issues
search
webmachinelearning
/
webnn
🧠Web Neural Network API
https://www.w3.org/TR/webnn/
Other
397
stars
48
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
MLOperandDataType and ArrayBufferView compatibility table missing int64/uint64
#693
inexorabletash
closed
6 months ago
0
Rename the logical operation "not" to "logicalNot"
#692
shiyi9801
closed
6 months ago
5
Divide-by-zero outcome should be standardized
#691
huningxin
opened
6 months ago
2
Add missing validation for pad(), slice(), and split()
#690
inexorabletash
closed
6 months ago
2
Consider removing `lstm` and `gru` operators
#689
a-sully
opened
6 months ago
15
[MLBuffer] Support interop with WebGPU
#688
bbernhar
closed
2 months ago
3
Validate layerNormalization options.axes
#687
inexorabletash
closed
6 months ago
1
`layerNormalization()` method steps should validate the values of `options.axes`
#686
huningxin
closed
6 months ago
0
Support int32 data type for 'indices' operand of 'gather' operator
#685
huningxin
closed
6 months ago
0
Remove note about interop issues with clamp()'s minValue == maxValue
#684
inexorabletash
closed
6 months ago
1
Validate no duplicate axes for reduction ops
#683
inexorabletash
closed
6 months ago
2
Process: Add "interop" label
#682
anssiko
closed
6 months ago
1
Shall we update the spec to add constraints on the axes and the input rank for the reduction operator?
#681
mei1127
closed
6 months ago
5
Add missing definitions of inputShape to conv2d algorithms
#680
inexorabletash
closed
6 months ago
2
Build fix: Correct link for "transferred" term
#679
inexorabletash
closed
6 months ago
2
Specifies scalar values casted to match input type.
#678
philloooo
closed
4 months ago
6
Missing TypeScript Type Declaration
#677
egalli
closed
4 months ago
6
Rename MLOperandDescriptor's "dimensions" key to "shape"
#676
a-sully
closed
2 months ago
10
why `gather` indices only accept uint32 or int64?
#675
philloooo
closed
6 months ago
2
Use consistent phrasing for operator creation
#674
inexorabletash
closed
6 months ago
3
Meta: Introduce "Interop" label?
#673
inexorabletash
closed
6 months ago
4
Handful of algorithm and convention fixes
#672
inexorabletash
closed
7 months ago
3
Remove uses of null as parameter to reshape() in code samples
#671
a-sully
closed
7 months ago
2
Make `MLOperandDescriptor.dimensions` a `FrozenArray`
#670
a-sully
closed
5 months ago
3
Rename `MLOperandDescriptor`'s `dimensions` to `shape`
#669
a-sully
closed
2 months ago
2
Do we need an `MLConstantOperand`?
#668
a-sully
closed
2 months ago
14
Remove use of `null` as parameter to `reshape` in code examples
#667
a-sully
closed
7 months ago
2
Reconsider `MLOperand` methods
#666
a-sully
closed
1 month ago
11
Add note about no-op graphs
#665
inexorabletash
closed
7 months ago
1
Remove MLActivation parameters used for op fusion
#664
a-sully
closed
7 months ago
2
Fix unidirectionally broadcast the shapes steps
#663
huningxin
closed
5 months ago
7
"Unidirectionally broadcast the shapes" steps have issues
#662
huningxin
closed
5 months ago
2
Bugfix: prelu should validate slope operand
#661
huningxin
closed
7 months ago
2
prelu steps should validate the data type of slope is equal to the data type of input
#660
huningxin
closed
7 months ago
0
Simplify, correct, and add validation for GRU/LSTM and friends
#659
inexorabletash
closed
7 months ago
4
Consider removing `MLActivation` parameters used for op fusion
#658
a-sully
closed
7 months ago
6
Make operand data type and rank validation table-driven
#657
inexorabletash
closed
2 weeks ago
14
Remove sequential filling overload of constant()
#656
a-sully
closed
7 months ago
4
gru/lstm: Fix missing/incorrect variables and options use
#655
inexorabletash
closed
7 months ago
2
Consider dropping the support of uint64/int64 data type for some operators
#654
lisa0314
opened
7 months ago
5
Consider changing output type of ArgMax/Argmin to int32, or allow passing output_type
#653
philloooo
closed
4 months ago
2
ArgMax/Min `selectLastIndex` is not supported on CoreML
#652
philloooo
closed
4 months ago
11
Remove softplus() steepness option
#651
a-sully
closed
7 months ago
5
Swap parameters of scalar constant() operand method
#650
inexorabletash
closed
6 months ago
7
Add axis argument to softmax()
#649
inexorabletash
closed
7 months ago
12
Editorial: Add note about keepDimensions of reduction ops
#648
inexorabletash
closed
7 months ago
2
Introduce MLNumber for specifying numeric inputs of any type
#647
inexorabletash
closed
4 months ago
7
Specify the operand data type constraints of operations
#646
inexorabletash
closed
7 months ago
2
Consider removing `steepness` parameter of softplus
#645
shiyi9801
closed
7 months ago
6
Validate the hidden size of GRU and LSTM operators
#644
inexorabletash
closed
7 months ago
2
Previous
Next