wulfdewolf

Wulfdewolf's group workspace

Group: signal-unfreezing-until-8-SD1OWD

1-1

of 1

Tags

Notes

Author

wulfdewolf

State

Finished

Start time

May 17th, 2022 5:01:58 AM

Runtime

2h 38m 50s

Tracked hours

2h 38m 47s

Run path

wulfdewolf/lpt-for-eeg/117e862t

Linux-3.10.0-1160.49.1.el7.x86_64-x86_64-with-glibc2.2.5

Python version

3.8.6

Git repository

git clone git@github.com:wulfdewolf/lpt-for-eeg.git

Git state

git checkout -b "test-subject-9" 72ee4c121804b77906fbd5feb7532c406e3fd6f5

Command

run.py --cluster /data/brussel/102/vsc10248 --wandb --name signal-unfreezing-from-8 --pretrained-transformer --batch-size 64 --epochs 32 --learning-rate 0.0004 --freeze-ff --freeze-attn --freeze-between 8 12

System Hardware

CPU count	32
GPU count	1
GPU type	NVIDIA A100-PCIE-40GB

Group

signal-unfreezing-until-8-SD1OWD

Job Type

COVC3G

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 16 keys
- cluster:
  "/data/brussel/102/vsc10248"
- data:
  "time-series"
- features:
  false
- freeze_attn:
  true
- ▶
  freeze_between:[] 2 items
  - 0:
    8
  - 1:
    12
- freeze_ff:
  true
- freeze_ln:
  false
- freeze_lower:
  8
- freeze_pos:
  false
- freeze_upper:
  12
- ▶
  hyperparams:{} 6 keys
- name:
  "signal-unfreezing-until-8"
- optimise:
  null
- pretrained_transformer:
  true
- run_type:
  "COVC3G"
- wandb:
  true

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 6 keys
- Test Accuracy:
  0.49826388888888895
- Test Loss:
  7.1406049198574495
- Train Accuracy:
  0.9876815352697106
- Train Loss:
  0.037623791294454985
- Validation Accuracy:
  0.4496527777777778
- Validation Loss:
  6.754942841000026

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

wandb-history

run-117e862t-history:v0