Merge pull request #47 from espnet/add_tts_models

Add JSUT VITS model

.github/workflows/model_test.yaml

@@ -9,14 +9,8 @@ on:
       - master
 
 jobs:
-  check_skip:
-    runs-on: ubuntu-18.04
-    if: "! contains(github.event.head_commit.message, '[ci skip]')"
-    steps:
-      - run: echo "${{ github.event.head_commit.message }}"
   test:
     runs-on: ${{ matrix.os }}
-    needs: check_skip
     strategy:
       max-parallel: 20
       matrix:

.github/workflows/unittest.yaml

@@ -9,20 +9,14 @@ on:
       - master
 
 jobs:
-  check_skip:
-    runs-on: ubuntu-18.04
-    if: "! contains(github.event.head_commit.message, '[ci skip]')"
-    steps:
-      - run: echo "${{ github.event.head_commit.message }}"
   linter_and_test:
     runs-on: ${{ matrix.os }}
-    needs: check_skip
     strategy:
       max-parallel: 20
       matrix:
         # os: [ubuntu-16.04, ubuntu-18.04]
-        os: [ubuntu-18.04]
-        python-version: [3.8]
+        os: [ubuntu-20.04]
+        python-version: [3.9]
         espnet-version: ["espnet", "git+https://github.com/espnet/espnet.git"]
     steps:
       - uses: actions/checkout@master

README.md

@@ -24,6 +24,7 @@ pip install espnet_model_zoo
 
 ## Python API for inference
 `model_name` in the following section should be `huggingface_id` or one of the tags in the [table.csv](espnet_model_zoo/table.csv).
+Or you can directly provide zenodo URL (e.g., `https://zenodo.org/record/xxxxxxx/files/hogehoge.zip?download=1`).
 
 ### ASR
 

ci/test_model.py

@@ -23,8 +23,12 @@ def _tts(model_name):
     inputs = {"text": "foo"}
     if text2speech.use_speech:
         inputs["speech"] = np.zeros((10000,), dtype=np.float32)
-    if text2speech.tts.spk_embed_dim is not None:
+    if text2speech.use_spembs:
         inputs["spembs"] = np.zeros((text2speech.tts.spk_embed_dim,), dtype=np.float32)
+    if text2speech.use_sids:
+        inputs["sids"] = np.ones((1,), dtype=np.int64)
+    if text2speech.use_lids:
+        inputs["lids"] = np.ones((1,), dtype=np.int64)
     text2speech(**inputs)
 
 

espnet_model_zoo/table.csv

@@ -115,3 +115,11 @@ su_openslr36,asr,su_openslr36,https://zenodo.org/record/5090135/files/asr_train_
 ksponspeech,asr,Yushi Ueda/ksponspeech_asr_train_asr_conformer8_n_fft512_hop_length256_raw_kr_bpe2309_valid.acc.best,https://zenodo.org/record/5154341/files/asr_train_asr_conformer8_n_fft512_hop_length256_raw_kr_bpe2309_valid.acc.best.zip?download=1,16000,kr,,1.8.1,0.10.0,538393c,true
 librispeech,asr,byan/librispeech_asr_train_asr_conformer_raw_bpe_batch_bins30000000_accum_grad3_optim_conflr0.001_sp,https://huggingface.co/,16000,en,,,,,true
 mini_librispeech,diar,Yushi Ueda/mini_librispeech_diar_train_diar_raw_max_epoch20_valid.acc.best,https://zenodo.org/record/5264020/files/diar_train_diar_raw_max_epoch20_valid.acc.best.zip?download=1,8000,,,1.9.1,0.10.2,,true
+jsut,tts,kan-bayashi/jsut_tts_train_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_train.total_count.ave,https://zenodo.org/record/5414980/files/tts_train_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_train.total_count.ave.zip?download=1,22050,jp,female,1.7.1,0.10.3a1,dee65,true
+jsut,tts,kan-bayashi/jsut_tts_train_full_band_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_train.total_count.ave,https://zenodo.org/record/5431984/files/tts_train_full_band_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_train.total_count.ave.zip?download=1,44100,jp,female,1.7.1,0.10.3a1,dee65,true
+jsut,tts,kan-bayashi/jsut_vits_accent_with_pause,https://zenodo.org/record/5414980/files/tts_train_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_train.total_count.ave.zip?download=1,22050,jp,female,1.7.1,0.10.3a1,dee65,true
+jsut,tts,kan-bayashi/jsut_full_band_vits_accent_with_pause,https://zenodo.org/record/5431984/files/tts_train_full_band_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_train.total_count.ave.zip?download=1,44100,jp,female,1.7.1,0.10.3a1,dee65,true
+jvs,tts,kan-bayashi/jvs_tts_finetune_jvs001_jsut_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_latest,https://zenodo.org/record/5432540/files/tts_finetune_jvs001_jsut_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_latest.zip?download=1,22050,jp,male,1.7.1,0.10.3a1,dee65,true
+jvs,tts,kan-bayashi/jvs_tts_finetune_jvs010_jsut_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_latest,https://zenodo.org/record/5432566/files/tts_finetune_jvs010_jsut_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_latest.zip?download=1,22050,jp,female,1.7.1,0.10.3a1,dee65,true
+jvs,tts,kan-bayashi/jvs_jvs001_vits_accent_with_pause,https://zenodo.org/record/5432540/files/tts_finetune_jvs001_jsut_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_latest.zip?download=1,22050,jp,male,1.7.1,0.10.3a1,dee65,true
+jvs,tts,kan-bayashi/jvs_jvs010_vits_accent_with_pause,https://zenodo.org/record/5432566/files/tts_finetune_jvs010_jsut_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause_latest.zip?download=1,22050,jp,female,1.7.1,0.10.3a1,dee65,true

setup.py

@@ -37,7 +37,7 @@
 dirname = os.path.dirname(__file__)
 setup(
     name="espnet_model_zoo",
-    version="0.1.0",
+    version="0.1.1a1",
     url="http://github.com/espnet/espnet_model_zoo",
     description="ESPnet Model Zoo",
     long_description=open(os.path.join(dirname, "README.md"), encoding="utf-8").read(),
@@ -63,6 +63,7 @@
         "Programming Language :: Python :: 3.6",
         "Programming Language :: Python :: 3.7",
         "Programming Language :: Python :: 3.8",
+        "Programming Language :: Python :: 3.9",
         "Development Status :: 5 - Production/Stable",
         "Intended Audience :: Science/Research",
         "Operating System :: POSIX :: Linux",