.15-hfd0df8a_0
210
+ Linking mpc-1.3.1-hfe3b2da_0
211
+ Linking tbb-2021.12.0-h297d8ca_1
212
+ Linking mkl-2022.1.0-h84fe81f_915
213
+ Linking mkl-devel-2022.1.0-ha770c72_916
214
+ Linking libblas-3.9.0-16_linux64_mkl
215
+ Linking liblapack-3.9.0-16_linux64_mkl
216
+ Linking libcblas-3.9.0-16_linux64_mkl
217
+ Linking liblapacke-3.9.0-16_linux64_mkl
218
+ Linking blas-devel-3.9.0-16_linux64_mkl
219
+ Linking blas-2.116-mkl
220
+ Linking cuda-version-12.5-3
221
+ Linking tzdata-2024a-h0c530f3_0
222
+ Linking libjpeg-turbo-2.0.0-h9bf148f_0
223
+ warning libmamba [libjpeg-turbo-2.0.0-h9bf148f_0] The following files were already present in the environment:
224
+ - bin/cjpeg
225
+ - bin/djpeg
226
+ - bin/jpegtran
227
+ - bin/rdjpgcom
228
+ - bin/wrjpgcom
229
+ - include/jconfig.h
230
+ - include/jerror.h
231
+ - include/jmorecfg.h
232
+ - include/jpeglib.h
233
+ - lib/libjpeg.a
234
+ - lib/libjpeg.so
235
+ - lib/pkgconfig/libjpeg.pc
236
+ - share/man/man1/cjpeg.1
237
+ - share/man/man1/djpeg.1
238
+ - share/man/man1/jpegtran.1
239
+ - share/man/man1/rdjpgcom.1
240
+ - share/man/man1/wrjpgcom.1
241
+ Linking ffmpeg-4.3-hf484d3e_0
242
+ Linking libcurand-10.3.6.39-0
243
+ Linking libcufile-1.10.0.4-0
244
+ Linking cuda-opencl-12.5.39-0
245
+ Linking cuda-libraries-12.1.0-0
246
+ Linking cuda-runtime-12.1.0-0
247
+ Linking python-3.11.9-hb806964_0_cpython
248
+ Linking pytorch-cuda-12.1-ha16c6d3_5
249
+ Linking wheel-0.43.0-pyhd8ed1ab_1
250
+ Linking setuptools-70.1.1-pyhd8ed1ab_0
251
+ Linking pip-24.0-pyhd8ed1ab_0
252
+ Linking pycparser-2.22-pyhd8ed1ab_0
253
+ Linking six-1.16.0-pyh6c4a22f_0
254
+ Linking hyperframe-6.0.1-pyhd8ed1ab_0
255
+ Linking pytz-2024.1-pyhd8ed1ab_0
256
+ Linking python-tzdata-2024.1-pyhd8ed1ab_0
257
+ Linking charset-normalizer-3.3.2-pyhd8ed1ab_0
258
+ Linking hpack-4.0.0-pyh9f0ad1d_0
259
+ Linking pysocks-1.7.1-pyha2e5f31_6
260
+ Linking idna-3.7-pyhd8ed1ab_0
261
+ Linking certifi-2024.6.2-pyhd8ed1ab_0
262
+ Linking mpmath-1.3.0-pyhd8ed1ab_0
263
+ Linking typing_extensions-4.12.2-pyha770c72_0
264
+ Linking networkx-3.3-pyhd8ed1ab_1
265
+ Linking filelock-3.15.4-pyhd8ed1ab_0
266
+ Linking python-dateutil-2.9.0-pyhd8ed1ab_0
267
+ Linking h2-4.1.0-pyhd8ed1ab_0
268
+ Linking brotli-python-1.1.0-py311hb755f60_1
269
+ Linking markupsafe-2.1.5-py311h459d7ec_0
270
+ Linking gmpy2-2.1.5-py311hc4f1f91_1
271
+ Linking pyyaml-6.0.1-py311h459d7ec_1
272
+ Linking pillow-9.4.0-py311h50def17_1
273
+ Linking numpy-2.0.0-py311h1461c94_0
274
+ Linking cffi-1.16.0-py311hb3a22ac_0
275
+ Linking pandas-2.2.2-py311h14de704_1
276
+ Linking zstandard-0.22.0-py311hb6f056b_1
277
+ Linking jinja2-3.1.4-pyhd8ed1ab_0
278
+ Linking sympy-1.12.1-pypyh2585a3b_103
279
+ Linking urllib3-2.2.2-pyhd8ed1ab_1
280
+ Linking requests-2.32.3-pyhd8ed1ab_0
281
+ Linking pytorch-2.3.1-py3.11_cuda12.1_cudnn8.9.2_0
282
+ Linking torchtriton-2.3.1-py311
283
+ Linking torchvision-0.18.1-py311_cu121
284
+
285
+ Transaction finished
286
+
287
+ To activate this environment, use:
288
+
289
+ mamba activate auto-zcubaqpyrbpe
290
+
291
+ Or to execute a single command in this environment, use:
292
+
293
+ mamba run -n auto-zcubaqpyrbpe mycommand
294
+
295
+ Installing pip packages
296
+ WARNING: The candidate selected for download or install is a yanked version: 'opencv-python' candidate (version 4.5.5.62 at https://files.pythonhosted.org/packages/9d/98/36bfcbff30da27dd6922ed73ca7802c37d87f77daf4c569da3dcb87b4296/opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (from https://pypi.org/simple/opencv-python/) (requires-python:>=3.6))
297
+ Reason for being yanked: deprecated, use 4.5.5.64
298
+ Traceback (most recent call last):
299
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/metatrain_CausalStyle_RN.py", line 124, in <module>
300
+ base_loader = base_datamgr.get_data_loader( base_file , aug = params.train_aug )
301
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
302
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/data/datamgr.py", line 137, in get_data_loader
303
+ dataset = SetDataset( data_file , self.batch_size, transform )
304
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
305
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/data/dataset.py", line 62, in __init__
306
+ with open(data_file, 'r') as f:
307
+ ^^^^^^^^^^^^^^^^^^^^
308
+ FileNotFoundError: [Errno 2] No such file or directory: '/scratch/yuqian_fu/Data/CDFSL/miniImagenet/base.json'
309
+ srun: error: gcp-eu-2: task 0: Exited with exit code 1
52715.log ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Collecting h5py>=2.9.0
2
+ Downloading h5py-3.11.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (2.5 kB)
3
+ Collecting ml-collections
4
+ Downloading ml_collections-0.1.1.tar.gz (77 kB)
5
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 77.9/77.9 kB 8.8 MB/s eta 0:00:00
6
+ Preparing metadata (setup.py): started
7
+ Preparing metadata (setup.py): finished with status 'done'
8
+ Collecting opencv-python==4.5.5.62
9
+ Downloading opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (18 kB)
10
+ Collecting scipy>=1.3.2
11
+ Downloading scipy-1.14.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (60 kB)
12
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 60.8/60.8 kB 8.1 MB/s eta 0:00:00
13
+ Collecting tensorboard
14
+ Downloading tensorboard-2.17.0-py3-none-any.whl.metadata (1.6 kB)
15
+ Collecting tensorboardX>=1.4
16
+ Downloading tensorboardX-2.6.2.2-py2.py3-none-any.whl.metadata (5.8 kB)
17
+ Collecting timm
18
+ Downloading timm-1.0.7-py3-none-any.whl.metadata (47 kB)
19
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 47.5/47.5 kB 16.0 MB/s eta 0:00:00
20
+ Requirement already satisfied: numpy>=1.21.2 in ./lib/python3.11/site-packages (from opencv-python==4.5.5.62) (2.0.0)
21
+ Collecting absl-py (from ml-collections)
22
+ Downloading absl_py-2.1.0-py3-none-any.whl.metadata (2.3 kB)
23
+ Requirement already satisfied: PyYAML in ./lib/python3.11/site-packages (from ml-collections) (6.0.1)
24
+ Requirement already satisfied: six in ./lib/python3.11/site-packages (from ml-collections) (1.16.0)
25
+ Collecting contextlib2 (from ml-collections)
26
+ Downloading contextlib2-21.6.0-py2.py3-none-any.whl.metadata (4.1 kB)
27
+ Collecting grpcio>=1.48.2 (from tensorboard)
28
+ Downloading grpcio-1.64.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.3 kB)
29
+ Collecting markdown>=2.6.8 (from tensorboard)
30
+ Downloading Markdown-3.6-py3-none-any.whl.metadata (7.0 kB)
31
+ Collecting protobuf!=4.24.0,<5.0.0,>=3.19.6 (from tensorboard)
32
+ Downloading protobuf-4.25.3-cp37-abi3-manylinux2014_x86_64.whl.metadata (541 bytes)
33
+ Requirement already satisfied: setuptools>=41.0.0 in ./lib/python3.11/site-packages (from tensorboard) (70.1.1)
34
+ Collecting tensorboard-data-server<0.8.0,>=0.7.0 (from tensorboard)
35
+ Downloading tensorboard_data_server-0.7.2-py3-none-manylinux_2_31_x86_64.whl.metadata (1.1 kB)
36
+ Collecting werkzeug>=1.0.1 (from tensorboard)
37
+ Downloading werkzeug-3.0.3-py3-none-any.whl.metadata (3.7 kB)
38
+ Collecting packaging (from tensorboardX>=1.4)
39
+ Downloading packaging-24.1-py3-none-any.whl.metadata (3.2 kB)
40
+ Requirement already satisfied: torch in ./lib/python3.11/site-packages (from timm) (2.3.1)
41
+ Requirement already satisfied: torchvision in ./lib/python3.11/site-packages (from timm) (0.18.1)
42
+ Collecting huggingface_hub (from timm)
43
+ Downloading huggingface_hub-0.23.4-py3-none-any.whl.metadata (12 kB)
44
+ Collecting safetensors (from timm)
45
+ Downloading safetensors-0.4.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.8 kB)
46
+ Requirement already satisfied: MarkupSafe>=2.1.1 in ./lib/python3.11/site-packages (from werkzeug>=1.0.1->tensorboard) (2.1.5)
47
+ Requirement already satisfied: filelock in ./lib/python3.11/site-packages (from huggingface_hub->timm) (3.15.4)
48
+ Collecting fsspec>=2023.5.0 (from huggingface_hub->timm)
49
+ Downloading fsspec-2024.6.0-py3-none-any.whl.metadata (11 kB)
50
+ Requirement already satisfied: requests in ./lib/python3.11/site-packages (from huggingface_hub->timm) (2.32.3)
51
+ Collecting tqdm>=4.42.1 (from huggingface_hub->timm)
52
+ Downloading tqdm-4.66.4-py3-none-any.whl.metadata (57 kB)
53
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 57.6/57.6 kB 22.1 MB/s eta 0:00:00
54
+ Requirement already satisfied: typing-extensions>=3.7.4.3 in ./lib/python3.11/site-packages (from huggingface_hub->timm) (4.12.2)
55
+ Requirement already satisfied: sympy in ./lib/python3.11/site-packages (from torch->timm) (1.12.1)
56
+ Requirement already satisfied: networkx in ./lib/python3.11/site-packages (from torch->timm) (3.3)
57
+ Requirement already satisfied: jinja2 in ./lib/python3.11/site-packages (from torch->timm) (3.1.4)
58
+ Requirement already satisfied: pillow!=8.3.*,>=5.3.0 in ./lib/python3.11/site-packages (from torchvision->timm) (9.4.0)
59
+ Requirement already satisfied: charset-normalizer<4,>=2 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (3.3.2)
60
+ Requirement already satisfied: idna<4,>=2.5 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (3.7)
61
+ Requirement already satisfied: urllib3<3,>=1.21.1 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (2.2.2)
62
+ Requirement already satisfied: certifi>=2017.4.17 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (2024.6.2)
63
+ Requirement already satisfied: mpmath<1.4.0,>=1.1.0 in ./lib/python3.11/site-packages (from sympy->torch->timm) (1.3.0)
64
+ Downloading opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (60.4 MB)
65
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 60.4/60.4 MB 91.0 MB/s eta 0:00:00
66
+ Downloading h5py-3.11.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.4 MB)
67
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.4/5.4 MB 183.1 MB/s eta 0:00:00
68
+ Downloading scipy-1.14.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (41.1 MB)
69
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 41.1/41.1 MB 107.5 MB/s eta 0:00:00
70
+ Downloading tensorboard-2.17.0-py3-none-any.whl (5.5 MB)
71
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.5/5.5 MB 30.9 MB/s eta 0:00:00
72
+ Downloading tensorboardX-2.6.2.2-py2.py3-none-any.whl (101 kB)
73
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 101.7/101.7 kB 36.0 MB/s eta 0:00:00
74
+ Downloading timm-1.0.7-py3-none-any.whl (2.3 MB)
75
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.3/2.3 MB 137.1 MB/s eta 0:00:00
76
+ Downloading absl_py-2.1.0-py3-none-any.whl (133 kB)
77
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 133.7/133.7 kB 51.3 MB/s eta 0:00:00
78
+ Downloading grpcio-1.64.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.6 MB)
79
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.6/5.6 MB 88.2 MB/s eta 0:00:00
80
+ Downloading Markdown-3.6-py3-none-any.whl (105 kB)
81
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 105.4/105.4 kB 38.1 MB/s eta 0:00:00
82
+ Downloading protobuf-4.25.3-cp37-abi3-manylinux2014_x86_64.whl (294 kB)
83
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 294.6/294.6 kB 50.7 MB/s eta 0:00:00
84
+ Downloading tensorboard_data_server-0.7.2-py3-none-manylinux_2_31_x86_64.whl (6.6 MB)
85
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 6.6/6.6 MB 173.4 MB/s eta 0:00:00
86
+ Downloading werkzeug-3.0.3-py3-none-any.whl (227 kB)
87
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 227.3/227.3 kB 73.1 MB/s eta 0:00:00
88
+ Downloading contextlib2-21.6.0-py2.py3-none-any.whl (13 kB)
89
+ Downloading huggingface_hub-0.23.4-py3-none-any.whl (402 kB)
90
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 402.6/402.6 kB 61.1 MB/s eta 0:00:00
91
+ Downloading packaging-24.1-py3-none-any.whl (53 kB)
92
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 54.0/54.0 kB 21.8 MB/s eta 0:00:00
93
+ Downloading safetensors-0.4.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.2 MB)
94
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.2/1.2 MB 122.2 MB/s eta 0:00:00
95
+ Downloading fsspec-2024.6.0-py3-none-any.whl (176 kB)
96
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 176.9/176.9 kB 62.7 MB/s eta 0:00:00
97
+ Downloading tqdm-4.66.4-py3-none-any.whl (78 kB)
98
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 78.3/78.3 kB 27.8 MB/s eta 0:00:00
99
+ Building wheels for collected packages: ml-collections
100
+ Building wheel for ml-collections (setup.py): started
101
+ Building wheel for ml-collections (setup.py): finished with status 'done'
102
+ Created wheel for ml-collections: filename=ml_collections-0.1.1-py3-none-any.whl size=94508 sha256=d89d1d746d60ee7c5ccd906afd932a6369bd5c90b009d4e595ac300929458aa5
103
+ Stored in directory: /scratch/yuqian_fu/.cache/pip/wheels/28/82/ef/a6971b09a96519d55ce6efef66f0cbcdef2ae9cc1e6b41daf7
104
+ Successfully built ml-collections
105
+ Installing collected packages: werkzeug, tqdm, tensorboard-data-server, scipy, safetensors, protobuf, packaging, opencv-python, markdown, h5py, grpcio, fsspec, contextlib2, absl-py, tensorboardX, tensorboard, ml-collections, huggingface_hub, timm
106
+ Successfully installed absl-py-2.1.0 contextlib2-21.6.0 fsspec-2024.6.0 grpcio-1.64.1 h5py-3.11.0 huggingface_hub-0.23.4 markdown-3.6 ml-collections-0.1.1 opencv-python-4.5.5.62 packaging-24.1 protobuf-4.25.3 safetensors-0.4.3 scipy-1.14.0 tensorboard-2.17.0 tensorboard-data-server-0.7.2 tensorboardX-2.6.2.2 timm-1.0.7 tqdm-4.66.4 werkzeug-3.0.3
107
+ backbone: maml: False
108
+ hi this is causal style
109
+ set seed = 0
110
+
111
+ --- prepare dataloader ---
112
+ train with single seen domain miniImagenet
113
+
114
+ --- build model ---
52729.error ADDED
@@ -0,0 +1,309 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Solving dependencies
2
+ Installing conda packages
3
+ Empty environment created at prefix: /scratch/yuqian_fu/micromamba/envs/auto-zcubaqpyrbpe
4
+ error libmamba Could not lock non-existing path '/scratch/yuqian_fu/micromamba/pkgs'
5
+ Transaction
6
+
7
+ Prefix: /scratch/yuqian_fu/micromamba/envs/auto-zcubaqpyrbpe
8
+
9
+
10
+
11
+ No specs added or removed.
12
+
13
+ Package Version Build Channel Size
14
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
15
+ Install:
16
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
17
+
18
+ + _libgcc_mutex 0.1 conda_forge conda-forge
19
+ + _openmp_mutex 4.5 2_kmp_llvm conda-forge
20
+ + blas 2.116 mkl conda-forge
21
+ + blas-devel 3.9.0 16_linux64_mkl conda-forge
22
+ + brotli-python 1.1.0 py311hb755f60_1 conda-forge
23
+ + bzip2 1.0.8 hd590300_5 conda-forge
24
+ + ca-certificates 2024.6.2 hbcca054_0 conda-forge
25
+ + certifi 2024.6.2 pyhd8ed1ab_0 conda-forge
26
+ + cffi 1.16.0 py311hb3a22ac_0 conda-forge
27
+ + charset-normalizer 3.3.2 pyhd8ed1ab_0 conda-forge
28
+ + cuda-cudart 12.1.105 0 nvidia
29
+ + cuda-cupti 12.1.105 0 /work/conda/cache/nvidia
30
+ + cuda-libraries 12.1.0 0 nvidia
31
+ + cuda-nvrtc 12.1.105 0 /work/conda/cache/nvidia
32
+ + cuda-nvtx 12.1.105 0 nvidia
33
+ + cuda-opencl 12.5.39 0 nvidia
34
+ + cuda-runtime 12.1.0 0 nvidia
35
+ + cuda-version 12.5 3 nvidia
36
+ + ffmpeg 4.3 hf484d3e_0 /work/conda/cache/pytorch
37
+ + filelock 3.15.4 pyhd8ed1ab_0 conda-forge
38
+ + freetype 2.12.1 h267a509_2 conda-forge
39
+ + gmp 6.3.0 hac33072_2 conda-forge
40
+ + gmpy2 2.1.5 py311hc4f1f91_1 conda-forge
41
+ + gnutls 3.6.13 h85f3911_1 /work/conda/cache/conda-forge
42
+ + h2 4.1.0 pyhd8ed1ab_0 conda-forge
43
+ + hpack 4.0.0 pyh9f0ad1d_0 conda-forge
44
+ + hyperframe 6.0.1 pyhd8ed1ab_0 conda-forge
45
+ + icu 73.2 h59595ed_0 /work/conda/cache/conda-forge
46
+ + idna 3.7 pyhd8ed1ab_0 conda-forge
47
+ + jinja2 3.1.4 pyhd8ed1ab_0 conda-forge
48
+ + jpeg 9e h166bdaf_2 conda-forge
49
+ + lame 3.100 h166bdaf_1003 conda-forge
50
+ + lcms2 2.15 hfd0df8a_0 conda-forge
51
+ + ld_impl_linux-64 2.40 hf3520f5_7 conda-forge
52
+ + lerc 4.0.0 h27087fc_0 conda-forge
53
+ + libblas 3.9.0 16_linux64_mkl conda-forge
54
+ + libcblas 3.9.0 16_linux64_mkl conda-forge
55
+ + libcublas 12.1.0.26 0 /work/conda/cache/nvidia
56
+ + libcufft 11.0.2.4 0 /work/conda/cache/nvidia
57
+ + libcufile 1.10.0.4 0 nvidia
58
+ + libcurand 10.3.6.39 0 nvidia
59
+ + libcusolver 11.4.4.55 0 /work/conda/cache/nvidia
60
+ + libcusparse 12.0.2.55 0 /work/conda/cache/nvidia
61
+ + libdeflate 1.17 h0b41bf4_0 conda-forge
62
+ + libexpat 2.6.2 h59595ed_0 conda-forge
63
+ + libffi 3.4.2 h7f98852_5 conda-forge
64
+ + libgcc-ng 13.2.0 h77fa898_13 conda-forge
65
+ + libgfortran-ng 13.2.0 h69a702a_13 conda-forge
66
+ + libgfortran5 13.2.0 h3d2ce59_13 conda-forge
67
+ + libhwloc 2.10.0 default_h5622ce7_1001 conda-forge
68
+ + libiconv 1.17 hd590300_2 conda-forge
69
+ + libjpeg-turbo 2.0.0 h9bf148f_0 pytorch
70
+ + liblapack 3.9.0 16_linux64_mkl conda-forge
71
+ + liblapacke 3.9.0 16_linux64_mkl conda-forge
72
+ + libnpp 12.0.2.50 0 /work/conda/cache/nvidia
73
+ + libnsl 2.0.1 hd590300_0 conda-forge
74
+ + libnvjitlink 12.1.105 0 /work/conda/cache/nvidia
75
+ + libnvjpeg 12.1.1.14 0 /work/conda/cache/nvidia
76
+ + libpng 1.6.43 h2797004_0 conda-forge
77
+ + libsqlite 3.46.0 hde9e2c9_0 conda-forge
78
+ + libstdcxx-ng 13.2.0 hc0a3c3a_13 conda-forge
79
+ + libtiff 4.5.0 h6adf6a1_2 conda-forge
80
+ + libuuid 2.38.1 h0b41bf4_0 conda-forge
81
+ + libwebp-base 1.4.0 hd590300_0 conda-forge
82
+ + libxcb 1.13 h7f98852_1004 conda-forge
83
+ + libxcrypt 4.4.36 hd590300_1 conda-forge
84
+ + libxml2 2.12.7 hc051c1a_1 conda-forge
85
+ + libzlib 1.2.13 h4ab18f5_6 conda-forge
86
+ + llvm-openmp 15.0.7 h0cdce71_0 /work/conda/cache/conda-forge
87
+ + markupsafe 2.1.5 py311h459d7ec_0 conda-forge
88
+ + mkl 2022.1.0 h84fe81f_915 /work/conda/cache/conda-forge
89
+ + mkl-devel 2022.1.0 ha770c72_916 conda-forge
90
+ + mkl-include 2022.1.0 h84fe81f_915 conda-forge
91
+ + mpc 1.3.1 hfe3b2da_0 conda-forge
92
+ + mpfr 4.2.1 h9458935_1 conda-forge
93
+ + mpmath 1.3.0 pyhd8ed1ab_0 conda-forge
94
+ + ncurses 6.5 h59595ed_0 conda-forge
95
+ + nettle 3.6 he412f7d_0 /work/conda/cache/conda-forge
96
+ + networkx 3.3 pyhd8ed1ab_1 /work/conda/cache/conda-forge
97
+ + numpy 2.0.0 py311h1461c94_0 conda-forge
98
+ + openh264 2.1.1 h780b84a_0 /work/conda/cache/conda-forge
99
+ + openjpeg 2.5.0 hfec8fc6_2 conda-forge
100
+ + openssl 3.3.1 h4ab18f5_1 conda-forge
101
+ + pandas 2.2.2 py311h14de704_1 conda-forge
102
+ + pillow 9.4.0 py311h50def17_1 conda-forge
103
+ + pip 24.0 pyhd8ed1ab_0 /work/conda/cache/conda-forge
104
+ + pthread-stubs 0.4 h36c2ea0_1001 conda-forge
105
+ + pycparser 2.22 pyhd8ed1ab_0 conda-forge
106
+ + pysocks 1.7.1 pyha2e5f31_6 conda-forge
107
+ + python 3.11.9 hb806964_0_cpython /work/conda/cache/conda-forge
108
+ + python-dateutil 2.9.0 pyhd8ed1ab_0 conda-forge
109
+ + python-tzdata 2024.1 pyhd8ed1ab_0 conda-forge
110
+ + python_abi 3.11 4_cp311 conda-forge
111
+ + pytorch 2.3.1 py3.11_cuda12.1_cudnn8.9.2_0 pytorch
112
+ + pytorch-cuda 12.1 ha16c6d3_5 pytorch
113
+ + pytorch-mutex 1.0 cuda pytorch
114
+ + pytz 2024.1 pyhd8ed1ab_0 conda-forge
115
+ + pyyaml 6.0.1 py311h459d7ec_1 conda-forge
116
+ + readline 8.2 h8228510_1 conda-forge
117
+ + requests 2.32.3 pyhd8ed1ab_0 conda-forge
118
+ + setuptools 70.1.1 pyhd8ed1ab_0 conda-forge
119
+ + six 1.16.0 pyh6c4a22f_0 conda-forge
120
+ + sympy 1.12.1 pypyh2585a3b_103 conda-forge
121
+ + tbb 2021.12.0 h297d8ca_1 conda-forge
122
+ + tk 8.6.13 noxft_h4845f30_101 /work/conda/cache/conda-forge
123
+ + torchtriton 2.3.1 py311 pytorch
124
+ + torchvision 0.18.1 py311_cu121 pytorch
125
+ + typing_extensions 4.12.2 pyha770c72_0 conda-forge
126
+ + tzdata 2024a h0c530f3_0 conda-forge
127
+ + urllib3 2.2.2 pyhd8ed1ab_1 conda-forge
128
+ + wheel 0.43.0 pyhd8ed1ab_1 conda-forge
129
+ + xorg-libxau 1.0.11 hd590300_0 conda-forge
130
+ + xorg-libxdmcp 1.1.3 h516909a_0 conda-forge
131
+ + xz 5.2.6 h166bdaf_0 conda-forge
132
+ + yaml 0.2.5 h7f98852_2 conda-forge
133
+ + zlib 1.2.13 h4ab18f5_6 conda-forge
134
+ + zstandard 0.22.0 py311hb6f056b_1 conda-forge
135
+ + zstd 1.5.6 ha6fb4c9_0 conda-forge
136
+
137
+ Summary:
138
+
139
+ Install: 118 packages
140
+
141
+ Total download: 0 B
142
+
143
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
144
+
145
+
146
+
147
+ Transaction starting
148
+ Linking libcublas-12.1.0.26-0
149
+ Linking libcufft-11.0.2.4-0
150
+ Linking libcusolver-11.4.4.55-0
151
+ Linking libcusparse-12.0.2.55-0
152
+ Linking libnpp-12.0.2.50-0
153
+ Linking libnvjitlink-12.1.105-0
154
+ Linking cuda-cudart-12.1.105-0
155
+ Linking cuda-nvrtc-12.1.105-0
156
+ Linking libnvjpeg-12.1.1.14-0
157
+ Linking cuda-cupti-12.1.105-0
158
+ Linking cuda-nvtx-12.1.105-0
159
+ Linking pytorch-mutex-1.0-cuda
160
+ Linking _libgcc_mutex-0.1-conda_forge
161
+ Linking mkl-include-2022.1.0-h84fe81f_915
162
+ Linking python_abi-3.11-4_cp311
163
+ Linking ld_impl_linux-64-2.40-hf3520f5_7
164
+ Linking ca-certificates-2024.6.2-hbcca054_0
165
+ Linking libgcc-ng-13.2.0-h77fa898_13
166
+ Linking libzlib-1.2.13-h4ab18f5_6
167
+ Linking llvm-openmp-15.0.7-h0cdce71_0
168
+ Linking _openmp_mutex-4.5-2_kmp_llvm
169
+ Linking xorg-libxdmcp-1.1.3-h516909a_0
170
+ Linking pthread-stubs-0.4-h36c2ea0_1001
171
+ Linking xorg-libxau-1.0.11-hd590300_0
172
+ Linking libwebp-base-1.4.0-hd590300_0
173
+ Linking libdeflate-1.17-h0b41bf4_0
174
+ Linking jpeg-9e-h166bdaf_2
175
+ Linking libffi-3.4.2-h7f98852_5
176
+ Linking tk-8.6.13-noxft_h4845f30_101
177
+ Linking openssl-3.3.1-h4ab18f5_1
178
+ Linking libxcrypt-4.4.36-hd590300_1
179
+ Linking libsqlite-3.46.0-hde9e2c9_0
180
+ Linking yaml-0.2.5-h7f98852_2
181
+ Linking ncurses-6.5-h59595ed_0
182
+ Linking libgfortran5-13.2.0-h3d2ce59_13
183
+ Linking lame-3.100-h166bdaf_1003
184
+ Linking nettle-3.6-he412f7d_0
185
+ Linking zlib-1.2.13-h4ab18f5_6
186
+ Linking libstdcxx-ng-13.2.0-hc0a3c3a_13
187
+ Linking libiconv-1.17-hd590300_2
188
+ Linking bzip2-1.0.8-hd590300_5
189
+ Linking libpng-1.6.43-h2797004_0
190
+ Linking xz-5.2.6-h166bdaf_0
191
+ Linking libuuid-2.38.1-h0b41bf4_0
192
+ Linking libnsl-2.0.1-hd590300_0
193
+ Linking libexpat-2.6.2-h59595ed_0
194
+ Linking libxcb-1.13-h7f98852_1004
195
+ Linking readline-8.2-h8228510_1
196
+ Linking libgfortran-ng-13.2.0-h69a702a_13
197
+ Linking icu-73.2-h59595ed_0
198
+ Linking zstd-1.5.6-ha6fb4c9_0
199
+ Linking lerc-4.0.0-h27087fc_0
200
+ Linking openh264-2.1.1-h780b84a_0
201
+ Linking gnutls-3.6.13-h85f3911_1
202
+ Linking gmp-6.3.0-hac33072_2
203
+ Linking freetype-2.12.1-h267a509_2
204
+ Linking libxml2-2.12.7-hc051c1a_1
205
+ Linking libtiff-4.5.0-h6adf6a1_2
206
+ Linking mpfr-4.2.1-h9458935_1
207
+ Linking libhwloc-2.10.0-default_h5622ce7_1001
208
+ Linking openjpeg-2.5.0-hfec8fc6_2
209
+ Linking lcms2-2.15-hfd0df8a_0
210
+ Linking mpc-1.3.1-hfe3b2da_0
211
+ Linking tbb-2021.12.0-h297d8ca_1
212
+ Linking mkl-2022.1.0-h84fe81f_915
213
+ Linking mkl-devel-2022.1.0-ha770c72_916
214
+ Linking libblas-3.9.0-16_linux64_mkl
215
+ Linking liblapack-3.9.0-16_linux64_mkl
216
+ Linking libcblas-3.9.0-16_linux64_mkl
217
+ Linking liblapacke-3.9.0-16_linux64_mkl
218
+ Linking blas-devel-3.9.0-16_linux64_mkl
219
+ Linking blas-2.116-mkl
220
+ Linking cuda-version-12.5-3
221
+ Linking tzdata-2024a-h0c530f3_0
222
+ Linking libjpeg-turbo-2.0.0-h9bf148f_0
223
+ warning libmamba [libjpeg-turbo-2.0.0-h9bf148f_0] The following files were already present in the environment:
224
+ - bin/cjpeg
225
+ - bin/djpeg
226
+ - bin/jpegtran
227
+ - bin/rdjpgcom
228
+ - bin/wrjpgcom
229
+ - include/jconfig.h
230
+ - include/jerror.h
231
+ - include/jmorecfg.h
232
+ - include/jpeglib.h
233
+ - lib/libjpeg.a
234
+ - lib/libjpeg.so
235
+ - lib/pkgconfig/libjpeg.pc
236
+ - share/man/man1/cjpeg.1
237
+ - share/man/man1/djpeg.1
238
+ - share/man/man1/jpegtran.1
239
+ - share/man/man1/rdjpgcom.1
240
+ - share/man/man1/wrjpgcom.1
241
+ Linking ffmpeg-4.3-hf484d3e_0
242
+ Linking libcurand-10.3.6.39-0
243
+ Linking libcufile-1.10.0.4-0
244
+ Linking cuda-opencl-12.5.39-0
245
+ Linking cuda-libraries-12.1.0-0
246
+ Linking cuda-runtime-12.1.0-0
247
+ Linking python-3.11.9-hb806964_0_cpython
248
+ Linking pytorch-cuda-12.1-ha16c6d3_5
249
+ Linking wheel-0.43.0-pyhd8ed1ab_1
250
+ Linking setuptools-70.1.1-pyhd8ed1ab_0
251
+ Linking pip-24.0-pyhd8ed1ab_0
252
+ Linking pycparser-2.22-pyhd8ed1ab_0
253
+ Linking six-1.16.0-pyh6c4a22f_0
254
+ Linking hyperframe-6.0.1-pyhd8ed1ab_0
255
+ Linking pytz-2024.1-pyhd8ed1ab_0
256
+ Linking python-tzdata-2024.1-pyhd8ed1ab_0
257
+ Linking charset-normalizer-3.3.2-pyhd8ed1ab_0
258
+ Linking hpack-4.0.0-pyh9f0ad1d_0
259
+ Linking pysocks-1.7.1-pyha2e5f31_6
260
+ Linking idna-3.7-pyhd8ed1ab_0
261
+ Linking certifi-2024.6.2-pyhd8ed1ab_0
262
+ Linking mpmath-1.3.0-pyhd8ed1ab_0
263
+ Linking typing_extensions-4.12.2-pyha770c72_0
264
+ Linking networkx-3.3-pyhd8ed1ab_1
265
+ Linking filelock-3.15.4-pyhd8ed1ab_0
266
+ Linking python-dateutil-2.9.0-pyhd8ed1ab_0
267
+ Linking h2-4.1.0-pyhd8ed1ab_0
268
+ Linking brotli-python-1.1.0-py311hb755f60_1
269
+ Linking markupsafe-2.1.5-py311h459d7ec_0
270
+ Linking gmpy2-2.1.5-py311hc4f1f91_1
271
+ Linking pyyaml-6.0.1-py311h459d7ec_1
272
+ Linking pillow-9.4.0-py311h50def17_1
273
+ Linking numpy-2.0.0-py311h1461c94_0
274
+ Linking cffi-1.16.0-py311hb3a22ac_0
275
+ Linking pandas-2.2.2-py311h14de704_1
276
+ Linking zstandard-0.22.0-py311hb6f056b_1
277
+ Linking jinja2-3.1.4-pyhd8ed1ab_0
278
+ Linking sympy-1.12.1-pypyh2585a3b_103
279
+ Linking urllib3-2.2.2-pyhd8ed1ab_1
280
+ Linking requests-2.32.3-pyhd8ed1ab_0
281
+ Linking pytorch-2.3.1-py3.11_cuda12.1_cudnn8.9.2_0
282
+ Linking torchtriton-2.3.1-py311
283
+ Linking torchvision-0.18.1-py311_cu121
284
+
285
+ Transaction finished
286
+
287
+ To activate this environment, use:
288
+
289
+ mamba activate auto-zcubaqpyrbpe
290
+
291
+ Or to execute a single command in this environment, use:
292
+
293
+ mamba run -n auto-zcubaqpyrbpe mycommand
294
+
295
+ Installing pip packages
296
+ WARNING: The candidate selected for download or install is a yanked version: 'opencv-python' candidate (version 4.5.5.62 at https://files.pythonhosted.org/packages/9d/98/36bfcbff30da27dd6922ed73ca7802c37d87f77daf4c569da3dcb87b4296/opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (from https://pypi.org/simple/opencv-python/) (requires-python:>=3.6))
297
+ Reason for being yanked: deprecated, use 4.5.5.64
298
+ Traceback (most recent call last):
299
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/metatrain_CausalStyle_RN.py", line 124, in <module>
300
+ base_loader = base_datamgr.get_data_loader( base_file , aug = params.train_aug )
301
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
302
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/data/datamgr.py", line 137, in get_data_loader
303
+ dataset = SetDataset( data_file , self.batch_size, transform )
304
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
305
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/data/dataset.py", line 62, in __init__
306
+ with open(data_file, 'r') as f:
307
+ ^^^^^^^^^^^^^^^^^^^^
308
+ FileNotFoundError: [Errno 2] No such file or directory: '/scratch/yuqian_fu/Data/CDFSL/miniImagenet/base.json'
309
+ srun: error: gcpl4-eu-2: task 0: Exited with exit code 1
52729.log ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Collecting h5py>=2.9.0
2
+ Downloading h5py-3.11.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (2.5 kB)
3
+ Collecting ml-collections
4
+ Downloading ml_collections-0.1.1.tar.gz (77 kB)
5
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 77.9/77.9 kB 8.5 MB/s eta 0:00:00
6
+ Preparing metadata (setup.py): started
7
+ Preparing metadata (setup.py): finished with status 'done'
8
+ Collecting opencv-python==4.5.5.62
9
+ Downloading opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (18 kB)
10
+ Collecting scipy>=1.3.2
11
+ Downloading scipy-1.14.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (60 kB)
12
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 60.8/60.8 kB 10.4 MB/s eta 0:00:00
13
+ Collecting tensorboard
14
+ Downloading tensorboard-2.17.0-py3-none-any.whl.metadata (1.6 kB)
15
+ Collecting tensorboardX>=1.4
16
+ Downloading tensorboardX-2.6.2.2-py2.py3-none-any.whl.metadata (5.8 kB)
17
+ Collecting timm
18
+ Downloading timm-1.0.7-py3-none-any.whl.metadata (47 kB)
19
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 47.5/47.5 kB 16.7 MB/s eta 0:00:00
20
+ Requirement already satisfied: numpy>=1.21.2 in ./lib/python3.11/site-packages (from opencv-python==4.5.5.62) (2.0.0)
21
+ Collecting absl-py (from ml-collections)
22
+ Downloading absl_py-2.1.0-py3-none-any.whl.metadata (2.3 kB)
23
+ Requirement already satisfied: PyYAML in ./lib/python3.11/site-packages (from ml-collections) (6.0.1)
24
+ Requirement already satisfied: six in ./lib/python3.11/site-packages (from ml-collections) (1.16.0)
25
+ Collecting contextlib2 (from ml-collections)
26
+ Downloading contextlib2-21.6.0-py2.py3-none-any.whl.metadata (4.1 kB)
27
+ Collecting grpcio>=1.48.2 (from tensorboard)
28
+ Downloading grpcio-1.64.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.3 kB)
29
+ Collecting markdown>=2.6.8 (from tensorboard)
30
+ Downloading Markdown-3.6-py3-none-any.whl.metadata (7.0 kB)
31
+ Collecting protobuf!=4.24.0,<5.0.0,>=3.19.6 (from tensorboard)
32
+ Downloading protobuf-4.25.3-cp37-abi3-manylinux2014_x86_64.whl.metadata (541 bytes)
33
+ Requirement already satisfied: setuptools>=41.0.0 in ./lib/python3.11/site-packages (from tensorboard) (70.1.1)
34
+ Collecting tensorboard-data-server<0.8.0,>=0.7.0 (from tensorboard)
35
+ Downloading tensorboard_data_server-0.7.2-py3-none-manylinux_2_31_x86_64.whl.metadata (1.1 kB)
36
+ Collecting werkzeug>=1.0.1 (from tensorboard)
37
+ Downloading werkzeug-3.0.3-py3-none-any.whl.metadata (3.7 kB)
38
+ Collecting packaging (from tensorboardX>=1.4)
39
+ Downloading packaging-24.1-py3-none-any.whl.metadata (3.2 kB)
40
+ Requirement already satisfied: torch in ./lib/python3.11/site-packages (from timm) (2.3.1)
41
+ Requirement already satisfied: torchvision in ./lib/python3.11/site-packages (from timm) (0.18.1)
42
+ Collecting huggingface_hub (from timm)
43
+ Downloading huggingface_hub-0.23.4-py3-none-any.whl.metadata (12 kB)
44
+ Collecting safetensors (from timm)
45
+ Downloading safetensors-0.4.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.8 kB)
46
+ Requirement already satisfied: MarkupSafe>=2.1.1 in ./lib/python3.11/site-packages (from werkzeug>=1.0.1->tensorboard) (2.1.5)
47
+ Requirement already satisfied: filelock in ./lib/python3.11/site-packages (from huggingface_hub->timm) (3.15.4)
48
+ Collecting fsspec>=2023.5.0 (from huggingface_hub->timm)
49
+ Downloading fsspec-2024.6.0-py3-none-any.whl.metadata (11 kB)
50
+ Requirement already satisfied: requests in ./lib/python3.11/site-packages (from huggingface_hub->timm) (2.32.3)
51
+ Collecting tqdm>=4.42.1 (from huggingface_hub->timm)
52
+ Downloading tqdm-4.66.4-py3-none-any.whl.metadata (57 kB)
53
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 57.6/57.6 kB 24.9 MB/s eta 0:00:00
54
+ Requirement already satisfied: typing-extensions>=3.7.4.3 in ./lib/python3.11/site-packages (from huggingface_hub->timm) (4.12.2)
55
+ Requirement already satisfied: sympy in ./lib/python3.11/site-packages (from torch->timm) (1.12.1)
56
+ Requirement already satisfied: networkx in ./lib/python3.11/site-packages (from torch->timm) (3.3)
57
+ Requirement already satisfied: jinja2 in ./lib/python3.11/site-packages (from torch->timm) (3.1.4)
58
+ Requirement already satisfied: pillow!=8.3.*,>=5.3.0 in ./lib/python3.11/site-packages (from torchvision->timm) (9.4.0)
59
+ Requirement already satisfied: charset-normalizer<4,>=2 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (3.3.2)
60
+ Requirement already satisfied: idna<4,>=2.5 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (3.7)
61
+ Requirement already satisfied: urllib3<3,>=1.21.1 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (2.2.2)
62
+ Requirement already satisfied: certifi>=2017.4.17 in ./lib/python3.11/site-packages (from requests->huggingface_hub->timm) (2024.6.2)
63
+ Requirement already satisfied: mpmath<1.4.0,>=1.1.0 in ./lib/python3.11/site-packages (from sympy->torch->timm) (1.3.0)
64
+ Downloading opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (60.4 MB)
65
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 60.4/60.4 MB 72.8 MB/s eta 0:00:00
66
+ Downloading h5py-3.11.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.4 MB)
67
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.4/5.4 MB 209.0 MB/s eta 0:00:00
68
+ Downloading scipy-1.14.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (41.1 MB)
69
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 41.1/41.1 MB 115.1 MB/s eta 0:00:00
70
+ Downloading tensorboard-2.17.0-py3-none-any.whl (5.5 MB)
71
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.5/5.5 MB 211.7 MB/s eta 0:00:00
72
+ Downloading tensorboardX-2.6.2.2-py2.py3-none-any.whl (101 kB)
73
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 101.7/101.7 kB 38.7 MB/s eta 0:00:00
74
+ Downloading timm-1.0.7-py3-none-any.whl (2.3 MB)
75
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.3/2.3 MB 181.5 MB/s eta 0:00:00
76
+ Downloading absl_py-2.1.0-py3-none-any.whl (133 kB)
77
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 133.7/133.7 kB 52.8 MB/s eta 0:00:00
78
+ Downloading grpcio-1.64.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.6 MB)
79
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 5.6/5.6 MB 212.4 MB/s eta 0:00:00
80
+ Downloading Markdown-3.6-py3-none-any.whl (105 kB)
81
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 105.4/105.4 kB 40.3 MB/s eta 0:00:00
82
+ Downloading protobuf-4.25.3-cp37-abi3-manylinux2014_x86_64.whl (294 kB)
83
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 294.6/294.6 kB 101.1 MB/s eta 0:00:00
84
+ Downloading tensorboard_data_server-0.7.2-py3-none-manylinux_2_31_x86_64.whl (6.6 MB)
85
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 6.6/6.6 MB 214.7 MB/s eta 0:00:00
86
+ Downloading werkzeug-3.0.3-py3-none-any.whl (227 kB)
87
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 227.3/227.3 kB 74.9 MB/s eta 0:00:00
88
+ Downloading contextlib2-21.6.0-py2.py3-none-any.whl (13 kB)
89
+ Downloading huggingface_hub-0.23.4-py3-none-any.whl (402 kB)
90
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 402.6/402.6 kB 109.9 MB/s eta 0:00:00
91
+ Downloading packaging-24.1-py3-none-any.whl (53 kB)
92
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 54.0/54.0 kB 21.8 MB/s eta 0:00:00
93
+ Downloading safetensors-0.4.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.2 MB)
94
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.2/1.2 MB 170.9 MB/s eta 0:00:00
95
+ Downloading fsspec-2024.6.0-py3-none-any.whl (176 kB)
96
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 176.9/176.9 kB 62.5 MB/s eta 0:00:00
97
+ Downloading tqdm-4.66.4-py3-none-any.whl (78 kB)
98
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 78.3/78.3 kB 30.5 MB/s eta 0:00:00
99
+ Building wheels for collected packages: ml-collections
100
+ Building wheel for ml-collections (setup.py): started
101
+ Building wheel for ml-collections (setup.py): finished with status 'done'
102
+ Created wheel for ml-collections: filename=ml_collections-0.1.1-py3-none-any.whl size=94508 sha256=2e320bb7bf02566bf671fd943ea8dfe7cb6c35a1fab523a080d4ab487706ca51
103
+ Stored in directory: /scratch/yuqian_fu/.cache/pip/wheels/28/82/ef/a6971b09a96519d55ce6efef66f0cbcdef2ae9cc1e6b41daf7
104
+ Successfully built ml-collections
105
+ Installing collected packages: werkzeug, tqdm, tensorboard-data-server, scipy, safetensors, protobuf, packaging, opencv-python, markdown, h5py, grpcio, fsspec, contextlib2, absl-py, tensorboardX, tensorboard, ml-collections, huggingface_hub, timm
106
+ Successfully installed absl-py-2.1.0 contextlib2-21.6.0 fsspec-2024.6.0 grpcio-1.64.1 h5py-3.11.0 huggingface_hub-0.23.4 markdown-3.6 ml-collections-0.1.1 opencv-python-4.5.5.62 packaging-24.1 protobuf-4.25.3 safetensors-0.4.3 scipy-1.14.0 tensorboard-2.17.0 tensorboard-data-server-0.7.2 tensorboardX-2.6.2.2 timm-1.0.7 tqdm-4.66.4 werkzeug-3.0.3
107
+ backbone: maml: False
108
+ hi this is causal style
109
+ set seed = 0
110
+
111
+ --- prepare dataloader ---
112
+ train with single seen domain miniImagenet
113
+
114
+ --- build model ---
Meta-causal/code-stage1-pipeline/56451.error ADDED
@@ -0,0 +1,297 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Solving dependencies
2
+ Installing conda packages
3
+ Empty environment created at prefix: /scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem
4
+ error libmamba Could not lock non-existing path '/scratch/yuqian_fu/micromamba/pkgs'
5
+ Transaction
6
+
7
+ Prefix: /scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem
8
+
9
+
10
+
11
+ No specs added or removed.
12
+
13
+ Package Version Build Channel Size
14
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
15
+ Install:
16
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
17
+
18
+ + _libgcc_mutex 0.1 conda_forge conda-forge
19
+ + _openmp_mutex 4.5 2_kmp_llvm conda-forge
20
+ + blas 2.116 mkl conda-forge
21
+ + blas-devel 3.9.0 16_linux64_mkl conda-forge
22
+ + brotli-python 1.1.0 py311hb755f60_1 conda-forge
23
+ + bzip2 1.0.8 hd590300_5 conda-forge
24
+ + ca-certificates 2024.6.2 hbcca054_0 conda-forge
25
+ + certifi 2024.6.2 pyhd8ed1ab_0 conda-forge
26
+ + cffi 1.16.0 py311hb3a22ac_0 conda-forge
27
+ + charset-normalizer 3.3.2 pyhd8ed1ab_0 conda-forge
28
+ + click 8.1.7 unix_pyh707e725_0 conda-forge
29
+ + cuda-cudart 12.1.105 0 nvidia
30
+ + cuda-cupti 12.1.105 0 /work/conda/cache/nvidia
31
+ + cuda-libraries 12.1.0 0 nvidia
32
+ + cuda-nvrtc 12.1.105 0 /work/conda/cache/nvidia
33
+ + cuda-nvtx 12.1.105 0 nvidia
34
+ + cuda-opencl 12.5.39 0 nvidia
35
+ + cuda-runtime 12.1.0 0 nvidia
36
+ + cuda-version 12.5 3 nvidia
37
+ + ffmpeg 4.3 hf484d3e_0 /work/conda/cache/pytorch
38
+ + filelock 3.15.4 pyhd8ed1ab_0 conda-forge
39
+ + freetype 2.12.1 h267a509_2 conda-forge
40
+ + gmp 6.3.0 hac33072_2 conda-forge
41
+ + gmpy2 2.1.5 py311hc4f1f91_1 conda-forge
42
+ + gnutls 3.6.13 h85f3911_1 /work/conda/cache/conda-forge
43
+ + h2 4.1.0 pyhd8ed1ab_0 conda-forge
44
+ + hpack 4.0.0 pyh9f0ad1d_0 conda-forge
45
+ + hyperframe 6.0.1 pyhd8ed1ab_0 conda-forge
46
+ + icu 73.2 h59595ed_0 /work/conda/cache/conda-forge
47
+ + idna 3.7 pyhd8ed1ab_0 conda-forge
48
+ + jinja2 3.1.4 pyhd8ed1ab_0 conda-forge
49
+ + jpeg 9e h166bdaf_2 conda-forge
50
+ + lame 3.100 h166bdaf_1003 conda-forge
51
+ + lcms2 2.15 hfd0df8a_0 conda-forge
52
+ + ld_impl_linux-64 2.40 hf3520f5_7 conda-forge
53
+ + lerc 4.0.0 h27087fc_0 conda-forge
54
+ + libblas 3.9.0 16_linux64_mkl conda-forge
55
+ + libcblas 3.9.0 16_linux64_mkl conda-forge
56
+ + libcublas 12.1.0.26 0 /work/conda/cache/nvidia
57
+ + libcufft 11.0.2.4 0 /work/conda/cache/nvidia
58
+ + libcufile 1.10.0.4 0 nvidia
59
+ + libcurand 10.3.6.39 0 nvidia
60
+ + libcusolver 11.4.4.55 0 /work/conda/cache/nvidia
61
+ + libcusparse 12.0.2.55 0 /work/conda/cache/nvidia
62
+ + libdeflate 1.17 h0b41bf4_0 conda-forge
63
+ + libexpat 2.6.2 h59595ed_0 conda-forge
64
+ + libffi 3.4.2 h7f98852_5 conda-forge
65
+ + libgcc-ng 14.1.0 h77fa898_0 conda-forge
66
+ + libgfortran-ng 14.1.0 h69a702a_0 conda-forge
67
+ + libgfortran5 14.1.0 hc5f4f2c_0 conda-forge
68
+ + libhwloc 2.10.0 default_h5622ce7_1001 conda-forge
69
+ + libiconv 1.17 hd590300_2 conda-forge
70
+ + libjpeg-turbo 2.0.0 h9bf148f_0 pytorch
71
+ + liblapack 3.9.0 16_linux64_mkl conda-forge
72
+ + liblapacke 3.9.0 16_linux64_mkl conda-forge
73
+ + libnpp 12.0.2.50 0 /work/conda/cache/nvidia
74
+ + libnsl 2.0.1 hd590300_0 conda-forge
75
+ + libnvjitlink 12.1.105 0 /work/conda/cache/nvidia
76
+ + libnvjpeg 12.1.1.14 0 /work/conda/cache/nvidia
77
+ + libpng 1.6.43 h2797004_0 conda-forge
78
+ + libsqlite 3.46.0 hde9e2c9_0 conda-forge
79
+ + libstdcxx-ng 14.1.0 hc0a3c3a_0 conda-forge
80
+ + libtiff 4.5.0 h6adf6a1_2 conda-forge
81
+ + libuuid 2.38.1 h0b41bf4_0 conda-forge
82
+ + libwebp-base 1.4.0 hd590300_0 conda-forge
83
+ + libxcb 1.13 h7f98852_1004 conda-forge
84
+ + libxcrypt 4.4.36 hd590300_1 conda-forge
85
+ + libxml2 2.12.7 hc051c1a_1 conda-forge
86
+ + libzlib 1.2.13 h4ab18f5_6 conda-forge
87
+ + llvm-openmp 15.0.7 h0cdce71_0 /work/conda/cache/conda-forge
88
+ + markupsafe 2.1.5 py311h459d7ec_0 conda-forge
89
+ + mkl 2022.1.0 h84fe81f_915 /work/conda/cache/conda-forge
90
+ + mkl-devel 2022.1.0 ha770c72_916 conda-forge
91
+ + mkl-include 2022.1.0 h84fe81f_915 conda-forge
92
+ + mpc 1.3.1 hfe3b2da_0 conda-forge
93
+ + mpfr 4.2.1 h9458935_1 conda-forge
94
+ + mpmath 1.3.0 pyhd8ed1ab_0 conda-forge
95
+ + ncurses 6.5 h59595ed_0 conda-forge
96
+ + nettle 3.6 he412f7d_0 /work/conda/cache/conda-forge
97
+ + networkx 3.3 pyhd8ed1ab_1 /work/conda/cache/conda-forge
98
+ + numpy 2.0.0 py311h1461c94_0 conda-forge
99
+ + openh264 2.1.1 h780b84a_0 /work/conda/cache/conda-forge
100
+ + openjpeg 2.5.0 hfec8fc6_2 conda-forge
101
+ + openssl 3.3.1 h4ab18f5_1 conda-forge
102
+ + pandas 2.2.2 py311h14de704_1 conda-forge
103
+ + pillow 9.4.0 py311h50def17_1 conda-forge
104
+ + pip 24.0 pyhd8ed1ab_0 /work/conda/cache/conda-forge
105
+ + pthread-stubs 0.4 h36c2ea0_1001 conda-forge
106
+ + pycparser 2.22 pyhd8ed1ab_0 conda-forge
107
+ + pysocks 1.7.1 pyha2e5f31_6 conda-forge
108
+ + python 3.11.9 hb806964_0_cpython /work/conda/cache/conda-forge
109
+ + python-dateutil 2.9.0 pyhd8ed1ab_0 conda-forge
110
+ + python-tzdata 2024.1 pyhd8ed1ab_0 conda-forge
111
+ + python_abi 3.11 4_cp311 conda-forge
112
+ + pytorch 2.3.1 py3.11_cuda12.1_cudnn8.9.2_0 pytorch
113
+ + pytorch-cuda 12.1 ha16c6d3_5 pytorch
114
+ + pytorch-mutex 1.0 cuda pytorch
115
+ + pytz 2024.1 pyhd8ed1ab_0 conda-forge
116
+ + pyyaml 6.0.1 py311h459d7ec_1 conda-forge
117
+ + readline 8.2 h8228510_1 conda-forge
118
+ + requests 2.32.3 pyhd8ed1ab_0 conda-forge
119
+ + setuptools 70.1.1 pyhd8ed1ab_0 conda-forge
120
+ + six 1.16.0 pyh6c4a22f_0 conda-forge
121
+ + sympy 1.12.1 pypyh2585a3b_103 conda-forge
122
+ + tbb 2021.12.0 h297d8ca_1 conda-forge
123
+ + tk 8.6.13 noxft_h4845f30_101 /work/conda/cache/conda-forge
124
+ + torchtriton 2.3.1 py311 pytorch
125
+ + torchvision 0.18.1 py311_cu121 pytorch
126
+ + typing_extensions 4.12.2 pyha770c72_0 conda-forge
127
+ + tzdata 2024a h0c530f3_0 conda-forge
128
+ + urllib3 2.2.2 pyhd8ed1ab_1 conda-forge
129
+ + wheel 0.43.0 pyhd8ed1ab_1 conda-forge
130
+ + xorg-libxau 1.0.11 hd590300_0 conda-forge
131
+ + xorg-libxdmcp 1.1.3 h516909a_0 conda-forge
132
+ + xz 5.2.6 h166bdaf_0 conda-forge
133
+ + yaml 0.2.5 h7f98852_2 conda-forge
134
+ + zlib 1.2.13 h4ab18f5_6 conda-forge
135
+ + zstandard 0.22.0 py311hb6f056b_1 conda-forge
136
+ + zstd 1.5.6 ha6fb4c9_0 conda-forge
137
+
138
+ Summary:
139
+
140
+ Install: 119 packages
141
+
142
+ Total download: 0 B
143
+
144
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
145
+
146
+
147
+
148
+ Transaction starting
149
+ Linking libcublas-12.1.0.26-0
150
+ Linking libcufft-11.0.2.4-0
151
+ Linking libcusolver-11.4.4.55-0
152
+ Linking libcusparse-12.0.2.55-0
153
+ Linking libnpp-12.0.2.50-0
154
+ Linking libnvjitlink-12.1.105-0
155
+ Linking cuda-cudart-12.1.105-0
156
+ Linking cuda-nvrtc-12.1.105-0
157
+ Linking libnvjpeg-12.1.1.14-0
158
+ Linking cuda-cupti-12.1.105-0
159
+ Linking cuda-nvtx-12.1.105-0
160
+ Linking pytorch-mutex-1.0-cuda
161
+ Linking _libgcc_mutex-0.1-conda_forge
162
+ Linking mkl-include-2022.1.0-h84fe81f_915
163
+ Linking python_abi-3.11-4_cp311
164
+ Linking ld_impl_linux-64-2.40-hf3520f5_7
165
+ Linking ca-certificates-2024.6.2-hbcca054_0
166
+ Linking libgcc-ng-14.1.0-h77fa898_0
167
+ Linking libzlib-1.2.13-h4ab18f5_6
168
+ Linking llvm-openmp-15.0.7-h0cdce71_0
169
+ Linking _openmp_mutex-4.5-2_kmp_llvm
170
+ Linking xorg-libxdmcp-1.1.3-h516909a_0
171
+ Linking pthread-stubs-0.4-h36c2ea0_1001
172
+ Linking xorg-libxau-1.0.11-hd590300_0
173
+ Linking libwebp-base-1.4.0-hd590300_0
174
+ Linking libdeflate-1.17-h0b41bf4_0
175
+ Linking jpeg-9e-h166bdaf_2
176
+ Linking libffi-3.4.2-h7f98852_5
177
+ Linking tk-8.6.13-noxft_h4845f30_101
178
+ Linking openssl-3.3.1-h4ab18f5_1
179
+ Linking libxcrypt-4.4.36-hd590300_1
180
+ Linking libsqlite-3.46.0-hde9e2c9_0
181
+ Linking yaml-0.2.5-h7f98852_2
182
+ Linking ncurses-6.5-h59595ed_0
183
+ Linking libgfortran5-14.1.0-hc5f4f2c_0
184
+ Linking lame-3.100-h166bdaf_1003
185
+ Linking nettle-3.6-he412f7d_0
186
+ Linking zlib-1.2.13-h4ab18f5_6
187
+ Linking libstdcxx-ng-14.1.0-hc0a3c3a_0
188
+ Linking libiconv-1.17-hd590300_2
189
+ Linking bzip2-1.0.8-hd590300_5
190
+ Linking libpng-1.6.43-h2797004_0
191
+ Linking xz-5.2.6-h166bdaf_0
192
+ Linking libuuid-2.38.1-h0b41bf4_0
193
+ Linking libnsl-2.0.1-hd590300_0
194
+ Linking libexpat-2.6.2-h59595ed_0
195
+ Linking libxcb-1.13-h7f98852_1004
196
+ Linking readline-8.2-h8228510_1
197
+ Linking libgfortran-ng-14.1.0-h69a702a_0
198
+ Linking icu-73.2-h59595ed_0
199
+ Linking zstd-1.5.6-ha6fb4c9_0
200
+ Linking lerc-4.0.0-h27087fc_0
201
+ Linking openh264-2.1.1-h780b84a_0
202
+ Linking gnutls-3.6.13-h85f3911_1
203
+ Linking gmp-6.3.0-hac33072_2
204
+ Linking freetype-2.12.1-h267a509_2
205
+ Linking libxml2-2.12.7-hc051c1a_1
206
+ Linking libtiff-4.5.0-h6adf6a1_2
207
+ Linking mpfr-4.2.1-h9458935_1
208
+ Linking libhwloc-2.10.0-default_h5622ce7_1001
209
+ Linking openjpeg-2.5.0-hfec8fc6_2
210
+ Linking lcms2-2.15-hfd0df8a_0
211
+ Linking mpc-1.3.1-hfe3b2da_0
212
+ Linking tbb-2021.12.0-h297d8ca_1
213
+ Linking mkl-2022.1.0-h84fe81f_915
214
+ Linking mkl-devel-2022.1.0-ha770c72_916
215
+ Linking libblas-3.9.0-16_linux64_mkl
216
+ Linking liblapack-3.9.0-16_linux64_mkl
217
+ Linking libcblas-3.9.0-16_linux64_mkl
218
+ Linking liblapacke-3.9.0-16_linux64_mkl
219
+ Linking blas-devel-3.9.0-16_linux64_mkl
220
+ Linking blas-2.116-mkl
221
+ Linking cuda-version-12.5-3
222
+ Linking tzdata-2024a-h0c530f3_0
223
+ Linking libjpeg-turbo-2.0.0-h9bf148f_0
224
+ warning libmamba [libjpeg-turbo-2.0.0-h9bf148f_0] The following files were already present in the environment:
225
+ - bin/cjpeg
226
+ - bin/djpeg
227
+ - bin/jpegtran
228
+ - bin/rdjpgcom
229
+ - bin/wrjpgcom
230
+ - include/jconfig.h
231
+ - include/jerror.h
232
+ - include/jmorecfg.h
233
+ - include/jpeglib.h
234
+ - lib/libjpeg.a
235
+ - lib/libjpeg.so
236
+ - lib/pkgconfig/libjpeg.pc
237
+ - share/man/man1/cjpeg.1
238
+ - share/man/man1/djpeg.1
239
+ - share/man/man1/jpegtran.1
240
+ - share/man/man1/rdjpgcom.1
241
+ - share/man/man1/wrjpgcom.1
242
+ Linking ffmpeg-4.3-hf484d3e_0
243
+ Linking libcurand-10.3.6.39-0
244
+ Linking libcufile-1.10.0.4-0
245
+ Linking cuda-opencl-12.5.39-0
246
+ Linking cuda-libraries-12.1.0-0
247
+ Linking cuda-runtime-12.1.0-0
248
+ Linking python-3.11.9-hb806964_0_cpython
249
+ Linking pytorch-cuda-12.1-ha16c6d3_5
250
+ Linking wheel-0.43.0-pyhd8ed1ab_1
251
+ Linking setuptools-70.1.1-pyhd8ed1ab_0
252
+ Linking pip-24.0-pyhd8ed1ab_0
253
+ Linking pycparser-2.22-pyhd8ed1ab_0
254
+ Linking six-1.16.0-pyh6c4a22f_0
255
+ Linking hyperframe-6.0.1-pyhd8ed1ab_0
256
+ Linking pytz-2024.1-pyhd8ed1ab_0
257
+ Linking python-tzdata-2024.1-pyhd8ed1ab_0
258
+ Linking charset-normalizer-3.3.2-pyhd8ed1ab_0
259
+ Linking hpack-4.0.0-pyh9f0ad1d_0
260
+ Linking pysocks-1.7.1-pyha2e5f31_6
261
+ Linking idna-3.7-pyhd8ed1ab_0
262
+ Linking certifi-2024.6.2-pyhd8ed1ab_0
263
+ Linking mpmath-1.3.0-pyhd8ed1ab_0
264
+ Linking typing_extensions-4.12.2-pyha770c72_0
265
+ Linking networkx-3.3-pyhd8ed1ab_1
266
+ Linking filelock-3.15.4-pyhd8ed1ab_0
267
+ Linking click-8.1.7-unix_pyh707e725_0
268
+ Linking python-dateutil-2.9.0-pyhd8ed1ab_0
269
+ Linking h2-4.1.0-pyhd8ed1ab_0
270
+ Linking brotli-python-1.1.0-py311hb755f60_1
271
+ Linking markupsafe-2.1.5-py311h459d7ec_0
272
+ Linking gmpy2-2.1.5-py311hc4f1f91_1
273
+ Linking pyyaml-6.0.1-py311h459d7ec_1
274
+ Linking pillow-9.4.0-py311h50def17_1
275
+ Linking numpy-2.0.0-py311h1461c94_0
276
+ Linking cffi-1.16.0-py311hb3a22ac_0
277
+ Linking pandas-2.2.2-py311h14de704_1
278
+ Linking zstandard-0.22.0-py311hb6f056b_1
279
+ Linking jinja2-3.1.4-pyhd8ed1ab_0
280
+ Linking sympy-1.12.1-pypyh2585a3b_103
281
+ Linking urllib3-2.2.2-pyhd8ed1ab_1
282
+ Linking requests-2.32.3-pyhd8ed1ab_0
283
+ Linking pytorch-2.3.1-py3.11_cuda12.1_cudnn8.9.2_0
284
+ Linking torchtriton-2.3.1-py311
285
+ Linking torchvision-0.18.1-py311_cu121
286
+
287
+ Transaction finished
288
+
289
+ To activate this environment, use:
290
+
291
+ mamba activate auto-uvapqvk3mmem
292
+
293
+ Or to execute a single command in this environment, use:
294
+
295
+ mamba run -n auto-uvapqvk3mmem mycommand
296
+
297
+ slurmstepd: error: *** JOB 56451 ON gcpl4-eu-1 CANCELLED AT 2024-07-03T18:51:16 ***
Meta-causal/code-stage1-pipeline/56451.log ADDED
File without changes
Meta-causal/code-stage1-pipeline/56452.error ADDED
@@ -0,0 +1,302 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0
  0%| | 0.00/44.7M [00:00<?, ?B/s]
1
  77%|███████▋ | 34.4M/44.7M [00:00<00:00, 360MB/s]
 
 
 
1
+ Solving dependencies
2
+ Installing conda packages
3
+ Empty environment created at prefix: /scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem
4
+ Transaction
5
+
6
+ Prefix: /scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem
7
+
8
+
9
+
10
+ No specs added or removed.
11
+
12
+ Package Version Build Channel Size
13
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
14
+ Install:
15
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
16
+
17
+ + _libgcc_mutex 0.1 conda_forge conda-forge
18
+ + _openmp_mutex 4.5 2_kmp_llvm conda-forge
19
+ + blas 2.116 mkl conda-forge
20
+ + blas-devel 3.9.0 16_linux64_mkl conda-forge
21
+ + brotli-python 1.1.0 py311hb755f60_1 conda-forge
22
+ + bzip2 1.0.8 hd590300_5 conda-forge
23
+ + ca-certificates 2024.6.2 hbcca054_0 conda-forge
24
+ + certifi 2024.6.2 pyhd8ed1ab_0 conda-forge
25
+ + cffi 1.16.0 py311hb3a22ac_0 conda-forge
26
+ + charset-normalizer 3.3.2 pyhd8ed1ab_0 conda-forge
27
+ + click 8.1.7 unix_pyh707e725_0 conda-forge
28
+ + cuda-cudart 12.1.105 0 nvidia
29
+ + cuda-cupti 12.1.105 0 /work/conda/cache/nvidia
30
+ + cuda-libraries 12.1.0 0 nvidia
31
+ + cuda-nvrtc 12.1.105 0 /work/conda/cache/nvidia
32
+ + cuda-nvtx 12.1.105 0 nvidia
33
+ + cuda-opencl 12.5.39 0 nvidia
34
+ + cuda-runtime 12.1.0 0 nvidia
35
+ + cuda-version 12.5 3 nvidia
36
+ + ffmpeg 4.3 hf484d3e_0 /work/conda/cache/pytorch
37
+ + filelock 3.15.4 pyhd8ed1ab_0 conda-forge
38
+ + freetype 2.12.1 h267a509_2 conda-forge
39
+ + gmp 6.3.0 hac33072_2 conda-forge
40
+ + gmpy2 2.1.5 py311hc4f1f91_1 conda-forge
41
+ + gnutls 3.6.13 h85f3911_1 /work/conda/cache/conda-forge
42
+ + h2 4.1.0 pyhd8ed1ab_0 conda-forge
43
+ + hpack 4.0.0 pyh9f0ad1d_0 conda-forge
44
+ + hyperframe 6.0.1 pyhd8ed1ab_0 conda-forge
45
+ + icu 73.2 h59595ed_0 /work/conda/cache/conda-forge
46
+ + idna 3.7 pyhd8ed1ab_0 conda-forge
47
+ + jinja2 3.1.4 pyhd8ed1ab_0 conda-forge
48
+ + jpeg 9e h166bdaf_2 conda-forge
49
+ + lame 3.100 h166bdaf_1003 conda-forge
50
+ + lcms2 2.15 hfd0df8a_0 conda-forge
51
+ + ld_impl_linux-64 2.40 hf3520f5_7 conda-forge
52
+ + lerc 4.0.0 h27087fc_0 conda-forge
53
+ + libblas 3.9.0 16_linux64_mkl conda-forge
54
+ + libcblas 3.9.0 16_linux64_mkl conda-forge
55
+ + libcublas 12.1.0.26 0 /work/conda/cache/nvidia
56
+ + libcufft 11.0.2.4 0 /work/conda/cache/nvidia
57
+ + libcufile 1.10.0.4 0 nvidia
58
+ + libcurand 10.3.6.39 0 nvidia
59
+ + libcusolver 11.4.4.55 0 /work/conda/cache/nvidia
60
+ + libcusparse 12.0.2.55 0 /work/conda/cache/nvidia
61
+ + libdeflate 1.17 h0b41bf4_0 conda-forge
62
+ + libexpat 2.6.2 h59595ed_0 conda-forge
63
+ + libffi 3.4.2 h7f98852_5 conda-forge
64
+ + libgcc-ng 14.1.0 h77fa898_0 conda-forge
65
+ + libgfortran-ng 14.1.0 h69a702a_0 conda-forge
66
+ + libgfortran5 14.1.0 hc5f4f2c_0 conda-forge
67
+ + libhwloc 2.10.0 default_h5622ce7_1001 conda-forge
68
+ + libiconv 1.17 hd590300_2 conda-forge
69
+ + libjpeg-turbo 2.0.0 h9bf148f_0 pytorch
70
+ + liblapack 3.9.0 16_linux64_mkl conda-forge
71
+ + liblapacke 3.9.0 16_linux64_mkl conda-forge
72
+ + libnpp 12.0.2.50 0 /work/conda/cache/nvidia
73
+ + libnsl 2.0.1 hd590300_0 conda-forge
74
+ + libnvjitlink 12.1.105 0 /work/conda/cache/nvidia
75
+ + libnvjpeg 12.1.1.14 0 /work/conda/cache/nvidia
76
+ + libpng 1.6.43 h2797004_0 conda-forge
77
+ + libsqlite 3.46.0 hde9e2c9_0 conda-forge
78
+ + libstdcxx-ng 14.1.0 hc0a3c3a_0 conda-forge
79
+ + libtiff 4.5.0 h6adf6a1_2 conda-forge
80
+ + libuuid 2.38.1 h0b41bf4_0 conda-forge
81
+ + libwebp-base 1.4.0 hd590300_0 conda-forge
82
+ + libxcb 1.13 h7f98852_1004 conda-forge
83
+ + libxcrypt 4.4.36 hd590300_1 conda-forge
84
+ + libxml2 2.12.7 hc051c1a_1 conda-forge
85
+ + libzlib 1.2.13 h4ab18f5_6 conda-forge
86
+ + llvm-openmp 15.0.7 h0cdce71_0 /work/conda/cache/conda-forge
87
+ + markupsafe 2.1.5 py311h459d7ec_0 conda-forge
88
+ + mkl 2022.1.0 h84fe81f_915 /work/conda/cache/conda-forge
89
+ + mkl-devel 2022.1.0 ha770c72_916 conda-forge
90
+ + mkl-include 2022.1.0 h84fe81f_915 conda-forge
91
+ + mpc 1.3.1 hfe3b2da_0 conda-forge
92
+ + mpfr 4.2.1 h9458935_1 conda-forge
93
+ + mpmath 1.3.0 pyhd8ed1ab_0 conda-forge
94
+ + ncurses 6.5 h59595ed_0 conda-forge
95
+ + nettle 3.6 he412f7d_0 /work/conda/cache/conda-forge
96
+ + networkx 3.3 pyhd8ed1ab_1 /work/conda/cache/conda-forge
97
+ + numpy 2.0.0 py311h1461c94_0 conda-forge
98
+ + openh264 2.1.1 h780b84a_0 /work/conda/cache/conda-forge
99
+ + openjpeg 2.5.0 hfec8fc6_2 conda-forge
100
+ + openssl 3.3.1 h4ab18f5_1 conda-forge
101
+ + pandas 2.2.2 py311h14de704_1 conda-forge
102
+ + pillow 9.4.0 py311h50def17_1 conda-forge
103
+ + pip 24.0 pyhd8ed1ab_0 /work/conda/cache/conda-forge
104
+ + pthread-stubs 0.4 h36c2ea0_1001 conda-forge
105
+ + pycparser 2.22 pyhd8ed1ab_0 conda-forge
106
+ + pysocks 1.7.1 pyha2e5f31_6 conda-forge
107
+ + python 3.11.9 hb806964_0_cpython /work/conda/cache/conda-forge
108
+ + python-dateutil 2.9.0 pyhd8ed1ab_0 conda-forge
109
+ + python-tzdata 2024.1 pyhd8ed1ab_0 conda-forge
110
+ + python_abi 3.11 4_cp311 conda-forge
111
+ + pytorch 2.3.1 py3.11_cuda12.1_cudnn8.9.2_0 pytorch
112
+ + pytorch-cuda 12.1 ha16c6d3_5 pytorch
113
+ + pytorch-mutex 1.0 cuda pytorch
114
+ + pytz 2024.1 pyhd8ed1ab_0 conda-forge
115
+ + pyyaml 6.0.1 py311h459d7ec_1 conda-forge
116
+ + readline 8.2 h8228510_1 conda-forge
117
+ + requests 2.32.3 pyhd8ed1ab_0 conda-forge
118
+ + setuptools 70.1.1 pyhd8ed1ab_0 conda-forge
119
+ + six 1.16.0 pyh6c4a22f_0 conda-forge
120
+ + sympy 1.12.1 pypyh2585a3b_103 conda-forge
121
+ + tbb 2021.12.0 h297d8ca_1 conda-forge
122
+ + tk 8.6.13 noxft_h4845f30_101 /work/conda/cache/conda-forge
123
+ + torchtriton 2.3.1 py311 pytorch
124
+ + torchvision 0.18.1 py311_cu121 pytorch
125
+ + typing_extensions 4.12.2 pyha770c72_0 conda-forge
126
+ + tzdata 2024a h0c530f3_0 conda-forge
127
+ + urllib3 2.2.2 pyhd8ed1ab_1 conda-forge
128
+ + wheel 0.43.0 pyhd8ed1ab_1 conda-forge
129
+ + xorg-libxau 1.0.11 hd590300_0 conda-forge
130
+ + xorg-libxdmcp 1.1.3 h516909a_0 conda-forge
131
+ + xz 5.2.6 h166bdaf_0 conda-forge
132
+ + yaml 0.2.5 h7f98852_2 conda-forge
133
+ + zlib 1.2.13 h4ab18f5_6 conda-forge
134
+ + zstandard 0.22.0 py311hb6f056b_1 conda-forge
135
+ + zstd 1.5.6 ha6fb4c9_0 conda-forge
136
+
137
+ Summary:
138
+
139
+ Install: 119 packages
140
+
141
+ Total download: 0 B
142
+
143
+ ─────────────────────────────────────────────────────────────────────────────────────────────────────────
144
+
145
+
146
+
147
+ Transaction starting
148
+ Linking libcublas-12.1.0.26-0
149
+ Linking libcufft-11.0.2.4-0
150
+ Linking libcusolver-11.4.4.55-0
151
+ Linking libcusparse-12.0.2.55-0
152
+ Linking libnpp-12.0.2.50-0
153
+ Linking libnvjitlink-12.1.105-0
154
+ Linking cuda-cudart-12.1.105-0
155
+ Linking cuda-nvrtc-12.1.105-0
156
+ Linking libnvjpeg-12.1.1.14-0
157
+ Linking cuda-cupti-12.1.105-0
158
+ Linking cuda-nvtx-12.1.105-0
159
+ Linking pytorch-mutex-1.0-cuda
160
+ Linking _libgcc_mutex-0.1-conda_forge
161
+ Linking mkl-include-2022.1.0-h84fe81f_915
162
+ Linking python_abi-3.11-4_cp311
163
+ Linking ld_impl_linux-64-2.40-hf3520f5_7
164
+ Linking ca-certificates-2024.6.2-hbcca054_0
165
+ Linking libgcc-ng-14.1.0-h77fa898_0
166
+ Linking libzlib-1.2.13-h4ab18f5_6
167
+ Linking llvm-openmp-15.0.7-h0cdce71_0
168
+ Linking _openmp_mutex-4.5-2_kmp_llvm
169
+ Linking xorg-libxdmcp-1.1.3-h516909a_0
170
+ Linking pthread-stubs-0.4-h36c2ea0_1001
171
+ Linking xorg-libxau-1.0.11-hd590300_0
172
+ Linking libwebp-base-1.4.0-hd590300_0
173
+ Linking libdeflate-1.17-h0b41bf4_0
174
+ Linking jpeg-9e-h166bdaf_2
175
+ Linking libffi-3.4.2-h7f98852_5
176
+ Linking tk-8.6.13-noxft_h4845f30_101
177
+ Linking openssl-3.3.1-h4ab18f5_1
178
+ Linking libxcrypt-4.4.36-hd590300_1
179
+ Linking libsqlite-3.46.0-hde9e2c9_0
180
+ Linking yaml-0.2.5-h7f98852_2
181
+ Linking ncurses-6.5-h59595ed_0
182
+ Linking libgfortran5-14.1.0-hc5f4f2c_0
183
+ Linking lame-3.100-h166bdaf_1003
184
+ Linking nettle-3.6-he412f7d_0
185
+ Linking zlib-1.2.13-h4ab18f5_6
186
+ Linking libstdcxx-ng-14.1.0-hc0a3c3a_0
187
+ Linking libiconv-1.17-hd590300_2
188
+ Linking bzip2-1.0.8-hd590300_5
189
+ Linking libpng-1.6.43-h2797004_0
190
+ Linking xz-5.2.6-h166bdaf_0
191
+ Linking libuuid-2.38.1-h0b41bf4_0
192
+ Linking libnsl-2.0.1-hd590300_0
193
+ Linking libexpat-2.6.2-h59595ed_0
194
+ Linking libxcb-1.13-h7f98852_1004
195
+ Linking readline-8.2-h8228510_1
196
+ Linking libgfortran-ng-14.1.0-h69a702a_0
197
+ Linking icu-73.2-h59595ed_0
198
+ Linking zstd-1.5.6-ha6fb4c9_0
199
+ Linking lerc-4.0.0-h27087fc_0
200
+ Linking openh264-2.1.1-h780b84a_0
201
+ Linking gnutls-3.6.13-h85f3911_1
202
+ Linking gmp-6.3.0-hac33072_2
203
+ Linking freetype-2.12.1-h267a509_2
204
+ Linking libxml2-2.12.7-hc051c1a_1
205
+ Linking libtiff-4.5.0-h6adf6a1_2
206
+ Linking mpfr-4.2.1-h9458935_1
207
+ Linking libhwloc-2.10.0-default_h5622ce7_1001
208
+ Linking openjpeg-2.5.0-hfec8fc6_2
209
+ Linking lcms2-2.15-hfd0df8a_0
210
+ Linking mpc-1.3.1-hfe3b2da_0
211
+ Linking tbb-2021.12.0-h297d8ca_1
212
+ Linking mkl-2022.1.0-h84fe81f_915
213
+ Linking mkl-devel-2022.1.0-ha770c72_916
214
+ Linking libblas-3.9.0-16_linux64_mkl
215
+ Linking liblapack-3.9.0-16_linux64_mkl
216
+ Linking libcblas-3.9.0-16_linux64_mkl
217
+ Linking liblapacke-3.9.0-16_linux64_mkl
218
+ Linking blas-devel-3.9.0-16_linux64_mkl
219
+ Linking blas-2.116-mkl
220
+ Linking cuda-version-12.5-3
221
+ Linking tzdata-2024a-h0c530f3_0
222
+ Linking libjpeg-turbo-2.0.0-h9bf148f_0
223
+ warning libmamba [libjpeg-turbo-2.0.0-h9bf148f_0] The following files were already present in the environment:
224
+ - bin/cjpeg
225
+ - bin/djpeg
226
+ - bin/jpegtran
227
+ - bin/rdjpgcom
228
+ - bin/wrjpgcom
229
+ - include/jconfig.h
230
+ - include/jerror.h
231
+ - include/jmorecfg.h
232
+ - include/jpeglib.h
233
+ - lib/libjpeg.a
234
+ - lib/libjpeg.so
235
+ - lib/pkgconfig/libjpeg.pc
236
+ - share/man/man1/cjpeg.1
237
+ - share/man/man1/djpeg.1
238
+ - share/man/man1/jpegtran.1
239
+ - share/man/man1/rdjpgcom.1
240
+ - share/man/man1/wrjpgcom.1
241
+ Linking ffmpeg-4.3-hf484d3e_0
242
+ Linking libcurand-10.3.6.39-0
243
+ Linking libcufile-1.10.0.4-0
244
+ Linking cuda-opencl-12.5.39-0
245
+ Linking cuda-libraries-12.1.0-0
246
+ Linking cuda-runtime-12.1.0-0
247
+ Linking python-3.11.9-hb806964_0_cpython
248
+ Linking pytorch-cuda-12.1-ha16c6d3_5
249
+ Linking wheel-0.43.0-pyhd8ed1ab_1
250
+ Linking setuptools-70.1.1-pyhd8ed1ab_0
251
+ Linking pip-24.0-pyhd8ed1ab_0
252
+ Linking pycparser-2.22-pyhd8ed1ab_0
253
+ Linking six-1.16.0-pyh6c4a22f_0
254
+ Linking hyperframe-6.0.1-pyhd8ed1ab_0
255
+ Linking pytz-2024.1-pyhd8ed1ab_0
256
+ Linking python-tzdata-2024.1-pyhd8ed1ab_0
257
+ Linking charset-normalizer-3.3.2-pyhd8ed1ab_0
258
+ Linking hpack-4.0.0-pyh9f0ad1d_0
259
+ Linking pysocks-1.7.1-pyha2e5f31_6
260
+ Linking idna-3.7-pyhd8ed1ab_0
261
+ Linking certifi-2024.6.2-pyhd8ed1ab_0
262
+ Linking mpmath-1.3.0-pyhd8ed1ab_0
263
+ Linking typing_extensions-4.12.2-pyha770c72_0
264
+ Linking networkx-3.3-pyhd8ed1ab_1
265
+ Linking filelock-3.15.4-pyhd8ed1ab_0
266
+ Linking click-8.1.7-unix_pyh707e725_0
267
+ Linking python-dateutil-2.9.0-pyhd8ed1ab_0
268
+ Linking h2-4.1.0-pyhd8ed1ab_0
269
+ Linking brotli-python-1.1.0-py311hb755f60_1
270
+ Linking markupsafe-2.1.5-py311h459d7ec_0
271
+ Linking gmpy2-2.1.5-py311hc4f1f91_1
272
+ Linking pyyaml-6.0.1-py311h459d7ec_1
273
+ Linking pillow-9.4.0-py311h50def17_1
274
+ Linking numpy-2.0.0-py311h1461c94_0
275
+ Linking cffi-1.16.0-py311hb3a22ac_0
276
+ Linking pandas-2.2.2-py311h14de704_1
277
+ Linking zstandard-0.22.0-py311hb6f056b_1
278
+ Linking jinja2-3.1.4-pyhd8ed1ab_0
279
+ Linking sympy-1.12.1-pypyh2585a3b_103
280
+ Linking urllib3-2.2.2-pyhd8ed1ab_1
281
+ Linking requests-2.32.3-pyhd8ed1ab_0
282
+ Linking pytorch-2.3.1-py3.11_cuda12.1_cudnn8.9.2_0
283
+ Linking torchtriton-2.3.1-py311
284
+ Linking torchvision-0.18.1-py311_cu121
285
+
286
+ Transaction finished
287
+
288
+ To activate this environment, use:
289
+
290
+ mamba activate auto-uvapqvk3mmem
291
+
292
+ Or to execute a single command in this environment, use:
293
+
294
+ mamba run -n auto-uvapqvk3mmem mycommand
295
+
296
+ Installing pip packages
297
+ WARNING: The candidate selected for download or install is a yanked version: 'opencv-python' candidate (version 4.5.5.62 at https://files.pythonhosted.org/packages/9d/98/36bfcbff30da27dd6922ed73ca7802c37d87f77daf4c569da3dcb87b4296/opencv_python-4.5.5.62-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (from https://pypi.org/simple/opencv-python/) (requires-python:>=3.6))
298
+ Reason for being yanked: deprecated, use 4.5.5.64
299
+ Downloading: "https://download.pytorch.org/models/resnet18-5c106cde.pth" to /home/yuqian_fu/.cache/torch/hub/checkpoints/resnet18-5c106cde.pth
300
+
301
  0%| | 0.00/44.7M [00:00<?, ?B/s]
302
  77%|███████▋ | 34.4M/44.7M [00:00<00:00, 360MB/s]
303
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
304
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
Meta-causal/code-stage1-pipeline/56452.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56454.error ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ slurmstepd: error: *** JOB 56454 ON gcpl4-eu-1 CANCELLED AT 2024-07-04T06:57:02 DUE TO TIME LIMIT ***
2
+ slurmstepd: error: *** STEP 56454.0 ON gcpl4-eu-1 CANCELLED AT 2024-07-04T06:57:02 DUE TO TIME LIMIT ***
3
+ srun: Job step aborted: Waiting up to 32 seconds for job step to finish.
Meta-causal/code-stage1-pipeline/56454.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56455.error ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
2
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
3
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
4
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
Meta-causal/code-stage1-pipeline/56455.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56456.error ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ slurmstepd: error: *** JOB 56456 ON gcpl4-eu-1 CANCELLED AT 2024-07-04T07:05:01 DUE TO TIME LIMIT ***
2
+ slurmstepd: error: *** STEP 56456.0 ON gcpl4-eu-1 CANCELLED AT 2024-07-04T07:05:01 DUE TO TIME LIMIT ***
3
+ srun: Job step aborted: Waiting up to 32 seconds for job step to finish.
Meta-causal/code-stage1-pipeline/56456.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56457.error ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
2
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
3
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
4
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
Meta-causal/code-stage1-pipeline/56457.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56458.error ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ slurmstepd: error: *** JOB 56458 ON gcpl4-eu-1 CANCELLED AT 2024-07-04T07:07:32 DUE TO TIME LIMIT ***
2
+ slurmstepd: error: *** STEP 56458.0 ON gcpl4-eu-1 CANCELLED AT 2024-07-04T07:07:32 DUE TO TIME LIMIT ***
3
+ srun: Job step aborted: Waiting up to 32 seconds for job step to finish.
Meta-causal/code-stage1-pipeline/56458.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56526.error ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Traceback (most recent call last):
2
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/main_test_pacs_v13.py", line 86, in <module>
3
+ main()
4
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 1157, in __call__
5
+ return self.main(*args, **kwargs)
6
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^
7
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 1078, in main
8
+ rv = self.invoke(ctx)
9
+ ^^^^^^^^^^^^^^^^
10
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 1434, in invoke
11
+ return ctx.invoke(self.callback, **ctx.params)
12
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
13
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 783, in invoke
14
+ return __callback(*args, **kwargs)
15
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^
16
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/main_test_pacs_v13.py", line 29, in main
17
+ evaluate_pacs(gpu, svroot, source_domain, svpath, factor_num, epoch, stride,eval_mapping, network)
18
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/main_test_pacs_v13.py", line 42, in evaluate_pacs
19
+ saved_weight = torch.load(os.path.join(svroot, 'best_cls_net.pkl'))
20
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
21
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/torch/serialization.py", line 997, in load
22
+ with _open_file_like(f, 'rb') as opened_file:
23
+ ^^^^^^^^^^^^^^^^^^^^^^^^
24
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/torch/serialization.py", line 444, in _open_file_like
25
+ return _open_file(name_or_buffer, mode)
26
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
27
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/torch/serialization.py", line 425, in __init__
28
+ super().__init__(open(name, mode))
29
+ ^^^^^^^^^^^^^^^^
30
+ FileNotFoundError: [Errno 2] No such file or directory: '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS//art_painting/CA_multiple_16fa_v2_ep30_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5/best_cls_net.pkl'
31
+ srun: error: gcpl4-eu-1: task 0: Exited with exit code 1
Meta-causal/code-stage1-pipeline/56526.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ /home/yuqian_fu
2
+ {'gpu': '0', 'svroot': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS//art_painting/CA_multiple_16fa_v2_ep30_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5', 'source_domain': 'art_painting', 'svpath': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS//art_painting/CA_multiple_16fa_v2_ep30_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5/art_painting_16factor_best_test_check.csv', 'factor_num': 16, 'epoch': 'best', 'stride': 5, 'eval_mapping': False, 'network': 'resnet18'}
3
+ -------------------------------------loading pretrain weights----------------------------------
4
+ loading weight of best
Meta-causal/code-stage1-pipeline/56527.error ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Traceback (most recent call last):
2
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/main_test_digit_v13.py", line 84, in <module>
3
+ main()
4
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 1157, in __call__
5
+ return self.main(*args, **kwargs)
6
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^
7
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 1078, in main
8
+ rv = self.invoke(ctx)
9
+ ^^^^^^^^^^^^^^^^
10
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 1434, in invoke
11
+ return ctx.invoke(self.callback, **ctx.params)
12
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
13
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/click/core.py", line 783, in invoke
14
+ return __callback(*args, **kwargs)
15
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^
16
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/main_test_digit_v13.py", line 28, in main
17
+ evaluate_digit(gpu, svroot, svpath, channels, factor_num, stride,epoch, eval_mapping)
18
+ File "/home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/main_test_digit_v13.py", line 42, in evaluate_digit
19
+ saved_weight = torch.load(os.path.join(svroot, 'best_cls_net.pkl'))
20
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
21
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/torch/serialization.py", line 997, in load
22
+ with _open_file_like(f, 'rb') as opened_file:
23
+ ^^^^^^^^^^^^^^^^^^^^^^^^
24
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/torch/serialization.py", line 444, in _open_file_like
25
+ return _open_file(name_or_buffer, mode)
26
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
27
+ File "/scratch/yuqian_fu/micromamba/envs/auto-uvapqvk3mmem/lib/python3.11/site-packages/torch/serialization.py", line 425, in __init__
28
+ super().__init__(open(name, mode))
29
+ ^^^^^^^^^^^^^^^^
30
+ FileNotFoundError: [Errno 2] No such file or directory: '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit/CA_multiple_14fa_all_ep100_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/best_cls_net.pkl'
31
+ srun: error: gcpl4-eu-1: task 0: Exited with exit code 1
Meta-causal/code-stage1-pipeline/56527.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ /home/yuqian_fu
2
+ {'gpu': '0', 'svroot': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit/CA_multiple_14fa_all_ep100_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3', 'svpath': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit/CA_multiple_14fa_all_ep100_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/14factor_best.csv', 'channels': 3, 'factor_num': 14, 'stride': 3, 'epoch': 'best', 'eval_mapping': True}
3
+ loading weight of best
Meta-causal/code-stage1-pipeline/56528.error ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ run_my_joint_v13_test.sh: line 25: ndm: command not found
2
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
3
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
Meta-causal/code-stage1-pipeline/56528.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56529.error ADDED
@@ -0,0 +1 @@
 
 
1
+ run_my_joint_test.sh: line 24: randm: command not found
Meta-causal/code-stage1-pipeline/56529.log ADDED
The diff for this file is too large to render. See raw diff
 
Meta-causal/code-stage1-pipeline/56540.error ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
2
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
3
+ /home/yuqian_fu/Projects/CausalStyleAdv/Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py:426: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
4
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
Meta-causal/code-stage1-pipeline/56540.log ADDED
@@ -0,0 +1,151 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ /home/yuqian_fu
2
+ {'gpu': '0', 'data': 'art_painting', 'ntr': None, 'translate': None, 'autoaug': 'CA_multiple', 'n': 3, 'stride': 5, 'factor_num': 16, 'epochs': 30, 'nbatch': 100, 'batchsize': 6, 'lr': 0.01, 'lr_scheduler': 'cosine', 'svroot': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS//art_painting/CA_multiple_16fa_v2_ep30_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5_pipelineAugWoNorm', 'clsadapt': True, 'lambda_causal': 1.0, 'lambda_re': 1.0, 'randm': True, 'randn': True, 'network': 'resnet18'}
3
+ /data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/art_painting_train.hdf5 torch.Size([1840, 3, 227, 227]) torch.Size([1840])
4
+ --------------------------CA_multiple--------------------------
5
+ ---------------------------16 factors-----------------
6
+ randm: True
7
+ randn: True
8
+ n: 3
9
+ randm: False
10
+ /data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/art_painting_val.hdf5 torch.Size([208, 3, 227, 227]) torch.Size([208])
11
+ -------------------------------------loading pretrain weights----------------------------------
12
+ 306
13
+ 0.0001
14
+ changing lr
15
+ ---------------------saving model at epoch 0----------------------------------------------------
16
+ epoch 0, time 328.72, cls_loss 8.5176
17
+ 306
18
+ 9.972609476841367e-05
19
+ changing lr
20
+ ---------------------saving model at epoch 1----------------------------------------------------
21
+ epoch 1, time 323.89, cls_loss 4.1852
22
+ 306
23
+ 9.890738003669029e-05
24
+ changing lr
25
+ ---------------------saving model at epoch 2----------------------------------------------------
26
+ epoch 2, time 321.19, cls_loss 1.7210
27
+ 306
28
+ 9.755282581475769e-05
29
+ changing lr
30
+ ---------------------saving model at epoch 3----------------------------------------------------
31
+ epoch 3, time 321.77, cls_loss 0.8687
32
+ 306
33
+ 9.567727288213003e-05
34
+ changing lr
35
+ epoch 4, time 322.73, cls_loss 0.5508
36
+ 306
37
+ 9.330127018922194e-05
38
+ changing lr
39
+ epoch 5, time 321.65, cls_loss 0.4191
40
+ 306
41
+ 9.045084971874738e-05
42
+ changing lr
43
+ epoch 6, time 323.65, cls_loss 0.3904
44
+ 306
45
+ 8.715724127386972e-05
46
+ changing lr
47
+ ---------------------saving model at epoch 7----------------------------------------------------
48
+ epoch 7, time 324.01, cls_loss 0.2998
49
+ 306
50
+ 8.345653031794292e-05
51
+ changing lr
52
+ epoch 8, time 327.47, cls_loss 0.2084
53
+ 306
54
+ 7.938926261462366e-05
55
+ changing lr
56
+ epoch 9, time 326.22, cls_loss 0.1815
57
+ 306
58
+ 7.500000000000001e-05
59
+ changing lr
60
+ epoch 10, time 325.03, cls_loss 0.1476
61
+ 306
62
+ 7.033683215379003e-05
63
+ changing lr
64
+ epoch 11, time 325.92, cls_loss 0.1094
65
+ 306
66
+ 6.545084971874738e-05
67
+ changing lr
68
+ ---------------------saving model at epoch 12----------------------------------------------------
69
+ epoch 12, time 322.71, cls_loss 0.0653
70
+ 306
71
+ 6.039558454088797e-05
72
+ changing lr
73
+ epoch 13, time 321.20, cls_loss 0.0639
74
+ 306
75
+ 5.522642316338269e-05
76
+ changing lr
77
+ ---------------------saving model at epoch 14----------------------------------------------------
78
+ epoch 14, time 322.13, cls_loss 0.0464
79
+ 306
80
+ 5.000000000000002e-05
81
+ changing lr
82
+ epoch 15, time 324.52, cls_loss 0.0402
83
+ 306
84
+ 4.4773576836617344e-05
85
+ changing lr
86
+ epoch 16, time 322.56, cls_loss 0.0452
87
+ 306
88
+ 3.9604415459112035e-05
89
+ changing lr
90
+ epoch 17, time 322.39, cls_loss 0.0403
91
+ 306
92
+ 3.4549150281252636e-05
93
+ changing lr
94
+ epoch 18, time 324.36, cls_loss 0.0190
95
+ 306
96
+ 2.966316784621e-05
97
+ changing lr
98
+ epoch 19, time 327.79, cls_loss 0.0250
99
+ 306
100
+ 2.5000000000000015e-05
101
+ changing lr
102
+ epoch 20, time 322.28, cls_loss 0.0416
103
+ 306
104
+ 2.0610737385376352e-05
105
+ changing lr
106
+ epoch 21, time 322.98, cls_loss 0.0203
107
+ 306
108
+ 1.654346968205711e-05
109
+ changing lr
110
+ ---------------------saving model at epoch 22----------------------------------------------------
111
+ epoch 22, time 325.56, cls_loss 0.0271
112
+ 306
113
+ 1.2842758726130304e-05
114
+ changing lr
115
+ epoch 23, time 321.58, cls_loss 0.0190
116
+ 306
117
+ 9.549150281252636e-06
118
+ changing lr
119
+ epoch 24, time 327.21, cls_loss 0.0236
120
+ 306
121
+ 6.698729810778068e-06
122
+ changing lr
123
+ epoch 25, time 301.36, cls_loss 0.0107
124
+ 306
125
+ 4.322727117869953e-06
126
+ changing lr
127
+ epoch 26, time 295.79, cls_loss 0.0165
128
+ 306
129
+ 2.447174185242324e-06
130
+ changing lr
131
+ epoch 27, time 296.53, cls_loss 0.0218
132
+ 306
133
+ 1.092619963309716e-06
134
+ changing lr
135
+ epoch 28, time 297.90, cls_loss 0.0198
136
+ 306
137
+ 2.7390523158633003e-07
138
+ changing lr
139
+ epoch 29, time 299.73, cls_loss 0.0094
140
+ ---------------------saving last model at epoch 29----------------------------------------------------
141
+ /home/yuqian_fu
142
+ {'gpu': '0', 'svroot': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS//art_painting/CA_multiple_16fa_v2_ep30_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5_pipelineAugWoNorm', 'source_domain': 'art_painting', 'svpath': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS//art_painting/CA_multiple_16fa_v2_ep30_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5_pipelineAugWoNorm/art_painting_16factor_best_test_check.csv', 'factor_num': 16, 'epoch': 'best', 'stride': 5, 'eval_mapping': False, 'network': 'resnet18'}
143
+ -------------------------------------loading pretrain weights----------------------------------
144
+ loading weight of best
145
+ columns: ['art_painting', 'cartoon', 'photo', 'sketch']
146
+ /data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/art_painting_test.hdf5 torch.Size([2048, 3, 227, 227]) torch.Size([2048])
147
+ /data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/cartoon_test.hdf5 torch.Size([2344, 3, 227, 227]) torch.Size([2344])
148
+ /data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/photo_test.hdf5 torch.Size([1670, 3, 227, 227]) torch.Size([1670])
149
+ /data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/sketch_test.hdf5 torch.Size([3929, 3, 227, 227]) torch.Size([3929])
150
+ art_painting cartoon photo sketch Avg
151
+ w/o do (original x) 93.017578 53.412969 88.203593 45.838636 62.485066
Meta-causal/code-stage1-pipeline/56541.error ADDED
File without changes
Meta-causal/code-stage1-pipeline/56541.log ADDED
@@ -0,0 +1,432 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ /home/yuqian_fu
2
+ {'gpu': '0', 'data': 'mnist', 'ntr': None, 'translate': None, 'autoaug': 'CA_multiple', 'n': 3, 'stride': 3, 'factor_num': 14, 'epochs': 100, 'nbatch': 100, 'batchsize': 32, 'lr': 0.0001, 'lr_scheduler': 'Step', 'svroot': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit/CA_multiple_14fa_all_ep100_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3_pipelineAugWoNorm', 'clsadapt': True, 'lambda_causal': 1.0, 'lambda_re': 1.0, 'randm': True, 'randn': True, 'network': 'resnet18'}
3
+ --------------------------CA_multiple--------------------------
4
+ ---------------------------14 factors-----------------
5
+ randm: True
6
+ randn: True
7
+ n: 3
8
+ randm: False
9
+ 100
10
+ 0.0001
11
+ changing lr
12
+ ---------------------saving model at epoch 0----------------------------------------------------
13
+ epoch 0, time 183.25, cls_loss 2.1515
14
+ 100
15
+ 0.0001
16
+ changing lr
17
+ ---------------------saving model at epoch 1----------------------------------------------------
18
+ epoch 1, time 183.70, cls_loss 1.7865
19
+ 100
20
+ 0.0001
21
+ changing lr
22
+ epoch 2, time 183.59, cls_loss 1.5733
23
+ 100
24
+ 0.0001
25
+ changing lr
26
+ ---------------------saving model at epoch 3----------------------------------------------------
27
+ epoch 3, time 183.31, cls_loss 1.4407
28
+ 100
29
+ 0.0001
30
+ changing lr
31
+ ---------------------saving model at epoch 4----------------------------------------------------
32
+ epoch 4, time 183.01, cls_loss 1.3369
33
+ 100
34
+ 0.0001
35
+ changing lr
36
+ ---------------------saving model at epoch 5----------------------------------------------------
37
+ epoch 5, time 183.85, cls_loss 1.3080
38
+ 100
39
+ 0.0001
40
+ changing lr
41
+ epoch 6, time 182.43, cls_loss 1.2082
42
+ 100
43
+ 0.0001
44
+ changing lr
45
+ ---------------------saving model at epoch 7----------------------------------------------------
46
+ epoch 7, time 182.60, cls_loss 1.1517
47
+ 100
48
+ 0.0001
49
+ changing lr
50
+ ---------------------saving model at epoch 8----------------------------------------------------
51
+ epoch 8, time 183.05, cls_loss 1.0938
52
+ 100
53
+ 0.0001
54
+ changing lr
55
+ ---------------------saving model at epoch 9----------------------------------------------------
56
+ epoch 9, time 183.04, cls_loss 1.0485
57
+ 100
58
+ 0.0001
59
+ changing lr
60
+ epoch 10, time 182.31, cls_loss 1.0636
61
+ 100
62
+ 0.0001
63
+ changing lr
64
+ epoch 11, time 182.08, cls_loss 0.9913
65
+ 100
66
+ 0.0001
67
+ changing lr
68
+ epoch 12, time 182.44, cls_loss 0.9240
69
+ 100
70
+ 0.0001
71
+ changing lr
72
+ ---------------------saving model at epoch 13----------------------------------------------------
73
+ epoch 13, time 182.56, cls_loss 0.8962
74
+ 100
75
+ 0.0001
76
+ changing lr
77
+ epoch 14, time 182.83, cls_loss 0.8474
78
+ 100
79
+ 0.0001
80
+ changing lr
81
+ epoch 15, time 182.24, cls_loss 0.8730
82
+ 100
83
+ 0.0001
84
+ changing lr
85
+ ---------------------saving model at epoch 16----------------------------------------------------
86
+ epoch 16, time 182.40, cls_loss 0.8184
87
+ 100
88
+ 0.0001
89
+ changing lr
90
+ epoch 17, time 182.12, cls_loss 0.8083
91
+ 100
92
+ 0.0001
93
+ changing lr
94
+ epoch 18, time 182.02, cls_loss 0.7381
95
+ 100
96
+ 0.0001
97
+ changing lr
98
+ epoch 19, time 182.19, cls_loss 0.7326
99
+ 100
100
+ 0.0001
101
+ changing lr
102
+ epoch 20, time 181.69, cls_loss 0.6649
103
+ 100
104
+ 0.0001
105
+ changing lr
106
+ epoch 21, time 181.62, cls_loss 0.6849
107
+ 100
108
+ 0.0001
109
+ changing lr
110
+ epoch 22, time 181.68, cls_loss 0.6675
111
+ 100
112
+ 0.0001
113
+ changing lr
114
+ ---------------------saving model at epoch 23----------------------------------------------------
115
+ epoch 23, time 182.29, cls_loss 0.6101
116
+ 100
117
+ 0.0001
118
+ changing lr
119
+ epoch 24, time 182.13, cls_loss 0.6237
120
+ 100
121
+ 0.0001
122
+ changing lr
123
+ epoch 25, time 182.23, cls_loss 0.6229
124
+ 100
125
+ 0.0001
126
+ changing lr
127
+ epoch 26, time 182.24, cls_loss 0.5664
128
+ 100
129
+ 0.0001
130
+ changing lr
131
+ epoch 27, time 182.13, cls_loss 0.5588
132
+ 100
133
+ 0.0001
134
+ changing lr
135
+ epoch 28, time 182.14, cls_loss 0.5539
136
+ 100
137
+ 0.0001
138
+ changing lr
139
+ epoch 29, time 182.35, cls_loss 0.5198
140
+ 100
141
+ 0.0001
142
+ changing lr
143
+ epoch 30, time 182.22, cls_loss 0.5153
144
+ 100
145
+ 0.0001
146
+ changing lr
147
+ epoch 31, time 182.36, cls_loss 0.4764
148
+ 100
149
+ 0.0001
150
+ changing lr
151
+ epoch 32, time 182.13, cls_loss 0.4748
152
+ 100
153
+ 0.0001
154
+ changing lr
155
+ epoch 33, time 181.83, cls_loss 0.4448
156
+ 100
157
+ 0.0001
158
+ changing lr
159
+ epoch 34, time 182.32, cls_loss 0.4358
160
+ 100
161
+ 0.0001
162
+ changing lr
163
+ epoch 35, time 181.92, cls_loss 0.4201
164
+ 100
165
+ 0.0001
166
+ changing lr
167
+ epoch 36, time 181.91, cls_loss 0.3949
168
+ 100
169
+ 0.0001
170
+ changing lr
171
+ epoch 37, time 182.01, cls_loss 0.3818
172
+ 100
173
+ 0.0001
174
+ changing lr
175
+ ---------------------saving model at epoch 38----------------------------------------------------
176
+ epoch 38, time 182.02, cls_loss 0.3651
177
+ 100
178
+ 0.0001
179
+ changing lr
180
+ epoch 39, time 182.07, cls_loss 0.3656
181
+ 100
182
+ 0.0001
183
+ changing lr
184
+ epoch 40, time 181.87, cls_loss 0.3864
185
+ 100
186
+ 0.0001
187
+ changing lr
188
+ epoch 41, time 182.33, cls_loss 0.3647
189
+ 100
190
+ 0.0001
191
+ changing lr
192
+ epoch 42, time 182.58, cls_loss 0.3301
193
+ 100
194
+ 0.0001
195
+ changing lr
196
+ ---------------------saving model at epoch 43----------------------------------------------------
197
+ epoch 43, time 182.56, cls_loss 0.3279
198
+ 100
199
+ 0.0001
200
+ changing lr
201
+ epoch 44, time 185.15, cls_loss 0.3470
202
+ 100
203
+ 0.0001
204
+ changing lr
205
+ epoch 45, time 182.28, cls_loss 0.2938
206
+ 100
207
+ 0.0001
208
+ changing lr
209
+ epoch 46, time 182.03, cls_loss 0.2920
210
+ 100
211
+ 0.0001
212
+ changing lr
213
+ epoch 47, time 182.53, cls_loss 0.2780
214
+ 100
215
+ 0.0001
216
+ changing lr
217
+ epoch 48, time 182.87, cls_loss 0.2592
218
+ 100
219
+ 0.0001
220
+ changing lr
221
+ epoch 49, time 182.61, cls_loss 0.2725
222
+ 100
223
+ 0.0001
224
+ changing lr
225
+ epoch 50, time 182.34, cls_loss 0.2344
226
+ 100
227
+ 0.0001
228
+ changing lr
229
+ epoch 51, time 182.13, cls_loss 0.2686
230
+ 100
231
+ 0.0001
232
+ changing lr
233
+ epoch 52, time 183.03, cls_loss 0.2475
234
+ 100
235
+ 0.0001
236
+ changing lr
237
+ epoch 53, time 182.25, cls_loss 0.2359
238
+ 100
239
+ 0.0001
240
+ changing lr
241
+ epoch 54, time 182.39, cls_loss 0.2279
242
+ 100
243
+ 0.0001
244
+ changing lr
245
+ epoch 55, time 182.38, cls_loss 0.2340
246
+ 100
247
+ 0.0001
248
+ changing lr
249
+ epoch 56, time 182.19, cls_loss 0.2217
250
+ 100
251
+ 0.0001
252
+ changing lr
253
+ epoch 57, time 182.01, cls_loss 0.2188
254
+ 100
255
+ 0.0001
256
+ changing lr
257
+ epoch 58, time 182.23, cls_loss 0.2269
258
+ 100
259
+ 0.0001
260
+ changing lr
261
+ epoch 59, time 182.47, cls_loss 0.2212
262
+ 100
263
+ 0.0001
264
+ changing lr
265
+ epoch 60, time 182.34, cls_loss 0.1887
266
+ 100
267
+ 0.0001
268
+ changing lr
269
+ epoch 61, time 182.11, cls_loss 0.1859
270
+ 100
271
+ 0.0001
272
+ changing lr
273
+ epoch 62, time 182.40, cls_loss 0.2021
274
+ 100
275
+ 0.0001
276
+ changing lr
277
+ epoch 63, time 182.09, cls_loss 0.1756
278
+ 100
279
+ 0.0001
280
+ changing lr
281
+ epoch 64, time 182.38, cls_loss 0.1737
282
+ 100
283
+ 0.0001
284
+ changing lr
285
+ epoch 65, time 182.21, cls_loss 0.1648
286
+ 100
287
+ 0.0001
288
+ changing lr
289
+ epoch 66, time 182.02, cls_loss 0.1613
290
+ 100
291
+ 0.0001
292
+ changing lr
293
+ epoch 67, time 182.29, cls_loss 0.1569
294
+ 100
295
+ 0.0001
296
+ changing lr
297
+ epoch 68, time 182.29, cls_loss 0.1487
298
+ 100
299
+ 0.0001
300
+ changing lr
301
+ ---------------------saving model at epoch 69----------------------------------------------------
302
+ epoch 69, time 182.61, cls_loss 0.1538
303
+ 100
304
+ 0.0001
305
+ changing lr
306
+ epoch 70, time 182.28, cls_loss 0.1653
307
+ 100
308
+ 0.0001
309
+ changing lr
310
+ epoch 71, time 181.94, cls_loss 0.1639
311
+ 100
312
+ 0.0001
313
+ changing lr
314
+ epoch 72, time 181.84, cls_loss 0.1784
315
+ 100
316
+ 0.0001
317
+ changing lr
318
+ epoch 73, time 181.70, cls_loss 0.1843
319
+ 100
320
+ 0.0001
321
+ changing lr
322
+ epoch 74, time 180.53, cls_loss 0.1832
323
+ 100
324
+ 0.0001
325
+ changing lr
326
+ epoch 75, time 180.51, cls_loss 0.1421
327
+ 100
328
+ 0.0001
329
+ changing lr
330
+ epoch 76, time 180.07, cls_loss 0.1224
331
+ 100
332
+ 0.0001
333
+ changing lr
334
+ epoch 77, time 180.21, cls_loss 0.1187
335
+ 100
336
+ 0.0001
337
+ changing lr
338
+ epoch 78, time 180.07, cls_loss 0.1058
339
+ 100
340
+ 0.0001
341
+ changing lr
342
+ epoch 79, time 180.76, cls_loss 0.1301
343
+ 100
344
+ 1e-05
345
+ changing lr
346
+ ---------------------saving model at epoch 80----------------------------------------------------
347
+ epoch 80, time 181.07, cls_loss 0.0915
348
+ 100
349
+ 1e-05
350
+ changing lr
351
+ epoch 81, time 180.00, cls_loss 0.0845
352
+ 100
353
+ 1e-05
354
+ changing lr
355
+ epoch 82, time 180.09, cls_loss 0.0767
356
+ 100
357
+ 1e-05
358
+ changing lr
359
+ epoch 83, time 180.14, cls_loss 0.0711
360
+ 100
361
+ 1e-05
362
+ changing lr
363
+ epoch 84, time 180.25, cls_loss 0.0698
364
+ 100
365
+ 1e-05
366
+ changing lr
367
+ epoch 85, time 180.12, cls_loss 0.0682
368
+ 100
369
+ 1e-05
370
+ changing lr
371
+ epoch 86, time 179.91, cls_loss 0.0590
372
+ 100
373
+ 1e-05
374
+ changing lr
375
+ epoch 87, time 179.84, cls_loss 0.0607
376
+ 100
377
+ 1e-05
378
+ changing lr
379
+ epoch 88, time 179.82, cls_loss 0.0634
380
+ 100
381
+ 1e-05
382
+ changing lr
383
+ epoch 89, time 180.04, cls_loss 0.0718
384
+ 100
385
+ 1e-05
386
+ changing lr
387
+ epoch 90, time 179.62, cls_loss 0.0704
388
+ 100
389
+ 1e-05
390
+ changing lr
391
+ epoch 91, time 179.77, cls_loss 0.0669
392
+ 100
393
+ 1e-05
394
+ changing lr
395
+ epoch 92, time 179.87, cls_loss 0.0574
396
+ 100
397
+ 1e-05
398
+ changing lr
399
+ epoch 93, time 179.66, cls_loss 0.0556
400
+ 100
401
+ 1e-05
402
+ changing lr
403
+ epoch 94, time 179.87, cls_loss 0.0631
404
+ 100
405
+ 1e-05
406
+ changing lr
407
+ epoch 95, time 179.67, cls_loss 0.0525
408
+ 100
409
+ 1e-05
410
+ changing lr
411
+ epoch 96, time 179.69, cls_loss 0.0473
412
+ 100
413
+ 1e-05
414
+ changing lr
415
+ epoch 97, time 179.39, cls_loss 0.0470
416
+ 100
417
+ 1e-05
418
+ changing lr
419
+ epoch 98, time 179.75, cls_loss 0.0529
420
+ 100
421
+ 1e-05
422
+ changing lr
423
+ epoch 99, time 180.06, cls_loss 0.0541
424
+ ---------------------saving last model at epoch 99----------------------------------------------------
425
+ /home/yuqian_fu
426
+ {'gpu': '0', 'svroot': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit/CA_multiple_14fa_all_ep100_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3_pipelineAugWoNorm', 'svpath': '/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit/CA_multiple_14fa_all_ep100_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3_pipelineAugWoNorm/14factor_best.csv', 'channels': 3, 'factor_num': 14, 'stride': 3, 'epoch': 'best', 'eval_mapping': True}
427
+ loading weight of best
428
+ Using downloaded and verified file: /home/yuqian_fu/.pytorch/SVHN/test_32x32.mat
429
+ mnist svhn ... usps Avg
430
+ w/o do (original x) 93.89 13.579441 ... 89.436971 40.16719
431
+
432
+ [1 rows x 6 columns]
Meta-causal/code-stage1-pipeline/AllEpochs_test_digit_v13.py ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+ from torch.utils.data import DataLoader
6
+
7
+ import os
8
+ import numpy as np
9
+ import click
10
+ import pandas as pd
11
+
12
+ from network import mnist_net_my as mnist_net
13
+ from network import adaptor_v2
14
+ from tools import causalaugment_v3 as causalaugment
15
+ from main_my_joint_v13_auto import evaluate
16
+ import data_loader_joint_v3 as data_loader
17
+
18
+ @click.command()
19
+ @click.option('--gpu', type=str, default='0', help='选择GPU编号')
20
+ @click.option('--svroot', type=str, default='./saved')
21
+ @click.option('--svpath', type=str, default=None, help='保存日志的路径')
22
+ @click.option('--channels', type=int, default=3)
23
+ @click.option('--factor_num', type=int, default=16)
24
+ @click.option('--stride', type=int, default=16)
25
+ @click.option('--epoch', type=str, default='best')
26
+ @click.option('--eval_mapping', type=bool, default=True, help='是否查看mapping学习效果')
27
+ def main(gpu, svroot, svpath, channels, factor_num,stride, epoch, eval_mapping):
28
+ evaluate_digit(gpu, svroot, svpath, channels, factor_num, stride,epoch, eval_mapping)
29
+
30
+ def evaluate_digit(gpu, svroot, svpath, channels=3, factor_num=16,stride=5,epoch='best', eval_mapping=True):
31
+ settings = locals().copy()
32
+ print(settings)
33
+ os.environ['CUDA_VISIBLE_DEVICES'] = gpu
34
+
35
+ # 加载分类模型
36
+ if channels == 3:
37
+ cls_net = mnist_net.ConvNet().cuda()
38
+ elif channels == 1:
39
+ cls_net = mnist_net.ConvNet(imdim=channels).cuda()
40
+
41
+
42
+ epoch_list = []
43
+ file_list = os.listdir(svroot)
44
+ for file in file_list:
45
+ if('.pkl' in file):
46
+ epoch_list.append(file)
47
+ print('epoch_list:', epoch_list)
48
+
49
+ '''
50
+ if epoch == 'best':
51
+ print("loading weight of %s"%(epoch))
52
+ saved_weight = torch.load(os.path.join(svroot, 'best_cls_net.pkl'))
53
+ elif epoch == 'last':
54
+ print("loading weight of %s"%(epoch))
55
+ saved_weight = torch.load(os.path.join(svroot, 'last_cls_net.pkl'))
56
+ '''
57
+
58
+ for epoch_file in epoch_list:
59
+ print("loading weight of %s"%(epoch_file))
60
+ saved_weight = torch.load(os.path.join(svroot, epoch_file))
61
+
62
+ cls_net.load_state_dict(saved_weight)
63
+ cls_net.eval()
64
+
65
+ # 测试
66
+ str2fun = {
67
+ 'mnist': data_loader.load_mnist,
68
+ 'mnist_m': data_loader.load_mnist_m,
69
+ 'usps': data_loader.load_usps,
70
+ 'svhn': data_loader.load_svhn,
71
+ 'syndigit': data_loader.load_syndigit,
72
+ }
73
+
74
+ columns = ['mnist', 'svhn', 'mnist_m', 'syndigit','usps']
75
+ target = ['svhn', 'mnist_m', 'syndigit','usps']
76
+
77
+ index = ['w/o do (original x)']
78
+ data_result = {}
79
+
80
+ for idx, data in enumerate(columns):
81
+ teset = str2fun[data]('test', channels=channels)
82
+ teloader = DataLoader(teset, batch_size=8, num_workers=0)
83
+ # 计算评价指标
84
+ teacc = evaluate(cls_net, teloader)
85
+ if data == 'mnist':
86
+ acc_avg = np.zeros(teacc.shape)
87
+ else:
88
+ acc_avg = acc_avg + teacc
89
+ data_result[data] = teacc
90
+ acc_avg = acc_avg/float(len(target))
91
+
92
+ data_result['Avg'] = acc_avg
93
+
94
+ df = pd.DataFrame(data_result,index = index)
95
+ print(df)
96
+ if svpath is not None:
97
+ df.to_csv(svpath)
98
+
99
+ if __name__=='__main__':
100
+ main()
101
+
Meta-causal/code-stage1-pipeline/AllEpochs_test_pacs_v13.py ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+ from torch.utils.data import DataLoader
6
+
7
+ import os
8
+ import numpy as np
9
+ import click
10
+ import pandas as pd
11
+
12
+ from network import resnet as resnet
13
+ from network import adaptor_v2
14
+ from tools import causalaugment_v3 as causalaugment
15
+ from main_my_joint_v13_auto import evaluate
16
+ import data_loader_joint_v3 as data_loader
17
+
18
+ @click.command()
19
+ @click.option('--gpu', type=str, default='0', help='选择GPU编号')
20
+ @click.option('--svroot', type=str, default='./saved')
21
+ @click.option('--source_domain', type=str, default='art_painting', help='source domain')
22
+ @click.option('--svpath', type=str, default=None, help='保存日志的路径')
23
+ @click.option('--factor_num', type=int, default=16)
24
+ @click.option('--epoch', type=str, default='best')
25
+ @click.option('--stride', type=int, default=5)
26
+ @click.option('--eval_mapping', type=bool, default=False, help='是否查看mapping学习效果')
27
+ @click.option('--network', type=str, default='resnet18', help='项目文件保存路径')
28
+ def main(gpu, svroot, source_domain, svpath, factor_num, epoch, stride,eval_mapping, network):
29
+ evaluate_pacs(gpu, svroot, source_domain, svpath, factor_num, epoch, stride,eval_mapping, network)
30
+
31
+ def evaluate_pacs(gpu, svroot, source_domain, svpath, factor_num=16, epoch='best', stride=5,eval_mapping=False, network='resnet18'):
32
+ settings = locals().copy()
33
+ print(settings)
34
+ os.environ['CUDA_VISIBLE_DEVICES'] = gpu
35
+
36
+ # 加载分类模型
37
+ if network == 'resnet18':
38
+ cls_net = resnet.resnet18(classes=7,c_dim=2048).cuda()
39
+ input_dim = 2048
40
+
41
+ epoch_list = []
42
+ file_list = os.listdir(svroot)
43
+ for file in file_list:
44
+ if('.pkl' in file):
45
+ epoch_list.append(file)
46
+ print('epoch_list:', epoch_list)
47
+
48
+ '''
49
+ if epoch == 'best':
50
+ print("loading weight of %s"%(epoch))
51
+ saved_weight = torch.load(os.path.join(svroot, 'best_cls_net.pkl'))
52
+ elif epoch == 'last':
53
+ print("loading weight of %s"%(epoch))
54
+ saved_weight = torch.load(os.path.join(svroot, 'last_cls_net.pkl'))
55
+ '''
56
+
57
+ for epoch_file in epoch_list:
58
+ print("loading weight of %s"%(epoch_file))
59
+ saved_weight = torch.load(os.path.join(svroot, epoch_file))
60
+
61
+ cls_net.load_state_dict(saved_weight)
62
+ cls_net.eval()
63
+
64
+
65
+ columns = ['art_painting', 'cartoon', 'photo', 'sketch']
66
+ target = [i for i in columns if i!=source_domain]
67
+ columns = [source_domain] + target
68
+ print("columns:",columns)
69
+
70
+
71
+ index = ['w/o do (original x)']
72
+
73
+ data_result = {}
74
+ data_result_ours = {}
75
+
76
+ for idx, data in enumerate(columns):
77
+ teset = data_loader.load_pacs(data, 'test')
78
+ teloader = DataLoader(teset, batch_size=4, num_workers=0)
79
+ # 计算评价指标
80
+ acc = evaluate(cls_net, teloader)
81
+ data_result_ours[data] = acc
82
+
83
+ teacc = evaluate(cls_net, teloader)
84
+ if data == source_domain:
85
+ acc_avg = np.zeros(teacc.shape)
86
+ else:
87
+ acc_avg = acc_avg + teacc
88
+ data_result[data] = teacc
89
+ acc_avg = acc_avg/float(len(target))
90
+
91
+ data_result['Avg'] = acc_avg
92
+
93
+ df = pd.DataFrame(data_result,index = index)
94
+ print(df)
95
+
96
+ if svpath is not None:
97
+ df.to_csv(svpath)
98
+
99
+ if __name__=='__main__':
100
+ main()
101
+
102
+
103
+
Meta-causal/code-stage1-pipeline/data_loader_joint_v3.py ADDED
@@ -0,0 +1,861 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ''' Digit 实验
2
+ '''
3
+ import torch
4
+ import torch.nn.functional as F
5
+ from torch.utils.data import Dataset, TensorDataset
6
+ from torchvision import transforms
7
+ from torchvision.datasets import MNIST, SVHN, CIFAR10, STL10, USPS
8
+
9
+ import os
10
+ import pickle
11
+ import numpy as np
12
+ import h5py
13
+ #import cv2
14
+ from scipy.io import loadmat
15
+ from PIL import Image
16
+
17
+ from tools.autoaugment import SVHNPolicy, CIFAR10Policy
18
+ from tools.randaugment import RandAugment
19
+ from tools.causalaugment_v3 import RandAugment_incausal, FactualAugment_incausal, CounterfactualAugment_incausal, MultiCounterfactualAugment_incausal
20
+
21
+ from PIL import ImageEnhance
22
+
23
+
24
+ transformtypedict=dict(Brightness=ImageEnhance.Brightness, Contrast=ImageEnhance.Contrast, Sharpness=ImageEnhance.Sharpness, Color=ImageEnhance.Color)
25
+
26
+ class ImageJitterforX(object):
27
+ '''
28
+ from StyleAdv dataaug
29
+ '''
30
+ def __init__(self, transformdict):
31
+ self.transforms = [(transformtypedict[k], transformdict[k]) for k in transformdict]
32
+
33
+
34
+ def __call__(self, img):
35
+ out = img
36
+ randtensor = torch.rand(len(self.transforms))
37
+
38
+ for i, (transformer, alpha) in enumerate(self.transforms):
39
+ r = alpha*(randtensor[i]*2.0 -1.0) + 1
40
+ out = transformer(out).enhance(r).convert('RGB')
41
+
42
+ return out
43
+
44
+ class TransformLoaderforX:
45
+ '''
46
+ from StyleAdv dataaug
47
+ '''
48
+ def __init__(self, image_size,
49
+ normalize_param = dict(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
50
+ jitter_param = dict(Brightness=0.4, Contrast=0.4, Color=0.4)):
51
+ self.image_size = image_size
52
+ self.normalize_param = normalize_param
53
+ self.jitter_param = jitter_param
54
+
55
+ def parse_transform(self, transform_type):
56
+ if transform_type=='ImageJitter':
57
+ method = ImageJitterforX( self.jitter_param )
58
+ return method
59
+ method = getattr(transforms, transform_type)
60
+
61
+ if transform_type=='RandomResizedCrop':
62
+ return method(self.image_size)
63
+ elif transform_type=='CenterCrop':
64
+ return method(self.image_size)
65
+ elif transform_type=='Resize':
66
+ return method([int(self.image_size*1.15), int(self.image_size*1.15)])
67
+ elif transform_type=='Normalize':
68
+ return method(**self.normalize_param )
69
+ else:
70
+ return method()
71
+
72
+
73
+ def get_composed_transform(self, aug = False):
74
+ if aug:
75
+ #transform_list = ['RandomResizedCrop', 'ImageJitter', 'RandomHorizontalFlip', 'ToTensor', 'Normalize']
76
+ transform_list = ['RandomResizedCrop', 'ImageJitter', 'RandomHorizontalFlip', 'ToTensor']
77
+ else:
78
+ #transform_list = ['Resize','CenterCrop', 'ToTensor', 'Normalize']
79
+ #transform_list = ['ToTensor', 'Normalize']
80
+ transform_list = ['ToTensor']
81
+
82
+ tranform0 = [transforms.ToPILImage()]
83
+ transform_funcs = [ self.parse_transform(x) for x in transform_list]
84
+ tranform_all = tranform0 + transform_funcs
85
+ transform = transforms.Compose(tranform_all)
86
+ return transform
87
+
88
+
89
+ class myTensorDataset(Dataset):
90
+ def __init__(self, x, y, transform=None, transform2=None, transform3=None, twox=False):
91
+ self.x = x
92
+ self.y = y
93
+ self.transform = transform
94
+ self.transform2 = transform2
95
+ self.transform3 = transform3
96
+ self.twox = twox
97
+ def __len__(self):
98
+ return len(self.x)
99
+
100
+ def __getitem__(self, index):
101
+ x = self.x[index]
102
+ y = self.y[index]
103
+ c, h, w =x.shape
104
+ if self.transform is not None:
105
+ x_RA = self.transform(x)
106
+ # print("x_RA.shape:",x_RA.shape)
107
+ if self.transform3 is not None:
108
+ x_CA = self.transform3(x_RA)
109
+ x_CA = x_CA.reshape(-1,c,h,w)
110
+ # print("x_CA.shape:",x_CA.shape)
111
+ if self.transform2 is not None:
112
+ x_FA = self.transform2(x)
113
+ # x_FA = x_FA.view(c,13,h,w)
114
+ x_FA = x_FA.reshape(-1,c,h,w)
115
+ # print("x_FA_in getitem.shape:",x_FA.shape)
116
+ # print("x_FA.shape:",x_FA.shape)
117
+
118
+ return (x, x_RA, x_FA, x_CA), y
119
+ else:
120
+ return (x, x_RA, x_CA), y
121
+ else:
122
+ if self.transform2 is not None:
123
+ x_FA = self.transform2(x)
124
+ x_FA = x_FA.reshape(-1,c,h,w)
125
+ return (x, x_RA, x_FA), y
126
+ else:
127
+ if self.twox:
128
+ return (x, x_RA), y
129
+ else:
130
+ x_RA = self.transform(x)
131
+ return x_RA, y
132
+
133
+
134
+ HOME = os.environ['HOME']
135
+ print(HOME)
136
+ def resize_imgs(x, size):
137
+ ''' 目前只能处理单通道
138
+ x [n, 28, 28]
139
+ size int
140
+ '''
141
+ resize_x = np.zeros([x.shape[0], size, size])
142
+ for i, im in enumerate(x):
143
+ im = Image.fromarray(im)
144
+ im = im.resize([size, size], Image.ANTIALIAS)
145
+ resize_x[i] = np.asarray(im)
146
+ return resize_x
147
+
148
+ def load_mnist(split='train', translate=None, twox=False, ntr=None, autoaug=None, factor_num=16, randm=False,randn=False,channels=3,n=3,stride=5):
149
+ '''
150
+ autoaug == 'AA', AutoAugment
151
+ 'FastAA', Fast AutoAugment
152
+ 'RA', RandAugment
153
+ channels == 3 默认返回 rgb 3通道图像
154
+ 1 返回单通道图像
155
+ '''
156
+ #path = f'data/mnist-{split}.pkl'
157
+ path = f'/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/minst-{split}.pkl'
158
+ if not os.path.exists(path):
159
+ dataset = MNIST(f'{HOME}/.pytorch/MNIST', train=(split=='train'), download=True)
160
+ x, y = dataset.data, dataset.targets
161
+ if split=='train':
162
+ x, y = x[0:10000], y[0:10000]
163
+ x = torch.tensor(resize_imgs(x.numpy(), 32))
164
+ x = (x.float()/255.).unsqueeze(1).repeat(1,3,1,1)
165
+ with open(path, 'wb') as f:
166
+ pickle.dump([x, y], f)
167
+ with open(path, 'rb') as f:
168
+ # print("reading!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!")
169
+ x, y = pickle.load(f)
170
+ if channels == 1:
171
+ x = x[:,0:1,:,:]
172
+
173
+ if ntr is not None:
174
+ x, y = x[0:ntr], y[0:ntr]
175
+
176
+ # 如果没有数据增强
177
+ if (translate is None) and (autoaug is None):
178
+ dataset = TensorDataset(x, y)
179
+ return dataset
180
+
181
+
182
+ #fuyuqian: add styleadv-style aug
183
+ transform_x_train = TransformLoaderforX((x.shape[-2], x.shape[-1])).get_composed_transform(aug=True)
184
+ transform_x_test = TransformLoaderforX((x.shape[-2], x.shape[-1])).get_composed_transform(aug=False)
185
+ if(split == 'train'):
186
+ transformed_images = []
187
+ for img in x:
188
+ img = transform_x_train(img) # Apply transform to each image
189
+ transformed_images.append(img)
190
+ x = torch.stack(transformed_images)
191
+ #print('x_aug train here', x.shape)
192
+ else:
193
+ transformed_images = []
194
+ for img in x:
195
+ img = transform_x_test(img) # Apply transform to each image
196
+ transformed_images.append(img)
197
+ x = torch.stack(transformed_images)
198
+ #print('x_aug test here', x.shape)
199
+
200
+
201
+
202
+ transform = [transforms.ToPILImage()]
203
+ transform_single_factor = [transforms.ToPILImage()]
204
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
205
+ transform_CA = [transforms.ToPILImage()]
206
+ if translate is not None:
207
+ transform.append(transforms.RandomAffine(0, [translate, translate]))
208
+ transform_single_factor.append(transforms.RandomAffine(0, [translate, translate]))
209
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
210
+ transform_CA.append(transforms.RandomAffine(0, [translate, translate]))
211
+ if autoaug is not None:
212
+ if autoaug == 'CA':
213
+ print("--------------------------CA--------------------------")
214
+ print("n:",n)
215
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
216
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
217
+ transform_CA.append(CounterfactualAugment_incausal(factor_num))
218
+ elif autoaug == 'CA_multiple':
219
+ print("--------------------------CA_multiple--------------------------")
220
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
221
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
222
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
223
+ elif autoaug == 'Ours_A':
224
+ print("--------------------------Ours_Augment--------------------------")
225
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
226
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
227
+
228
+ transform.append(transforms.ToTensor())
229
+ transform = transforms.Compose(transform)
230
+ transform_single_factor.append(transforms.ToTensor())
231
+ transform_single_factor = transforms.Compose(transform_single_factor)
232
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
233
+ transform_CA.append(transforms.ToTensor())
234
+ transform_CA = transforms.Compose(transform_CA)
235
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, transform3=transform_CA,twox=twox)
236
+ else:
237
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, twox=twox)
238
+ # print(x.shape)
239
+ # print(y.shape)
240
+ return dataset
241
+
242
+ def load_cifar10(split='train', translate=None, twox=False, autoaug=None, factor_num=16, randm=False,randn=False,channels=3,n=3,stride=5):
243
+ dataset = CIFAR10(f'{HOME}/.pytorch/CIFAR10', train=(split=='train'), download=True)
244
+ x, y = dataset.data, dataset.targets
245
+ x = x.transpose(0,3,1,2)
246
+
247
+ x, y = torch.tensor(x), torch.tensor(y)
248
+ x = x.float()/255.
249
+ print(x.shape,y.shape)
250
+ if (translate is None) and (autoaug is None):
251
+ dataset = TensorDataset(x, y)
252
+ return dataset
253
+ #x.transpose(0,3,1,2)
254
+
255
+ # 数据增强管道
256
+ transform = [transforms.ToPILImage()]
257
+ transform_single_factor = [transforms.ToPILImage()]
258
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
259
+ transform_CA = [transforms.ToPILImage()]
260
+ if translate is not None:
261
+ transform.append(transforms.RandomAffine(0, [translate, translate]))
262
+ transform_single_factor.append(transforms.RandomAffine(0, [translate, translate]))
263
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
264
+ transform_CA.append(transforms.RandomAffine(0, [translate, translate]))
265
+ if autoaug is not None:
266
+ if autoaug == 'CA':
267
+ print("--------------------------CA--------------------------")
268
+ print("n:",n)
269
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
270
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
271
+ transform_CA.append(CounterfactualAugment_incausal(factor_num))
272
+ elif autoaug == 'CA_multiple':
273
+ print("--------------------------CA_multiple--------------------------")
274
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
275
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
276
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
277
+ elif autoaug == 'Ours_A':
278
+ print("--------------------------Ours_Augment--------------------------")
279
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
280
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
281
+
282
+ transform.append(transforms.ToTensor())
283
+ transform = transforms.Compose(transform)
284
+ transform_single_factor.append(transforms.ToTensor())
285
+ transform_single_factor = transforms.Compose(transform_single_factor)
286
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
287
+ transform_CA.append(transforms.ToTensor())
288
+ transform_CA = transforms.Compose(transform_CA)
289
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, transform3=transform_CA,twox=twox)
290
+ else:
291
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, twox=twox)
292
+ # print(x.shape)
293
+ # print(y.shape)
294
+ return dataset
295
+ def load_IMG(task='S-U', translate=None, twox=False, autoaug=None, factor_num=16, randm=False,randn=False,channels=3,n=3,stride=5):
296
+ # path = f'data/img2vid/{domain}/stanford40_12.npz'
297
+ if task == 'S-U':
298
+ path = f'data/img2vid/{task}/stanford40_12.npz'
299
+ elif task == 'E-H':
300
+ path = f'data/img2vid/{task}/EAD50_13.npz'
301
+ print(path)
302
+ dataset = np.load(path)
303
+ x, y = dataset['x'], dataset['y']
304
+ b, g, r = np.split(x,3,axis=-1)
305
+ x = np.concatenate((r,g,b),axis=-1)
306
+ x = x.transpose(0,3,1,2)
307
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
308
+ x = x.float()/255.
309
+ print(path,x.shape,y.shape)
310
+ # for i in range(20):
311
+ # img_temp = transforms.ToPILImage()(x[i])
312
+ # img_temp.save('data/PACS/debug_images/img_pil_'+domain+'_'+split+'_'+str(i)+'.png')
313
+ if (translate is None) and (autoaug is None):
314
+ dataset = TensorDataset(x, y)
315
+ return dataset
316
+
317
+ #x.transpose(0,3,1,2)
318
+
319
+ # 数据增强管道
320
+ transform = [transforms.ToPILImage()]
321
+ if autoaug != 'CA_multiple_noSingle':
322
+ transform_single_factor = [transforms.ToPILImage()]
323
+ if autoaug == 'CA' or autoaug == 'CA_multiple' or autoaug == 'CA_multiple_noSingle':
324
+ transform_CA = [transforms.ToPILImage()]
325
+ if translate is not None:
326
+ transform.append(transforms.RandomAffine(0, [translate, translate]))
327
+ if autoaug != 'CA_multiple_noSingle':
328
+ transform_single_factor.append(transforms.RandomAffine(0, [translate, translate]))
329
+ if autoaug == 'CA' or autoaug == 'CA_multiple' or autoaug == 'CA_multiple_noSingle':
330
+ transform_CA.append(transforms.RandomAffine(0, [translate, translate]))
331
+ if autoaug is not None:
332
+ if autoaug == 'CA':
333
+ print("--------------------------CA--------------------------")
334
+ print("n:",n)
335
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
336
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
337
+ transform_CA.append(CounterfactualAugment_incausal(factor_num))
338
+ elif autoaug == 'CA_multiple':
339
+ print("--------------------------CA_multiple--------------------------")
340
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
341
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
342
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
343
+ elif autoaug == 'CA_multiple_noSingle':
344
+ print("--------------------------CA_multiple_noSingle--------------------------")
345
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
346
+ # transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
347
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
348
+ elif autoaug == 'Ours_A':
349
+ print("--------------------------Ours_Augment--------------------------")
350
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
351
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
352
+
353
+ transform.append(transforms.ToTensor())
354
+ transform = transforms.Compose(transform)
355
+ if autoaug != 'CA_multiple_noSingle':
356
+ transform_single_factor.append(transforms.ToTensor())
357
+ transform_single_factor = transforms.Compose(transform_single_factor)
358
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
359
+ transform_CA.append(transforms.ToTensor())
360
+ transform_CA = transforms.Compose(transform_CA)
361
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, transform3=transform_CA,twox=twox)
362
+ elif autoaug == 'CA_multiple_noSingle':
363
+ transform_CA.append(transforms.ToTensor())
364
+ transform_CA = transforms.Compose(transform_CA)
365
+ dataset = myTensorDataset(x, y, transform=transform, transform3=transform_CA,twox=twox)
366
+ else:
367
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, twox=twox)
368
+ # print(x.shape)
369
+ # print(y.shape)
370
+ return dataset
371
+
372
+ def load_VID(task='S-U',split='1'):
373
+ if task == 'S-U':
374
+ path = f'data/img2vid/{task}/ucf101_12_frame_sample8_{split}.npz'
375
+ elif task == 'E-H':
376
+ path = f'data/img2vid/{task}/hmdb51_13_frame_sample8_{split}.npz'
377
+ dataset = np.load(path)
378
+ print(path)
379
+ x, y = dataset['x'], dataset['y']
380
+ b, g, r = np.split(x,3,axis=-1)
381
+ x = np.concatenate((r,g,b),axis=-1)
382
+ x = x.transpose(0,3,1,2)
383
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
384
+ x = x.float()/255.
385
+ print(path,x.shape,y.shape)
386
+ # for i in range(20):
387
+ # img_temp = transforms.ToPILImage()(x[i])
388
+ # img_temp.save('data/PACS/debug_images/img_pil_'+domain+'_'+split+'_'+str(i)+'.png')
389
+ dataset = TensorDataset(x, y)
390
+ return dataset
391
+
392
+ def load_pacs(domain='photo', split='train', translate=None, twox=False, autoaug=None, factor_num=16, randm=False,randn=False,channels=3,n=3,stride=5):
393
+ #path = f'data/PACS/{domain}_{split}.hdf5'
394
+ path = f'/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/{domain}_{split}.hdf5'
395
+ dataset = h5py.File(path, 'r')
396
+ x, y = dataset['images'], dataset['labels']
397
+ #for i in range(20):
398
+ # cv2.imwrite('data/PACS/debug_images/img_cv2_'+domain+'_'+split+'_'+str(i)+'.png', x[i])
399
+ b, g, r = np.split(x,3,axis=-1)
400
+ x = np.concatenate((r,g,b),axis=-1)
401
+
402
+ #x = x.transpose(0,3,1,2)
403
+ # Convert image data to uint8
404
+
405
+
406
+ #fuyuqian: add styleadv-style aug
407
+ x = x.astype(np.uint8)
408
+ transform_x_train = TransformLoaderforX((x.shape[-3], x.shape[-2])).get_composed_transform(aug=True)
409
+ transform_x_test = TransformLoaderforX((x.shape[-3], x.shape[-2])).get_composed_transform(aug=False)
410
+ if(split == 'train'):
411
+ transformed_images = []
412
+ for img in x:
413
+ img = transform_x_train(img) # Apply transform to each image
414
+ transformed_images.append(img)
415
+ x = torch.stack(transformed_images)
416
+ #print('x_aug train here', x.shape)
417
+ else:
418
+ transformed_images = []
419
+ for img in x:
420
+ img = transform_x_test(img) # Apply transform to each image
421
+ transformed_images.append(img)
422
+ x = torch.stack(transformed_images)
423
+ #print('x_aug test here', x.shape)
424
+
425
+
426
+ x, y = torch.tensor(x), torch.tensor(y, dtype=torch.long)
427
+
428
+ y = y - 1
429
+ x = x.float()/255.
430
+ print(path,x.shape,y.shape)
431
+ # for i in range(20):
432
+ # img_temp = transforms.ToPILImage()(x[i])
433
+ # img_temp.save('data/PACS/debug_images/img_pil_'+domain+'_'+split+'_'+str(i)+'.png')
434
+ if (translate is None) and (autoaug is None):
435
+ dataset = TensorDataset(x, y)
436
+ return dataset
437
+ #x.transpose(0,3,1,2)
438
+
439
+ # 数据增强管道
440
+ transform = [transforms.ToPILImage()]
441
+ if autoaug != 'CA_multiple_noSingle':
442
+ transform_single_factor = [transforms.ToPILImage()]
443
+ if autoaug == 'CA' or autoaug == 'CA_multiple' or autoaug == 'CA_multiple_noSingle':
444
+ transform_CA = [transforms.ToPILImage()]
445
+ if translate is not None:
446
+ transform.append(transforms.RandomAffine(0, [translate, translate]))
447
+ if autoaug != 'CA_multiple_noSingle':
448
+ transform_single_factor.append(transforms.RandomAffine(0, [translate, translate]))
449
+ if autoaug == 'CA' or autoaug == 'CA_multiple' or autoaug == 'CA_multiple_noSingle':
450
+ transform_CA.append(transforms.RandomAffine(0, [translate, translate]))
451
+ if autoaug is not None:
452
+ if autoaug == 'CA':
453
+ print("--------------------------CA--------------------------")
454
+ print("n:",n)
455
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
456
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
457
+ transform_CA.append(CounterfactualAugment_incausal(factor_num))
458
+ elif autoaug == 'CA_multiple':
459
+ print("--------------------------CA_multiple--------------------------")
460
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
461
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
462
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
463
+ elif autoaug == 'CA_multiple_noSingle':
464
+ print("--------------------------CA_multiple_noSingle--------------------------")
465
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
466
+ # transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
467
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
468
+ elif autoaug == 'Ours_A':
469
+ print("--------------------------Ours_Augment--------------------------")
470
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
471
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
472
+
473
+ transform.append(transforms.ToTensor())
474
+ transform = transforms.Compose(transform)
475
+ if autoaug != 'CA_multiple_noSingle':
476
+ transform_single_factor.append(transforms.ToTensor())
477
+ transform_single_factor = transforms.Compose(transform_single_factor)
478
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
479
+ transform_CA.append(transforms.ToTensor())
480
+ transform_CA = transforms.Compose(transform_CA)
481
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, transform3=transform_CA,twox=twox)
482
+ elif autoaug == 'CA_multiple_noSingle':
483
+ transform_CA.append(transforms.ToTensor())
484
+ transform_CA = transforms.Compose(transform_CA)
485
+ dataset = myTensorDataset(x, y, transform=transform, transform3=transform_CA,twox=twox)
486
+ else:
487
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, twox=twox)
488
+ # print(x.shape)
489
+ # print(y.shape)
490
+ return dataset
491
+
492
+ def read_dataset(domain, split):
493
+ path = f'/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/PACS/{domain}_{split}.hdf5'
494
+ dataset = h5py.File(path, 'r')
495
+ x_temp, y_temp = dataset['images'], dataset['labels']
496
+ b, g, r = np.split(x_temp,3,axis=-1)
497
+ x_temp = np.concatenate((r,g,b),axis=-1)
498
+ x_temp = x_temp.transpose(0,3,1,2)
499
+ x_temp, y_temp = torch.tensor(x_temp), torch.tensor(y_temp, dtype=torch.long)
500
+ y_temp = y_temp - 1
501
+ x_temp = x_temp.float()/255.
502
+ return x_temp, y_temp
503
+
504
+ def load_pacs_multi(target_domain=['photo'], split='train', translate=None, twox=False, autoaug=None, factor_num=16, randm=False,randn=False,channels=3,n=3,stride=5):
505
+ domains = ['art_painting', 'cartoon', 'photo', 'sketch']
506
+ source_domain = [i for i in domains if i != target_domain]
507
+ for i in range(len(source_domain)):
508
+ x_temp, y_temp = read_dataset(source_domain[i],split=split)
509
+ print(x_temp.shape,y_temp.shape)
510
+ if i == 0:
511
+ x = x_temp.clone()
512
+ y = y_temp.clone()
513
+ else:
514
+ x = torch.cat([x,x_temp],0)
515
+ y = torch.cat([y,y_temp],0)
516
+ print(x.shape,y.shape)
517
+ if (translate is None) and (autoaug is None):
518
+ dataset = TensorDataset(x, y)
519
+ return dataset
520
+ #x.transpose(0,3,1,2)
521
+
522
+ # 数据增强管道
523
+ transform = [transforms.ToPILImage()]
524
+ if autoaug != 'CA_multiple_noSingle':
525
+ transform_single_factor = [transforms.ToPILImage()]
526
+ if autoaug == 'CA' or autoaug == 'CA_multiple' or autoaug == 'CA_multiple_noSingle':
527
+ transform_CA = [transforms.ToPILImage()]
528
+ if translate is not None:
529
+ transform.append(transforms.RandomAffine(0, [translate, translate]))
530
+ if autoaug != 'CA_multiple_noSingle':
531
+ transform_single_factor.append(transforms.RandomAffine(0, [translate, translate]))
532
+ if autoaug == 'CA' or autoaug == 'CA_multiple' or autoaug == 'CA_multiple_noSingle':
533
+ transform_CA.append(transforms.RandomAffine(0, [translate, translate]))
534
+ if autoaug is not None:
535
+ if autoaug == 'CA':
536
+ print("--------------------------CA--------------------------")
537
+ print("n:",n)
538
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
539
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
540
+ transform_CA.append(CounterfactualAugment_incausal(factor_num))
541
+ elif autoaug == 'CA_multiple':
542
+ print("--------------------------CA_multiple--------------------------")
543
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
544
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
545
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
546
+ elif autoaug == 'CA_multiple_noSingle':
547
+ print("--------------------------CA_multiple_noSingle--------------------------")
548
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
549
+ # transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
550
+ transform_CA.append(MultiCounterfactualAugment_incausal(factor_num, stride))
551
+ elif autoaug == 'Ours_A':
552
+ print("--------------------------Ours_Augment--------------------------")
553
+ transform.append(RandAugment_incausal(n,4,factor_num, randm=randm,randn=randn))
554
+ transform_single_factor.append(FactualAugment_incausal(4, factor_num, randm=False))
555
+
556
+ transform.append(transforms.ToTensor())
557
+ transform = transforms.Compose(transform)
558
+ if autoaug != 'CA_multiple_noSingle':
559
+ transform_single_factor.append(transforms.ToTensor())
560
+ transform_single_factor = transforms.Compose(transform_single_factor)
561
+ if autoaug == 'CA' or autoaug == 'CA_multiple':
562
+ transform_CA.append(transforms.ToTensor())
563
+ transform_CA = transforms.Compose(transform_CA)
564
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, transform3=transform_CA,twox=twox)
565
+ elif autoaug == 'CA_multiple_noSingle':
566
+ transform_CA.append(transforms.ToTensor())
567
+ transform_CA = transforms.Compose(transform_CA)
568
+ dataset = myTensorDataset(x, y, transform=transform, transform3=transform_CA,twox=twox)
569
+ else:
570
+ dataset = myTensorDataset(x, y, transform=transform, transform2=transform_single_factor, twox=twox)
571
+ # print(x.shape)
572
+ # print(y.shape)
573
+ return dataset
574
+
575
+
576
+ def load_cifar10_c_level1(dataroot):
577
+ path = f'data/cifar10_c_level1.pkl'
578
+ if not os.path.exists(path):
579
+ print("genenrating cifar10_c_level1")
580
+ labels = np.load(os.path.join(dataroot, 'labels.npy'))
581
+ y_single = labels[0:10000]
582
+ x = torch.zeros((190000,3,32,32))
583
+ for j in range(19):
584
+ if j == 0:
585
+ y = y_single
586
+ else:
587
+ y = np.hstack((y,y_single))
588
+ index = 0
589
+ for filename in os.listdir(dataroot):
590
+ if filename=='labels.npy':
591
+ continue
592
+ else:
593
+ imgs = np.load(os.path.join(dataroot,filename))
594
+ imgs = imgs.transpose(0,3,1,2)
595
+ imgs = torch.tensor(imgs)
596
+ imgs = imgs.float()/255.
597
+ print(imgs.shape)
598
+ x[index*10000:(index+1)*10000] = imgs[0:10000]
599
+ index = index + 1
600
+ y = torch.tensor(y)
601
+ with open(path, 'wb') as f:
602
+ pickle.dump([x, y], f)
603
+ else:
604
+ print("reading cifar10_c_level1")
605
+ with open(path, 'rb') as f:
606
+ x, y = pickle.load(f)
607
+ dataset = TensorDataset(x, y)
608
+ return dataset
609
+ def load_cifar10_c_level2(dataroot):
610
+ path = f'data/cifar10_c_level2.pkl'
611
+ if not os.path.exists(path):
612
+ print("genenrating cifar10_c_level2")
613
+ labels = np.load(os.path.join(dataroot, 'labels.npy'))
614
+ y_single = labels[0:10000]
615
+ x = torch.zeros((190000,3,32,32))
616
+ for j in range(19):
617
+ if j == 0:
618
+ y = y_single
619
+ else:
620
+ y = np.hstack((y,y_single))
621
+ index = 0
622
+ for filename in os.listdir(dataroot):
623
+ if filename=='labels.npy':
624
+ continue
625
+ else:
626
+ imgs = np.load(os.path.join(dataroot,filename))
627
+ imgs = imgs.transpose(0,3,1,2)
628
+ imgs = torch.tensor(imgs)
629
+ imgs = imgs.float()/255.
630
+ print(imgs.shape)
631
+ x[index*10000:(index+1)*10000] = imgs[10000:20000]
632
+ index = index + 1
633
+ y = torch.tensor(y)
634
+ with open(path, 'wb') as f:
635
+ pickle.dump([x, y], f)
636
+ else:
637
+ print("reading cifar10_c_level2")
638
+ with open(path, 'rb') as f:
639
+ x, y = pickle.load(f)
640
+ dataset = TensorDataset(x, y)
641
+ return dataset
642
+ def load_cifar10_c_level3(dataroot):
643
+ path = f'data/cifar10_c_level3.pkl'
644
+ if not os.path.exists(path):
645
+ print("generating cifar10_c_level3")
646
+ labels = np.load(os.path.join(dataroot, 'labels.npy'))
647
+ y_single = labels[0:10000]
648
+ x = torch.zeros((190000,3,32,32))
649
+ for j in range(19):
650
+ if j == 0:
651
+ y = y_single
652
+ else:
653
+ y = np.hstack((y,y_single))
654
+ index = 0
655
+ for filename in os.listdir(dataroot):
656
+ if filename=='labels.npy':
657
+ continue
658
+ else:
659
+ imgs = np.load(os.path.join(dataroot,filename))
660
+ imgs = imgs.transpose(0,3,1,2)
661
+ imgs = torch.tensor(imgs)
662
+ imgs = imgs.float()/255.
663
+ print(imgs.shape)
664
+ x[index*10000:(index+1)*10000] = imgs[20000:30000]
665
+ index = index + 1
666
+ y = torch.tensor(y)
667
+ with open(path, 'wb') as f:
668
+ pickle.dump([x, y], f)
669
+ else:
670
+ print("reading cifar10_c_level3")
671
+ with open(path, 'rb') as f:
672
+ x, y = pickle.load(f)
673
+ dataset = TensorDataset(x, y)
674
+ return dataset
675
+ def load_cifar10_c_level4(dataroot):
676
+ path = f'data/cifar10_c_level4.pkl'
677
+ if not os.path.exists(path):
678
+ print("genenrating cifar10_c_level4")
679
+ labels = np.load(os.path.join(dataroot, 'labels.npy'))
680
+ y_single = labels[0:10000]
681
+ x = torch.zeros((190000,3,32,32))
682
+ for j in range(19):
683
+ if j == 0:
684
+ y = y_single
685
+ else:
686
+ y = np.hstack((y,y_single))
687
+ index = 0
688
+ for filename in os.listdir(dataroot):
689
+ if filename=='labels.npy':
690
+ continue
691
+ else:
692
+ imgs = np.load(os.path.join(dataroot,filename))
693
+ imgs = imgs.transpose(0,3,1,2)
694
+ imgs = torch.tensor(imgs)
695
+ imgs = imgs.float()/255.
696
+ print(imgs.shape)
697
+ x[index*10000:(index+1)*10000] = imgs[30000:40000]
698
+ index = index + 1
699
+ y = torch.tensor(y)
700
+ with open(path, 'wb') as f:
701
+ pickle.dump([x, y], f)
702
+ else:
703
+ print("reading cifar10_c_level4")
704
+ with open(path, 'rb') as f:
705
+ x, y = pickle.load(f)
706
+ dataset = TensorDataset(x, y)
707
+ return dataset
708
+ def load_cifar10_c_level5(dataroot):
709
+ path = f'data/cifar10_c_level5.pkl'
710
+ if not os.path.exists(path):
711
+ print("genenrating cifar10_c_level5")
712
+ labels = np.load(os.path.join(dataroot, 'labels.npy'))
713
+ y_single = labels[0:10000]
714
+ x = torch.zeros((190000,3,32,32))
715
+ for j in range(19):
716
+ if j == 0:
717
+ y = y_single
718
+ else:
719
+ y = np.hstack((y,y_single))
720
+ index = 0
721
+ for filename in os.listdir(dataroot):
722
+ if filename=='labels.npy':
723
+ continue
724
+ else:
725
+ imgs = np.load(os.path.join(dataroot,filename))
726
+ imgs = imgs.transpose(0,3,1,2)
727
+ imgs = torch.tensor(imgs)
728
+ imgs = imgs.float()/255.
729
+ print(imgs.shape)
730
+ x[index*10000:(index+1)*10000] = imgs[40000:50000]
731
+ index = index + 1
732
+ y = torch.tensor(y)
733
+ with open(path, 'wb') as f:
734
+ pickle.dump([x, y], f)
735
+ else:
736
+ print("reading cifar10_c_level5")
737
+ with open(path, 'rb') as f:
738
+ x, y = pickle.load(f)
739
+ dataset = TensorDataset(x, y)
740
+ return dataset
741
+ def load_cifar10_c(dataroot):
742
+ y = np.load(os.path.join(dataroot, 'labels.npy'))
743
+ print("y.shape:",y.shape)
744
+ y_single = y[0:10000]
745
+ x1 = torch.zeros((190000,3,32,32))
746
+ x2 = torch.zeros((190000,3,32,32))
747
+ x3 = torch.zeros((190000,3,32,32))
748
+ x4 = torch.zeros((190000,3,32,32))
749
+ x5 = torch.zeros((190000,3,32,32))
750
+ for j in range(19):
751
+ if j == 0:
752
+ y_total = y_single
753
+ else:
754
+ y_total = np.hstack((y_total,y_single))
755
+ print("y_total.shape:",y_total.shape)
756
+ index = 0
757
+ for filename in os.listdir(dataroot):
758
+ if filename=='labels.npy':
759
+ continue
760
+ else:
761
+ x = np.load(os.path.join(dataroot,filename))
762
+ x = x.transpose(0,3,1,2)
763
+ x = torch.tensor(x)
764
+ x = x.float()/255.
765
+ print(x.shape)
766
+ x1[index*10000:(index+1)*10000] = x[0:10000]
767
+ x2[index*10000:(index+1)*10000] = x[10000:20000]
768
+ x3[index*10000:(index+1)*10000] = x[20000:30000]
769
+ x4[index*10000:(index+1)*10000] = x[30000:40000]
770
+ x5[index*10000:(index+1)*10000] = x[40000:50000]
771
+ index = index + 1
772
+ # x1, x2, x3, x4, x5, y_total = torch.tensor(x1), torch.tensor(x2), torch.tensor(x3),\
773
+ # torch.tensor(x4),torch.tensor(x5),torch.tensor(y_total)
774
+ y_total = torch.tensor(y_total)
775
+ dataset1 = TensorDataset(x1, y_total)
776
+ dataset2 = TensorDataset(x2, y_total)
777
+ dataset3 = TensorDataset(x3, y_total)
778
+ dataset4 = TensorDataset(x4, y_total)
779
+ dataset5 = TensorDataset(x5, y_total)
780
+ return dataset1,dataset2,dataset3,dataset4,dataset5
781
+
782
+ def load_cifar10_c_class(dataroot,CORRUPTIONS):
783
+ y = np.load(os.path.join(dataroot, 'labels.npy'))
784
+ y_single = y[0:10000]
785
+ y_single = torch.tensor(y_single)
786
+ print("y.shape:",y.shape)
787
+ x = np.load(os.path.join(dataroot,CORRUPTIONS+'.npy'))
788
+ print("loading data of",os.path.join(dataroot,CORRUPTIONS+'.npy'))
789
+ x = x.transpose(0,3,1,2)
790
+ x = torch.tensor(x)
791
+ x = x.float()/255.
792
+ dataset = []
793
+ for i in range(5):
794
+ x_single = x[i*10000:(i+1)*10000]
795
+ dataset.append(TensorDataset(x_single, y_single))
796
+ return dataset
797
+
798
+ def load_usps(split='train', channels=3):
799
+ path = f'/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/usps-{split}.pkl'
800
+ if not os.path.exists(path):
801
+ dataset = USPS(f'{HOME}/.pytorch/USPS', train=(split=='train'), download=True)
802
+ x, y = dataset.data, dataset.targets
803
+ x = torch.tensor(resize_imgs(x, 32))
804
+ x = (x.float()/255.).unsqueeze(1).repeat(1,3,1,1)
805
+ y = torch.tensor(y)
806
+ with open(path, 'wb') as f:
807
+ pickle.dump([x, y], f)
808
+ with open(path, 'rb') as f:
809
+ x, y = pickle.load(f)
810
+ if channels == 1:
811
+ x = x[:,0:1,:,:]
812
+ dataset = TensorDataset(x, y)
813
+ return dataset
814
+
815
+ def load_svhn(split='train', channels=3):
816
+ dataset = SVHN(f'{HOME}/.pytorch/SVHN', split=split, download=True)
817
+ x, y = dataset.data, dataset.labels
818
+ x = x.astype('float32')/255.
819
+ x, y = torch.tensor(x), torch.tensor(y)
820
+ if channels == 1:
821
+ x = x.mean(1, keepdim=True)
822
+ dataset = TensorDataset(x, y)
823
+ return dataset
824
+
825
+
826
+ def load_syndigit(split='train', channels=3):
827
+ path = f'/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/synth_{split}_32x32.mat'
828
+ data = loadmat(path)
829
+ x, y = data['X'], data['y']
830
+ x = np.transpose(x, [3, 2, 0, 1]).astype('float32')/255.
831
+ y = y.squeeze()
832
+ x, y = torch.tensor(x), torch.tensor(y)
833
+ if channels == 1:
834
+ x = x.mean(1, keepdim=True)
835
+ dataset = TensorDataset(x, y)
836
+ return dataset
837
+
838
+ def load_mnist_m(split='train', channels=3):
839
+ path = f'/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/data/mnist_m-{split}.pkl'
840
+ with open(path, 'rb') as f:
841
+ x, y = pickle.load(f)
842
+ x, y = torch.tensor(x.astype('float32')/255.), torch.tensor(y)
843
+ if channels==1:
844
+ x = x.mean(1, keepdim=True)
845
+ dataset = TensorDataset(x, y)
846
+ return dataset
847
+
848
+ if __name__=='__main__':
849
+ dataset = load_mnist(split='train')
850
+ print('mnist train', len(dataset))
851
+ dataset = load_mnist('test')
852
+ print('mnist test', len(dataset))
853
+ dataset = load_mnist_m('test')
854
+ print('mnsit_m test', len(dataset))
855
+ dataset = load_svhn(split='test')
856
+ print('svhn', len(dataset))
857
+ dataset = load_usps(split='test')
858
+ print('usps', len(dataset))
859
+ dataset = load_syndigit(split='test')
860
+ print('syndigit', len(dataset))
861
+
Meta-causal/code-stage1-pipeline/env.yaml ADDED
@@ -0,0 +1,119 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ name: Py3.7_torch1.8
2
+ channels:
3
+ - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge
4
+ - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
5
+ - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/bioconda/
6
+ - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge/
7
+ - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
8
+ - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
9
+ - conda-forge
10
+ - bioconda
11
+ - defaults
12
+ dependencies:
13
+ - _libgcc_mutex=0.1=main
14
+ - asn1crypto=1.2.0=py37_0
15
+ - blas=1.0=mkl
16
+ - bottleneck=1.3.2=py37heb32a55_1
17
+ - bzip2=1.0.8=h7b6447c_0
18
+ - ca-certificates=2021.10.8=ha878542_0
19
+ - cairo=1.14.12=h8948797_3
20
+ - certifi=2021.10.8=py37h89c1867_1
21
+ - cffi=1.13.0=py37h2e261b9_0
22
+ - chardet=3.0.4=py37_1003
23
+ - click=8.0.3=pyhd3eb1b0_0
24
+ - conda-package-handling=1.6.0=py37h7b6447c_0
25
+ - cryptography=2.8=py37h1ba5d50_0
26
+ - ffmpeg=4.0=hcdf2ecd_0
27
+ - fontconfig=2.13.0=h9420a91_0
28
+ - freeglut=3.0.0=hf484d3e_5
29
+ - freetype=2.11.0=h70c0345_0
30
+ - glib=2.63.1=h5a9c865_0
31
+ - graphite2=1.3.14=h23475e2_0
32
+ - h5py=2.8.0=py37h3010b51_1003
33
+ - harfbuzz=1.8.8=hffaf4a1_0
34
+ - hdf5=1.10.2=hba1933b_1
35
+ - icu=58.2=he6710b0_3
36
+ - idna=2.8=py37_0
37
+ - intel-openmp=2021.3.0=h06a4308_3350
38
+ - jasper=2.0.14=hd8c5072_2
39
+ - jpeg=9d=h7f8727e_0
40
+ - libedit=3.1.20181209=hc058e9b_0
41
+ - libffi=3.2.1=hd88cf55_4
42
+ - libgcc-ng=9.1.0=hdf63c60_0
43
+ - libgfortran-ng=7.5.0=ha8ba4b0_17
44
+ - libgfortran4=7.5.0=ha8ba4b0_17
45
+ - libglu=9.0.0=hf484d3e_1
46
+ - libopencv=3.4.2=hb342d67_1
47
+ - libopus=1.3.1=h7b6447c_0
48
+ - libpng=1.6.37=hbc83047_0
49
+ - libprotobuf=3.17.2=h4ff587b_1
50
+ - libstdcxx-ng=9.1.0=hdf63c60_0
51
+ - libtiff=4.1.0=h2733197_0
52
+ - libuuid=1.0.3=h7f8727e_2
53
+ - libvpx=1.7.0=h439df22_0
54
+ - libxcb=1.14=h7b6447c_0
55
+ - libxml2=2.9.9=hea5a465_1
56
+ - mkl=2021.3.0=h06a4308_520
57
+ - mkl-service=2.4.0=py37h7f8727e_0
58
+ - mkl_fft=1.3.1=py37hd3c417c_0
59
+ - mkl_random=1.2.2=py37h51133e4_0
60
+ - ncurses=6.1=he6710b0_1
61
+ - numexpr=2.7.3=py37h22e1b3c_1
62
+ - numpy-base=1.21.2=py37h79a1101_0
63
+ - opencv=3.4.2=py37h6fd60c2_1
64
+ - openssl=1.1.1h=h516909a_0
65
+ - pandas=1.3.3=py37h8c16a72_0
66
+ - pcre=8.45=h295c915_0
67
+ - pip=19.3.1=py37_0
68
+ - pixman=0.40.0=h7f8727e_1
69
+ - protobuf=3.17.2=py37h295c915_0
70
+ - py-opencv=3.4.2=py37hb342d67_1
71
+ - pycosat=0.6.3=py37h14c3975_0
72
+ - pycparser=2.19=py37_0
73
+ - pyopenssl=19.0.0=py37_0
74
+ - pysocks=1.7.1=py37_0
75
+ - python=3.7.4=h265db76_1
76
+ - python-dateutil=2.8.2=pyhd3eb1b0_0
77
+ - python_abi=3.7=2_cp37m
78
+ - pytz=2021.3=pyhd3eb1b0_0
79
+ - readline=7.0=h7b6447c_5
80
+ - requests=2.22.0=py37_0
81
+ - ruamel_yaml=0.15.46=py37h14c3975_0
82
+ - scipy=1.7.1=py37h292c36d_2
83
+ - setuptools=41.4.0=py37_0
84
+ - six=1.12.0=py37_0
85
+ - sqlite=3.30.0=h7b6447c_0
86
+ - tensorboardx=2.2=pyhd3eb1b0_0
87
+ - tk=8.6.8=hbc83047_0
88
+ - tqdm=4.36.1=py_0
89
+ - urllib3=1.24.2=py37_0
90
+ - wheel=0.33.6=py37_0
91
+ - xz=5.2.4=h14c3975_4
92
+ - yaml=0.1.7=had09818_2
93
+ - zlib=1.2.11=h7b6447c_3
94
+ - zstd=1.3.7=h0b5b093_0
95
+ - pip:
96
+ - absl-py==1.0.0
97
+ - cachetools==4.2.4
98
+ - conda-pack==0.6.0
99
+ - google-auth==2.3.3
100
+ - google-auth-oauthlib==0.4.6
101
+ - grpcio==1.42.0
102
+ - importlib-metadata==4.8.2
103
+ - markdown==3.3.6
104
+ - numpy==1.21.3
105
+ - oauthlib==3.1.1
106
+ - pillow==8.4.0
107
+ - pyasn1==0.4.8
108
+ - pyasn1-modules==0.2.8
109
+ - requests-oauthlib==1.3.0
110
+ - rsa==4.8
111
+ - tensorboard==2.7.0
112
+ - tensorboard-data-server==0.6.1
113
+ - tensorboard-plugin-wit==1.8.0
114
+ - torch==1.8.1+cu111
115
+ - torchvision==0.9.1+cu111
116
+ - typing-extensions==3.10.0.2
117
+ - werkzeug==2.0.2
118
+ - zipp==3.6.0
119
+ prefix: /home/chenjin/miniconda3/envs/Py3.7_torch1.8
Meta-causal/code-stage1-pipeline/main_my_joint_v13_auto.py ADDED
@@ -0,0 +1,279 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ '''
3
+ 训练 base 模型
4
+ '''
5
+
6
+ import torch
7
+ import torch.nn as nn
8
+ import torch.nn.functional as F
9
+ import itertools
10
+ from torch import optim
11
+ from torch.utils.data import DataLoader, RandomSampler
12
+ from torchvision import models
13
+ from torchvision.datasets import CIFAR10
14
+ from torchvision.utils import make_grid
15
+ import torchvision.transforms as transforms
16
+ from tensorboardX import SummaryWriter
17
+ from torch.cuda.amp import autocast,GradScaler
18
+
19
+ import os
20
+ import click
21
+ import time
22
+ import numpy as np
23
+
24
+ from network import mnist_net_my as mnist_net
25
+ from network import wideresnet as wideresnet
26
+ from network import resnet as resnet
27
+ from network import adaptor_v2
28
+
29
+ from tools import causalaugment_v3 as causalaugment
30
+ import data_loader_joint_v3 as data_loader
31
+ # from utils import set_requires_grad
32
+
33
+ HOME = os.environ['HOME']
34
+
35
+ @click.command()
36
+ @click.option('--gpu', type=str, default='0', help='选择gpu')
37
+ @click.option('--data', type=str, default='mnist', help='数据集名称')
38
+ @click.option('--ntr', type=int, default=None, help='选择训练集前ntr个样本')
39
+ @click.option('--translate', type=float, default=None, help='随机平移数据增强')
40
+ @click.option('--autoaug', type=str, default=None, help='AA FastAA RA')
41
+ @click.option('--n', type=int, default=3, help='选择多少个factor生成RA')
42
+ @click.option('--stride', type=int, default=5, help='if autoaug==CA_multiple, stride is used')
43
+ @click.option('--factor_num', type=int, default=16, help='the first n factors')
44
+ @click.option('--epochs', type=int, default=100)
45
+ @click.option('--nbatch', type=int, default=100, help='每个epoch中batch的数量')
46
+ @click.option('--batchsize', type=int, default=128, help='每个batch中样本的数量')
47
+ @click.option('--lr', type=float, default=1e-3)
48
+ @click.option('--lr_scheduler', type=str, default='none', help='是否选择学习率衰减策略')
49
+ @click.option('--svroot', type=str, default='./saved', help='项目文件保存路径')
50
+ @click.option('--clsadapt', type=bool, default=True, help='映射后是否用分类损失')
51
+ @click.option('--lambda_causal', type=float, default=1, help='the weight of reconstruction during mapping and causal ')
52
+ @click.option('--lambda_re', type=float, default=1, help='the weight of reconstruction during mapping and causal ')
53
+ @click.option('--randm', type=bool, default=True, help='m取值是否randm')
54
+ @click.option('--randn', type=bool, default=False, help='原始特征是否detach')
55
+ @click.option('--network', type=str, default='resnet18', help='项目文件保存路径')
56
+
57
+
58
+ def experiment(gpu, data, ntr, translate, autoaug, n, stride, factor_num, epochs, nbatch, batchsize, lr, lr_scheduler, svroot, clsadapt, lambda_causal,lambda_re,randm,randn,network):
59
+ settings = locals().copy()
60
+ print(settings)
61
+
62
+ # 全局设置
63
+ os.environ['CUDA_VISIBLE_DEVICES'] = gpu
64
+ if not os.path.exists(svroot):
65
+ os.makedirs(svroot)
66
+ log_file = open(svroot+os.sep+'log.log',"w")
67
+ log_file.write(str(settings)+'\n')
68
+ writer = SummaryWriter(svroot)
69
+
70
+ # 加载数据集和模型
71
+ if data in ['mnist', 'mnist_t']:
72
+ if data == 'mnist':
73
+ trset = data_loader.load_mnist('train', translate=translate,twox=True, ntr=ntr, factor_num=factor_num,autoaug=autoaug,randm=randm,randn=randn,n=n,stride=stride)
74
+ elif data == 'mnist_t':
75
+ trset = data_loader.load_mnist_t('train', translate=translate, ntr=ntr)
76
+ teset = data_loader.load_mnist('test')
77
+ trloader = DataLoader(trset, batch_size=batchsize, num_workers=0, \
78
+ sampler=RandomSampler(trset, True, nbatch*batchsize))
79
+ teloader = DataLoader(teset, batch_size=batchsize, num_workers=0, shuffle=False)
80
+ cls_net = mnist_net.ConvNet().cuda()
81
+
82
+ parameter_list = []
83
+ parameter_list.append({'params':cls_net.parameters(),'lr':lr})
84
+ opt = optim.Adam(parameter_list, lr=lr)
85
+ if lr_scheduler == 'cosine':
86
+ scheduler = optim.lr_scheduler.CosineAnnealingLR(opt, epochs)
87
+ elif lr_scheduler == 'Exp':
88
+ scheduler = optim.lr_scheduler.ExponentialLR(opt, gamma=0.95)
89
+ elif lr_scheduler == 'Step':
90
+ scheduler = optim.lr_scheduler.StepLR(opt, step_size=int(epochs*0.8))
91
+
92
+ elif data == 'cifar10':
93
+ # 加载数据集
94
+ trset = data_loader.load_cifar10(split='train',twox=True, factor_num=factor_num,autoaug=autoaug,randm=randm,randn=randn,n=n,stride=stride)
95
+ teset = data_loader.load_cifar10(split='test')
96
+ trloader = DataLoader(trset, batch_size=batchsize, num_workers=4, shuffle=True, drop_last=True)
97
+ teloader = DataLoader(teset, batch_size=batchsize, num_workers=4, shuffle=False)
98
+ cls_net = wideresnet.WideResNet(16, 10, 4).cuda()
99
+ # cls_opt = optim.SGD(cls_net.parameters(), lr=lr, momentum=0.9, nesterov=True, weight_decay=5e-4)
100
+ AdaptNet = []
101
+ parameter_list = []
102
+ for i in range(factor_num):
103
+ mapping = adaptor_v2.mapping(256,512,256,4).cuda()
104
+ AdaptNet.append(mapping)
105
+ parameter_list.append({'params':mapping.parameters(),'lr':lr})
106
+ if autoaug == 'CA_multiple':
107
+ var_num = len(list(range(0, 31, stride)))
108
+ E_to_W = adaptor_v2.effect_to_weight(10,100,1).cuda()
109
+ else:
110
+ E_to_W = adaptor_v2.effect_to_weight(10,100,1).cuda()
111
+ parameter_list.append({'params':cls_net.parameters(),'lr':lr})
112
+ parameter_list.append({'params':E_to_W.parameters(),'lr':lr})
113
+ #print("---------------------------------------------------------------------------------------")
114
+ # opt = optim.Adam(parameter_list)
115
+ opt = optim.SGD(parameter_list, lr=lr, momentum=0.9, nesterov=True, weight_decay=5e-4)
116
+ if lr_scheduler == 'cosine':
117
+ scheduler = optim.lr_scheduler.CosineAnnealingLR(opt, epochs)
118
+ elif lr_scheduler == 'Exp':
119
+ scheduler = optim.lr_scheduler.ExponentialLR(opt, gamma=0.95)
120
+ elif lr_scheduler == 'Step':
121
+ scheduler = optim.lr_scheduler.StepLR(opt, step_size=int(epochs*0.8))
122
+
123
+
124
+ elif data in ['art_painting', 'cartoon', 'photo', 'sketch']:
125
+ # 加载数据集
126
+ trset = data_loader.load_pacs(domain=data, split='train', twox=True, factor_num=factor_num,autoaug=autoaug,randm=randm,randn=randn,n=n,stride=stride)
127
+ teset = data_loader.load_pacs(domain=data, split='val')
128
+ trloader = DataLoader(trset, batch_size=batchsize, num_workers=4, shuffle=True, drop_last=True)
129
+ teloader = DataLoader(teset, batch_size=batchsize, num_workers=4, shuffle=False)
130
+ if network == 'resnet18':
131
+ cls_net = resnet.resnet18(classes=7,c_dim=2048).cuda()
132
+
133
+ classifier_param = list(map(id, cls_net.class_classifier.parameters()))
134
+ backbone_param = filter(lambda p: id(p) not in classifier_param and p.requires_grad, cls_net.parameters())
135
+
136
+ parameter_list = []
137
+ parameter_list.append({'params':backbone_param,'lr':0.01*lr})
138
+ parameter_list.append({'params':cls_net.class_classifier.parameters(),'lr':lr})
139
+
140
+ opt = optim.SGD(parameter_list, momentum=0.9, nesterov=True, weight_decay=5e-4)
141
+ if lr_scheduler == 'cosine':
142
+ scheduler = optim.lr_scheduler.CosineAnnealingLR(opt, epochs)
143
+ elif lr_scheduler == 'Exp':
144
+ scheduler = optim.lr_scheduler.ExponentialLR(opt, gamma=0.99999)
145
+ elif lr_scheduler == 'Step':
146
+ scheduler = optim.lr_scheduler.StepLR(opt, step_size=15)
147
+
148
+ cls_criterion = nn.CrossEntropyLoss()
149
+
150
+ # 开始训练
151
+ best_acc = 0
152
+ best_acc_t = 0
153
+ scaler = GradScaler()
154
+ for epoch in range(epochs):
155
+ t1 = time.time()
156
+ loss_list = []
157
+ cls_net.train()
158
+ print(len(trloader))
159
+ for i, (x_four,y) in enumerate(trloader):
160
+ x, x_RA, x_FA, x_CA, y = x_four[0].cuda(), x_four[1].cuda(), x_four[2].cuda(), x_four[3].cuda(), y.cuda()
161
+ #print('x:', x.shape, 'x_RA:', x_RA.shape, 'x_FA:', x_FA.shape, 'x_CA:', x_CA.shape, 'y:', y.shape)
162
+ b, c, h, w = x.shape
163
+ with autocast():
164
+ p,f = cls_net(x)
165
+ #print('p:', p.size(), 'f:', f.size())
166
+
167
+ cls_loss = cls_criterion(p, y)
168
+ #print('cls_loss:', cls_loss)
169
+
170
+ loss = cls_loss
171
+
172
+ opt.zero_grad()
173
+ scaler.scale(loss).backward()
174
+ scaler.step(opt)
175
+ scaler.update()
176
+ #loss_list.append([cls_loss.item(), cls_loss_mapping.item(),cls_loss_causal.item(), re_mapping.item(), re_causal.item()])
177
+ loss_list.append(cls_loss.item())
178
+
179
+ # 调整学习率
180
+ if lr_scheduler in ['cosine', 'Exp', 'Step']:
181
+ writer.add_scalar('scalar/lr', opt.param_groups[0]["lr"], epoch)
182
+ print(opt.param_groups[0]["lr"])
183
+ print("changing lr")
184
+ scheduler.step()
185
+ #cls_loss, cls_loss_mapping, cls_loss_causal, re_mapping, re_causal = np.mean(loss_list, 0)
186
+ cls_loss = np.mean(loss_list)
187
+
188
+ # 测试，并保存最优模型
189
+ cls_net.eval()
190
+ if data in ['mnist', 'mnist_t', 'cifar10', 'mnistvis', 'art_painting', 'cartoon', 'photo', 'sketch']:
191
+ teacc = evaluate(cls_net, teloader)
192
+
193
+ if best_acc < teacc:
194
+ print(f'---------------------saving model at epoch {epoch}----------------------------------------------------')
195
+ log_file.write(f'saving model at epoch {epoch}\n')
196
+
197
+ best_acc = teacc
198
+ torch.save(cls_net.state_dict(),os.path.join(svroot, 'best_cls_net.pkl'))
199
+
200
+ if ((epoch+1)%5==0):
201
+ torch.save(cls_net.state_dict(),os.path.join(svroot, f'epoch{epoch}_cls_net.pkl'))
202
+
203
+ # 保存日志
204
+ t2 = time.time()
205
+ #print(f'epoch {epoch}, time {t2-t1:.2f}, cls_loss {cls_loss:.4f} cls_loss_mapping {cls_loss_mapping:.4f} cls_loss_causal {cls_loss_causal:.4f} re_mapping {re_mapping:.4f} re_causal {re_causal:.4f} /// teacc {teacc:2.2f} lr {opt.param_groups[0]["lr"]:.8f}')
206
+ print(f'epoch {epoch}, time {t2-t1:.2f}, cls_loss {cls_loss:.4f}')
207
+
208
+ #log_file.write(f'epoch {epoch}, time {t2-t1:.2f}, cls_loss {cls_loss:.4f} cls_loss_mapping {cls_loss_mapping:.4f} cls_loss_causal {cls_loss_causal:.4f} re_mapping {re_mapping:.4f} re_causal {re_causal:.4f} /// teacc {teacc:2.2f} lr {opt.param_groups[0]["lr"]:.8f} \n')
209
+ log_file.write(f'epoch {epoch}, time {t2-t1:.2f}, cls_loss {cls_loss:.4f}')
210
+
211
+ writer.add_scalar('scalar/cls_loss', cls_loss, epoch)
212
+ #writer.add_scalar('scalar/cls_loss_mapping', cls_loss_mapping, epoch)
213
+ #writer.add_scalar('scalar/cls_loss_causal', cls_loss_causal, epoch)
214
+ #writer.add_scalar('scalar/re_mapping', re_mapping, epoch)
215
+ #writer.add_scalar('scalar/re_causal', re_causal, epoch)
216
+ writer.add_scalar('scalar/teacc', teacc, epoch)
217
+
218
+ print(f'---------------------saving last model at epoch {epoch}----------------------------------------------------')
219
+ log_file.write(f'saving last model at epoch {epoch}\n')
220
+ torch.save(cls_net.state_dict(),os.path.join(svroot, 'last_cls_net.pkl'))
221
+ writer.close()
222
+
223
+
224
+ def evalute_pacs(source_domain,cls_net):
225
+ cls_net.eval()
226
+ data_total = ['art_painting', 'cartoon', 'photo', 'sketch']
227
+ target = [i for i in data_total if i!=source_domain]
228
+ acc = np.zeros(len(target))
229
+ for idx, data in enumerate(target):
230
+ teset = data_loader.load_pacs(data, 'test')
231
+ teloader = DataLoader(teset, batch_size=6, num_workers=0)
232
+ # 计算评价指标
233
+ acc[idx] = evaluate(cls_net, teloader)
234
+ acc_avg = sum(acc)/len(target)
235
+ return acc_avg,acc
236
+
237
+ def evaluate(net, teloader):
238
+ ps = []
239
+ ys = []
240
+ for i,(x1, y1) in enumerate(teloader):
241
+ with torch.no_grad():
242
+ x1 = x1.cuda()
243
+ p1,_ = net(x1, mode='fc')
244
+ p1 = p1.argmax(dim=1)
245
+ ps.append(p1.detach().cpu().numpy())
246
+ ys.append(y1.numpy())
247
+ # 计算评价指标
248
+ ps = np.concatenate(ps)
249
+ ys = np.concatenate(ys)
250
+ acc = np.mean(ys==ps)*100
251
+ return acc
252
+
253
+ def extract_feature(net, teloader, savedir):
254
+ ps = []
255
+ ys = []
256
+ for i,(x1, y1) in enumerate(teloader):
257
+ img_class = y1[0].cpu().numpy()
258
+ save_path = os.path.join(savedir,str(img_class))
259
+ if not os.path.exists(save_path):
260
+ os.makedirs(save_path)
261
+
262
+ with torch.no_grad():
263
+ x1 = x1.cuda()
264
+ p1,f1 = net(x1, mode='fc')
265
+ save_name = save_path+os.sep+str(i)+'.npy'
266
+ np.save(save_name,f1.cpu())
267
+ p1 = p1.argmax(dim=1)
268
+ ps.append(p1.detach().cpu().numpy())
269
+ ys.append(y1.numpy())
270
+ # 计算评价指标
271
+ ps = np.concatenate(ps)
272
+ ys = np.concatenate(ys)
273
+ acc = np.mean(ys==ps)*100
274
+ return acc
275
+
276
+
277
+
278
+ if __name__=='__main__':
279
+ experiment()
Meta-causal/code-stage1-pipeline/main_test_digit_v13.py ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+ from torch.utils.data import DataLoader
6
+
7
+ import os
8
+ import numpy as np
9
+ import click
10
+ import pandas as pd
11
+
12
+ from network import mnist_net_my as mnist_net
13
+ from network import adaptor_v2
14
+ from tools import causalaugment_v3 as causalaugment
15
+ from main_my_joint_v13_auto import evaluate
16
+ import data_loader_joint_v3 as data_loader
17
+
18
+ @click.command()
19
+ @click.option('--gpu', type=str, default='0', help='选择GPU编号')
20
+ @click.option('--svroot', type=str, default='./saved')
21
+ @click.option('--svpath', type=str, default=None, help='保存日志的路径')
22
+ @click.option('--channels', type=int, default=3)
23
+ @click.option('--factor_num', type=int, default=16)
24
+ @click.option('--stride', type=int, default=16)
25
+ @click.option('--epoch', type=str, default='best')
26
+ @click.option('--eval_mapping', type=bool, default=True, help='是否查看mapping学习效果')
27
+ def main(gpu, svroot, svpath, channels, factor_num,stride, epoch, eval_mapping):
28
+ evaluate_digit(gpu, svroot, svpath, channels, factor_num, stride,epoch, eval_mapping)
29
+
30
+ def evaluate_digit(gpu, svroot, svpath, channels=3, factor_num=16,stride=5,epoch='best', eval_mapping=True):
31
+ settings = locals().copy()
32
+ print(settings)
33
+ os.environ['CUDA_VISIBLE_DEVICES'] = gpu
34
+
35
+ # 加载分类模型
36
+ if channels == 3:
37
+ cls_net = mnist_net.ConvNet().cuda()
38
+ elif channels == 1:
39
+ cls_net = mnist_net.ConvNet(imdim=channels).cuda()
40
+ if epoch == 'best':
41
+ print("loading weight of %s"%(epoch))
42
+ saved_weight = torch.load(os.path.join(svroot, 'best_cls_net.pkl'))
43
+ elif epoch == 'last':
44
+ print("loading weight of %s"%(epoch))
45
+ saved_weight = torch.load(os.path.join(svroot, 'last_cls_net.pkl'))
46
+ cls_net.load_state_dict(saved_weight)
47
+ cls_net.eval()
48
+
49
+ # 测试
50
+ str2fun = {
51
+ 'mnist': data_loader.load_mnist,
52
+ 'mnist_m': data_loader.load_mnist_m,
53
+ 'usps': data_loader.load_usps,
54
+ 'svhn': data_loader.load_svhn,
55
+ 'syndigit': data_loader.load_syndigit,
56
+ }
57
+
58
+ columns = ['mnist', 'svhn', 'mnist_m', 'syndigit','usps']
59
+ target = ['svhn', 'mnist_m', 'syndigit','usps']
60
+
61
+ index = ['w/o do (original x)']
62
+ data_result = {}
63
+
64
+ for idx, data in enumerate(columns):
65
+ teset = str2fun[data]('test', channels=channels)
66
+ teloader = DataLoader(teset, batch_size=8, num_workers=0)
67
+ # 计算评价指标
68
+ teacc = evaluate(cls_net, teloader)
69
+ if data == 'mnist':
70
+ acc_avg = np.zeros(teacc.shape)
71
+ else:
72
+ acc_avg = acc_avg + teacc
73
+ data_result[data] = teacc
74
+ acc_avg = acc_avg/float(len(target))
75
+
76
+ data_result['Avg'] = acc_avg
77
+
78
+ df = pd.DataFrame(data_result,index = index)
79
+ print(df)
80
+ if svpath is not None:
81
+ df.to_csv(svpath)
82
+
83
+ if __name__=='__main__':
84
+ main()
85
+
Meta-causal/code-stage1-pipeline/main_test_pacs_v13.py ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+ from torch.utils.data import DataLoader
6
+
7
+ import os
8
+ import numpy as np
9
+ import click
10
+ import pandas as pd
11
+
12
+ from network import resnet as resnet
13
+ from network import adaptor_v2
14
+ from tools import causalaugment_v3 as causalaugment
15
+ from main_my_joint_v13_auto import evaluate
16
+ import data_loader_joint_v3 as data_loader
17
+
18
+ @click.command()
19
+ @click.option('--gpu', type=str, default='0', help='选择GPU编号')
20
+ @click.option('--svroot', type=str, default='./saved')
21
+ @click.option('--source_domain', type=str, default='art_painting', help='source domain')
22
+ @click.option('--svpath', type=str, default=None, help='保存日志的路径')
23
+ @click.option('--factor_num', type=int, default=16)
24
+ @click.option('--epoch', type=str, default='best')
25
+ @click.option('--stride', type=int, default=5)
26
+ @click.option('--eval_mapping', type=bool, default=False, help='是否查看mapping学习效果')
27
+ @click.option('--network', type=str, default='resnet18', help='项目文件保存路径')
28
+ def main(gpu, svroot, source_domain, svpath, factor_num, epoch, stride,eval_mapping, network):
29
+ evaluate_pacs(gpu, svroot, source_domain, svpath, factor_num, epoch, stride,eval_mapping, network)
30
+
31
+ def evaluate_pacs(gpu, svroot, source_domain, svpath, factor_num=16, epoch='best', stride=5,eval_mapping=False, network='resnet18'):
32
+ settings = locals().copy()
33
+ print(settings)
34
+ os.environ['CUDA_VISIBLE_DEVICES'] = gpu
35
+
36
+ # 加载分类模型
37
+ if network == 'resnet18':
38
+ cls_net = resnet.resnet18(classes=7,c_dim=2048).cuda()
39
+ input_dim = 2048
40
+ if epoch == 'best':
41
+ print("loading weight of %s"%(epoch))
42
+ saved_weight = torch.load(os.path.join(svroot, 'best_cls_net.pkl'))
43
+ elif epoch == 'last':
44
+ print("loading weight of %s"%(epoch))
45
+ saved_weight = torch.load(os.path.join(svroot, 'last_cls_net.pkl'))
46
+
47
+ cls_net.load_state_dict(saved_weight)
48
+ cls_net.eval()
49
+
50
+
51
+ columns = ['art_painting', 'cartoon', 'photo', 'sketch']
52
+ target = [i for i in columns if i!=source_domain]
53
+ columns = [source_domain] + target
54
+ print("columns:",columns)
55
+
56
+
57
+ index = ['w/o do (original x)']
58
+
59
+ data_result = {}
60
+ data_result_ours = {}
61
+
62
+ for idx, data in enumerate(columns):
63
+ teset = data_loader.load_pacs(data, 'test')
64
+ teloader = DataLoader(teset, batch_size=4, num_workers=0)
65
+ # 计算评价指标
66
+ acc = evaluate(cls_net, teloader)
67
+ data_result_ours[data] = acc
68
+
69
+ teacc = evaluate(cls_net, teloader)
70
+ if data == source_domain:
71
+ acc_avg = np.zeros(teacc.shape)
72
+ else:
73
+ acc_avg = acc_avg + teacc
74
+ data_result[data] = teacc
75
+ acc_avg = acc_avg/float(len(target))
76
+
77
+ data_result['Avg'] = acc_avg
78
+
79
+ df = pd.DataFrame(data_result,index = index)
80
+ print(df)
81
+
82
+ if svpath is not None:
83
+ df.to_csv(svpath)
84
+
85
+ if __name__=='__main__':
86
+ main()
87
+
88
+
89
+
Meta-causal/code-stage1-pipeline/network/adaptor_v2.py ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+ import numpy as np
6
+
7
+ class mapping(nn.Module):
8
+ def __init__(self, input_dim=1024, hidden_dim = 512, out_dim=1024, layernum=4):
9
+ '''
10
+ '''
11
+ super().__init__()
12
+ self.layernum = layernum
13
+ if layernum == 4:
14
+ self.fc1 = nn.Linear(input_dim, hidden_dim)
15
+ self.fc2 = nn.Linear(hidden_dim, hidden_dim)
16
+ self.fc3 = nn.Linear(hidden_dim, hidden_dim)
17
+ self.fc4 = nn.Linear(hidden_dim, out_dim)
18
+ elif layernum == 2:
19
+ self.fc1 = nn.Linear(input_dim, hidden_dim)
20
+ self.fc2 = nn.Linear(hidden_dim, out_dim)
21
+ self.relu = nn.ReLU(inplace=True)
22
+
23
+ def forward(self, x):
24
+ ''' x '''
25
+ if self.layernum == 4:
26
+ x = self.relu(self.fc1(x))
27
+ x = self.relu(self.fc2(x))
28
+ x = self.relu(self.fc3(x))
29
+ x = self.fc4(x)
30
+ elif self.layernum == 2:
31
+ x = self.relu(self.fc1(x))
32
+ x = self.fc2(x)
33
+ return x
34
+
35
+
36
+ class effect_to_weight(nn.Module):
37
+ def __init__(self, input_dim = 512, hidden_dim = 256, out_dim = 1, layernum=2, hidden_dim2 = 128):
38
+ '''
39
+ '''
40
+ super().__init__()
41
+
42
+ self.layernum = layernum
43
+ if layernum == 2:
44
+ self.fc1 = nn.Linear(input_dim, hidden_dim)
45
+ self.fc2 = nn.Linear(hidden_dim, out_dim)
46
+ elif layernum == 3:
47
+ self.fc1 = nn.Linear(input_dim, hidden_dim)
48
+ self.fc2 = nn.Linear(hidden_dim, hidden_dim2)
49
+ self.fc3 = nn.Linear(hidden_dim2, out_dim)
50
+ self.relu = nn.ReLU(inplace=True)
51
+
52
+ def forward(self, x):
53
+ ''' x '''
54
+ if self.layernum == 2:
55
+ x = self.relu(self.fc1(x))
56
+ x = self.fc2(x)
57
+ else:
58
+ x = self.relu(self.fc1(x))
59
+ x = self.relu(self.fc2(x))
60
+ x = self.fc3(x)
61
+ return x
62
+
63
+
Meta-causal/code-stage1-pipeline/network/mnist_net_my.py ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+
6
+ class ConvNet(nn.Module):
7
+ ''' 网络结构和cvpr2020的 M-ADA 方法一致 '''
8
+ def __init__(self, imdim=3):
9
+ super(ConvNet, self).__init__()
10
+
11
+ self.conv1 = nn.Conv2d(imdim, 64, kernel_size=5, stride=1, padding=0)
12
+ self.mp = nn.MaxPool2d(2)
13
+ self.relu1 = nn.ReLU(inplace=True)
14
+ self.conv2 = nn.Conv2d(64, 128, kernel_size=5, stride=1, padding=0)
15
+ self.relu2 = nn.ReLU(inplace=True)
16
+ self.fc1 = nn.Linear(128*5*5, 1024)
17
+ self.relu3 = nn.ReLU(inplace=True)
18
+ self.fc2 = nn.Linear(1024, 1024)
19
+ self.relu4 = nn.ReLU(inplace=True)
20
+
21
+ self.cls_head_src = nn.Linear(1024, 10)
22
+ # self.cls_head_tgt = nn.Linear(1024, 10)
23
+ # self.pro_head = nn.Linear(1024, 128)
24
+
25
+ def forward(self, x, mode='fc'):
26
+ if mode == 'c':
27
+ out4 = self.relu4(x)
28
+ p = self.cls_head_src(out4)
29
+ return p
30
+ elif mode == 'fc':
31
+ in_size = x.size(0)
32
+ out1 = self.mp(self.relu1(self.conv1(x)))
33
+ out2 = self.mp(self.relu2(self.conv2(out1)))
34
+ out2 = out2.view(in_size, -1)
35
+ out3 = self.relu3(self.fc1(out2))
36
+ out4_worelu = self.fc2(out3)
37
+ out4 = self.relu4(out4_worelu)
38
+ p = self.cls_head_src(out4)
39
+ return p, out4_worelu
40
+
41
+ # if mode == 'test':
42
+ # p = self.cls_head_src(out4)
43
+ # return p
44
+ # elif mode == 'train':
45
+ # p = self.cls_head_src(out4)
46
+ # # z = self.pro_head(out4)
47
+ # # z = F.normalize(z)
48
+ # return p,out4_worelu
49
+ # elif mode == 'p_f':
50
+ # p = self.cls_head_src(out4)
51
+ # return p, out4
52
+ #elif mode == 'target':
53
+ # p = self.cls_head_tgt(out4)
54
+ # z = self.pro_head(out4)
55
+ # z = F.normalize(z)
56
+ # return p,z
57
+
58
+ class ConvNetVis(nn.Module):
59
+ ''' 方便可视化，特征提取器输出2-d特征
60
+ '''
61
+ def __init__(self, imdim=3):
62
+ super(ConvNetVis, self).__init__()
63
+
64
+ self.conv1 = nn.Conv2d(imdim, 64, kernel_size=5, stride=1, padding=0)
65
+ self.mp = nn.MaxPool2d(2)
66
+ self.relu1 = nn.ReLU(inplace=True)
67
+ self.conv2 = nn.Conv2d(64, 128, kernel_size=5, stride=1, padding=0)
68
+ self.relu2 = nn.ReLU(inplace=True)
69
+ self.fc1 = nn.Linear(128*5*5, 1024)
70
+ self.relu3 = nn.ReLU(inplace=True)
71
+ self.fc2 = nn.Linear(1024, 2)
72
+ self.relu4 = nn.ReLU(inplace=True)
73
+
74
+ self.cls_head_src = nn.Linear(2, 10)
75
+ self.cls_head_tgt = nn.Linear(2, 10)
76
+ self.pro_head = nn.Linear(2, 128)
77
+
78
+ def forward(self, x, mode='test'):
79
+
80
+ in_size = x.size(0)
81
+ out1 = self.mp(self.relu1(self.conv1(x)))
82
+ out2 = self.mp(self.relu2(self.conv2(out1)))
83
+ out2 = out2.view(in_size, -1)
84
+ out3 = self.relu3(self.fc1(out2))
85
+ out4 = self.relu4(self.fc2(out3))
86
+
87
+ if mode == 'test':
88
+ p = self.cls_head_src(out4)
89
+ return p
90
+ elif mode == 'train':
91
+ p = self.cls_head_src(out4)
92
+ z = self.pro_head(out4)
93
+ z = F.normalize(z)
94
+ return p,z
95
+ elif mode == 'p_f':
96
+ p = self.cls_head_src(out4)
97
+ return p, out4
98
+ #elif mode == 'target':
99
+ # p = self.cls_head_tgt(out4)
100
+ # z = self.pro_head(out4)
101
+ # z = F.normalize(z)
102
+ # return p,z
103
+
104
+
Meta-causal/code-stage1-pipeline/network/resnet.py ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from torch import nn
2
+ from torch.utils import model_zoo
3
+ #from torchvision.models.resnet import BasicBlock, model_urls, Bottleneck
4
+ from torchvision.models.resnet import BasicBlock, Bottleneck
5
+
6
+ import torch
7
+ import ssl
8
+ # from torch import nn as nn
9
+ # from utils.util import *
10
+
11
+ ssl._create_default_https_context = ssl._create_unverified_context
12
+
13
+ all = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101','resnet152']
14
+
15
+ model_urls = {
16
+ 'resnet18': 'https://download.pytorch.org/models/resnet18-5c106cde.pth',
17
+ 'resnet34': 'https://download.pytorch.org/models/resnet34-333f7ec4.pth',
18
+ 'resnet50': 'https://download.pytorch.org/models/resnet50-19c8e357.pth',
19
+ 'resnet101': 'https://download.pytorch.org/models/resnet101-5d3b4d8f.pth',
20
+ 'resnet152': 'https://download.pytorch.org/models/resnet152-b121ed2d.pth',
21
+ }
22
+
23
+
24
+ class ResNet(nn.Module):
25
+ def __init__(self, block, layers,classes=7,c_dim=512):
26
+ self.inplanes = 64
27
+ super(ResNet, self).__init__()
28
+ self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3,
29
+ bias=False)
30
+ self.bn1 = nn.BatchNorm2d(64)
31
+ self.relu = nn.ReLU(inplace=True)
32
+ self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
33
+ self.layer1 = self._make_layer(block, 64, layers[0])
34
+ self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
35
+ self.layer3 = self._make_layer(block, 256, layers[2], stride=2)
36
+ self.layer4 = self._make_layer(block, 512, layers[3], stride=2)
37
+ self.avgpool = nn.AvgPool2d(7, stride=1)
38
+ self.class_classifier = nn.Linear(c_dim, classes)
39
+ for m in self.modules():
40
+ if isinstance(m, nn.Conv2d):
41
+ nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
42
+ elif isinstance(m, nn.BatchNorm2d):
43
+ nn.init.constant_(m.weight, 1)
44
+ nn.init.constant_(m.bias, 0)
45
+
46
+ def _make_layer(self, block, planes, blocks, stride=1):
47
+ downsample = None
48
+ if stride != 1 or self.inplanes != planes * block.expansion:
49
+ downsample = nn.Sequential(
50
+ nn.Conv2d(self.inplanes, planes * block.expansion,
51
+ kernel_size=1, stride=stride, bias=False),
52
+ nn.BatchNorm2d(planes * block.expansion),
53
+ )
54
+
55
+ layers = []
56
+ layers.append(block(self.inplanes, planes, stride, downsample))
57
+ self.inplanes = planes * block.expansion
58
+ for i in range(1, blocks):
59
+ layers.append(block(self.inplanes, planes))
60
+
61
+ return nn.Sequential(*layers)
62
+ def forward(self, x, mode='fc'):
63
+ if mode == 'c':
64
+ return self.class_classifier(x)
65
+ else:
66
+ x = self.conv1(x)
67
+ x = self.bn1(x)
68
+ x = self.relu(x)
69
+ x = self.maxpool(x)
70
+
71
+ x = self.layer1(x)
72
+ x = self.layer2(x)
73
+ x = self.layer3(x)
74
+ x = self.layer4(x)
75
+ x = self.avgpool(x)
76
+ x = x.view(x.size(0), -1)
77
+ # print("x.shape:",x.shape)
78
+ return self.class_classifier(x), x
79
+
80
+
81
+ def resnet18(pretrained=True, **kwargs):
82
+ """Constructs a ResNet-18 model.
83
+ Args:
84
+ pretrained (bool): If True, returns a model pre-trained on ImageNet
85
+ """
86
+ model = ResNet(BasicBlock, [2, 2, 2, 2], **kwargs)
87
+ if pretrained:
88
+ print("-------------------------------------loading pretrain weights----------------------------------")
89
+ model.load_state_dict(model_zoo.load_url(model_urls['resnet18']), strict=False)
90
+ return model
91
+
92
+ def resnet50(pretrained=True, **kwargs):
93
+ """Constructs a ResNet-50 model.
94
+ Args:
95
+ pretrained (bool): If True, returns a model pre-trained on ImageNet
96
+ """
97
+ model = ResNet(Bottleneck, [3, 4, 6, 3], **kwargs)
98
+ if pretrained:
99
+ print("-------------------------------------loading pretrain weights----------------------------------")
100
+ model.load_state_dict(model_zoo.load_url(model_urls['resnet50']), strict=False)
101
+ return model
Meta-causal/code-stage1-pipeline/network/wideresnet.py ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import math
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+
6
+
7
+ class BasicBlock(nn.Module):
8
+ def __init__(self, in_planes, out_planes, stride, dropRate=0.0):
9
+ super(BasicBlock, self).__init__()
10
+ self.bn1 = nn.BatchNorm2d(in_planes)
11
+ self.relu1 = nn.ReLU(inplace=True)
12
+ self.conv1 = nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
13
+ padding=1, bias=False)
14
+ self.bn2 = nn.BatchNorm2d(out_planes)
15
+ self.relu2 = nn.ReLU(inplace=True)
16
+ self.conv2 = nn.Conv2d(out_planes, out_planes, kernel_size=3, stride=1,
17
+ padding=1, bias=False)
18
+ self.droprate = dropRate
19
+ self.equalInOut = (in_planes == out_planes)
20
+ self.convShortcut = (not self.equalInOut) and nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride,
21
+ padding=0, bias=False) or None
22
+ def forward(self, x):
23
+ if not self.equalInOut:
24
+ x = self.relu1(self.bn1(x))
25
+ else:
26
+ out = self.relu1(self.bn1(x))
27
+ out = self.relu2(self.bn2(self.conv1(out if self.equalInOut else x)))
28
+ if self.droprate > 0:
29
+ out = F.dropout(out, p=self.droprate, training=self.training)
30
+ out = self.conv2(out)
31
+ return torch.add(x if self.equalInOut else self.convShortcut(x), out)
32
+
33
+ class NetworkBlock(nn.Module):
34
+ def __init__(self, nb_layers, in_planes, out_planes, block, stride, dropRate=0.0):
35
+ super(NetworkBlock, self).__init__()
36
+ self.layer = self._make_layer(block, in_planes, out_planes, nb_layers, stride, dropRate)
37
+ def _make_layer(self, block, in_planes, out_planes, nb_layers, stride, dropRate):
38
+ layers = []
39
+ for i in range(int(nb_layers)):
40
+ layers.append(block(i == 0 and in_planes or out_planes, out_planes, i == 0 and stride or 1, dropRate))
41
+ return nn.Sequential(*layers)
42
+ def forward(self, x):
43
+ return self.layer(x)
44
+
45
+ class WideResNet(nn.Module):
46
+ def __init__(self, depth, num_classes, widen_factor=1, dropRate=0.0):
47
+ super(WideResNet, self).__init__()
48
+ nChannels = [16, 16*widen_factor, 32*widen_factor, 64*widen_factor]
49
+ assert((depth - 4) % 6 == 0)
50
+ n = (depth - 4) / 6
51
+ block = BasicBlock
52
+ # 1st conv before any network block
53
+ self.conv1 = nn.Conv2d(3, nChannels[0], kernel_size=3, stride=1,
54
+ padding=1, bias=False)
55
+ # 1st block
56
+ self.block1 = NetworkBlock(n, nChannels[0], nChannels[1], block, 1, dropRate)
57
+ # 2nd block
58
+ self.block2 = NetworkBlock(n, nChannels[1], nChannels[2], block, 2, dropRate)
59
+ # 3rd block
60
+ self.block3 = NetworkBlock(n, nChannels[2], nChannels[3], block, 2, dropRate)
61
+ # global average pooling and classifier
62
+ self.bn1 = nn.BatchNorm2d(nChannels[3])
63
+ self.relu = nn.ReLU(inplace=True)
64
+ self.fc = nn.Linear(nChannels[3], num_classes)
65
+ self.nChannels = nChannels[3]
66
+
67
+ for m in self.modules():
68
+ if isinstance(m, nn.Conv2d):
69
+ nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
70
+ elif isinstance(m, nn.BatchNorm2d):
71
+ m.weight.data.fill_(1)
72
+ m.bias.data.zero_()
73
+ elif isinstance(m, nn.Linear):
74
+ m.bias.data.zero_()
75
+ def forward(self, x, mode='fc'):
76
+ if mode == 'c':
77
+ return self.fc(x)
78
+ else:
79
+ out = self.conv1(x)
80
+ out = self.block1(out)
81
+ out = self.block2(out)
82
+ out = self.block3(out)
83
+ out = self.relu(self.bn1(out))
84
+ out = F.avg_pool2d(out, 8)
85
+ out = out.view(-1, self.nChannels)
86
+ return self.fc(out), out
Meta-causal/code-stage1-pipeline/run_PACS/run_my_joint_v13_test.sh ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ # $1 gpuid
3
+ # $2 runid
4
+
5
+ # base方法
6
+ cd ..
7
+ epochs=30
8
+ clsadapt=True
9
+ lr=0.01
10
+ factor_num=16
11
+ lr_scheduler=cosine
12
+ lambda_causal=1
13
+ lambda_re=1
14
+ batchsize=6
15
+ stride=5
16
+ randm=True
17
+ randn=True
18
+ autoaug=CA_multiple
19
+ network=resnet18
20
+ UniqueExpName=pipelineAugWoNorm
21
+
22
+ root=/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-PACS/
23
+ data=art_painting
24
+ svroot=$root/${data}/${autoaug}_${factor_num}fa_v2_ep${epochs}_lr${lr}_${lr_scheduler}_base0.01_bs${batchsize}_lamCa_${lambda_causal}_lamRe${lambda_re}_adt4_cls1_EW2_70_rm${randm}_rn${randn}_str${stride}_${UniqueExpName}
25
+ #python3 main_my_joint_v13_auto.py --gpu $1 --data ${data} --epochs $epochs --autoaug $autoaug --lambda_causal ${lambda_causal} --lambda_re ${lambda_re} --lr $lr --svroot $svroot --clsadapt $clsadapt --factor_num $factor_num --lr_scheduler ${lr_scheduler} --batchsize ${batchsize} --network ${network} --randm ${randm} --randn ${randn} --stride ${stride}
26
+
27
+ test_epoch=best
28
+ #python3 main_test_pacs_v13.py --gpu $1 --source_domain $data --svroot $svroot --svpath $svroot/${data}_${factor_num}factor_${test_epoch}_test_check.csv --factor_num $factor_num --epoch $test_epoch --network ${network} --stride ${stride}
29
+
30
+ python3 AllEpochs_test_pacs_v13.py --gpu $1 --source_domain $data --svroot $svroot --svpath $svroot/${data}_${factor_num}factor_${test_epoch}_test_check.csv --factor_num $factor_num --epoch $test_epoch --network ${network} --stride ${stride}
31
+
32
+
33
+
34
+
35
+
Meta-causal/code-stage1-pipeline/run_digits/run_my_joint_test.sh ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ # $1 gpuid
3
+
4
+ cd ..
5
+ epochs=100
6
+ clsadapt=True
7
+ lr=1e-4
8
+ lr_scheduler=Step
9
+ factor_num=14
10
+ test_epoch=best
11
+ lambda_causal=1
12
+ lambda_re=1
13
+ batchsize=32
14
+ stride=3
15
+ randm=True
16
+ randn=True
17
+ autoaug=CA_multiple
18
+ UniqueExpName='pipelineAugWoNorm'
19
+
20
+
21
+ root=/data/work-gcp-europe-west4-a/yuqian_fu/datasets/SingleSourceDG/saved-digit
22
+ svroot=$root/${autoaug}_${factor_num}fa_all_ep${epochs}_lr${lr}_lr_scheduler${lr_scheduler}0.8_bs${batchsize}_lamCa_${lambda_causal}_lamRe_${lambda_re}_cls1_adt2_EW2_100_rm${randm}_rn${randn}_str${stride}_${UniqueExpName}
23
+
24
+ #python3 main_my_joint_v13_auto.py --gpu $1 --data mnist --epochs $epochs --autoaug $autoaug --lambda_causal ${lambda_causal} --lambda_re ${lambda_re} --lr $lr --lr_scheduler $lr_scheduler --svroot $svroot --clsadapt $clsadapt --factor_num $factor_num --batchsize ${batchsize} --randm ${randm} --randn ${randn} --stride ${stride}
25
+
26
+ #python3 main_test_digit_v13.py --gpu $1 --svroot $svroot --svpath $svroot/${factor_num}factor_${test_epoch}.csv --factor_num $factor_num --epoch $test_epoch --stride ${stride}
27
+
28
+ python3 AllEpochs_test_digit_v13.py --gpu $1 --svroot $svroot --svpath $svroot/${factor_num}factor_${test_epoch}.csv --factor_num $factor_num --epoch $test_epoch --stride ${stride}
29
+
30
+
31
+
32
+
33
+
34
+
Meta-causal/code-stage1-pipeline/saved-PACS/art_painting/CA_multiple_16fa_v2_ep70_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5/events.out.tfevents.1719926752.hala ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a44a49f0a1b3c59b9763c67ea85708ef8b56cae5fe4336f0383f5f71ba0dac84
3
+ size 40
Meta-causal/code-stage1-pipeline/saved-PACS/art_painting/CA_multiple_16fa_v2_ep70_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5/log.log ADDED
@@ -0,0 +1 @@
 
 
1
+ {'gpu': '0', 'data': 'art_painting', 'ntr': None, 'translate': None, 'autoaug': 'CA_multiple', 'n': 3, 'stride': 5, 'factor_num': 16, 'epochs': 70, 'nbatch': 100, 'batchsize': 6, 'lr': 0.01, 'lr_scheduler': 'cosine', 'svroot': 'saved-PACS//art_painting/CA_multiple_16fa_v2_ep70_lr0.01_cosine_base0.01_bs6_lamCa_1_lamRe1_adt4_cls1_EW2_70_rmTrue_rnTrue_str5', 'clsadapt': True, 'lambda_causal': 1.0, 'lambda_re': 1.0, 'randm': True, 'randn': True, 'network': 'resnet18'}
Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/events.out.tfevents.1719925086.hala ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7946f93077ec2136f75fc090a5762ce810be71cc78d5201e8a671217a678c563
3
+ size 40
Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/events.out.tfevents.1719925314.hala ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2021c61739fbe1f9c066067b4e5903d8d2d6c1c44865e1e9c61449eb3d90327
3
+ size 40
Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/events.out.tfevents.1719925652.hala ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fafb4b17d350157735eb6174ff44bafcea7ab8bf86948df3421447ef45ffcae3
3
+ size 40
Meta-causal/code-stage1-pipeline/saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3/log.log ADDED
@@ -0,0 +1 @@
 
 
1
+ {'gpu': '0çç', 'data': 'mnist', 'ntr': None, 'translate': None, 'autoaug': 'CA_multiple', 'n': 3, 'stride': 3, 'factor_num': 14, 'epochs': 500, 'nbatch': 100, 'batchsize': 32, 'lr': 0.0001, 'lr_scheduler': 'Step', 'svroot': 'saved-digit/CA_multiple_14fa_all_ep500_lr1e-4_lr_schedulerStep0.8_bs32_lamCa_1_lamRe_1_cls1_adt2_EW2_100_rmTrue_rnTrue_str3', 'clsadapt': True, 'lambda_causal': 1.0, 'lambda_re': 1.0, 'randm': True, 'randn': True, 'network': 'resnet18'}