Porting TF fake_quant_with_min_max functions #20641

doncarlos999 · 2024-12-13T09:57:27Z

Based on the discussion here: #20319 I started porting the fake_quant_with_min_max functions from tensorflow to keras3.
This PR contains those ported functions and the relevant tests from https://github.com/tensorflow/tensorflow/blob/master/tensorflow/compiler/tests/fake_quant_ops_test.py.

I didn't implement tf.quantization.fake_quant_with_min_max_vars as it looks the same as tf.quantization.fake_quant_with_min_max_args. But, I can add this one too if required.

~~For the CLA I am waiting on our CTO to add me to the Edge Impulse <-> Google CLA. But I figured that I can work on revisions to the PR in the meantime.~~

CC: @matpalm, @dansitu, @james77777778

* adds fake_quant_with_min_max functions from TF to keras3

google-cla · 2024-12-13T09:57:31Z

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

codecov-commenter · 2024-12-13T10:06:12Z

Codecov Report

Attention: Patch coverage is 97.91667% with 1 line in your changes missing coverage. Please review.

Project coverage is 81.98%. Comparing base (e345cbd) to head (0da24e2).

Files with missing lines	Patch %	Lines
keras/api/_tf_keras/keras/quantizers/__init__.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #20641      +/-   ##
==========================================
+ Coverage   81.96%   81.98%   +0.01%     
==========================================
  Files         554      554              
  Lines       51656    51704      +48     
  Branches     7996     8000       +4     
==========================================
+ Hits        42342    42389      +47     
- Misses       7367     7368       +1     
  Partials     1947     1947

Flag	Coverage Δ
keras	`81.80% <95.83%> (+0.01%)`	⬆️
keras-jax	`64.05% <87.50%> (+0.02%)`	⬆️
keras-numpy	`58.92% <66.66%> (+<0.01%)`	⬆️
keras-openvino	`29.94% <12.50%> (-0.02%)`	⬇️
keras-tensorflow	`64.76% <95.83%> (+0.02%)`	⬆️
keras-torch	`64.09% <85.41%> (+0.01%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

fchollet

Thanks for the PR!

keras/src/quantizers/quantizers.py

james77777778

Hi @doncarlos999
I have left some comments.

Additionally, I think we still need fake_quant_with_min_max_vars, as it is used in TFMOT:
https://github.com/tensorflow/model-optimization/blob/master/tensorflow_model_optimization/python/core/quantization/keras/quant_ops.py#L340

keras/api/_tf_keras/keras/quantizers/__init__.py

keras/api/quantizers/__init__.py

keras/src/quantizers/__init__.py

keras/src/quantizers/quantizers.py

james77777778 · 2024-12-19T05:50:03Z

keras/src/quantizers/quantizers_test.py

@@ -100,3 +100,759 @@ def test_quantize_and_dequantize(self):
        )
        # A loose assertion due to an expected quantization error
        self.assertAllClose(qdq_values, values, atol=5e-1)
+
+    def _TestOp(


We can use @parameterized.named_parameters and named_product to organize similar tests like this one:
https://github.com/keras-team/keras/blob/master/keras/src/ops/nn_test.py#L2355-L2365

I have reduced the number of tests by merging the logic into a single function but if you still think it would help to use named_parameters to organize the tests then I can add it.

I think we should follow the convention in the codebase by using @parameterized.named_parameters to organize the tests.

keras/src/quantizers/quantizers_test.py

doncarlos999 · 2024-12-19T08:28:37Z

@james77777778 thank you for the review. I'm working on revisions now.

Regarding the *_gradient functions, I added those as a way to test that the gradients that come from the main functions were being calculated correctly. Should we keep them just for testing purposes but not expose them in the public facing API? If not I will remove them.

james77777778 · 2024-12-19T08:38:36Z

Regarding the *_gradient functions, I added those as a way to test that the gradients that come from the main functions were being calculated correctly. Should we keep them just for testing purposes but not expose them in the public facing API? If not I will remove them.

We can test the gradients of fake_* functions using:

tensorflow: tf.GradientTape() + tape.gradient
torch: loss.backward() + variable.grad
jax: jax.grad

You can refer to this test for an example:
https://github.com/keras-team/keras/blob/master/keras/src/layers/core/dense_test.py#L584-L649

Using a different function, separate from the user-facing function, for testing purposes seems redundant and fragile to me. However, we should wait for calls from @fchollet

doncarlos999 · 2024-12-19T08:42:58Z

I agree that having two separate functions is fragile I simply kept the functions separate as that was how they were tested in the Tensorflow repo.
I will start adding tests based on your example in the meantime. Thank you.

…nt_with_min_max_vars function

… and torch

Refactor to use backend specific gradient functions in tests and merges logic into single function

doncarlos999 · 2025-01-06T15:29:03Z

@james77777778 I have addressed your previous comments.
I refactored the code into a single function to avoid duplication as the logic for the per_channel function on a 1 length array is the same as the other separate functions. If we need a separate function for handling parameters being passed in as float/int rather than single length arrays of floats/ints, then I can add the other function back (I have a previous commit that I can pull it back from).

fchollet · 2025-01-12T00:50:27Z

Thanks for the updates!

@james77777778 should we merge?

james77777778 · 2025-01-12T06:10:42Z

Thanks for the updates!

@james77777778 should we merge?

Sorry for the late reply. I missed the updates.

Let me take a look...

james77777778

@doncarlos999
It’s great to merge three functions into one!

However, the implementation might be slightly different from TF’s because I had to set a higher tolerance to pass this simple test:

import numpy as np
import tensorflow as tf

from keras.src.quantizers.quantizers import fake_quant_with_min_max_args
from keras.src.quantizers.quantizers import fake_quant_with_min_max_vars

# tf.quantization.fake_quant_with_min_max_args
x = np.random.uniform(low=-10, high=10, size=[10, 32, 32, 64]).astype("float32")
inp = tf.convert_to_tensor(x)
with tf.GradientTape() as tape:
    tape.watch(inp)
    tf_out = tf.quantization.fake_quant_with_min_max_args(
        inp, min=-6.0, max=6.0, num_bits=8
    )
tf_gradients = tape.gradient(tf_out, inp)

inp = tf.convert_to_tensor(x)
with tf.GradientTape() as tape:
    tape.watch(inp)
    keras_out = fake_quant_with_min_max_args(
        inp, min_vals=-6.0, max_vals=6.0, num_bits=8
    )
keras_gradients = tape.gradient(keras_out, inp)

np.testing.assert_allclose(tf_out.numpy(), keras_out.numpy(), atol=1e-5)
np.testing.assert_allclose(
    tf_gradients.numpy(), keras_gradients.numpy(), atol=1e-5
)

# tf.quantization.fake_quant_with_min_max_vars
x = np.random.uniform(low=-10, high=10, size=[10, 32, 32, 64]).astype("float32")
min = np.array(-5.5, dtype="float32")
max = np.array(7.5, dtype="float32")
tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    tf_out = tf.quantization.fake_quant_with_min_max_vars(
        tf_x, tf_min, tf_max, num_bits=8
    )
tf_gradients = tape.gradient(tf_out, [tf_x, tf_min, tf_max])

tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    keras_out = fake_quant_with_min_max_vars(tf_x, tf_min, tf_max, num_bits=8)
keras_gradients = tape.gradient(keras_out, [tf_x, tf_min, tf_max])

for tf_g, keras_g in zip(tf_gradients, keras_gradients):
    np.testing.assert_allclose(tf_g.numpy(), keras_g.numpy())

print("All passed!")

Could you please verify the impl?
Thanks!

keras/src/quantizers/quantizers.py

james77777778 · 2025-01-13T13:48:27Z

keras/src/quantizers/quantizers_test.py

@@ -100,3 +100,759 @@ def test_quantize_and_dequantize(self):
        )
        # A loose assertion due to an expected quantization error
        self.assertAllClose(qdq_values, values, atol=5e-1)
+
+    def _TestOp(


I think we should follow the convention in the codebase by using @parameterized.named_parameters to organize the tests.

doncarlos999 · 2025-01-13T13:57:14Z

@james77777778 Thanks for the comments. I will work on revisions.

This PR addresses review feedback to fix implementation and to move tests to using named_parameters rather than individual functions.

doncarlos999 · 2025-01-14T16:55:17Z

@james77777778 I have fixed the implementation so that the simple test code that you provided works. I also updated the test to use parameterized.named_parameters() rather than individual functions. However, I couldn't see a way to use named_product() due to the tests requiring the expected values to be provided.

Let me know if there are any further revisions are required.

james77777778

Hey @doncarlos999 Thanks for the updates!

I think this PR is almost ready to be merged.

However, I failed to pass this test using tf.quantization.fake_quant_with_min_max_vars_per_channel:

import numpy as np
import tensorflow as tf

from keras.src.quantizers.quantizers import (
    fake_quant_with_min_max_vars_per_channel,
)


# tf.quantization.fake_quant_with_min_max_vars_per_channel
x = np.random.uniform(low=-10, high=10, size=[10, 32, 32, 3]).astype("float32")
min = np.array([-5.5] * 3, dtype="float32")
max = np.array([7.5] * 3, dtype="float32")
tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    tf_out = tf.quantization.fake_quant_with_min_max_vars_per_channel(
        tf_x, tf_min, tf_max, num_bits=8
    )
tf_gradients = tape.gradient(tf_out, [tf_x, tf_min, tf_max])

tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    keras_out = fake_quant_with_min_max_vars_per_channel(
        tf_x, tf_min, tf_max, num_bits=8, narrow_range=False
    )
keras_gradients = tape.gradient(keras_out, [tf_x, tf_min, tf_max])

for tf_g, keras_g in zip(tf_gradients, keras_gradients):
    np.testing.assert_allclose(tf_g.numpy(), keras_g.numpy())

print("All passed!")

Could you confirm the impl in fake_quant_with_min_max_vars_per_channel?
Thanks!

keras/src/quantizers/quantizers.py

Adds axis param and fixes logic for per channel function

doncarlos999 · 2025-01-15T11:29:52Z

@james77777778 Thanks for the feedback.
I went ahead and added the axis parameter in this PR as it helped fix the discrepancy with tf.quantization.fake_quant_with_min_max_vars_per_channel. I also addressed the other comments.

james77777778

@doncarlos999
On second thought, I think we only need to export one API since you’ve already combined the logic into a single impl.

Here’s my impl based on yours:

...

from keras.src.backend.common.backend_utils import canonicalize_axis

...

@keras_export("keras.quantizers.fake_quant_with_min_max_vars")
def fake_quant_with_min_max_vars(
    inputs,
    min_vals,
    max_vals,
    num_bits,
    narrow_range=False,
    axis=None,
):
    """Perform per-tensor or per-channel fake quantization.

    `[min_vals, max_vals]` define the clamping range for the `inputs`.

    The `inputs` are quantized into the quantization range:
    - `[0, 2^num_bits - 1]` when `narrow_range=False`
    - `[1, 2^num_bits - 1]` when `narrow_range=True`

    After quantization, the values are dequantized and output as floats within
    the `[min_vals, max_vals]` interval.

    This operation supports gradient computation, allowing `min_vals` and
    `max_vals` to be trained.

    Args:
        inputs: Input tensor of float dtype.
        min_vals: A global minimum scalar or a per-channel minimum tensor.
        max_vals: A global maximum scalar or a per-channel maximum tensor.
        num_bits: Quantization bit width (e.g., `8` for int8).
        narrow_range: Whether to use narrow quantization range.
        axis: Axis along which to perform per-channel quantization. If `None`,
              per-tensor quantization is performed. Defaults to `None`.

    Returns:
        Fake-quantized tensor
    """
    inputs = ops.convert_to_tensor(inputs)
    min_vals = ops.convert_to_tensor(min_vals)
    max_vals = ops.convert_to_tensor(max_vals)

    if axis is not None:
        axis = canonicalize_axis(axis, inputs.ndim)

    @ops.custom_gradient
    def _fake_quant_with_min_max_vars_per_channel(x, min_val, max_val):
        # Calculate quantization parameters for all channels at once
        nudged_min, nudged_max, scale, inv_scale = adjust_and_nudge(
            min_val, max_val, num_bits, narrow_range
        )

        quant_zero = ops.floor(
            ops.add(ops.multiply(-nudged_min, inv_scale), 0.5)
        )
        x_clamped = ops.clip(x, nudged_min, nudged_max)
        x_clamped_shifted = ops.subtract(x_clamped, nudged_min)
        result = ops.multiply(
            ops.floor(
                ops.add(
                    ops.subtract(
                        ops.multiply(x_clamped_shifted, inv_scale), quant_zero
                    ),
                    0.5,
                )
            ),
            scale,
        )

        # Create gradient mask for all channels
        masks = ops.cast(
            (x >= nudged_min) & (x <= nudged_max),
            dtype="float32",
        )

        def grad(*args, upstream=None):
            if upstream is None:
                (upstream,) = args

            # Gradient for x
            dx = ops.multiply(upstream, masks)
            axes = [i for i in range(len(dx.shape)) if i != axis]
            # Gradient for min_val
            # When x is clipped to min, the gradient flows to min_val
            min_mask = ops.cast(x <= nudged_min, dtype="float32")
            grad_min = ops.multiply(upstream, min_mask)
            if axis is not None:
                grad_min = ops.sum(grad_min, axis=axes)
            else:
                grad_min = ops.sum(grad_min)

            # Gradient for max_val
            # When x is clipped to max, the gradient flows to max_val
            max_mask = ops.cast(x >= nudged_max, dtype="float32")
            grad_max = ops.multiply(upstream, max_mask)
            if axis is not None:
                grad_max = ops.sum(grad_max, axis=axes)
            else:
                grad_max = ops.sum(grad_max)

            return dx, grad_min, grad_max

        return result, grad

    return _fake_quant_with_min_max_vars_per_channel(inputs, min_vals, max_vals)

The basic usage:

import numpy as np
import tensorflow as tf

from keras.src.quantizers.quantizers import fake_quant_with_min_max_vars

# tf.quantization.fake_quant_with_min_max_vars
x = np.random.uniform(low=-10, high=10, size=[10, 32, 32, 64]).astype("float32")
min = np.array(-5.5, dtype="float32")
max = np.array(7.5, dtype="float32")
tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    tf_out = tf.quantization.fake_quant_with_min_max_vars(
        tf_x, tf_min, tf_max, num_bits=8
    )
tf_gradients = tape.gradient(tf_out, [tf_x, tf_min, tf_max])

tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    keras_out = fake_quant_with_min_max_vars(tf_x, tf_min, tf_max, num_bits=8)
keras_gradients = tape.gradient(keras_out, [tf_x, tf_min, tf_max])

for tf_g, keras_g in zip(tf_gradients, keras_gradients):
    np.testing.assert_allclose(tf_g.numpy(), keras_g.numpy())

# tf.quantization.fake_quant_with_min_max_vars_per_channel
x = np.random.uniform(low=-10, high=10, size=[10, 32, 32, 3]).astype("float32")
min = np.array([-5.5] * 3, dtype="float32")
max = np.array([7.5] * 3, dtype="float32")
tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    tf_out = tf.quantization.fake_quant_with_min_max_vars_per_channel(
        tf_x, tf_min, tf_max, num_bits=8
    )
tf_gradients = tape.gradient(tf_out, [tf_x, tf_min, tf_max])

tf_x = tf.convert_to_tensor(x)
tf_min = tf.Variable(min)
tf_max = tf.Variable(max)
with tf.GradientTape() as tape:
    tape.watch(tf_x)
    tape.watch(tf_min)
    tape.watch(tf_max)
    keras_out = fake_quant_with_min_max_vars(
        tf_x, tf_min, tf_max, num_bits=8, narrow_range=False, axis=-1
    )
keras_gradients = tape.gradient(keras_out, [tf_x, tf_min, tf_max])

for tf_g, keras_g in zip(tf_gradients, keras_gradients):
    np.testing.assert_allclose(tf_g.numpy(), keras_g.numpy())

print("All passed!")

WDYT?

doncarlos999 · 2025-01-15T14:52:17Z

@james77777778 That looks good to me, do you want me to update the implementation? Or will you push those changes?

james77777778 · 2025-01-15T14:54:46Z

@james77777778 That looks good to me, do you want me to update the implementation? Or will you push those changes?

Feel free to update your impl if you think it looks good!

doncarlos999 · 2025-01-15T15:06:08Z

@james77777778 Done.

james77777778 · 2025-01-15T15:08:49Z

@james77777778 Done.

Shouldn't fake_quant_with_min_max_args and fake_quant_with_min_max_vars_per_channel be removed?

EDITED:
Once the CI passes, this should be ready to merge.

doncarlos999 · 2025-01-15T15:25:42Z

Shouldn't fake_quant_with_min_max_args and fake_quant_with_min_max_vars_per_channel be removed?

Ah yes, my bad. Fixed

james77777778 · 2025-01-15T15:41:57Z

LGTM!
Let's wait for @fchollet 's decision

fchollet

Thanks to both of you! LGTM

QAT (squashed this time) (#1)

fef281e

* adds fake_quant_with_min_max functions from TF to keras3

google-ml-butler bot added the size:XL label Dec 13, 2024

google-ml-butler bot assigned gbaned Dec 13, 2024

fchollet reviewed Dec 14, 2024

View reviewed changes

keras/src/quantizers/quantizers.py Outdated Show resolved Hide resolved

keras/src/quantizers/quantizers.py Outdated Show resolved Hide resolved

keras/src/quantizers/quantizers.py Outdated Show resolved Hide resolved

doncarlos999 and others added 3 commits December 16, 2024 11:19

Merge branch 'keras-team:master' into master

9043966

Addresses PR review comments

b66e9c3

drops another type hint

84338f2

doncarlos999 requested a review from fchollet December 18, 2024 08:34

google-ml-butler bot added the awaiting review label Dec 18, 2024

james77777778 reviewed Dec 19, 2024

View reviewed changes

doncarlos999 and others added 5 commits December 19, 2024 09:39

swaps out if statements, change float() to ops.cast and adds fake_qua…

170f31e

…nt_with_min_max_vars function

fix missed if statement, adds gradient tests via main function for tf…

c193d16

… and torch

Merge branch 'keras-team:master' into master

5c48be2

Merge branch 'keras-team:master' into master

4d97345

fix unbound variable error when not using torch or tf backend (#2)

4007769

Refactor to use backend specific gradient functions in tests and merges logic into single function

Merge branch 'keras-team:master' into master

34954b6

james77777778 reviewed Jan 13, 2025

View reviewed changes

doncarlos999 added 2 commits January 14, 2025 14:55

Merge branch 'keras-team:master' into master

3ed51df

More QAT function revisions (#3)

bfb98b9

This PR addresses review feedback to fix implementation and to move tests to using named_parameters rather than individual functions.

james77777778 mentioned this pull request Jan 15, 2025

Question regarding mode deployment james77777778/keras-image-models#59

Open

james77777778 reviewed Jan 15, 2025

View reviewed changes

keras/src/quantizers/quantizers.py Outdated Show resolved Hide resolved

keras/src/quantizers/quantizers.py Outdated Show resolved Hide resolved

keras/src/quantizers/quantizers.py Outdated Show resolved Hide resolved

doncarlos999 added 2 commits January 15, 2025 11:21

Qat revisions (#4)

42cd701

Adds axis param and fixes logic for per channel function

Merge branch 'keras-team:master' into master

9f96d3e

james77777778 reviewed Jan 15, 2025

View reviewed changes

updated implementation

018ccaf

removed redundant functions

0da24e2

fchollet approved these changes Jan 15, 2025

View reviewed changes

google-ml-butler bot added kokoro:force-run ready to pull Ready to be merged into the codebase labels Jan 15, 2025

fchollet merged commit 6557961 into keras-team:master Jan 15, 2025
7 checks passed

google-ml-butler bot removed awaiting review ready to pull Ready to be merged into the codebase kokoro:force-run labels Jan 15, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Porting TF fake_quant_with_min_max functions #20641

Porting TF fake_quant_with_min_max functions #20641

doncarlos999 commented Dec 13, 2024 •

edited

Loading

google-cla bot commented Dec 13, 2024

codecov-commenter commented Dec 13, 2024 •

edited

Loading

fchollet left a comment

james77777778 left a comment

james77777778 Dec 19, 2024

doncarlos999 Jan 6, 2025

james77777778 Jan 13, 2025

doncarlos999 commented Dec 19, 2024

james77777778 commented Dec 19, 2024

doncarlos999 commented Dec 19, 2024

doncarlos999 commented Jan 6, 2025

fchollet commented Jan 12, 2025

james77777778 commented Jan 12, 2025

james77777778 left a comment

james77777778 Jan 13, 2025

doncarlos999 commented Jan 13, 2025

doncarlos999 commented Jan 14, 2025 •

edited

Loading

james77777778 left a comment

doncarlos999 commented Jan 15, 2025

james77777778 left a comment •

edited

Loading

doncarlos999 commented Jan 15, 2025

james77777778 commented Jan 15, 2025

doncarlos999 commented Jan 15, 2025

james77777778 commented Jan 15, 2025 •

edited

Loading

doncarlos999 commented Jan 15, 2025

james77777778 commented Jan 15, 2025

fchollet left a comment

Porting TF fake_quant_with_min_max functions #20641

Porting TF fake_quant_with_min_max functions #20641

Conversation

doncarlos999 commented Dec 13, 2024 • edited Loading

google-cla bot commented Dec 13, 2024

codecov-commenter commented Dec 13, 2024 • edited Loading

Codecov Report

fchollet left a comment

Choose a reason for hiding this comment

james77777778 left a comment

Choose a reason for hiding this comment

james77777778 Dec 19, 2024

Choose a reason for hiding this comment

doncarlos999 Jan 6, 2025

Choose a reason for hiding this comment

james77777778 Jan 13, 2025

Choose a reason for hiding this comment

doncarlos999 commented Dec 19, 2024

james77777778 commented Dec 19, 2024

doncarlos999 commented Dec 19, 2024

doncarlos999 commented Jan 6, 2025

fchollet commented Jan 12, 2025

james77777778 commented Jan 12, 2025

james77777778 left a comment

Choose a reason for hiding this comment

james77777778 Jan 13, 2025

Choose a reason for hiding this comment

doncarlos999 commented Jan 13, 2025

doncarlos999 commented Jan 14, 2025 • edited Loading

james77777778 left a comment

Choose a reason for hiding this comment

doncarlos999 commented Jan 15, 2025

james77777778 left a comment • edited Loading

Choose a reason for hiding this comment

doncarlos999 commented Jan 15, 2025

james77777778 commented Jan 15, 2025

doncarlos999 commented Jan 15, 2025

james77777778 commented Jan 15, 2025 • edited Loading

doncarlos999 commented Jan 15, 2025

james77777778 commented Jan 15, 2025

fchollet left a comment

Choose a reason for hiding this comment

doncarlos999 commented Dec 13, 2024 •

edited

Loading

codecov-commenter commented Dec 13, 2024 •

edited

Loading

doncarlos999 commented Jan 14, 2025 •

edited

Loading

james77777778 left a comment •

edited

Loading

james77777778 commented Jan 15, 2025 •

edited

Loading