sam_dfki

Arne's group workspace

Group: unknown

1-1

of 1

Tags

Notes

Author

bverma

State

Finished

Start time

May 6th, 2022 9:44:27 PM

Runtime

1h 13m 24s

Tracked hours

1h 13m 18s

Run path

sam_dfki/best_adu_uncased/10r5ge6a

Linux-5.4.0-109-generic-x86_64-with-glibc2.10

Python version

3.8.10

Git repository

git clone git@github.com:ArneBinder/SAM.git

Git state

git checkout -b "faithful-glitter-1-4" b761d672e4cc2773e6cb2223c6e2e12bc8b43370

Command

/home/verma/Github/SAM/training_scripts/cross_validation/train-cross-validation.py train -f allennlp_configs/adu_best_uncased.jsonnet -s experiments/rel/uncased_best/10r5ge6a -o "{\"dataset_reader.num_shards\": 5, \"dataset_reader.dev_shard_idx\": 4, \"trainer.callbacks\": [{\"group\": \"unknown\", \"name\": \"faithful-glitter-1-4\", \"wandb_kwargs\": {\"job_type\": \"faithful-glitter-1\", \"id\": \"10r5ge6a\"}, \"type\": \"custom_wandb\", \"entity\": \"sam_dfki\", \"project\": \"uncased_best_adu\", \"files_to_save\": [\"config.json\", \"out.log\", \"metrics.json\"]}]}"

System Hardware

CPU count	96
GPU count	1
GPU type	NVIDIA RTX A6000

W&B CLI Version

0.12.0

Group

unknown

Job Type

faithful-glitter-1

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 6 keys
- ▶
  data_loader:{} 1 key
  - batch_size:
    8
- ▶
  dataset_reader:{} 10 keys
- ▶
  model:{} 8 keys
- train_data_path:
  "./dataset_scripts/sciarg.py@train"
- ▶
  trainer:{} 6 keys
- validation_data_path:
  "./dataset_scripts/sciarg.py@dev"

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 557 keys
- best_epoch:
  23
- best_validation_loss:
  1,225.8265838623047
- best_validation_span/overall/f1:
  0.4951182469081731
- best_validation_span/overall/precision:
  0.4956559513466551
- best_validation_span/overall/recall:
  0.49458170784568706
- best_validation_token/accuracy:
  0.7802674430820489
- best_validation_token/accuracy3:
  0.7873662409389023
- best_validation_token/overall-macro/fscore:
  0.5465525388717651
- best_validation_token/overall-macro/precision:
  0.5663408637046814
- best_validation_token/overall-macro/recall:
  0.5567867755889893
- best_validation_token/overall-micro/fscore:
  0.626060962677002
- best_validation_token/overall-micro/precision:
  0.5909266471862793
- best_validation_token/overall-micro/recall:
  0.6656373739242554
- epoch:
  43
- gradient_mean/crf.transitions:
  -0.00000003332922560162
- gradient_mean/encoder._module.bias_hh_l0:
  0.00007883506623329595
- gradient_mean/encoder._module.bias_hh_l0_reverse:
  -0.00017691975517664105
- gradient_mean/encoder._module.bias_hh_l1:
  0.0003050543018616736
- gradient_mean/encoder._module.bias_hh_l1_reverse:
  0.0002436250215396285
- gradient_mean/encoder._module.bias_ih_l0:
  0.00007883506623329595
- gradient_mean/encoder._module.bias_ih_l0_reverse:
  -0.00017691975517664105
- gradient_mean/encoder._module.bias_ih_l1:
  0.0003050543018616736
- gradient_mean/encoder._module.bias_ih_l1_reverse:
  0.0002436250215396285
- gradient_mean/encoder._module.weight_hh_l0:
  0.00000004630620864532
- gradient_mean/encoder._module.weight_hh_l0_reverse:
  -0.00000119287244615407
- gradient_mean/encoder._module.weight_hh_l1:
  -0.00000092489273129104
- gradient_mean/encoder._module.weight_hh_l1_reverse:
  0.00000183196766556648
- gradient_mean/encoder._module.weight_ih_l0:
  0.00000215039835893549
- gradient_mean/encoder._module.weight_ih_l0_reverse:
  -0.00000093604649009649
- gradient_mean/encoder._module.weight_ih_l1:
  0.00000197435974769178
- gradient_mean/encoder._module.weight_ih_l1_reverse:
  -0.00000076048326036471
- gradient_mean/tag_projection_layer._module.bias:
  0.00000475690922030481
- gradient_mean/tag_projection_layer._module.weight:
  0.00000002365845830354
- gradient_norm:
  758.7786865234375
- gradient_std/crf.transitions:
  0.2188877314329147
- gradient_std/encoder._module.bias_hh_l0:
  0.0034934720024466515
- gradient_std/encoder._module.bias_hh_l0_reverse:
  0.0020788321271538734
- gradient_std/encoder._module.bias_hh_l1:
  0.004545277450233698
- gradient_std/encoder._module.bias_hh_l1_reverse:
  0.0027462008874863386
- gradient_std/encoder._module.bias_ih_l0:
  0.0034934720024466515
- gradient_std/encoder._module.bias_ih_l0_reverse:
  0.0020788321271538734
- gradient_std/encoder._module.bias_ih_l1:
  0.004545277450233698
- gradient_std/encoder._module.bias_ih_l1_reverse:
  0.0027462008874863386
- gradient_std/encoder._module.weight_hh_l0:
  0.0005738005856983364
- gradient_std/encoder._module.weight_hh_l0_reverse:
  0.0003478081780485809
- gradient_std/encoder._module.weight_hh_l1:
  0.0004022366192657501
- validation/token/overall-macro/recall:
  0.5136938095092773
- validation/token/overall-micro/fscore:
  0.590088427066803
- validation/token/overall-micro/precision:
  0.5513905882835388
- validation/token/overall-micro/recall:
  0.6346279382705688

This run produced these artifacts as outputs. Total: 1. Learn more

wandb-history

run-10r5ge6a-history:v0