huanz

Huanz's group workspace

Group: composer_id-ATEPP-perfmidi-sequence-MIDILike-bpe4

1-3

of 3

Tags

Notes

Author

huanz

State

Crashed

Start time

March 21st, 2023 10:01:28 AM

Runtime

6h 56m 44s

Tracked hours

Run path

huanz/symrep/je0q0w7l

Linux-3.10.0-1160.11.1.el7.x86_64-x86_64-with-glibc2.17

Python version

3.9.7

Git repository

git clone https://github.com/anusfoil/SymRep.git

Git state

git checkout -b "composer_id-ATEPP-perfmidi-sequence-fold0" 7e0c86b8c3c48706312f7f33b136109d8f244962

Command

/homes/hz009/Research/SymRep/experiments/train.py experiment.fold_idx=0 experiment.grouped=1 experiment.dataset=ATEPP experiment.symrep=sequence experiment.input_format=perfmidi experiment.lr=1e-6 sequence.mid_encoding=MIDILike experiment.device=3 experiment.batch_size=2 sequence.BPE=4 sequence.max_seq_len=1000

System Hardware

CPU count	40
Logical CPU count	80
GPU count	8
GPU type	NVIDIA RTX A5000

W&B CLI Version

0.13.10

Group

composer_id-ATEPP-perfmidi-sequence-MIDILike-bpe4

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 73 keys
- cfg/dataset/ASAP/dataset_dir:
  "/homes/hz009/Research/Datasets/asap-dataset/"
- cfg/dataset/ASAP/metadata_file:
  "/homes/hz009/Research/Datasets/asap-dataset/metadata-v1.3.csv"
- cfg/dataset/ATEPP/dataset_dir:
  "/homes/hz009/Research/Datasets/ATEPP-1.1/"
- cfg/dataset/ATEPP/metadata_file:
  "/homes/hz009/Research/Datasets/ATEPP-1.1/ATEPP-metadata-1.5.csv"
- cfg/dataset/PD/dataset_dir:
  "/homes/hz009/Research/Datasets/piano_difficulty/"
- cfg/dataset/PD/metadata_file:
  "/homes/hz009/Research/Datasets/piano_difficulty/labels.csv"
- cfg/experiment/artifact_dir:
  "huanz/symrep/model-6gg9q4u6:v11"
- cfg/experiment/batch_size:
  2
- cfg/experiment/checkpoint_dir:
  "checkpoints/composer_id-ATEPP-perfmidi-sequence-fold0"
- cfg/experiment/checkpoint_file:
  "checkpoints/composer_id-ATEPP-perfmidi-sequence-fold0/epoch=49-step=1850.ckpt"
- cfg/experiment/data_save_dir:
  "/import/c4dm-datasets-ext/symrep_assets/sequence/ATEPP/perfmidi"
- cfg/experiment/dataset:
  "ATEPP"
- cfg/experiment/device:
  3
- cfg/experiment/emb_dim:
  32
- cfg/experiment/epoch:
  500
- cfg/experiment/es_patience:
  60
- cfg/experiment/es_threshold:
  0.005
- cfg/experiment/exp_name:
  "composer_id-ATEPP-perfmidi-sequence-fold0"
- cfg/experiment/feat_level:
  0
- cfg/experiment/fold_idx:
  0
- cfg/experiment/group_name:
  "composer_id-ATEPP-perfmidi-sequence-MIDILike-bpe4"
- cfg/experiment/grouped:
  1
- cfg/experiment/input_format:
  "perfmidi"
- cfg/experiment/load_data:
  true
- cfg/experiment/load_model:
  false
- cfg/experiment/lr:
  0.000001
- cfg/experiment/lr_gamma:
  0.998
- cfg/experiment/random_seed:
  13
- cfg/experiment/restore_training:
  false
- cfg/experiment/symrep:
  "sequence"
- cfg/experiment/task:
  "composer_id"
- cfg/experiment/testing_only:
  false
- ▶
  cfg/graph/basic_edges:[] 7 items
- cfg/graph/bi_dir:
  true
- cfg/graph/conv_type:
  "SAGEConv"
- cfg/graph/dropout:
  0
- cfg/graph/edge_agg:
  "sum"
- cfg/graph/hid_dim:
  64
- cfg/graph/homo:
  0
- cfg/graph/mid_window:
  0.03
- cfg/graph/n_layers:
  4
- cfg/graph/sage_agg:
  "lstm"
- cfg/matrix/bins:
  131
- cfg/matrix/n_channels:
  1
- ▶
  cfg/matrix/res_layers:[] 4 items
  - 0:
    1
  - 1:
    1
  - 2:
    1
  - 3:
    1
- cfg/matrix/resolution:
  800
- model_backend:
  null
- model_frontend:
  null
- ▶
  segmentation:{} 5 keys
- ▶
  sequence:{} 13 keys

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 16 keys
- epoch:
  19
- ▶
  graph_0:{} 4 keys
  - _type:
    "graph-file"
  - path:
    "media/graph/graph_0_summary_0fa7bd31b00b8698a591.graph.json"
  - sha256:
    "0fa7bd31b00b8698a591af1e5863bf4728b93bac3b52374dcc703806d9b87fcc"
  - size:
    7,282
- lr-Adam:
  0.00000096267630964563
- train_acc_epoch:
  0.8004984855651855
- train_acc_step:
  1
- train_auroc_epoch:
  0.8750051856040955
- train_auroc_step:
  0.11764705926179886
- train_fscore_epoch:
  0.3734315633773804
- train_fscore_step:
  0.11764705926179886
- train_loss_epoch:
  0.8199736475944519
- train_loss_step:
  0.21503682434558868
- trainer/global_step:
  83,399
- val_acc:
  0.7607973217964172
- val_auroc:
  0.8470929265022278
- val_fscore:
  0.35404711961746216
- val_loss:
  0.917222023010254

This run produced these artifacts as outputs. Total: 2. Learn more

model

model-je0q0w7l:v15

wandb-history

run-je0q0w7l-history:v0