SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Fix seamless TTS generate (#34968)

Browse Source 
* fix seamless tts generate

* apply same fix for v2

* [run-slow] seamless_m4t, seamless_m4t_v2

* remove TODO

* [run-slow] seamless_m4t, seamless_m4t_v2

* [run-slow] seamless_m4t, seamless_m4t_v2

* ignore failing test on multigpus

* [run-slow] seamless_m4t, seamless_m4t_v2

* [run-slow] seamless_m4t, seamless_m4t_v2
This commit is contained in:
Yoach Lacombe
2024-12-11 15:38:42 +01:00
committed by GitHub
parent 33c12e4d80
commit 6181c6b095
4 changed files with 9 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/pipelines/test_pipelines_text_to_audio.py
View File

@@ -27,7 +27,6 @@ from transformers.testing_utils import (
require_torch,
require_torch_accelerator,
require_torch_or_tf,
run_test_using_subprocess,
slow,
torch_device,
)
@@ -67,10 +66,8 @@ class TextToAudioPipelineTests(unittest.TestCase):
audio = [output["audio"] for output in outputs]
self.assertEqual([ANY(np.ndarray), ANY(np.ndarray)], audio)
# TODO: @ylacombe: `SeamlessM4TForTextToSpeech.generate` has issue with `generation_config`. See issue #34811
@slow
@require_torch
@run_test_using_subprocess
def test_medium_seamless_m4t_pt(self):
speech_generator = pipeline(task="text-to-audio", model="facebook/hf-seamless-m4t-medium", framework="pt")