Added leaky rectified linear algorithm by atomicsorcerer · Pull Request #6260 · TheAlgorithms/Python

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer · 2022-07-20T19:46:23Z

Describe your change:

Added the leaky rectified linear algorithm (also known as leaky ReLU). Leaky ReLU is an alternative to normal ReLU because it solves the dying ReLU problem, which is an issue in some neural networks.

Add an algorithm?
Fix a bug or typo in an existing algorithm?
Documentation change?

Checklist:

stale · 2022-11-02T06:25:37Z

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

tianyizheng02 · 2023-06-18T14:37:03Z

+Script inspired from its corresponding Wikipedia article
+https://en.wikipedia.org/wiki/Rectifier_(neural_networks)
+"""
+from __future__ import annotations


Is the __future__ import needed here?

tianyizheng02 · 2023-06-18T14:41:15Z

+def leaky_relu(
+ vector: float | list[float], negative_slope: float = 0.01
+) -> float | list[float]:


Suggested change
defleaky_relu(
vector: float|list[float], negative_slope: float=0.01
) ->float|list[float]:
defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:
I think just type hinting it as np.ndarray is fine for this function. Using numpy arrays is pretty standard when it comes to NN-related Python programming, and numpy functions that take in arrays generally also support scalars as well ("array_like")

tianyizheng02 · 2023-06-18T14:42:53Z

+ if isinstance(vector, int):
+ raise ValueError(
+ "leaky_relu() only accepts floats or a list of floats for vector"
+ )


Why do we not want to support ints as input? They can all be cast to floats as output

tianyizheng02 · 2023-06-18T14:45:40Z

+ if not isinstance(negative_slope, float):
+ raise ValueError("leaky_relu() only accepts a float value for negative_slope")


I think the constraints on the possible range for the negative slope should be clearer. Are we restricting it to a float between 0 and 1? If so, that should be the if-condition instead

tianyizheng02 · 2023-06-18T14:50:19Z

+ if isinstance(vector, float):
+ if vector < 0:
+ return vector * negative_slope
+ return vector
+
+ for index, value in enumerate(vector):
+ if value < 0:
+ vector[index] = value * negative_slope
+
+ return vector


Suggested change
ifisinstance(vector, float):
ifvector<0:
returnvector*negative_slope
returnvector
forindex, valueinenumerate(vector):
ifvalue<0:
vector[index] =value*negative_slope
returnvector
returnnp.maximum(vector, negative_slope*vector)
numpy functions can handle these cases very easily. Also, leaky ReLU is equivalent to $f(x) = \max(x, ax)$ for negative slopes $0 \leq a \leq 1$ (see the Wikipedia article you cited for more info)

tianyizheng02 · 2023-08-17T01:22:51Z

Closing in favor of #8962

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

atomicsorcerer added 2 commits July 19, 2022 23:29

Created leaky_relu.py and implemented the algorithm.
d91eb1c

Updated docstrings in leaky_relu
e1daa2b

atomicsorcerer requested a review from Kush1101 as a code owner July 20, 2022 19:46

ghost added the awaiting reviews This PR is ready to be reviewed label Jul 20, 2022

Changed slope to negative_slope to provide more info
97dfba6

staleBot added the stale Used to mark an issue or pull request stale. label Nov 2, 2022

tianyizheng02 requested changes Jun 18, 2023
View reviewed changes

staleBot removed stale Used to mark an issue or pull request stale. labels Jun 18, 2023

tianyizheng02 closed this Aug 17, 2023

-defleaky_relu(
-vector: float|list[float], negative_slope: float=0.01
-) ->float|list[float]:
+defleaky_relu(vector: np.ndarray, negative_slope: float=0.01) ->np.ndarray:

		if not isinstance(negative_slope, float):
		raise ValueError("leaky_relu() only accepts a float value for negative_slope")

-ifisinstance(vector, float):
-ifvector<0:
-returnvector*negative_slope
-returnvector
-forindex, valueinenumerate(vector):
-ifvalue<0:
-vector[index] =value*negative_slope
-returnvector
+returnnp.maximum(vector, negative_slope*vector)

Uh oh!

Uh oh!

Conversation

atomicsorcerer commented Jul 20, 2022• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Describe your change:

Checklist:

Uh oh!

staleBot commented Nov 2, 2022

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

tianyizheng02 commented Aug 17, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Uh oh!

Conversation

atomicsorcerer commented Jul 20, 2022• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Describe your change:

Checklist:

Uh oh!

staleBot commented Nov 2, 2022

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

tianyizheng02 commented Aug 17, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Uh oh!

Conversation

atomicsorcerer commented Jul 20, 2022• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Describe your change:

Checklist:

Uh oh!

staleBot commented Nov 2, 2022

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

atomicsorcerer commented Jul 20, 2022•
edited
Loading

atomicsorcerer commented Jul 20, 2022•
edited
Loading

atomicsorcerer commented Jul 20, 2022•
edited
Loading

atomicsorcerer commented Jul 20, 2022•
edited
Loading