Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 9, 2024

• 0.8.10

  lucidrains

  committed
  5a5974d

• Merge pull request #46 from johahi/main

  Show description for 31f61d7
  lucidrains

  authored
  31f61d7

• small fix to disable future warning with torch.load

  

  Johannes

  committed
  a94ec15

Commits on Oct 7, 2024

• 0.8.9

  lucidrains

  committed
  d4ae222

• Merge pull request #45 from johahi/main

  Show description for f585b81
  lucidrains

  authored
  f585b81

• fixed edge case in shift augmentation

  

  Johannes

  committed
  1c63dda

Commits on Dec 15, 2023

• 0.8.8

  lucidrains

  authored
  9ffeb8b

• Merge pull request #35 from yang-dongxu/main

  Show description for 6e0f16c
  lucidrains

  authored
  6e0f16c

• support for bfloat16

  yang-dongxu

  committed
  bc0218b

Commits on Nov 28, 2023

• used in a new paper!

  lucidrains

  authored
  a4a28f6

Commits on Nov 3, 2023

• Update setup.py

  lucidrains

  authored
  588cf3f

• Merge pull request #34 from Vance-Raiti/main

  Show description for e31cdb4
  lucidrains

  authored
  e31cdb4

• Remove depreciated torchmetrics __init__ argument and fix str_to_one_hot for CUDA devices

  Vance-Raiti

  committed
  a46cf5a

Commits on Nov 1, 2023

• automatically use SyncBatchNorm if doing distributed training

  lucidrains

  committed
  243151b

Commits on Oct 11, 2023

• only allow seq length of 1536 if using tf gamma

  lucidrains

  committed
  dce5709

• address variable sequence lengths while using tf gamma #32

  lucidrains

  committed
  87cc4c0

Commits on Sep 28, 2023

• directly set use_tf_gamma on Attention instances

  lucidrains

  committed
  d2dbc21

• adopt the fix proposed by @johahi for fixing the small numerical discrepancy in the pretrained model between tensorflow and pytorch #31

  lucidrains

  committed
  ece4306

Commits on Sep 27, 2023

• calculate gamma positions in float64 #31

  lucidrains

  committed
  05a3654

Commits on Jun 20, 2023

• update

  lucidrains

  committed
  5117062

Commits on May 29, 2023

• only torch transformers

  lucidrains

  committed
  07a7576

Commits on May 14, 2023

• address #26

  lucidrains

  committed
  3cbcd94

Commits on Apr 5, 2023

• address #25

  lucidrains

  committed
  e3aeba1

Commits on Feb 27, 2023

• readme

  lucidrains

  committed
  c35e742

Commits on Feb 21, 2023

• allow for activation in finetuning head to be customizable, addressing #23

  lucidrains

  committed
  e6526e3

Commits on Feb 11, 2023

• address #21

  lucidrains

  committed
  4e70710

Commits on Feb 9, 2023

• make fine tuning even easier by auto setting target length to crop embeddings by

  lucidrains

  committed
  abb50a0

Commits on Jan 31, 2023

• fix typo in kwargs for polars read_csv, thanks to @icdh99

  lucidrains

  committed
  2affe90

Commits on Jan 29, 2023

• add ability to use the embeddings from the output of the transformer block, and not from after the final pointwise, use layernorm if taking embeddings from post-transformer

  lucidrains

  committed
  7ada950

Commits on Jan 13, 2023

• add finetune_last_n_layers_only keyword argument on the forwards of all the finetune adapters, to specify the last N penultimate layers of transformer one wants to finetune for the enformer

  lucidrains

  committed
  a8d8d2f

Commits on Dec 21, 2022

• address logic that was indented too far in FastaInterval, thanks to @sofroniewn

  lucidrains

  committed
  8fe99cf

Commits on Aug 23, 2022

• fix pool function in attention pool to respect the pool_size, thanks to @cmlakhan

  lucidrains

  committed
  ab29196

Commits on Jul 31, 2022

• add discrete key value bottleneck for finetuning

  lucidrains

  committed
  588fc54

Commits on Jun 12, 2022

• 0.5.3

  lucidrains

  committed
  4a4a6fd

• remove experimental convnext blocks

  lucidrains

  committed
  2001380

Pagination

PreviousNext