|
|
--- |
|
|
language: en |
|
|
tags: |
|
|
- pythae |
|
|
- reproducibility |
|
|
license: apache-2.0 |
|
|
--- |
|
|
|
|
|
This model was trained with pythae. It can be downloaded or reloaded using the method `load_from_hf_hub` |
|
|
```python |
|
|
>>> from pythae.models import AutoModel |
|
|
>>> model = AutoModel.load_from_hf_hub(hf_hub_path="clementchadebec/reproduced_beta_tc_vae") |
|
|
``` |
|
|
|
|
|
## Reproducibility |
|
|
This trained model reproduces the results of the official implementation of [1]. |
|
|
|
|
|
| Model | Dataset | Metric | Obtained value | Reference value | |
|
|
|:---:|:---:|:---:|:---:|:---:| |
|
|
| BetaTCVAE | DSPRITES | ELBO/Modified ELBO (after 50 epochs) | 710.41/85.54 | 712.26/86.40 | |
|
|
|
|
|
[1] Ricky TQ Chen, Xuechen Li, Roger B Grosse, and David K Duvenaud. Isolating sources of disentanglement in variational autoencoders. Advances in neural information processing systems, 31, 2018. |
