Commit History

  1. Update geneformer/evaluation_utils.py to be compatible with different versions of the datasets package.⁠

    Verified signature
    fc7a1c0

  1. add input_identifier to tokenize specific matched files

    Christina Theodoris
    ac59e36

  1. plot umap for all labels in same view

    Christina Theodoris
    ff551ee
  2. Update README.md

    Verified signature
    13cd541

  1. silence tensor copy warning

    Christina Theodoris
    9d8d2f0
  2. clone rather than wrap tensors

    Christina Theodoris
    aea4988
  3. update collator with dummy save fn for transformers compatibility

    Christina Theodoris
    f17618a

  1. fix to properly move model after checking device

    Christina Theodoris
    f953554
  2. move model to device after checking

    Christina Theodoris
    efc403d
  3. update V1 token dict usage to self attr

    Christina Theodoris
    f26136d
  4. move V1 autoformatting to after validate_options

    Christina Theodoris
    4364d31

  1. add option for relabeling data from prior label class dict, update dict paths in manifest

    Christina Theodoris
    76a78a0

  1. delete first version of dist multitask example

    Verified signature
    a67c9c0

  1. update dist multitask example name

    Christina Theodoris
    596f1c2
  2. Add example notebook for distributed multi-task cell classification (#543)⁠

    ctheodoris +1: madhavanvvs
    Verified signature
    f3ff19d

  1. specify bitsandbites version

    Christina Theodoris
    6a50f33
  2. add V2 models

    Christina Theodoris
    7cf49a3
  3. update with V2 models

    Christina Theodoris
    d319fef

  1. edit docs to remove backend isp functions

    Christina Theodoris
    31bf641

  1. fix TypeError: TrainingArguments.__init__() got an unexpected keyword argument 'evaluation_strategy' (#539)⁠

    ctheodoris +1: jackkuo
    Verified signature
    42dbf0a

  1. add fallback task_type for LoraConfig to support different PEFT versions (#538)⁠

    ctheodoris +1: madhavanvvs
    Verified signature
    d471d1b

  1. update eval_strategy for new transformers version

    Christina Theodoris
    18a2ca6
  2. move import wandb to conditional

    Christina Theodoris
    7e4697d

  1. fully qualified imports to resolve name-space conflicts (#532)⁠

    ctheodoris +1: madhavanvvs
    Verified signature
    5f07ec8

  1. ddp_mtl_nccl_support (#524)⁠

    Verified signature
    e2efe2b

  1. edit imports in perturber_utils

    Christina Theodoris
    c6d04a6

  1. fix exact_mean subselection to be for index rather than col name

    Christina Theodoris
    a13a2cf
  2. fix exact_mean and exact_median to subselect dataframe emb cols, not cell rows

    Christina Theodoris
    cfc8cdb
  3. add quantization for pretrained model

    Christina Theodoris
    7b591f6

  1. Prepare data cell level pass cell state dict instead of genes (#483)⁠

    ctheodoris +1: hchen725
    Verified signature
    69e6887
  2. ensure gene class dict only inverted when using gene classifier

    Christina Theodoris
    54b408b

  1. make plot roc compatible with eval-only metrics

    Christina Theodoris
    ae4867d
  2. Merge remote-tracking branch 'origin/main'

    Christina Theodoris
    5760b30
  3. fix gene class dict labeling

    Christina Theodoris
    bfcada4

  1. Fix how mappings and train and val data are returned (#473)⁠

    ctheodoris +1: hchen725
    Verified signature
    c81f6f9

  1. Upload gene_name_id_dict_gc95M.pkl (#470)⁠

    ctheodoris +1: hchen725
    Verified signature
    01d3ea8

  1. Add checks for custom attributes and n_counts prior to sum ensembl id (#461)⁠

    ctheodoris +1: hchen725
    Verified signature
    09de197
  2. further emphasize different token dictionaries in examples

    Christina Theodoris
    e2ee685
  3. remove unused imports while no longer using distributed sampler

    Verified signature
    b9028ba
  4. Raise error for train and validation mismatch (#459)⁠

    ctheodoris +1: madhavanvenkatesh
    Verified signature
    9af94b8

  1. Upload in_silico_perturber.py (#432)⁠

    ctheodoris +1: davidjwen
    Verified signature
    cb89107

  1. Update geneformer/emb_extractor.py (#453)⁠

    ctheodoris +1: hchen725
    Verified signature
    ebc1e09
  2. Update geneformer/tokenizer.py (#450)⁠

    ctheodoris +1: hchen725
    Verified signature
    664f71e

  1. update isp default mode to cls

    Verified signature
    0c99403

  1. Adding tags to the model (#448)⁠

    ctheodoris +1: canergen
    Verified signature
    3a68669

  1. update transformers version to match pretrainer using accelerate

    Verified signature
    3d62bb9
  2. update pretrainer to not use distributed sampler (Trainer uses accelerate)

    Verified signature
    8140c51

  1. update function for N_Detections for mixture_model without anchor_token

    Christina Theodoris
    df297bc

  1. add check to ensure emb_label is None for getting state embs dict

    Christina Theodoris
    39b4444

  1. change default emb to cls

    Verified signature
    9ee140a