krasserm/perceiver-io

Issues

• NotImplementedError: Module [InputAdapter] is missing the required "forward" function

  #59 opened a month ago by jiveshkalra
  2

• About evaluation metrics

  #58 opened 9 months ago by Colezwhy
  0

• AttributeError: 'tuple' object has no attribute 'last_hidden_state'

  #56 opened a year ago by martin-minovski
  1

• Multimodal autoencoder

  #52 opened a year ago by lantiga
  3

• TypeError: __new__() missing 1 required positional argument: 'task'

  #55 opened a year ago by Norooa
  2

• Error while adding CIFAR10 and CIFAR100

  #54 opened a year ago by ArianKhorasani
  3

• ValueError: Could not load model krasserm/perceiver-io-optical-flow with any of the following classes: (<class 'perceiver.model.vision.optical_flow.huggingface.OpticalFlowPerceiver'>,).

  #53 opened a year ago by wscffaa
  2

• Sequence Modeling Examples

  #50 opened a year ago by awarebayes
  0

• Upgrade to PyTorch 2.0 and PyTorch Lightning 2.0

  #42 opened 2 years ago by krasserm
  0

• Replace example-wise with batch-wise random-truncation

  #39 opened 2 years ago by krasserm
  0

• Optional bias term for `TiedTextOutputAdapter`

  #40 opened 2 years ago by krasserm
  0

• Optional absolute position embedding for Perceiver AR

  #41 opened 2 years ago by krasserm
  0

• Parameterize `generate` method with `top_k` directly instead of `threshold`

  #38 opened 2 years ago by krasserm
  0

• Support tokenizers that do not have a `pad_token` configured

  #37 opened 2 years ago by krasserm
  0

• AttributeError: module 'numpy' has no attribute '_no_nep50_warning'

  #35 opened 2 years ago by 0scarJ1ang
  7

• Generate sequences from batch of prompts

  #31 opened 2 years ago by krasserm
  0

• Support user-defined prefix length at training and inference time

  #30 opened 2 years ago by krasserm
  0

• Perceiver AR FSDP training via command line

  #34 opened 2 years ago by krasserm
  0

• Optional random sequence truncation in C4 data module

  #33 opened 2 years ago by krasserm
  0

• Configurable number of attention heads to be processed in parallel

  #32 opened 2 years ago by krasserm
  0

• Compute-optimal Perceiver AR models

  #28 opened 2 years ago by krasserm
  0

• Use `OutputAdapter` in `PerceiverAR` subclasses

  #26 opened 2 years ago by krasserm
  0

• Share weights of embedding layer with output layer in `CausalLanguageModel`

  #27 opened 2 years ago by krasserm
  0

• Any plan for optical flow?

  #3 opened 2 years ago by kl2005ad
  1

• citing the repo

  #21 opened 2 years ago by je-santos
  2

• Implement Perceiver AR

  #20 opened 2 years ago by krasserm
  0

• Upgrade to PyTorch Lightning 1.7 and PyTorch 1.12

  #19 opened 2 years ago by krasserm
  0

• Set number of processes for parallel text dataset pre-processing to `num_workers` of data module

  #17 opened 2 years ago by krasserm
  0

• Enable external widgets in inference examples notebook

  #18 opened 2 years ago by krasserm
  0

• Genomic sequences

  #14 opened 2 years ago by ajv012
  7

• What is Q in the latent encoder layers?

  #2 opened 3 years ago by zhangyuygss
  1

• text encoding error

  #1 opened 3 years ago by batrlatom
  2