tests for concatenate op

forge/test/operators/pytorch/eltwise_nary/__init__.py

@@ -0,0 +1,3 @@
+# SPDX-FileCopyrightText: © 2024 Tenstorrent AI ULC
+
+# SPDX-License-Identifier: Apache-2.0

forge/test/operators/pytorch/eltwise_nary/test_concatenate.py

@@ -0,0 +1,346 @@
+# SPDX-FileCopyrightText: © 2024 Tenstorrent AI ULC
+
+# SPDX-License-Identifier: Apache-2.0
+#
+# Tests for testing of concatenate operators
+#
+# In this test we test pytorch concatenate operator
+
+# GENERAL OP SUPPORT TEST PLAN:
+# 1. Operand type - any supported type
+# 2. Operand source(s):
+# (+)  2.1 From another op
+#       - Operator -> input
+# (+)  2.2 From DRAM queue
+#       - Operator is first node in network
+#       - Input_queue flag = false
+# (+)  2.3 Const Inputs (const eval pass)
+#       - Operator where all inputs are constants.
+# (+)  2.4 From host
+#       - Input tensor as input of network
+#       - Operator is first node in network
+#       - Input_queue flag = true
+# 3 Operand shapes type(s):
+# (+)  3.1 Full tensor (i.e. full expected shape)
+#       - 3-4 by default P1 (high prioriy)
+#       - 2, 5, ++ include P2 (lower prioriy)
+# (+)  3.2 Tensor reduce on one or more dims to 1
+#       - Vector
+#       - Only one dim is not equal to 1
+# (+)  3.3 Scalar P2
+#       - Create tensor of dimension equal to 0 (tensor from scalar) or just to use scalar as simple value
+# 4. Operand / output size of dimensions (few examples of each, 10 values total)
+# (+)  4.1 Divisible by 32
+# (+)  4.2 Prime numbers
+# (+)  4.3 Very large (thousands, 10s of thousands)
+#       - 100x100, 100x1000
+#       - maybe nightly only
+# (+)  4.4 Extreme ratios between height/width
+#      4.5 ...probably many more interesting combinations here
+# 5. Data format - all supported formats
+# (/)  5.1 Output DF
+# (/)  5.2 Intermediate DF
+# (/)  5.3 Accumulation DF
+# (+)  5.4 Operand DFs
+#       - Fix HiFi4 for math fidelity value
+# (+) 6. Math fidelity - LoFi, HiFi2a, Hifi2b, Hifi3, Hifi4
+#       - Fix fp16b (default) for data format value
+# (/) 7. Special attributes - if applicable.. like approx_mode for Exp, for example
+# (/) 8. Special cases - if applicable
+# 9. Variable number of operands - if applicable
+# (/) Few representative values
+# (/) Reuse inputs for selected operators
+
+
+import pytest
+
+from typing import List, Dict, Type, Optional, Any
+from loguru import logger
+
+import torch
+import forge
+import forge.op
+
+
+from test.operators.utils import InputSourceFlags, VerifyUtils
+from test.operators.utils import ShapeUtils
+from test.operators.utils import InputSource
+from test.operators.utils import TestVector
+from test.operators.utils import TestPlan
+from test.operators.utils import FailingReasons
+from test.operators.utils.compat import TestDevice
+from test.operators.utils import TestCollection
+from test.operators.utils import TestCollectionCommon
+
+
+class ModelFromAnotherOp(torch.nn.Module):
+
+    model_name = "model_op_src_from_another_op"
+
+    def __init__(self, operator, opname, shape, number_of_operands, kwargs):
+        super(ModelFromAnotherOp, self).__init__()
+        self.testname = "Concatenate_pytorch_operator_" + opname + "_test_op_src_from_another_op"
+        self.operator = operator
+        self.opname = opname
+        self.shape = shape
+        self.number_of_operands = number_of_operands
+        self.kwargs = kwargs
+
+    def forward(self, *x: torch.Tensor):
+        # we use Add operator to create one operands which is input for the concatenate operator
+        add_tuple = ()
+        for i in range(len(x)):
+            add_tuple += (torch.add(x[i], x[i]),)
+        output = self.operator(add_tuple, **self.kwargs)
+        return output
+
+
+class ModelDirect(torch.nn.Module):
+
+    model_name = "model_op_src_from_host"
+
+    def __init__(self, operator, opname, shape, number_of_operands, kwargs):
+        super(ModelDirect, self).__init__()
+        self.testname = "Concatenate_pytorch_operator_" + opname + "_test_op_src_from_host"
+        self.operator = operator
+        self.opname = opname
+        self.shape = shape
+        self.number_of_operands = number_of_operands
+        self.kwargs = kwargs
+
+    def forward(self, *x: torch.Tensor):
+        output = self.operator(x, **self.kwargs)
+        return output
+
+
+class ModelConstEvalPass(torch.nn.Module):
+
+    model_name = "model_op_src_const_eval_pass"
+
+    def __init__(self, operator, opname, shape, number_of_operands, kwargs):
+        super(ModelConstEvalPass, self).__init__()
+        self.testname = "Concatenate_pytorch_operator_" + opname + "_test_op_src_const_eval_pass"
+        self.operator = operator
+        self.opname = opname
+        self.shape = shape
+        self.kwargs = kwargs
+
+        self.constants = tuple()
+        for _ in range(number_of_operands):
+            self.constants += (torch.rand(*self.shape) - 0.5,)
+
+    def forward(self, *x: torch.Tensor):
+        v1 = self.operator(self.constants, **self.kwargs)
+        v2 = torch.add(x[0], x[1])
+        for i in range(2, len(x)):
+            v3 = torch.add(v2, x[i])
+            v2 = v3
+        # add consume inputs
+        add = torch.add(v1, v2)
+        return add
+
+
+class TestVerification:
+
+    MODEL_TYPES = {
+        InputSource.FROM_ANOTHER_OP: ModelFromAnotherOp,
+        InputSource.FROM_HOST: ModelDirect,
+        InputSource.FROM_DRAM_QUEUE: ModelDirect,
+        InputSource.CONST_EVAL_PASS: ModelConstEvalPass,
+    }
+
+    @classmethod
+    def verify(
+        cls,
+        test_device: TestDevice,
+        test_vector: TestVector,
+        input_params: List[Dict] = [],
+        warm_reset: bool = False,
+    ):
+        """Common verification function for all tests"""
+
+        input_source_flag: InputSourceFlags = None
+        if test_vector.input_source in (InputSource.FROM_DRAM_QUEUE,):
+            input_source_flag = InputSourceFlags.FROM_DRAM
+
+        operator = getattr(torch, test_vector.operator)
+
+        kwargs = test_vector.kwargs if test_vector.kwargs else {}
+
+        model_type = cls.MODEL_TYPES[test_vector.input_source]
+        pytorch_model = model_type(
+            operator=operator,
+            opname=test_vector.operator,
+            shape=test_vector.input_shape,
+            number_of_operands=test_vector.number_of_operands,
+            kwargs=kwargs,
+        )
+
+        dim = test_vector.kwargs["dim"]
+        match test_vector.input_source:
+            case InputSource.CONST_EVAL_PASS:
+                input_shape = list(test_vector.input_shape)
+                input_shape[dim] = input_shape[dim] * test_vector.number_of_operands
+                input_shapes = tuple([input_shape for _ in range(test_vector.number_of_operands)])
+            case _:
+                input_shapes = tuple([test_vector.input_shape for _ in range(test_vector.number_of_operands)])
+
+        logger.trace(f"***input_shapes: {input_shapes}")
+
+        VerifyUtils.verify(
+            model=pytorch_model,
+            test_device=test_device,
+            input_shapes=input_shapes,
+            input_params=input_params,
+            input_source_flag=input_source_flag,
+            dev_data_format=test_vector.dev_data_format,
+            math_fidelity=test_vector.math_fidelity,
+            pcc=test_vector.pcc,
+            warm_reset=warm_reset,
+        )
+
+
+class TestParamsData:
+
+    __test__ = False  # Avoid collecting TestParamsData as a pytest test
+
+    test_plan: TestPlan = None
+
+    @classmethod
+    def generate_kwargs(cls, test_vector: TestVector):
+        shape_with_kwargs = cls.extend_shape_with_dims(test_vector.input_shape)
+        kwarg_list = []
+        for item in shape_with_kwargs:
+            kwargs = {}
+            kwargs["dim"] = item[1]
+            kwarg_list.append(kwargs)
+        return kwarg_list
+
+    @classmethod
+    def extend_shape_with_dims(cls, shape):
+        shape_with_dims = list()
+        for dim in list(range(0, len(shape), 1)):
+            shape_with_dims.append((shape, dim))
+        return shape_with_dims
+
+
+class TestCollectionData:
+
+    __test__ = False  # Avoid collecting TestCollectionData as a pytest test
+
+    all = TestCollection(
+        operators=[
+            "concatenate",  # 00
+        ],
+        input_sources=TestCollectionCommon.all.input_sources,
+        input_shapes=TestCollectionCommon.all.input_shapes,
+        numbers_of_operands=[
+            2,
+            3,
+            7,
+            # 15, # consume too much memory
+        ],
+        dev_data_formats=TestCollectionCommon.all.dev_data_formats,
+        math_fidelities=TestCollectionCommon.all.math_fidelities,
+    )
+
+    single = TestCollection(
+        input_sources=TestCollectionCommon.single.input_sources,
+        input_shapes=TestCollectionCommon.single.input_shapes,
+        numbers_of_operands=[
+            2,
+        ],
+        dev_data_formats=TestCollectionCommon.single.dev_data_formats,
+        math_fidelities=TestCollectionCommon.single.math_fidelities,
+    )
+
+
+TestParamsData.test_plan = TestPlan(
+    verify=lambda test_device, test_vector: TestVerification.verify(
+        test_device,
+        test_vector,
+    ),
+    collections=[
+        # Test plan:
+        # 2. Operand source(s):
+        # 3. Operand shapes type(s):
+        # 4. Operand / output size of dimensions
+        TestCollection(
+            operators=TestCollectionData.all.operators,
+            input_sources=TestCollectionData.all.input_sources,
+            input_shapes=TestCollectionData.all.input_shapes,
+            numbers_of_operands=TestCollectionData.all.numbers_of_operands,
+            kwargs=lambda test_vector: TestParamsData.generate_kwargs(test_vector),
+        ),
+        # Test plan:
+        # 5. Data format
+        TestCollection(
+            operators=TestCollectionData.all.operators,
+            input_sources=TestCollectionData.single.input_sources,
+            input_shapes=TestCollectionData.single.input_shapes,
+            numbers_of_operands=TestCollectionData.single.numbers_of_operands,
+            kwargs=lambda test_vector: TestParamsData.generate_kwargs(test_vector),
+            dev_data_formats=[
+                item
+                for item in TestCollectionData.all.dev_data_formats
+                if item not in TestCollectionData.single.dev_data_formats
+            ],
+            math_fidelities=TestCollectionData.single.math_fidelities,
+        ),
+        # Test plan:
+        # 6. Math fidelity
+        TestCollection(
+            operators=TestCollectionData.all.operators,
+            input_sources=TestCollectionData.single.input_sources,
+            input_shapes=TestCollectionData.single.input_shapes,
+            numbers_of_operands=TestCollectionData.single.numbers_of_operands,
+            kwargs=lambda test_vector: TestParamsData.generate_kwargs(test_vector),
+            dev_data_formats=TestCollectionData.single.dev_data_formats,
+            math_fidelities=TestCollectionData.all.math_fidelities,
+        ),
+    ],
+    failing_rules=[
+        # Skip all tests with input shapes with 2 dimensions
+        TestCollection(
+            criteria=lambda test_vector: len(test_vector.input_shape) == 2,
+            skip_reason=FailingReasons.NOT_IMPLEMENTED,
+        ),
+        TestCollection(
+            input_sources=[
+                InputSource.CONST_EVAL_PASS,
+            ],
+            input_shapes=[
+                (10, 10, 10000, 1),
+            ],
+            numbers_of_operands=[
+                7,
+            ],
+            kwargs=[
+                {"dim": 0},
+                {"dim": 1},
+                {"dim": 2},
+            ],
+            failing_reason=FailingReasons.ALLOCATION_FAILED,
+        ),
+        TestCollection(
+            criteria=lambda test_vector: test_vector.input_source != InputSource.CONST_EVAL_PASS
+            and len(test_vector.input_shape) == 3
+            and (test_vector.kwargs["dim"] == 1 or test_vector.kwargs["dim"] == 2)
+            and test_vector.input_shape[test_vector.kwargs["dim"]] % 32 != 0,
+            failing_reason=FailingReasons.NOT_IMPLEMENTED,
+        ),
+        TestCollection(
+            criteria=lambda test_vector: test_vector.input_source != InputSource.CONST_EVAL_PASS
+            and len(test_vector.input_shape) == 4
+            and (test_vector.kwargs["dim"] == 2 or test_vector.kwargs["dim"] == 3)
+            and test_vector.input_shape[test_vector.kwargs["dim"]] % 32 != 0,
+            failing_reason=FailingReasons.NOT_IMPLEMENTED,
+        ),
+    ],
+)
+
+
+def get_test_plans() -> List[TestPlan]:
+    return [
+        TestParamsData.test_plan,
+    ]

forge/test/operators/utils/failing_reasons.py

@@ -58,6 +58,10 @@ class FailingReasons:
 
     UNSUPPORTED_INPUT_SOURCE = "Unsupported input source"
 
+    # INFO     | forge.compiled_graph_state:__call__:247  Running model forward on device...
+    # Always | FATAL    | Out of Memory: Not enough space to allocate 896204800 B DRAM buffer across 12 banks, where each bank needs to store 74686464 B
+    ALLOCATION_FAILED = "Out of Memory"
+
 
 class FailingReasonsValidation:
     @classmethod
@@ -107,6 +111,14 @@ def validate_exception_message(
             lambda ex: isinstance(ex, RuntimeError) and " not implemented for " in f"{ex}",
             lambda ex: isinstance(ex, AssertionError)
             and f"{ex}" == "Encountered unsupported op types. Check error logs for more details",
+            lambda ex: isinstance(ex, RuntimeError)
+            and "!in_ref.get_shape().has_tile_padding(this->dim)"  # tt-forge-fe/third_party/tt-mlir/third_party/tt-metal/src/tt-metal/ttnn/cpp/ttnn/operations/data_movement/concat/device/concat_device_operation.cpp:47: !in_ref.get_shape().has_tile_padding(this->dim)
+            in f"{ex}",
+        ],
+        FailingReasons.ALLOCATION_FAILED: [
+            lambda ex: isinstance(ex, RuntimeError)
+            and "tt-forge-fe/third_party/tt-mlir/third_party/tt-metal/src/tt-metal/tt_metal/impl/allocator/allocator.cpp:143"
+            in f"{ex}",
         ],
     }