srush / annotated-s4

Implementation of https://srush.github.io/annotated-s4
https://srush.github.io/annotated-s4
MIT License
468 stars 60 forks source link

Segmentation fault! #80

Open muxiddin19 opened 5 months ago

muxiddin19 commented 5 months ago

Hi! Great job! Can you help me with this issue? (ann1_s4) wsl@DESKTOP-CJL0ECN:/mnt/h/muhiddin/annotated-s4$ python -m s4.train dataset=quickdraw layer=s4 train.epochs=10 train.bsz=128 model.d_model=128 model.layer.N=64 2024-05-21 13:17:39.628139: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT dataset: quickdraw layer: s4 seed: 0 model: d_model: 128 n_layers: 4 dropout: 0.0 prenorm: true embedding: false layer: 'N': 64 train: epochs: 10 bsz: 128 lr: 0.001 lr_schedule: false weight_decay: 0.01 checkpoint: false suffix: null sample: null wandb: mode: disabled project: s4 entity: null

[] Warning: models are not being checkpoint [] Setting Randomness... Segmentation fault

Whatever training code I run gives me this issue. I would appreciate any help!