long_small_scale_awac_chaotic_lstm_multiseed-v0 | tlab Overview

Howuhh's group workspace

Group: long_small_scale_awac_chaotic_lstm_multiseed-v0

1-1

of 1

Tags

Notes

Author

State

Failed

Start time

June 9th, 2023 5:48:37 PM

Sweep

2seu1a6w

Runtime

4d 15h 15m 56s

Tracked hours

4d 15h 11m 43s

Run path

tlab/NetHack/droldujs

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

Python version

3.8.10

Git repository

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

Git state

git checkout -b "awac-mon-hum-neu-c1b62a5c" c6d28bed00e5a35d8e55bacde8bf1e3d6a7c7f21

Command

algorithms/small_scale/awac_chaotic_lstm.py --character=mon-hum-neu --data_mode=memmap --group=long_small_scale_awac_chaotic_lstm_multiseed --train_seed=0 --update_steps=6600000 --version=0

System Hardware

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

W&B CLI Version

0.12.21

Group

long_small_scale_awac_chaotic_lstm_multiseed-v0

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 26 keys
- batch_size:
  64
- character:
  "mon-hum-neu"
- 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:
  "long_small_scale_awac_chaotic_lstm_multiseed"
- learning_rate:
  0.0003
- name:
  "awac-mon-hum-neu-c1b62a5c"
- project:
  "NetHack"
- rnn_dropout:
  0
- rnn_hidden_dim:
  2,048
- rnn_layers:
  2
- seq_len:
  16
- tau:
  0.005
- temperature:
  1
- train_seed:
  0
- update_steps:
  6,600,000
- use_prev_action:
  true
- version:
  0
- weight_decay:
  0

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 21 keys
- actor_loss:
  6.061090469360352
- depth_max:
  1
- depth_mean:
  0.72
- depth_median:
  1
- depth_min:
  0
- depth_std:
  0.448998886412873
- loss:
  2,088.849853515625
- q_target:
  9,064
- reward_max:
  0
- reward_mean:
  0
- reward_median:
  0
- reward_min:
  0
- reward_std:
  0
- td_loss:
  2,082.788818359375
- times/backward_pass:
  0.018397600173950195
- times/batch_loading_cpu:
  0.05579423904418945
- times/batch_loading_gpu:
  0.05574028778076172
- times/evaluation_cpu:
  11.147830963134766
- times/evaluation_gpu:
  11.1477646484375
- times/forward_pass:
  0.012670047760009764
- transitions:
  3,680,824,320

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

wandb-history

run-droldujs-history:v0