52 lines
716 B
Plaintext
52 lines
716 B
Plaintext
# Fundamentals
|
|
numpy
|
|
pyyaml
|
|
tb-nightly
|
|
future
|
|
scp
|
|
tqdm
|
|
matplotlib
|
|
scipy
|
|
munch
|
|
tqdm
|
|
scp
|
|
tensorboard
|
|
orjson
|
|
einops
|
|
lambda-networks
|
|
mup
|
|
|
|
# For image generation stuff
|
|
opencv-python
|
|
kornia
|
|
pytorch_ssim
|
|
gsa-pytorch
|
|
pytorch_fid
|
|
|
|
# For audio generation stuff
|
|
inflect
|
|
librosa
|
|
Unidecode
|
|
tgt
|
|
pyworld
|
|
audio2numpy
|
|
SoundFile
|
|
|
|
# For text stuff
|
|
transformers
|
|
tokenizers
|
|
jiwer # calculating WER
|
|
omegaconf
|
|
|
|
# lucidrains stuff
|
|
vector_quantize_pytorch
|
|
linear_attention_transformer
|
|
rotary-embedding-torch
|
|
axial_positional_embedding
|
|
g-mlp-pytorch
|
|
x-clip
|
|
x_transformers==1.0.4
|
|
|
|
# bitsandbytes
|
|
bitsandbytes==0.35.0 ; platform_system == "Windows"
|
|
bitsandbytes ; platform_system != "Windows" |