state-machine

State-machine's group workspace

Group: bc-eps-ad-sweep

1-100

of 355

Tags

Notes

Author

state-machine

State

Finished

Start time

March 22nd, 2024 7:45:17 PM

Sweep

fh7nkyr6

Runtime

4h 1m 17s

Tracked hours

4h 1m 10s

Run path

state-machine/xminigrid/ujkkb9iq

Linux-5.4.0-171-generic-x86_64-with-glibc2.35

Python version

3.10.12

Git repository

git clone git@github.com:corl-team/clean-algorithm-distillation.git

Git state

git checkout -b "eps-bc-ad-tuples-k2d-2k-0.75eps-7f3f46c6" aeaa57a2554edb648d3561ba0562da1c10196b9c

Command

/workspace/clean-algorithm-distillation/train_tuples_iterstyle.py --attention_dropout=0.1 --batch_size=1024 --config_path=train_config.yaml --embedding_dropout=0.1 --group=bc-eps-ad-sweep --label_smoothing=0.4 --learning_rate=0.0002442521371769573 --pre_norm=False --residual_dropout=0.2 --subsample=1 --update_steps=80000

System Hardware

CPU count	64
Logical CPU count	128
GPU count	1
GPU type	NVIDIA A100-PCIE-40GB

W&B CLI Version

0.16.3

Group

bc-eps-ad-sweep

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 34 keys
- attention_dropout:
  0.1
- batch_size:
  1,024
- ▶
  betas:[] 2 items
  - 0:
    0.9
  - 1:
    0.99
- checkpoints_path:
  null
- clip_grad:
  1
- config_path:
  "train_config.yaml"
- data_seed:
  0
- embedding_dropout:
  0.1
- env_name:
  "Dark-Key2Door-9x9-v0"
- eval_episodes:
  250
- eval_every:
  10,000
- eval_seed:
  0
- eval_test_goals:
  50
- eval_train_goals:
  20
- group:
  "bc-eps-ad-sweep"
- hidden_dim:
  512
- label_smoothing:
  0.4
- learning_histories_path:
  "data/09ebf"
- learning_rate:
  0.0002442521371769573
- mlc_job:
  "sl-a100-1x-ad-k2d-ugf1xj"
- name:
  "eps-bc-ad-tuples-k2d-2k-0.75eps-7f3f46c6"
- normalize_qk:
  false
- num_heads:
  4
- num_layers:
  4
- num_workers:
  0
- pre_norm:
  false
- project:
  "AD"
- residual_dropout:
  0.2
- seq_len:
  160
- subsample:
  1
- train_seed:
  0
- update_steps:
  80,000
- warmup_ratio:
  0.05
- weight_decay:
  0

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 11 keys
- accuracy:
  0.2841735780239105
- eval/eval-time:
  65.480734375
- ▶
  eval/test_graph:{} 7 keys
- eval/test_mean_return:
  1.46
- eval/test_median_return:
  1
- ▶
  eval/train_graph:{} 7 keys
- eval/train_mean_return:
  1.45
- eval/train_median_return:
  1.5
- loss:
  1.599437952041626
- lr:
  0.00000000000010434002
- udpate-time:
  0.15428115844726562

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

wandb-history

run-ujkkb9iq-history:v0