Arm backend: Upstreamed NFRU model testing. (pytorch#22887)

Change-Id: I6f1a93ff1b7c1b4064aba7f4531857c5272bd2b9

cc @digantdesai @freddan80 @per @zingo @oscarandersson8218 @mansnils
@Sebastian-Larsson @robell @rascani

Signed-off-by: Michiel Olieslagers <michiel.olieslagers@arm.com>

Author: Michiel-Olieslagers

backends/arm/test/models/test_nfru.py

@@ -4,14 +4,17 @@
 # LICENSE file in the root directory of this source tree.
 
 import os
+from typing import Tuple
 
 os.environ.setdefault("HF_HUB_DISABLE_XET", "1")
 
 import pytest
+
 import torch
 
 from executorch.backends.arm.test import common
 from executorch.backends.arm.test.tester.test_pipeline import (
+    TosaPipelineFP,
     TosaPipelineINT,
     VgfPipeline,
 )
@@ -20,6 +23,7 @@
     NFRUAutoEncoder,
 )
 
+input_t = Tuple[torch.Tensor]  # Input x
 
 _RELEASE_REFS = (
     os.environ.get("GITHUB_REF", ""),
@@ -38,6 +42,7 @@
 
 
 def nfru() -> NFRUAutoEncoder:
+    """Get an instance of NFRU with FP32 weights loaded."""
     weights = hf_hub_download(  # nosec B615
         repo_id="Arm/neural-frame-rate-upscaling",
         filename="nfru_v1_fp32.pt",
@@ -60,6 +65,85 @@ def nfru() -> NFRUAutoEncoder:
     return model
 
 
+def example_inputs(memory_format: torch.memory_format = torch.channels_last):
+    return (torch.randn((1, 16, 270, 480)).to(memory_format=memory_format),)
+
+
+def test_nfru_tosa_FP():
+    pipeline = TosaPipelineFP[input_t](
+        nfru().eval(),
+        example_inputs(),
+        aten_op=[],
+        exir_op=[],
+    )
+    pipeline.run()
+
+
+def test_nfru_tosa_INT():
+    pipeline = TosaPipelineINT[input_t](
+        nfru().eval(),
+        example_inputs(),
+        aten_op=[],
+        exir_op=[],
+        atol=0.2,
+    )
+    pipeline.run()
+
+
+def test_nfru_tosa_INT_a16w8():
+    pipeline = TosaPipelineINT[input_t](
+        nfru().eval(),
+        example_inputs(),
+        aten_op=[],
+        exir_op=[],
+        tosa_extensions=["int16"],
+        atol=0.1,
+    )
+    pipeline.run()
+
+
+@common.SkipIfNoModelConverter
+def test_nfru_vgf_no_quant():
+    pipeline = VgfPipeline[input_t](
+        nfru().eval(),
+        example_inputs(),
+        aten_op=[],
+        exir_op=[],
+        tosa_version="TOSA-1.0+FP",
+        quantize=False,
+    )
+    pipeline.run()
+
+
+@common.SkipIfNoModelConverter
+def test_nfru_vgf_quant():
+    pipeline = VgfPipeline[input_t](
+        nfru().eval(),
+        example_inputs(),
+        aten_op=[],
+        exir_op=[],
+        tosa_version="TOSA-1.0+INT",
+        symmetric_io_quantization=True,
+        atol=0.2,
+    )
+    pipeline.run()
+
+
+@common.SkipIfNoModelConverter
+def test_nfru_vgf_quant_a16w8():
+    pipeline = VgfPipeline[input_t](
+        nfru().eval(),
+        example_inputs(),
+        aten_op=[],
+        exir_op=[],
+        tosa_version="TOSA-1.0+INT",
+        tosa_extensions=["int16"],
+        symmetric_io_quantization=True,
+        atol=0.2,
+    )
+    pipeline.run()
+
+
 def test_nfru_qat_tosa_INT() -> None:
     pipeline = TosaPipelineINT(
         nfru(),