gpt_visual_memory / logs /imagenet_gimel_0.out
eminorhan's picture
Upload 26 files
fea8e2a
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
Namespace(data_path='/scratch/work/public/imagenet/train', vqconfig_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.yaml', vqmodel_path='/scratch/eo41/visual-recognition-memory/vqgan_pretrained_models/imagenet_16x16_16384.ckpt', num_workers=8, seed=0, save_dir='/scratch/eo41/visual-recognition-memory/gpt_pretrained_models', gpt_config='GPT_gimel', vocab_size=16384, block_size=255, batch_size=32, lr=0.0003, optimizer='Adam', epochs=1000, resume='', save_prefix='imagenet', gpu=None, world_size=-1, rank=-1, dist_url='env://', dist_backend='nccl', local_rank=-1)
model:
base_learning_rate: 4.5e-06
params:
ddconfig:
attn_resolutions:
- 16
ch: 128
ch_mult:
- 1
- 1
- 2
- 2
- 4
double_z: false
dropout: 0.0
in_channels: 3
num_res_blocks: 2
out_ch: 3
resolution: 256
z_channels: 256
embed_dim: 256
lossconfig:
params:
codebook_weight: 1.0
disc_conditional: false
disc_in_channels: 3
disc_num_layers: 2
disc_start: 0
disc_weight: 0.75
target: vqloss.VQLPIPSWithDiscriminator
monitor: val/rec_loss
n_embed: 16384
target: vqmodel.VQModel
Working with z of shape (1, 256, 16, 16) = 65536 dimensions.
loaded pretrained LPIPS loss from taming/modules/autoencoder/lpips/vgg.pth
VQLPIPSWithDiscriminator running with hinge loss.
Loaded VQ encoder.
Data loaded: dataset contains 1281167 images, and takes 5005 training iterations per epoch.
Number of parameters: 750659840
Running on 8 GPUs total
=> no checkpoint loaded, will train from scratch
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
/scratch/eo41/miniconda3/lib/python3.9/site-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='none' instead.
warnings.warn(warning.format(ret))
Epoch: 0 | Training loss: 6.135367824123813 | Elapsed time: 4453.789637804031
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_000_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 1 | Training loss: 5.798131484299392 | Elapsed time: 4448.56702375412
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_001_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 2 | Training loss: 5.7218508500319265 | Elapsed time: 4448.635702133179
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_002_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 3 | Training loss: 5.675868084475949 | Elapsed time: 4448.570371866226
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_003_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 4 | Training loss: 5.64415309231479 | Elapsed time: 4448.48592376709
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_004_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 5 | Training loss: 5.617594873083459 | Elapsed time: 4448.740148067474
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_005_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 6 | Training loss: 5.5994461619770615 | Elapsed time: 4449.281894683838
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_006_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 7 | Training loss: 5.580984299856943 | Elapsed time: 4448.514730215073
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_007_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 8 | Training loss: 5.568105853306545 | Elapsed time: 4448.5986959934235
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_008_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 9 | Training loss: 5.553497776713643 | Elapsed time: 4449.114318370819
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_009_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 10 | Training loss: 5.545027699884954 | Elapsed time: 4449.138834238052
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_010_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 11 | Training loss: 5.531521415519905 | Elapsed time: 4449.098812580109
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_011_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 12 | Training loss: 5.522672693498365 | Elapsed time: 4448.901001691818
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_012_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 13 | Training loss: 5.515013064823665 | Elapsed time: 4448.462759017944
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_013_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 14 | Training loss: 5.508660832556573 | Elapsed time: 4448.8206622600555
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_014_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 15 | Training loss: 5.500996496865561 | Elapsed time: 4448.373802423477
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_015_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 16 | Training loss: 5.494677847463053 | Elapsed time: 4449.025486946106
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_016_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 17 | Training loss: 5.488317275642753 | Elapsed time: 4448.813071966171
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_017_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 18 | Training loss: 5.482922156159575 | Elapsed time: 4448.179989337921
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_018_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 19 | Training loss: 5.480782058879688 | Elapsed time: 4448.937339067459
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_019_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 20 | Training loss: 5.471766509876384 | Elapsed time: 4448.567655324936
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_020_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 21 | Training loss: 5.468871520973228 | Elapsed time: 4448.808972358704
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_021_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 22 | Training loss: 5.463682885698743 | Elapsed time: 4447.909594774246
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_022_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 23 | Training loss: 5.459242056466483 | Elapsed time: 4447.975906133652
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_023_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 24 | Training loss: 5.454185632654242 | Elapsed time: 4447.988601446152
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_024_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 25 | Training loss: 5.451898510329849 | Elapsed time: 4448.234513998032
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_025_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 26 | Training loss: 5.446100732496569 | Elapsed time: 4448.3813943862915
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_026_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 27 | Training loss: 5.443605179839082 | Elapsed time: 4448.738905668259
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_027_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 28 | Training loss: 5.440151975633619 | Elapsed time: 4448.119179487228
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_028_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 29 | Training loss: 5.435839955123154 | Elapsed time: 4447.811242103577
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_029_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 30 | Training loss: 5.43510546612811 | Elapsed time: 4447.706588983536
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_030_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 31 | Training loss: 5.428354823505962 | Elapsed time: 4448.152802705765
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_031_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 32 | Training loss: 5.4250246925430226 | Elapsed time: 4448.008017539978
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_032_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 33 | Training loss: 5.425318639833372 | Elapsed time: 4448.51774430275
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_033_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 34 | Training loss: 5.419239971187565 | Elapsed time: 4447.93063378334
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_034_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 35 | Training loss: 5.418465005815565 | Elapsed time: 4447.893654823303
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_035_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 36 | Training loss: 5.416968753978566 | Elapsed time: 4447.303329944611
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_036_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
Epoch: 37 | Training loss: 5.414549265088854 | Elapsed time: 4447.923640966415
Saving model to: /scratch/eo41/visual-recognition-memory/gpt_pretrained_models/model_037_imagenet_GPT_gimel_256b_0.0003lr_Adamo_0s.pt
slurmstepd: error: *** JOB 25434829 ON ga007 CANCELLED AT 2022-09-27T05:16:34 DUE TO TIME LIMIT ***
slurmstepd: error: *** STEP 25434829.0 ON ga007 CANCELLED AT 2022-09-27T05:16:34 DUE TO TIME LIMIT ***
srun: Job step aborted: Waiting up to 32 seconds for job step to finish.