small_scale_cql_chaotic_lstm_multiseed-v0 | tlab Workspace

Howuhh's group workspace

Group: small_scale_cql_chaotic_lstm_multiseed-v0

114

1-1

of 1

Tags

Notes

Author

howuhh

State

Finished

Start time

May 31st, 2023 8:58:43 AM

Sweep

qvdw596f

Runtime

12h 23m 41s

Tracked hours

12h 23m 10s

Run path

tlab/NetHack/hpa4qcyk

Linux-5.4.0-146-generic-x86_64-with-glibc2.29

Python version

3.8.10

Git repository

git clone git@github.com:tinkoff-ai/katakomba.git

Git state

git checkout -b "cql-kni-hum-law-a8675b63" 82a20301b0fbb3f7baa77ca5891a91bcfbbd7346

Command

algorithms/small_scale/cql_chaotic_lstm.py --alpha=0.0001 --character=kni-hum-law --data_mode=memmap --group=small_scale_cql_chaotic_lstm_multiseed --train_seed=2 --version=0

System Hardware

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

W&B CLI Version

0.12.21

Group

small_scale_cql_chaotic_lstm_multiseed-v0

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 26 keys
- alpha:
  0.0001
- batch_size:
  64
- character:
  "kni-hum-law"
- checkpoints_path:
  null
- clip_grad_norm:
  null
- clip_range:
  10
- data_mode:
  "memmap"
- eval_episodes:
  50
- eval_every:
  10,000
- eval_processes:
  14
- eval_seed:
  50
- gamma:
  0.999
- group:
  "small_scale_cql_chaotic_lstm_multiseed"
- learning_rate:
  0.0003
- name:
  "cql-kni-hum-law-a8675b63"
- project:
  "NetHack"
- rnn_dropout:
  0
- rnn_hidden_dim:
  2,048
- rnn_layers:
  2
- seq_len:
  16
- tau:
  0.005
- train_seed:
  2
- update_steps:
  500,000
- use_prev_action:
  true
- version:
  0
- weight_decay:
  0

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 21 keys
- cql_loss:
  0.26471376419067383
- depth_max:
  1
- depth_mean:
  0.06
- depth_median:
  0
- depth_min:
  0
- depth_std:
  0.23748684174075835
- loss:
  0.26558417081832886
- q_target:
  6.134271144866943
- reward_max:
  1,039
- reward_mean:
  273.94
- reward_median:
  211.5
- reward_min:
  0
- reward_std:
  254.08427027267945
- td_loss:
  0.0008704036590643227
- times/backward_pass:
  0.013601984024047852
- times/batch_loading_cpu:
  0.0773019790649414
- times/batch_loading_gpu:
  0.0772652816772461
- times/evaluation_cpu:
  71.35113763809204
- times/evaluation_gpu:
  71.3508359375
- times/forward_pass:
  0.00831824016571045
- transitions:
  512,000,000

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

wandb-history

run-hpa4qcyk-history:v0