Twu31
/

TanyaoDojo

Reinforcement Learning

behavior-cloning

Model card Files Files and versions

89 MB

Ctrl+K

Ctrl+K

1 contributor

History: 23 commits

Twu31's picture

Card: fine-tuning from the baseline itself = +0.10 +/- 0.34 at 100k (point estimate only); bust excess is a property of our own base

4b17716 verified 10 days ago

.gitattributes

1.52 kB
initial commit about 1 month ago
README.md

34.4 kB
Card: fine-tuning from the baseline itself = +0.10 +/- 0.34 at 100k (point estimate only); bust excess is a property of our own base 10 days ago
bc_lean_g402.pkl
Detected Pickle imports (3)
- "numpy.dtype",
- "numpy._core.multiarray._reconstruct",
- "numpy.ndarray"
What is a pickle import?
25 MB
xet

TanyaoDojo checkpoints: BC lean/v2 champions + oracle-critic negative result about 1 month ago
bc_lean_w192_ep2.pkl
Detected Pickle imports (3)
- "numpy.dtype",
- "numpy.ndarray",
- "numpy._core.multiarray._reconstruct"
What is a pickle import?
14 MB
xet

TanyaoDojo checkpoints: BC lean/v2 champions + oracle-critic negative result about 1 month ago
bc_v2_g186.pkl
Detected Pickle imports (3)
- "numpy.dtype",
- "numpy.ndarray",
- "numpy._core.multiarray._reconstruct"
What is a pickle import?
25 MB
xet

TanyaoDojo checkpoints: BC lean/v2 champions + oracle-critic negative result about 1 month ago
rl_oracle_800m.pkl
Detected Pickle imports (3)
- "numpy.dtype",
- "numpy._core.multiarray._reconstruct",
- "numpy.ndarray"
What is a pickle import?
25 MB
xet

TanyaoDojo checkpoints: BC lean/v2 champions + oracle-critic negative result about 1 month ago