Closed gptq closed 9 months ago
ok
已适配,目前mindspore不支持weight_norm,因此暂不支持对标训练
$ python -c 'from mindnlp.transformers import SeamlessM4Tv2Model;model = SeamlessM4Tv2Model.from_pretrained("/home/fuq/ckpt/seamless-m4t-v2-large", from_pt=True)' Segmentation fault (core dumped) $ pytest -c pytest.ini tests/ut/transformers/models/seamless_m4t_v2/test_modeling_seamless_m4t_v2.py::SeamlessM4Tv2ModelIntegrationTest::test_to_rus_speech =========================================================================== test session starts ===========================================================================platform linux -- Python 3.7.5, pytest-7.4.3, pluggy-1.2.0 rootdir: /home/fuq/mindnlp configfile: pytest.ini plugins: anyio-3.7.1 collected 1 item
tests/ut/transformers/models/seamless_m4t_v2/test_modeling_seamless_m4t_v2.py::SeamlessM4Tv2ModelIntegrationTest::test_to_rus_speech Fatal Python error: Segmentation fault
OS:ubuntu 18.04.5 server CPU:X86_64 NPU:910B 32G MS:2.2.0 CANN:7.0.RC1.beta1
$ pytest -c pytest.ini tests/ut/transformers/models/seamless_m4t_v2/test_modeling_seamless_m4t_v2.py::SeamlessM4Tv2ModelIntegrationTest::test_to_rus_speech =========================================================================== test session starts ===========================================================================platform linux -- Python 3.9.18, pytest-7.4.3, pluggy-1.3.0 rootdir: /home/fuq/mindnlp configfile: pytest.ini collected 1 item
tests/ut/transformers/models/seamless_m4t_v2/test_modeling_seamless_m4t_v2.py::SeamlessM4Tv2ModelIntegrationTest::test_to_rus_speech Fatal Python error: Segmentation fault
https://huggingface.co/facebook/seamless-m4t-v2-large facebook推出业界领先的all-in-one翻译模型,建议适配。