Search results

Migration to PyTorch

enhancementNew feature or requestNew feature or request

Status: Open.

#225 In kermitt2/delft;

· lfoppiano opened on Jun 19, 2026

Handle sequences larger than the model context window

Status: Open.

#222 In kermitt2/delft;

· lfoppiano opened on May 11, 2026

Make embeddings optional

Status: Open.

#216 In kermitt2/delft;

· de-code opened on May 5, 2026

Experiment with more recent static embeddings

Status: Open.

#178 In kermitt2/delft;

· kermitt2 opened on Jan 16, 2025

Implement feature channel in classification

enhancementNew feature or requestNew feature or request

Status: Open.

#152 In kermitt2/delft;

· lfoppiano opened on Jan 6, 2023

Sub-tokenization with certain transformers

Status: Open.

#150 In kermitt2/delft;

· lfoppiano opened on Jan 6, 2023

Better support SentencePiece tokenizer(s)

Status: Open.

#128 In kermitt2/delft;

· kermitt2 opened on Mar 7, 2022

improve directory management for training/nfold-training

enhancementNew feature or requestNew feature or request

Status: Open.

#126 In kermitt2/delft;

· lfoppiano opened on Mar 7, 2022

add parameter to optionally output raw results in the evaluation

enhancementNew feature or requestNew feature or request

Status: Open.

#117 In kermitt2/delft;

· lfoppiano opened on Jul 10, 2021

sequence labelling, n-fold training should use separate preprocessors

Status: Open.

#102 In kermitt2/delft;

· de-code opened on Apr 23, 2020

Allow multiple tokens per feature data row

Status: Open.

#99 In kermitt2/delft;

· de-code opened on Apr 1, 2020

Implement sliding window

Status: Open.

#90 In kermitt2/delft;

· de-code opened on Feb 18, 2020