forked from allenai/RL4LMs
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathcommands.txt
More file actions
40 lines (35 loc) · 1.5 KB
/
Copy pathcommands.txt
File metadata and controls
40 lines (35 loc) · 1.5 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
# Fine-tune supervised
CUDA_VISIBLE_DEVICES=0,1,2,3 python scripts/training/train_text_generation.py \
--config_path scripts/training/task_configs/faithdial/t5_supervised.yml \
--project_name rl4lm_exps \
--experiment_name t5-base-fd-supervised-5epoch \
--entity_name wyu-du \
--log_to_wandb
# Supervised eval
CUDA_VISIBLE_DEVICES=0 python scripts/training/eval_text_generation.py \
--config_path scripts/training/task_configs/faithdial/t5_supervised_eval.yml \
--project_name rl4lm_exps \
--experiment_name t5-fd-supervised-full
# PPO-1k
CUDA_VISIBLE_DEVICES=0,1,2,3 python scripts/training/train_text_generation.py \
--config_path scripts/training/task_configs/faithdial/t5_ppo_1k.yml \
--project_name rl4lm_exps \
--experiment_name t5-fd-supervised-ppo-1k-reward \
--entity_name wyu-du \
--log_to_wandb
# MDD PPO eval
CUDA_VISIBLE_DEVICES=0 python scripts/training/eval_text_generation.py \
--config_path scripts/training/task_configs/multidoc2dial/t5_ppo_eval.yml \
--project_name rl4lm_exps \
--experiment_name t5-mdd-supervised-ppo-full-token_know_f1 \
--epoch 99
# FD PPO eval
CUDA_VISIBLE_DEVICES=3 python scripts/training/eval_text_generation.py \
--config_path scripts/training/task_configs/faithdial/t5_ppo_eval.yml \
--project_name rl4lm_exps \
--experiment_name t5-fd-supervised-ppo-full-combined_lr \
--epoch 2399
python scripts/training/get_best_dev_scores.py \
--config_path scripts/training/task_configs/faithdial/t5_ppo_eval.yml \
--project_name rl4lm_exps \
--experiment_name t5-fd-supervised-ppo-full-combined005