From ca57b45071d95ed08d498e23b102387524467c9e Mon Sep 17 00:00:00 2001 From: Patrick von Platen Date: Thu, 24 Feb 2022 19:08:54 +0100 Subject: [PATCH] [Unispeech] Fix slow tests (#15818) * remove soundfile old way of loading audio * Adapt slow test --- tests/unispeech/test_modeling_unispeech.py | 20 ++++--------- .../test_modeling_unispeech_sat.py | 28 +++++++------------ 2 files changed, 16 insertions(+), 32 deletions(-) diff --git a/tests/unispeech/test_modeling_unispeech.py b/tests/unispeech/test_modeling_unispeech.py index 03e1d6dacd..b118120e58 100644 --- a/tests/unispeech/test_modeling_unispeech.py +++ b/tests/unispeech/test_modeling_unispeech.py @@ -538,21 +538,13 @@ class UniSpeechRobustModelTest(ModelTesterMixin, unittest.TestCase): @slow class UniSpeechModelIntegrationTest(unittest.TestCase): def _load_datasamples(self, num_samples): - import soundfile as sf + ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation") + # automatic decoding with librispeech + speech_samples = ds.sort("id").filter( + lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)] + )[:num_samples]["audio"] - ids = [f"1272-141231-000{i}" for i in range(num_samples)] - - # map files to raw - def map_to_array(batch): - speech, _ = sf.read(batch["file"]) - batch["speech"] = speech - return batch - - ds = load_dataset("patrickvonplaten/librispeech_asr_dummy", "clean", split="validation") - - ds = ds.filter(lambda x: x["id"] in ids).sort("id").map(map_to_array) - - return ds["speech"][:num_samples] + return [x["array"] for x in speech_samples] def _load_superb(self, task, num_samples): diff --git a/tests/unispeech_sat/test_modeling_unispeech_sat.py b/tests/unispeech_sat/test_modeling_unispeech_sat.py index 46f862806e..bae49c7b4c 100644 --- a/tests/unispeech_sat/test_modeling_unispeech_sat.py +++ b/tests/unispeech_sat/test_modeling_unispeech_sat.py @@ -800,21 +800,13 @@ class UniSpeechSatRobustModelTest(ModelTesterMixin, unittest.TestCase): @slow class UniSpeechSatModelIntegrationTest(unittest.TestCase): def _load_datasamples(self, num_samples): - import soundfile as sf + ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation") + # automatic decoding with librispeech + speech_samples = ds.sort("id").filter( + lambda x: x["id"] in [f"1272-141231-000{i}" for i in range(num_samples)] + )[:num_samples]["audio"] - ids = [f"1272-141231-000{i}" for i in range(num_samples)] - - # map files to raw - def map_to_array(batch): - speech, _ = sf.read(batch["file"]) - batch["speech"] = speech - return batch - - ds = load_dataset("patrickvonplaten/librispeech_asr_dummy", "clean", split="validation") - - ds = ds.filter(lambda x: x["id"] in ids).sort("id").map(map_to_array) - - return ds["speech"][:num_samples] + return [x["array"] for x in speech_samples] def _load_superb(self, task, num_samples): ds = load_dataset("anton-l/superb_dummy", task, split="test") @@ -865,10 +857,10 @@ class UniSpeechSatModelIntegrationTest(unittest.TestCase): # fmt: off expected_hidden_states_slice = torch.tensor( - [[[-0.1172, -0.0797], - [-0.0012, 0.0213]], - [[-0.1225, -0.1277], - [-0.0668, -0.0585]]], + [[[-0.1192, -0.0825], + [-0.0012, 0.0235]], + [[-0.1240, -0.1332], + [-0.0658, -0.0565]]], device=torch_device, ) # fmt: on