Spaces:

chane335
/

permanence-training

Paused

App Files Files Community

permanence-training / training

206 kB

Ctrl+K

Ctrl+K

1 contributor

History: 23 commits

chane335's picture

PERMANENCE: reversibility-aware RL environment for training LLM agents

8f27137 verified about 1 month ago

stages
PERMANENCE: reversibility-aware RL environment for training LLM agents about 1 month ago
__init__.py

44 Bytes
Upload folder using huggingface_hub about 1 month ago
auto_upload.py

7.89 kB
PERMANENCE: reversibility-aware RL environment for training LLM agents about 1 month ago
config.py

4.01 kB
PERMANENCE: reversibility-aware RL environment for training LLM agents about 1 month ago
config.yaml

1.6 kB
PERMANENCE: reversibility-aware RL environment for training LLM agents about 1 month ago
dry_run.py

10.3 kB
Run 4: tech-only curriculum, 3B model, integrated deploy task about 1 month ago
evaluate.py

16.7 kB
Sync: new curriculum, 6 tasks, composable rubric, latent dynamics about 1 month ago
generate_demo.py

2.1 kB
Upload folder using huggingface_hub about 1 month ago
generate_warmup_traces.py

24.4 kB
PERMANENCE: reversibility-aware RL environment for training LLM agents about 1 month ago
pipeline.py

4.85 kB
Run 4: tech-only curriculum, 3B model, integrated deploy task about 1 month ago
reward_functions.py

5.72 kB
Upload folder using huggingface_hub about 1 month ago
reward_functions_legacy.py

5.72 kB
Run 4: tech-only curriculum, 3B model, integrated deploy task about 1 month ago
rewards.py

16.5 kB
PERMANENCE: reversibility-aware RL environment for training LLM agents about 1 month ago
train.py

15.6 kB
Run 4: tech-only curriculum, 3B model, integrated deploy task about 1 month ago
train_legacy.py

16 kB
Run 4: tech-only curriculum, 3B model, integrated deploy task about 1 month ago
train_trl.py

18.5 kB
Upload folder using huggingface_hub about 1 month ago
train_trl_legacy.py

18.7 kB
Run 4: tech-only curriculum, 3B model, integrated deploy task about 1 month ago