TabPFN
TabPFN copied to clipboard
automl
Official implementation of the TabPFN and the tabpfn package.
Hi there! While trying to reproduce the [training](https://github.com/carteakey/tabpfn-eval/tree/main) i thought of cleaning up some code, even though TabPFN 2.0 might be in the works (cant wait). - Removed the duplicate...
I am running tabpfn v0.0.9 within a Python 3.10 environment, and torch version 2.2.1. There's the following user warning: ```python /usr/local/lib/python3.10/dist-packages/torch/utils/checkpoint.py:460: UserWarning: torch.utils.checkpoint: please pass in use_reentrant=True or use_reentrant=False explicitly....
When I run the demo, I get the following error:ConnectionError: We have to download the TabPFN, as there is no checkpoint at G:\anaconda\Lib\site-packages\tabpfn\models_diff/prior_diff_real_checkpoint_n_0_epoch_100.cpkt It has about 100MB, so this might...
Hi! I wanted to check in about the weights in the bag prior. It seems they are 0.961 for the MLP and 0.038 for the GP. Does that seem right?...
My understanding of `mix_activations` is whether to mix activation functions within a batch when sampling from the MLP prior. Now there's ```python if not (('mix_activations' in hyperparameters) and hyperparameters['mix_activations']): s...
@noahho was so nice as to walk me through some of the prior code, but I have one more question. In the model builder part of the prior config is...
I use the TabPFNClassifier and set the no_grad=False. However, the model can not return a tensor with gradient. What should I do to get the gradient of the model output...
When I run the PriorFittingCustomPrior.ipynb, there is an error in line 160 of train.py file. RuntimeError: weight tensor should be defined either for all 2 classes or no classes but...
It would be great if the model would support Apple Silicon devices so that the 'mps' device could be used as follows: `model = TabPFNClassifier(device='mps', N_ensemble_configurations=16)`
Thank you very much for this paper and sharing the code! A small question please: In the paper you show the results of the SCM prior and the BNN prior....
