bloomRL / td3 /config.yaml
maxymoo2's picture
Upload pedagogical policy baseline checkpoints
6f81d73 verified
_wandb:
value:
cli_version: 0.24.2
e:
01gyu8q9oxvtak0e47y4lz0g5metwek3:
args:
- --critic_model=critic_dkt
- --tau=0.95
- --gamma=0.99
- --exploration_noise_sigma=0.1
- --use_wandb
- --wandb_project_name=task1_all_kc
- --kc_to_que_path=data/XES3G5M/metadata/kc_questions_map.json
- --kc_emb_path=data/XES3G5M_embeddings/kc_emb.json
- --cluster_to_kc_path=data/XES3G5M/metadata/kc_clusters.json
- --cluster_to_que_path=data/XES3G5M/metadata/cluster_to_que_ids_map.json
- --pretrained_model_path=data/pretrained_kt_model.ckpt
- --dataloader_num_workers=8
- --seed=4
cpu_count: 252
cpu_count_logical: 252
cudaVersion: "12.8"
disk:
/:
total: "1247017926656"
used: "907788820480"
email: maxonorris@gmail.com
executable: /mnt/pvc/Baselines/ExRec/.venv_task1/bin/python
gpu: NVIDIA A100-SXM4-40GB
gpu_count: 1
gpu_nvidia:
- architecture: Ampere
cudaCores: 6912
memoryTotal: "42949672960"
name: NVIDIA A100-SXM4-40GB
uuid: GPU-c22c8774-4afe-dc6c-6c63-88c853d4ee94
host: alphamcts-run-models-td3-mve-5jszf-xwzm6
memory:
total: "962415210496"
os: Linux-6.8.0-55-generic-x86_64-with-glibc2.31
program: -m train_test.all_kc_train_td3
python: CPython 3.10.13
root: /mnt/pvc/Baselines/ExRec
startedAt: "2026-02-12T13:56:46.616991Z"
writerId: 01gyu8q9oxvtak0e47y4lz0g5metwek3
m: []
python_version: 3.10.13
t:
"1":
- 1
- 5
- 53
"2":
- 1
- 5
- 53
"3":
- 2
- 13
- 15
- 16
"4": 3.10.13
"5": 0.24.2
"12": 0.24.2
"13": linux-x86_64
action_size:
value: 768
actor_lr:
value: 5e-05
batch_size:
value: 512
checkpoint_path:
value: ./td3_saved_models/70150a76-b4f6-4350-8e03-57f86dee3960
cluster_to_kc_path:
value: data/XES3G5M/metadata/kc_clusters.json
cluster_to_que_path:
value: data/XES3G5M/metadata/cluster_to_que_ids_map.json
critic_hidden_size:
value: 300
critic_lr:
value: 5e-05
critic_model:
value: critic_dkt
critic_up_projection_size:
value: 1200
dataloader_num_workers:
value: 8
exploration_noise_sigma:
value: 0.1
gamma:
value: 0.99
hidden_size:
value: 300
kc_emb_path:
value: data/XES3G5M_embeddings/kc_emb.json
kc_emb_size:
value: 768
kc_to_que_path:
value: data/XES3G5M/metadata/kc_questions_map.json
log_dir:
value: ./train_td3_logs
log_path:
value: ./train_td3_logs/70150a76-b4f6-4350-8e03-57f86dee3960
n_epoch:
value: 100
noise_clip:
value: 0.5
policy_noise:
value: 0.2
pretrained_model_path:
value: data/pretrained_kt_model.ckpt
sample_per_epoch:
value: 20
save_dir:
value: ./td3_saved_models
seed:
value: 4
student_state_size:
value: 300
tau:
value: 0.95
test_batch_size:
value: 2048
test_init_seq_size:
value: 100
test_last_n_steps:
value: 10
test_log_wandb:
value: false
test_max_steps:
value: 10
test_n_episode:
value: 2048
test_reward_func:
value: step_by_step
test_reward_scale:
value: 1000
train_batch_size:
value: 512
train_folds:
value: 2-3-4
train_init_seq_size:
value: 100
train_last_n_steps:
value: 10
train_max_steps:
value: 10
train_max_steps_until_student_change:
value: 10
train_n_episode:
value: 512
train_replay_buffer_size:
value: 100
train_reward_func:
value: step_by_step
train_reward_scale:
value: 1000
update_actor_freq:
value: 2
use_wandb:
value: true
wandb_project_name:
value: task1_all_kc
wandb_run_name:
value: td3