harvardnlp namedtensor issues

harvardnlp / namedtensor

Named Tensor implementation for Torch

http://nlp.seas.harvard.edu/NamedTensor

MIT License

441 stars 42 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Various fixes to bring notebook up to usable state.

#105 zoeqevans opened 2 years ago
7
Add support for tensor abbreviations and better addressing

#104 EndingCredits opened 3 years ago
0
Feature Request: Tensor Abbreviations

#103 EndingCredits opened 5 years ago
2
Dict-like interface isn't looking good!

#102 buriy opened 5 years ago
7
Comparison of tuples of namedtensors yields true

#101 samgoldman97 opened 5 years ago
0
Convert to tensor

#100 MilesCranmer opened 5 years ago
0
Add index_add_

#99 amspector100 opened 5 years ago
0
README documentation incorrect line 169 logit_dim

#98 tynanseltzer opened 5 years ago
0
Update distributions.py

#97 samgoldman97 opened 5 years ago
0
Added cumsum + tests

#96 amspector100 closed 5 years ago
1
Update Examples for ".spec()"

#95 Ricocotam closed 5 years ago
1
Understanding for new commers

#94 Ricocotam closed 5 years ago
1
0-dim named tensors crash on _force_order

#93 gtanzer opened 5 years ago
0
Fix unique

#92 sshen8 closed 5 years ago
4
Wrap `expand` and `expand_as`

#91 srush opened 5 years ago
0
Possible fix for NTorch.index_copy_

#90 JKinx closed 5 years ago
0
Possible fix NTorch.index_copy_

#89 JKinx closed 5 years ago
0
Update the format of torch distribution to be unordered

#88 srush closed 5 years ago
0
slice with step size

#87 srush opened 5 years ago
0
added the ability to iterate over named tensors

#86 edjah closed 5 years ago
2
Update cat

#85 sullivan-sean closed 5 years ago
3
TypeError: call() got multiple values for argument 'names'

#84 fepegar closed 5 years ago
4
Gather: broadcast + expand

#83 justinchiu opened 5 years ago
1
"reciprical" > "reciprocal"

#82 gtanzer closed 5 years ago
0
Torch distributions rewrite.

#81 srush opened 5 years ago
4
Index Copy fancy indexing

#80 srush opened 5 years ago
0
Make Tril/Triu take explicit dimension arguments.

#79 srush closed 5 years ago
0
PR for chunk, and fixing other functions to use tensor._new

#78 tynanseltzer closed 5 years ago
1
Remove dependency on einops

#77 srush closed 5 years ago
0
added chunk for named tensor

#76 tynanseltzer closed 5 years ago
1
Added equal, unique and deepcopy functions

#75 Sanyuan-Chen closed 5 years ago
2
Add JaggedInfo

#74 justinchiu closed 5 years ago
1
Triu / Tril Implemented wrong

#73 srush closed 5 years ago
0
Field lengths + mask

#72 justinchiu closed 5 years ago
0
Fixes fmod + adds test

#71 idk3 closed 5 years ago
1
Padpack

#70 justinchiu closed 5 years ago
0
fix namedfield for tuples

#69 justinchiu closed 5 years ago
0
Begin to convert transformer

#68 srush closed 5 years ago
0
Add a test suite for the NN modules

#67 srush closed 5 years ago
0
CrossEntropyLoss() - underlying order of dimensions of scores and targets need to match (they shouldn't)

#66 jasonren12 closed 5 years ago
1
1D convolution requires 3-dimensional input

#65 samgoldman97 closed 5 years ago
2
Removed accidental print() statement

#64 larslorch closed 5 years ago
0
Option for cuda added in masked tensor assertion

#63 JKinx closed 5 years ago
1
Wrapped ConstantPad

#62 samgoldman97 closed 5 years ago
3
Tried to add topk

#61 JoshFeldman95 closed 5 years ago
1
Changes to config

#60 srush closed 5 years ago
0
Rewrite of test suite

#59 srush closed 5 years ago
0
Update torch_nn.py

#58 da03 closed 5 years ago
0
ntorch.cat/stack will automatically reorder tensors

#57 amspector100 closed 5 years ago
1
Addition isn't commutative (if dimensions are ordered)

#56 zou3519 opened 5 years ago
1