Merge branch 'unifyai:main' into fix-paddle-test

.github/workflows/pre-release.yml

@@ -15,8 +15,40 @@ on:
 permissions:
   actions: read
 jobs:
+  display_test_results:
+    if: ${{ always() }}
+    runs-on: ubuntu-latest
+    needs:
+      - run_tests
+
+    steps:
+      - name: Download all test results
+        uses: actions/download-artifact@v3
+
+      - name: Combined Test Results
+        run: |
+          find . -name "test_results_*.txt" -exec cat {} + > combined_test_results.txt
+          echo "Test results summary:"
+          cat combined_test_results.txt
+
   run_tests:
     runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        branch: [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
+                  11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
+                  21, 22, 23, 24, 25, 26, 27, 28, 29, 30,
+                  31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
+                  41, 42, 43, 44, 45, 46, 47, 48, 49, 50,
+                  51, 52, 53, 54, 55, 56, 57, 58, 59, 60,
+                  61, 62, 63, 64, 65, 66, 67, 68, 69, 70,
+                  71, 72, 73, 74, 75, 76, 77, 78, 79, 80,
+                  81, 82, 83, 84, 85, 86, 87, 88, 89, 90,
+                  91, 92, 93, 94, 95, 96, 97, 98, 99, 100,
+                  101, 102, 103, 104, 105, 106, 107, 108, 109, 110,
+                  111, 112, 113, 114, 115, 116, 117, 118, 119, 120,
+                  121, 122, 123, 124, 125, 126, 127, 128 ]
     steps:
       - name: Checkout Ivy 🛎
         uses: actions/checkout@v3
@@ -41,10 +73,21 @@ jobs:
           mkdir .ivy
           touch .ivy/key.pem
           echo -n ${{ secrets.USER_API_KEY }} > .ivy/key.pem
-          python scripts/setup_tests/setup_priority_tests.py
+          python scripts/setup_tests/setup_priority_tests.py ${{ matrix.branch }}
           cd ..
 
       - name: Run CPU Tests
         run: |
           cd ivy
           python scripts/run_tests/run_tests.py ${{ secrets.REDIS_CONNECTION_URL }} ${{ secrets.REDIS_PASSWORD }} ${{ secrets.MONGODB_PASSWORD }} 'false' 'false' ${{ github.run_id }} 'true' ${{ github.event.inputs.tracer }} ${{ github.event.inputs.tracer_each }} ${{ steps.jobs.outputs.html_url }}
+        continue-on-error: true
+
+      - name: Upload test results
+        uses: actions/upload-artifact@v3
+        with:
+          name: test_results_${{ matrix.branch }}
+          path: ivy/test_results_${{ matrix.branch }}.txt
+
+      - name: Check on failures
+        if: steps.tests.outcome != 'success'
+        run: exit 1

README.md

@@ -1150,8 +1150,7 @@ class IvyNet(ivy.Module):
         self.output_channels = output_channels
         self.num_classes = num_classes
         self.data_format = data_format
-        self.device = device
-        super().__init__()
+        super().__init__(device=device)
 
     def _build(self, *args, **kwargs):
         self.extractor = ivy.Sequential(

docs/overview/volunteer_ranks.rst

@@ -15,6 +15,9 @@ Top Contributors
    * - samunder singh
      - `samthakur587 <https://github.com/samthakur587>`_
      - Merging Master Gold, Merging Wizard, Ivy Inspector Silver
+   * - BANAFSHEH HEZARDASTAN
+     - `NripeshN <https://github.com/NripeshN>`_
+     - Merging Master Bronze, Ivy Inspector Gold
    * - V\. Sai Suraj
      - `Sai-Suraj-27 <https://github.com/Sai-Suraj-27>`_
      - Merging Master Gold, Ivy Inspector Bronze
@@ -66,6 +69,12 @@ Contributors
    * - Jackson McClintock
      - `jacksondm33 <https://github.com/jacksondm33>`_
      - Merging Master, Merging Wizard, Ivy Inspector
+   * - Ayush Lokare
+     - `ayush111111 <https://github.com/ayush111111>`_
+     - Merging Master, Ivy Inspector
+   * - Chaitanya Lakhchaura
+     - `ZenithFlux <https://github.com/ZenithFlux>`_
+     - Debugging Dynamo, Merging Master
    * - David Adlai Nettey
      - `Adlai-1 <https://github.com/Adlai-1>`_
      - Merging Master, Ivy Inspector

ivy/__init__.py

@@ -304,59 +304,23 @@ def __rmul__(self, other):
         return self
 
     def __bool__(self):
+        if ivy.current_backend_str() == "tensorflow":
+            return builtins.bool(builtins.tuple(self._shape))
         return builtins.bool(self._shape)
 
-    def __div__(self, other):
-        return self._shape // other
-
-    def __floordiv__(self, other):
-        return self._shape // other
-
-    def __mod__(self, other):
-        return self._shape % other
-
-    def __rdiv__(self, other):
-        return other // self._shape
-
-    def __rmod__(self, other):
-        return other % self._shape
-
     def __reduce__(self):
         return (self.__class__, (self._shape,))
 
     def as_dimension(self, other):
         if isinstance(other, self._shape):
-            return other
+            return to_ivy(other)
         else:
             return self._shape
 
-    def __sub__(self, other):
-        try:
-            self._shape = self._shape - other
-        except TypeError:
-            self._shape = self._shape - list(other)
-        return self
-
-    def __rsub__(self, other):
-        try:
-            self._shape = other - self._shape
-        except TypeError:
-            self._shape = list(other) - self._shape
-        return self
-
     def __eq__(self, other):
         self._shape = Shape._shape_casting_helper(self._shape, other)
         return self._shape == other
 
-    def __int__(self):
-        if hasattr(self._shape, "__int__"):
-            res = self._shape.__int__()
-        else:
-            res = int(self._shape)
-        if res is NotImplemented:
-            return res
-        return to_ivy(res)
-
     def __ge__(self, other):
         self._shape = Shape._shape_casting_helper(self._shape, other)
         return self._shape >= other
@@ -378,7 +342,7 @@ def __getattribute__(self, item):
 
     def __getitem__(self, key):
         try:
-            return self._shape[key]
+            return to_ivy(self._shape[key])
         except (TypeError, IndexError):
             return None
 
@@ -416,11 +380,11 @@ def index(self, index):
         if self._shape.rank is None:
             return Shape(None)
         else:
-            return self._shape[index]
+            return to_ivy(self._shape[index])
 
     def as_dimension(self):
         if isinstance(self._shape, Shape):
-            return self._shape
+            return to_ivy(self._shape)
         else:
             return Shape(self._shape)
 

ivy/_version.py

@@ -1 +1 @@
-__version__ = "0.0.7.5"
+__version__ = "0.0.8.0"

ivy/func_wrapper.py

@@ -137,11 +137,12 @@ def cross_caster(intersect):
     dtype = ""
     valid_float = sorted(ivy.valid_float_dtypes)
     valid_int = sorted(ivy.valid_int_dtypes)
+    valid_bool = [ivy.bool]
     intersect = sorted(intersect)
     if set(valid_int).issubset(intersect):
         # make dtype equal to default float
         dtype = ivy.default_float_dtype()
-    elif set(valid_float).issubset(intersect):
+    elif set(valid_float).issubset(intersect) or set(valid_bool).issubset(intersect):
         # make dtype equal to default int
         dtype = ivy.default_int_dtype()
 

ivy/functional/backends/jax/__init__.py

@@ -102,7 +102,7 @@ def _array_unflatten(aux_data, children):
 
 # update these to add new dtypes
 valid_dtypes = {
-    "0.4.24 and below": (
+    "0.4.25 and below": (
         ivy.int8,
         ivy.int16,
         ivy.int32,
@@ -121,7 +121,7 @@ def _array_unflatten(aux_data, children):
     )
 }
 valid_numeric_dtypes = {
-    "0.4.24 and below": (
+    "0.4.25 and below": (
         ivy.int8,
         ivy.int16,
         ivy.int32,
@@ -140,7 +140,7 @@ def _array_unflatten(aux_data, children):
 }
 
 valid_int_dtypes = {
-    "0.4.24 and below": (
+    "0.4.25 and below": (
         ivy.int8,
         ivy.int16,
         ivy.int32,
@@ -153,12 +153,12 @@ def _array_unflatten(aux_data, children):
 }
 
 valid_uint_dtypes = {
-    "0.4.24 and below": (ivy.uint8, ivy.uint16, ivy.uint32, ivy.uint64)
+    "0.4.25 and below": (ivy.uint8, ivy.uint16, ivy.uint32, ivy.uint64)
 }
 valid_float_dtypes = {
-    "0.4.24 and below": (ivy.bfloat16, ivy.float16, ivy.float32, ivy.float64)
+    "0.4.25 and below": (ivy.bfloat16, ivy.float16, ivy.float32, ivy.float64)
 }
-valid_complex_dtypes = {"0.4.24 and below": (ivy.complex64, ivy.complex128)}
+valid_complex_dtypes = {"0.4.25 and below": (ivy.complex64, ivy.complex128)}
 
 
 # leave these untouched
@@ -173,12 +173,12 @@ def _array_unflatten(aux_data, children):
 # invalid data types
 
 # update these to add new dtypes
-invalid_dtypes = {"0.4.24 and below": ()}
-invalid_numeric_dtypes = {"0.4.24 and below": ()}
-invalid_int_dtypes = {"0.4.24 and below": ()}
-invalid_float_dtypes = {"0.4.24 and below": ()}
-invalid_uint_dtypes = {"0.4.24 and below": ()}
-invalid_complex_dtypes = {"0.4.24 and below": ()}
+invalid_dtypes = {"0.4.25 and below": ()}
+invalid_numeric_dtypes = {"0.4.25 and below": ()}
+invalid_int_dtypes = {"0.4.25 and below": ()}
+invalid_float_dtypes = {"0.4.25 and below": ()}
+invalid_uint_dtypes = {"0.4.25 and below": ()}
+invalid_complex_dtypes = {"0.4.25 and below": ()}
 
 # leave these untouched
 invalid_dtypes = _dtype_from_version(invalid_dtypes, backend_version)

ivy/functional/backends/paddle/experimental/layers.py

@@ -469,7 +469,35 @@ def interpolate(
     antialias: Optional[bool] = False,
     out: Optional[paddle.Tensor] = None,
 ):
-    raise IvyNotImplementedException()
+    if mode not in ["linear", "bilinear", "bicubic", "trilinear"]:
+        align_corners = None
+    return paddle.nn.functional.interpolate(
+        x,
+        size=size,
+        scale_factor=scale_factor,
+        mode=mode,
+        align_corners=align_corners,
+    )
+
+
+interpolate.partial_mixed_handler = (
+    lambda *args, **kwargs: kwargs.get("mode", "linear")
+    not in [
+        "tf_area",
+        "nd",
+        "tf_bicubic",
+        "mitchellcubic",
+        "lanczos3",
+        "lanczos5",
+        "gaussian",
+    ]
+    and (
+        kwargs.get("mode", "linear") in ["linear", "bilinear", "bicubic", "trilinear"]
+        or not kwargs.get("align_corners", False)
+    )
+    and not kwargs.get("antialias", False)
+    and not kwargs.get("recompute_scale_factor", False)
+)
 
 
 def adaptive_max_pool2d(

ivy/functional/backends/paddle/experimental/manipulation.py

@@ -598,6 +598,14 @@ def expand(
     copy: Optional[bool] = None,
     out: Optional[paddle.Tensor] = None,
 ) -> paddle.Tensor:
+    shape = list(shape)
+    n_extra_dims = len(shape) - x.ndim
+    if n_extra_dims > 0:
+        with ivy.ArrayMode(False):
+            x = paddle_backend.expand_dims(x, tuple(range(n_extra_dims)))
+    for i, dim in enumerate(shape):
+        if dim < 0:
+            shape[i] = x.shape[i]
     return paddle_backend.broadcast_to(x, shape)
 
 

ivy/functional/backends/tensorflow/creation.py

@@ -52,6 +52,15 @@ def arange(
             stop = float(start)
         else:
             stop = start
+
+    # convert builtin types to tf scalars, as is expected by tf.range
+    if isinstance(start, (float, int)):
+        start = tf.constant(start)
+    if isinstance(stop, (float, int)):
+        stop = tf.constant(stop)
+    if isinstance(step, (float, int)):
+        step = tf.constant(step)
+
     if dtype is None:
         if isinstance(start, int) and isinstance(stop, int) and isinstance(step, int):
             return tf.cast(tf.range(start, stop, delta=step, dtype=tf.int64), tf.int32)

ivy/functional/backends/tensorflow/elementwise.py

@@ -706,6 +706,7 @@ def sinh(
     return tf.sinh(x)
 
 
+@with_unsupported_dtypes({"2.15.0 and below": ("integer",)}, backend_version)
 def sqrt(
     x: Union[tf.Tensor, tf.Variable],
     /,
@@ -740,6 +741,7 @@ def subtract(
     return tf.subtract(x1, x2)
 
 
+@with_unsupported_dtypes({"2.15.0 and below": ("integer",)}, backend_version)
 def tan(
     x: Union[tf.Tensor, tf.Variable],
     /,
@@ -749,6 +751,7 @@ def tan(
     return tf.tan(x)
 
 
+@with_unsupported_dtypes({"2.15.0 and below": ("integer",)}, backend_version)
 def tanh(
     x: Union[tf.Tensor, tf.Variable],
     /,

ivy/functional/backends/tensorflow/experimental/manipulation.py

@@ -294,9 +294,10 @@ def pad(
     **kwargs: Optional[Any],
 ) -> Union[tf.Tensor, tf.Variable]:
     pad_width = _to_tf_padding(pad_width, len(input.shape))
-    if isinstance(constant_values, (tf.Variable, tf.Tensor)):
-        if constant_values.dtype != input.dtype:
-            constant_values = tf.cast(constant_values, input.dtype)
+    if not isinstance(constant_values, (tf.Variable, tf.Tensor)):
+        constant_values = tf.constant(constant_values)
+    if constant_values.dtype != input.dtype:
+        constant_values = tf.cast(constant_values, input.dtype)
     return tf.pad(
         input,
         pad_width,
@@ -578,6 +579,14 @@ def unflatten(
     name: Optional[str] = None,
 ) -> tf.Tensor:
     dim = abs(len(x.shape) + dim) if dim < 0 else dim
+
+    # infer the size of any dimensions that are -1
+    tf_shape = tf.constant(shape)
+    inferred_size = tf.reduce_prod(tf.shape(x)[dim]) // tf.reduce_prod(
+        tf.where(tf_shape != -1, x=shape, y=tf.constant(1))
+    )
+    shape = tf.where(tf_shape != -1, x=shape, y=inferred_size)
+
     res_shape = x.shape[:dim] + tf.TensorShape(shape) + x.shape[dim + 1 :]
     res = tf.reshape(x, res_shape, name)
     return res

ivy/functional/backends/tensorflow/layers.py

@@ -410,7 +410,9 @@ def conv3d_transpose(
     return res
 
 
-@with_unsupported_dtypes({"2.15.0 and below": ("bfloat16", "complex")}, backend_version)
+@with_unsupported_dtypes(
+    {"2.15.0 and below": ("bfloat16", "complex", "integer")}, backend_version
+)
 def conv_general_dilated(
     x: Union[tf.Tensor, tf.Variable],
     filters: Union[tf.Tensor, tf.Variable],