├── apex ├── .nojekyll ├── apex │ ├── amp │ │ ├── lists │ │ │ ├── __init__.py │ │ │ ├── tensor_overrides.py │ │ │ ├── torch_overrides.py │ │ │ └── functional_overrides.py │ │ ├── __version__.py │ │ ├── __init__.py │ │ ├── compat.py │ │ ├── rnn_compat.py │ │ ├── _amp_state.py │ │ └── README.md │ ├── RNN │ │ ├── README.md │ │ ├── __init__.py │ │ ├── models.py │ │ └── cells.py │ ├── reparameterization │ │ └── README.md │ ├── normalization │ │ └── __init__.py │ ├── optimizers │ │ └── __init__.py │ ├── multi_tensor_apply │ │ ├── __init__.py │ │ └── multi_tensor_apply.py │ ├── fp16_utils │ │ ├── __init__.py │ │ └── README.md │ ├── __init__.py │ └── parallel │ │ ├── multiproc.py │ │ └── README.md ├── tests │ ├── L0 │ │ ├── run_amp │ │ │ ├── __init__.py │ │ │ ├── utils.py │ │ │ ├── test_promotion.py │ │ │ └── test_multi_tensor_l2norm.py │ │ ├── run_fp16util │ │ │ ├── __init__.py │ │ │ └── test_fp16util.py │ │ ├── run_mixed_adam │ │ │ └── __init__.py │ │ ├── run_test.py │ │ └── run_fused_layer_norm │ │ │ └── test_fused_layer_norm.py │ ├── L1 │ │ ├── cross_product_distributed │ │ │ └── run.sh │ │ ├── cross_product │ │ │ └── run.sh │ │ └── common │ │ │ └── compare.py │ ├── distributed │ │ ├── amp_master_params │ │ │ ├── run.sh │ │ │ ├── compare.py │ │ │ └── amp_master_params.py │ │ ├── DDP │ │ │ ├── run_race_test.sh │ │ │ └── ddp_race_condition_test.py │ │ └── synced_batchnorm │ │ │ └── unit_test.sh │ └── docker_extension_builds │ │ └── run.sh ├── examples │ ├── dcgan │ │ └── README.md │ ├── simple │ │ └── distributed │ │ │ ├── run.sh │ │ │ ├── README.md │ │ │ └── distributed_data_parallel.py │ ├── README.md │ └── docker │ │ ├── Dockerfile │ │ └── README.md ├── .gitignore ├── docs │ ├── source │ │ ├── _static │ │ │ ├── img │ │ │ │ └── nv-pytorch2.png │ │ │ └── css │ │ │ │ └── pytorch_theme.css │ │ ├── optimizers.rst │ │ ├── layernorm.rst │ │ ├── parallel.rst │ │ ├── _templates │ │ │ └── layout.html │ │ ├── index.rst │ │ └── fp16_utils.rst │ └── Makefile ├── csrc │ ├── flatten_unflatten.cpp │ ├── fused_adam_cuda.cpp │ └── amp_C_frontend.cpp ├── LICENSE └── apex.patch ├── jukebox ├── __init__.py ├── data │ ├── __init__.py │ ├── text_processor.py │ └── ids │ │ └── v2_genre_ids.txt ├── prior │ └── __init__.py ├── utils │ ├── __init__.py │ ├── torch_utils.py │ ├── sample_utils.py │ ├── checkpoint.py │ ├── remote_utils.py │ └── dist_adapter.py ├── vqvae │ ├── __init__.py │ └── resnet.py └── transformer │ └── __init__.py ├── tensorboardX ├── examples │ ├── __init__.py │ ├── RUN_AFTER_PIP_INSTALL │ ├── tensorboardX │ ├── demo_purge.py │ ├── demo_matplotlib.py │ ├── demo_onnx.py │ ├── chainer │ │ ├── extension_logger │ │ │ ├── visualize.py │ │ │ └── updater.py │ │ └── plain_logger │ │ │ ├── net.py │ │ │ └── data.py │ ├── demo_hparams.py │ ├── demo_nvidia_smi.py │ ├── demo_custom_scalars.py │ ├── demo_beholder.py │ └── demo_multiple_embedding.py ├── tensorboardX │ ├── proto │ │ ├── __init__.py │ │ ├── plugin_mesh.proto │ │ ├── plugin_pr_curve.proto │ │ ├── resource_handle.proto │ │ ├── versions.proto │ │ ├── plugin_text.proto │ │ ├── tensor_shape.proto │ │ ├── types.proto │ │ ├── plugin_text_pb2.py │ │ ├── graph.proto │ │ ├── event.proto │ │ ├── attr_value.proto │ │ ├── node_def.proto │ │ ├── plugin_pr_curve_pb2.py │ │ └── tensor.proto │ ├── __init__.py │ ├── beholder │ │ ├── __init__.py │ │ ├── file_system_tools.py │ │ └── shared_config.py │ ├── x2num.py │ ├── onnx_graph.py │ ├── proto_graph.py │ └── torchvis.py ├── tests │ ├── __init__.py │ ├── test_test.py │ ├── expect │ │ ├── test_summary.test_text.expect │ │ ├── test_summary.test_image_with_one_channel.expect │ │ ├── test_summary.test_image_without_channel.expect │ │ ├── test_summary.test_image_with_four_channel.expect │ │ ├── test_summary.test_image_with_one_channel_batched.expect │ │ ├── test_summary.test_image_with_3_channel_batched.expect │ │ ├── test_summary.test_image_with_four_channel_batched.expect │ │ ├── test_pr_curve.test_pr_purve.expect │ │ ├── test_summary.test_custom_scalars.expect │ │ ├── test_summary.test_uint8_image.expect │ │ ├── test_summary.test_float32_image.expect │ │ ├── test_summary.test_histogram_fd.expect │ │ ├── test_summary.test_histogram_auto.expect │ │ ├── test_summary.test_histogram_doane.expect │ │ ├── test_summary.test_image_with_boxes.expect │ │ ├── test_summary.test_hparams.expect │ │ ├── test_summary.test_audio.expect │ │ ├── test_pr_curve.test_pr_purve_raw.expect │ │ └── test_summary.test_mesh.expect │ ├── test_crc32c.py │ ├── test_onnx_graph.py │ ├── test_record_writer.py │ ├── test_pytorch_graph.py │ ├── test_utils.py │ ├── test_visdom.py │ ├── test_numpy.py │ ├── test_chainer_np.py │ ├── expect_reader.py │ ├── test_summary_writer.py │ ├── test_figure.py │ ├── test_beholder.py │ ├── test_embedding.py │ ├── record_writer_test.py │ └── test_pytorch_np.py ├── setup.cfg ├── screenshots │ ├── Demo.gif │ ├── audio.png │ ├── graph.png │ ├── image.png │ ├── text.png │ ├── scalar.png │ ├── embedding.png │ ├── histogram.png │ └── distribution.png ├── docs │ ├── utils.rst │ ├── tensorboard.rst │ ├── index.rst │ └── Makefile ├── .flake8 ├── .gitignore ├── MANIFEST.in ├── .github │ └── ISSUE_TEMPLATE │ │ ├── feature-requests-or-general-questions.md │ │ └── bug_report.md ├── .codecov.yml ├── LICENSE ├── compile.sh ├── tensorboardX.patch ├── .travis.yml └── setup.py ├── .DS_Store ├── MANIFEST.in ├── requirements.txt ├── setup.py ├── LICENSE └── .gitignore /apex/.nojekyll: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /jukebox/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /jukebox/data/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /jukebox/prior/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /jukebox/utils/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /jukebox/vqvae/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /apex/apex/amp/lists/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /apex/tests/L0/run_amp/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /jukebox/transformer/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /tensorboardX/examples/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /apex/tests/L0/run_fp16util/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /apex/tests/L0/run_mixed_adam/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /tensorboardX/examples/RUN_AFTER_PIP_INSTALL: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /tensorboardX/tensorboardX/proto/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /apex/apex/RNN/README.md: -------------------------------------------------------------------------------- 1 | Under construction... 2 | -------------------------------------------------------------------------------- /tensorboardX/examples/tensorboardX: -------------------------------------------------------------------------------- 1 | ../tensorboardX/ -------------------------------------------------------------------------------- /apex/examples/dcgan/README.md: -------------------------------------------------------------------------------- 1 | Under construction... 2 | -------------------------------------------------------------------------------- /.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/.DS_Store -------------------------------------------------------------------------------- /apex/apex/reparameterization/README.md: -------------------------------------------------------------------------------- 1 | Under construction... 2 | -------------------------------------------------------------------------------- /apex/.gitignore: -------------------------------------------------------------------------------- 1 | apex.egg-info 2 | dist 3 | build 4 | docs/build 5 | *~ -------------------------------------------------------------------------------- /tensorboardX/tests/__init__.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import tensorboardX.proto 3 | -------------------------------------------------------------------------------- /MANIFEST.in: -------------------------------------------------------------------------------- 1 | recursive-include jukebox *.py 2 | recursive-include jukebox *.txt 3 | -------------------------------------------------------------------------------- /apex/apex/normalization/__init__.py: -------------------------------------------------------------------------------- 1 | from .fused_layer_norm import FusedLayerNorm 2 | -------------------------------------------------------------------------------- /apex/apex/amp/__version__.py: -------------------------------------------------------------------------------- 1 | VERSION = (0, 1, 0) 2 | __version__ = '.'.join(map(str, VERSION)) 3 | -------------------------------------------------------------------------------- /apex/apex/RNN/__init__.py: -------------------------------------------------------------------------------- 1 | from .models import LSTM, GRU, ReLU, Tanh, mLSTM 2 | 3 | __all__ = ['models'] 4 | -------------------------------------------------------------------------------- /tensorboardX/setup.cfg: -------------------------------------------------------------------------------- 1 | [metadata] 2 | license_file = LICENSE 3 | 4 | [bdist_wheel] 5 | universal = 1 6 | -------------------------------------------------------------------------------- /apex/apex/optimizers/__init__.py: -------------------------------------------------------------------------------- 1 | from .fused_adam import FusedAdam 2 | from .fp16_optimizer import FP16_Optimizer 3 | -------------------------------------------------------------------------------- /tensorboardX/screenshots/Demo.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/Demo.gif -------------------------------------------------------------------------------- /tensorboardX/screenshots/audio.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/audio.png -------------------------------------------------------------------------------- /tensorboardX/screenshots/graph.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/graph.png -------------------------------------------------------------------------------- /tensorboardX/screenshots/image.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/image.png -------------------------------------------------------------------------------- /tensorboardX/screenshots/text.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/text.png -------------------------------------------------------------------------------- /tensorboardX/screenshots/scalar.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/scalar.png -------------------------------------------------------------------------------- /apex/tests/L1/cross_product_distributed/run.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | 3 | cp ../common/* . 4 | bash run_test.sh distributed $1 5 | -------------------------------------------------------------------------------- /tensorboardX/screenshots/embedding.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/embedding.png -------------------------------------------------------------------------------- /tensorboardX/screenshots/histogram.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/histogram.png -------------------------------------------------------------------------------- /tensorboardX/screenshots/distribution.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/tensorboardX/screenshots/distribution.png -------------------------------------------------------------------------------- /apex/docs/source/_static/img/nv-pytorch2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Todo/jukebox/master/apex/docs/source/_static/img/nv-pytorch2.png -------------------------------------------------------------------------------- /apex/examples/simple/distributed/run.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | python -m torch.distributed.launch --nproc_per_node=2 distributed_data_parallel.py 3 | -------------------------------------------------------------------------------- /tensorboardX/docs/utils.rst: -------------------------------------------------------------------------------- 1 | Helper functions 2 | =================================== 3 | .. autofunction:: tensorboardX.utils.figure_to_image -------------------------------------------------------------------------------- /tensorboardX/tests/test_test.py: -------------------------------------------------------------------------------- 1 | def test_linting(): 2 | import subprocess 3 | # subprocess.check_output(['flake8', 'tensorboardX']) 4 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | fire==0.1.3 2 | tqdm==4.45.0 3 | soundfile==0.10.3.post1 4 | unidecode==1.1.1 5 | numba==0.48.0 6 | librosa==0.7.2 7 | mpi4py>=3.0.0 -------------------------------------------------------------------------------- /tensorboardX/.flake8: -------------------------------------------------------------------------------- 1 | [flake8] 2 | max-line-length = 120 3 | ignore = E305,E402,E721,E741,F401,F403,F405,F821,F841,F999 4 | exclude = tensorboardX/proto -------------------------------------------------------------------------------- /apex/apex/multi_tensor_apply/__init__.py: -------------------------------------------------------------------------------- 1 | from .multi_tensor_apply import MultiTensorApply 2 | 3 | multi_tensor_applier = MultiTensorApply(2048*32) 4 | 5 | -------------------------------------------------------------------------------- /tensorboardX/.gitignore: -------------------------------------------------------------------------------- 1 | proto_src/ 2 | protoc-*.zip 3 | protoc/ 4 | __pycache__ 5 | docs/_* 6 | build 7 | dist 8 | *.egg-info 9 | runs/* 10 | *.pyc 11 | -------------------------------------------------------------------------------- /apex/tests/distributed/amp_master_params/run.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | python -m torch.distributed.launch --nproc_per_node=2 amp_master_params.py 3 | 4 | python compare.py 5 | -------------------------------------------------------------------------------- /apex/tests/distributed/DDP/run_race_test.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | 3 | CUDA_VISIBLE_DEVICES=0,1 python -m torch.distributed.launch --nproc_per_node=2 ddp_race_condition_test.py 4 | -------------------------------------------------------------------------------- /tensorboardX/MANIFEST.in: -------------------------------------------------------------------------------- 1 | include HISTORY.rst 2 | include LICENSE 3 | include compile.sh 4 | recursive-include tensorboardX/proto * 5 | recursive-exclude test * 6 | recursive-exclude examples * 7 | recursive-include tensorboardX/beholder * -------------------------------------------------------------------------------- /tensorboardX/.github/ISSUE_TEMPLATE/feature-requests-or-general-questions.md: -------------------------------------------------------------------------------- 1 | --- 2 | name: Feature requests or General questions 3 | about: Feature requests or general questions 4 | title: '' 5 | labels: '' 6 | assignees: '' 7 | 8 | --- 9 | 10 | 11 | -------------------------------------------------------------------------------- /apex/tests/L1/cross_product/run.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | 3 | DATADIR="/home/mcarilli/Desktop/pt18data/apex_stale/examples/imagenet/bare_metal_train_val/" 4 | # DATADIR="/opt/home/apex/examples/imagenet/" 5 | cp ../common/* . 6 | bash run_test.sh single_gpu $1 $DATADIR yes 7 | -------------------------------------------------------------------------------- /tensorboardX/tensorboardX/__init__.py: -------------------------------------------------------------------------------- 1 | """A module for visualization with tensorboard 2 | """ 3 | 4 | from .record_writer import RecordWriter 5 | from .torchvis import TorchVis 6 | from .writer import FileWriter, SummaryWriter 7 | 8 | __version__ = "1.8" # will be overwritten if run setup.py 9 | -------------------------------------------------------------------------------- /tensorboardX/.codecov.yml: -------------------------------------------------------------------------------- 1 | coverage: 2 | status: 3 | project: # measuring the overall project coverage 4 | default: # context, you can create multiple ones with custom titles 5 | enabled: yes 6 | patch: 7 | default: 8 | enabled: no 9 | -------------------------------------------------------------------------------- /tensorboardX/docs/tensorboard.rst: -------------------------------------------------------------------------------- 1 | tensorboardX 2 | =================================== 3 | .. automodule:: tensorboardX 4 | 5 | .. autoclass:: SummaryWriter 6 | :members: 7 | 8 | .. automethod:: __init__ 9 | 10 | .. autoclass:: TorchVis 11 | :members: 12 | 13 | .. automethod:: __init__ -------------------------------------------------------------------------------- /apex/apex/amp/__init__.py: -------------------------------------------------------------------------------- 1 | from .amp import init, half_function, float_function, promote_function,\ 2 | register_half_function, register_float_function, register_promote_function 3 | from .handle import scale_loss, disable_casts 4 | from .frontend import initialize 5 | from ._amp_state import master_params, _amp_state 6 | -------------------------------------------------------------------------------- /apex/docs/source/optimizers.rst: -------------------------------------------------------------------------------- 1 | .. role:: hidden 2 | :class: hidden-section 3 | 4 | apex.optimizers 5 | =================================== 6 | 7 | .. automodule:: apex.optimizers 8 | .. currentmodule:: apex.optimizers 9 | 10 | .. FusedAdam 11 | ---------- 12 | 13 | .. autoclass:: FusedAdam 14 | :members: 15 | -------------------------------------------------------------------------------- /apex/docs/source/layernorm.rst: -------------------------------------------------------------------------------- 1 | .. role:: hidden 2 | :class: hidden-section 3 | 4 | apex.normalization.fused_layer_norm 5 | =================================== 6 | 7 | .. automodule:: apex.normalization 8 | .. currentmodule:: apex.normalization 9 | 10 | .. FusedAdam 11 | ---------- 12 | 13 | .. autoclass:: FusedLayerNorm 14 | :members: 15 | -------------------------------------------------------------------------------- /tensorboardX/tests/expect/test_summary.test_text.expect: -------------------------------------------------------------------------------- 1 | value { 2 | tag: "dummy/text_summary" 3 | tensor { 4 | dtype: DT_STRING 5 | tensor_shape { 6 | dim { 7 | size: 1 8 | } 9 | } 10 | string_val: "text 123" 11 | } 12 | metadata { 13 | plugin_data { 14 | plugin_name: "text" 15 | } 16 | } 17 | } 18 | -------------------------------------------------------------------------------- /apex/tests/distributed/synced_batchnorm/unit_test.sh: -------------------------------------------------------------------------------- 1 | python single_gpu_unit_test.py 2 | python -m torch.distributed.launch --nproc_per_node=2 two_gpu_unit_test.py 3 | python -m torch.distributed.launch --nproc_per_node=2 two_gpu_unit_test.py --fp64 4 | #beware, you need a system with at least 4 gpus to test group_size