nonenman

Nonenman's group workspace

Group: <null>

1-20

of 50

Tags

Notes

fine-tuning roberta-large on bipar

Author

nonenman

State

Finished

Start time

January 10th, 2024 11:35:02 AM

Runtime

10m 42s

Tracked hours

10m 39s

Run path

nonenman/RC_project/sj44n0lh

Linux-4.18.0-372.46.1.el8_6.x86_64-x86_64-with-glibc2.28

Python version

3.10.0

Command

/pfs/data5/home/as/as_as/as_nonenman/python_scripts/train.py --model_name roberta-large --model_checkpoint roberta-large-finetuned-squad_v2/epoch-1 --tokenizer_checkpoint roberta-large --batch_size 24 --num_train_epochs 4 --learning_rate 2e-5 --dataset bipar --i_dont_know

System Hardware

CPU count	64
Logical CPU count	128
GPU count	1
GPU type	NVIDIA A100 80GB PCIe

W&B CLI Version

0.14.0

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 6 keys
- batch_size:
  24
- learning_rate:
  0.00002
- num_train_epochs:
  4
- seq_length:
  384
- warmup_proportion:
  0.1
- weight_decay:
  0.01

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 14 keys
- eval/avg_loss:
  1.8336588975154993
- eval/exact_match:
  58.1473968897904
- eval/f1:
  73.22028144460998
- eval/i_dont_know_counter:
  21
- model/hidden_size:
  1,024
- model/max_position_embeddings:
  514
- model/num_attention_heads:
  16
- model/num_hidden_layers:
  24
- model/vocab_size:
  50,265
- train/avg_loss:
  0.7161948278488353
- train/learning_rate:
  0
- train/loss:
  0.9399263858795166
- train/samples_per_sec:
  87.3157596820747
- train/total_runtime:
  10.547160520466663

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

wandb-history

run-sj44n0lh-history:v0

Nonenman's group workspace

<null>

30_roberta-large_A100_24_384_2e-5_squad_v2