test: update test voices and notebook (#387)

voicepaw · Apr 18, 2023 · a798ed6 · a798ed6
1 parent d08b155
commit a798ed6
Show file tree

Hide file tree

Showing 25 changed files with 26 additions and 5 deletions.
diff --git a/.gitignore b/.gitignore
@@ -141,7 +141,7 @@ cython_debug/
 
 # additional files
 tests/**/*.wav
-!tests/dataset_raw/34j/**/*.wav
+!tests/dataset_raw/test/**/*.wav
 tests/**/*.npy
 tests/**/*.pt
 tests/**/*.txt

diff --git a/notebooks/so-vits-svc-fork-4.0.ipynb b/notebooks/so-vits-svc-fork-4.0.ipynb
@@ -194,8 +194,22 @@
    "source": [
     "#@title Get the author's voice as a source\n",
     "import random\n",
-    "NAME = str(random.randint(1, 100))\n",
-    "!wget -N \"https://github.com/34j/34j/raw/main/jvs-parallel100/{NAME}.wav\"\n",
+    "NAME = str(random.randint(1, 49))\n",
+    "TYPE = \"fsd50k\" #@param [\"\", \"digit\", \"dog\", \"fsd50k\"]\n",
+    "CUSTOM_FILEPATH = \"\" #@param {type: \"string\"}\n",
+    "if CUSTOM_FILEPATH != \"\":\n",
+    "    NAME = CUSTOM_FILEPATH\n",
+    "else:\n",
+    "    # it is extremely difficult to find a voice that can download from the internet directly\n",
+    "    if TYPE == \"dog\":\n",
+    "        !wget -N f\"https://huggingface.co/datasets/437aewuh/dog-dataset/resolve/main/dogs/dogs_{NAME:.0000}.wav\" -O {NAME}.wav\n",
+    "    elif TYPE == \"digit\":\n",
+    "        # george, jackson, lucas, nicolas, ...\n",
+    "        !wget -N f\"https://github.com/Jakobovski/free-spoken-digit-dataset/raw/master/recordings/0_george_{NAME}.wav\" -O {NAME}.wav\n",
+    "    elif TYPE == \"fsd50k\":\n",
+    "        !wget -N f\"https://huggingface.co/datasets/Fhrozen/FSD50k/blob/main/clips/dev/{10000+int(NAME)}.wav\" -O {NAME}.wav\n",
+    "    else:\n",
+    "        !wget -N f\"https://zunko.jp/sozai/utau/voice_{\"kiritan\" if NAME < 25 else \"itako\"}{NAME % 5 + 1}.wav\" -O {NAME}.wav\n",
     "from IPython.display import Audio, display\n",
     "display(Audio(f\"{NAME}.wav\"))"
    ]
@@ -209,7 +223,6 @@
     "#@title Use trained model\n",
     "#@markdown **Put your .wav file in `so-vits-svc-fork/audio` directory**\n",
     "from IPython.display import Audio, display\n",
-    "NAME = 'your audio file name' #@param {type: \"string\"}\n",
     "!svc infer drive/MyDrive/so-vits-svc-fork/audio/{NAME}.wav -m drive/MyDrive/so-vits-svc-fork/logs/44k/ -c drive/MyDrive/so-vits-svc-fork/logs/44k/config.json\n",
     "display(Audio(f\"drive/MyDrive/so-vits-svc-fork/audio/{NAME}.out.wav\", autoplay=True))"
    ]
@@ -286,7 +299,15 @@
    "name": "python3"
   },
   "language_info": {
-   "name": "python"
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3"
   }
  },
  "nbformat": 4,

diff --git a/tests/dataset_raw/34j/nested/1.wav b/tests/dataset_raw/34j/nested/1.wav
diff --git a/tests/dataset_raw/34j/nested/10.wav b/tests/dataset_raw/34j/nested/10.wav
diff --git a/tests/dataset_raw/34j/nested/2.wav b/tests/dataset_raw/34j/nested/2.wav
diff --git a/tests/dataset_raw/34j/nested/3.wav b/tests/dataset_raw/34j/nested/3.wav
diff --git a/tests/dataset_raw/34j/nested/4.wav b/tests/dataset_raw/34j/nested/4.wav
diff --git a/tests/dataset_raw/34j/nested/5.wav b/tests/dataset_raw/34j/nested/5.wav
diff --git a/tests/dataset_raw/34j/nested/6.wav b/tests/dataset_raw/34j/nested/6.wav
diff --git a/tests/dataset_raw/34j/nested/7.wav b/tests/dataset_raw/34j/nested/7.wav
diff --git a/tests/dataset_raw/34j/nested/8.wav b/tests/dataset_raw/34j/nested/8.wav
diff --git a/tests/dataset_raw/34j/nested/9.wav b/tests/dataset_raw/34j/nested/9.wav
diff --git a/tests/dataset_raw/34j/nested/あ.wav b/tests/dataset_raw/34j/nested/あ.wav
diff --git a/tests/dataset_raw/test/LJ001-0001.wav b/tests/dataset_raw/test/LJ001-0001.wav
diff --git a/tests/dataset_raw/test/LJ001-0002.wav b/tests/dataset_raw/test/LJ001-0002.wav
diff --git a/tests/dataset_raw/test/LJ001-0003.wav b/tests/dataset_raw/test/LJ001-0003.wav
diff --git a/tests/dataset_raw/test/LJ001-0004.wav b/tests/dataset_raw/test/LJ001-0004.wav
diff --git a/tests/dataset_raw/test/LJ001-0005.wav b/tests/dataset_raw/test/LJ001-0005.wav
diff --git a/tests/dataset_raw/test/LJ001-0006.wav b/tests/dataset_raw/test/LJ001-0006.wav
diff --git a/tests/dataset_raw/test/LJ001-0007.wav b/tests/dataset_raw/test/LJ001-0007.wav
diff --git a/tests/dataset_raw/test/LJ001-0008.wav b/tests/dataset_raw/test/LJ001-0008.wav
diff --git a/tests/dataset_raw/test/LJ001-0009.wav b/tests/dataset_raw/test/LJ001-0009.wav
diff --git a/tests/dataset_raw/test/LJ001-0010.wav b/tests/dataset_raw/test/LJ001-0010.wav
diff --git a/tests/dataset_raw/test/nested/LJ001-0001.wav b/tests/dataset_raw/test/nested/LJ001-0001.wav
diff --git a/tests/dataset_raw/test/nested/に.wav b/tests/dataset_raw/test/nested/に.wav