Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

update benchmark notebook

This commit is contained in:
Eren Golge 2019-09-24 16:58:52 +02:00
parent 5b6b1f354d
commit 113f5860b8
1 changed files with 138 additions and 306 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

444
notebooks/Benchmark.ipynb
View File

@ -19,7 +19,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 1, "execution_count": null,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -29,28 +29,11 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 2, "execution_count": null,
"metadata": { "metadata": {
"scrolled": true "scrolled": true
}, },
"outputs": [ "outputs": [],
{
"name": "stdout",
"output_type": "stream",
"text": [
"Populating the interactive namespace from numpy and matplotlib\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/erogol/miniconda3/lib/python3.7/site-packages/IPython/core/magics/pylab.py:160: UserWarning: pylab import has clobbered these variables: ['plt']\n",
"`%matplotlib` prevents importing * from pylab and numpy\n",
" \"\\n`%matplotlib` prevents importing * from pylab and numpy\"\n"
]
}
],
"source": [ "source": [
"%load_ext autoreload\n", "%load_ext autoreload\n",
"%autoreload 2\n", "%autoreload 2\n",
@ -59,6 +42,7 @@
"import io\n", "import io\n",
"import torch \n", "import torch \n",
"import time\n", "import time\n",
"import json\n",
"import numpy as np\n", "import numpy as np\n",
"from collections import OrderedDict\n", "from collections import OrderedDict\n",
"from matplotlib import pylab as plt\n", "from matplotlib import pylab as plt\n",
@ -86,23 +70,25 @@
"from IPython.display import Audio\n", "from IPython.display import Audio\n",
"\n", "\n",
"import os\n", "import os\n",
"os.environ['CUDA_VISIBLE_DEVICES']='1'\n", "os.environ['CUDA_VISIBLE_DEVICES']='1'"
"os.environ['OMP_NUM_THREADS']='1'\n"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": null,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"def tts(model, text, CONFIG, use_cuda, ap, use_gl, speaker_id=None, figures=True):\n", "def tts(model, text, CONFIG, use_cuda, ap, use_gl, figures=True):\n",
" t_1 = time.time()\n", " t_1 = time.time()\n",
" waveform, alignment, mel_spec, mel_postnet_spec, stop_tokens = synthesis(model, text, CONFIG, use_cuda, ap, truncated=False, speaker_id=speaker_id, enable_eos_bos_chars=CONFIG.enable_eos_bos_chars)\n", " waveform, alignment, mel_spec, mel_postnet_spec, stop_tokens = synthesis(model, text, CONFIG, use_cuda, ap, speaker_id, False, CONFIG.enable_eos_bos_chars)\n",
" if CONFIG.model == \"Tacotron\" and not use_gl:\n", " if CONFIG.model == \"Tacotron\" and not use_gl:\n",
" # coorect the normalization differences b/w TTS and the Vocoder.\n",
" mel_postnet_spec = ap.out_linear_to_mel(mel_postnet_spec.T).T\n", " mel_postnet_spec = ap.out_linear_to_mel(mel_postnet_spec.T).T\n",
" mel_postnet_spec = ap._denormalize(mel_postnet_spec)\n",
" mel_postnet_spec = ap_vocoder._normalize(mel_postnet_spec)\n",
" if not use_gl:\n", " if not use_gl:\n",
" waveform = wavernn.generate(torch.FloatTensor(mel_postnet_spec.T).unsqueeze(0).cuda(), batched=batched_wavernn, target=11000, overlap=550)\n", " waveform = wavernn.generate(torch.FloatTensor(mel_postnet_spec.T).unsqueeze(0).cuda(), batched=batched_wavernn, target=8000, overlap=400)\n",
"\n", "\n",
" print(\" > Run-time: {}\".format(time.time() - t_1))\n", " print(\" > Run-time: {}\".format(time.time() - t_1))\n",
" if figures: \n", " if figures: \n",
@ -117,31 +103,18 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 9, "execution_count": null,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"ename": "FileNotFoundError",
"evalue": "[Errno 2] No such file or directory: '/media/erogol/data_ssd/Data/models/wavernn/mozilla/mozilla-May24-4763/config.json'",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mFileNotFoundError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-9-3306702a6bbc>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 7\u001b[0m \u001b[0mVOCODER_MODEL_PATH\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"/media/erogol/data_ssd/Data/models/wavernn/mozilla/mozilla-May24-4763/model_checkpoints/best_model.pth.tar\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 8\u001b[0m \u001b[0mVOCODER_CONFIG_PATH\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"/media/erogol/data_ssd/Data/models/wavernn/mozilla/mozilla-May24-4763/config.json\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 9\u001b[0;31m \u001b[0mVOCODER_CONFIG\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mload_config\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mVOCODER_CONFIG_PATH\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 10\u001b[0m \u001b[0muse_cuda\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 11\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;32m~/projects/TTS/tts_namespace/TTS/utils/generic_utils.py\u001b[0m in \u001b[0;36mload_config\u001b[0;34m(config_path)\u001b[0m\n\u001b[1;32m 20\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mload_config\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mconfig_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 21\u001b[0m \u001b[0mconfig\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mAttrDict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 22\u001b[0;31m \u001b[0;32mwith\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mconfig_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"r\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 23\u001b[0m \u001b[0minput_str\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 24\u001b[0m \u001b[0minput_str\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mre\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msub\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mr'\\\\\\n'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m''\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput_str\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: '/media/erogol/data_ssd/Data/models/wavernn/mozilla/mozilla-May24-4763/config.json'"
]
}
],
"source": [ "source": [
"# Set constants\n", "# Set constants\n",
"ROOT_PATH = '/media/erogol/data_ssd/Models/libri_tts/5049/'\n", "ROOT_PATH = '/media/erogol/data_ssd/Models/libri_tts/5099/'\n",
"MODEL_PATH = ROOT_PATH + 'best_model.pth.tar'\n", "MODEL_PATH = ROOT_PATH + '/best_model.pth.tar'\n",
"CONFIG_PATH = ROOT_PATH + '/config.json'\n", "CONFIG_PATH = ROOT_PATH + '/config.json'\n",
"OUT_FOLDER = \"/home/erogol/Dropbox/AudioSamples/benchmark_samples/\"\n", "OUT_FOLDER = '/home/erogol/Dropbox/AudioSamples/benchmark_samples/'\n",
"CONFIG = load_config(CONFIG_PATH)\n", "CONFIG = load_config(CONFIG_PATH)\n",
"VOCODER_MODEL_PATH = \"/media/erogol/data_ssd/Data/models/wavernn/mozilla/mozilla-May24-4763/model_checkpoints/best_model.pth.tar\"\n", "VOCODER_MODEL_PATH = \"/media/erogol/data_ssd/Models/wavernn/ljspeech/mold_ljspeech_best_model/checkpoint_433000.pth.tar\"\n",
"VOCODER_CONFIG_PATH = \"/media/erogol/data_ssd/Data/models/wavernn/mozilla/mozilla-May24-4763/config.json\"\n", "VOCODER_CONFIG_PATH = \"/media/erogol/data_ssd/Models/wavernn/ljspeech/mold_ljspeech_best_model/config.json\"\n",
"VOCODER_CONFIG = load_config(VOCODER_CONFIG_PATH)\n", "VOCODER_CONFIG = load_config(VOCODER_CONFIG_PATH)\n",
"use_cuda = False\n", "use_cuda = False\n",
"\n", "\n",
@ -149,10 +122,12 @@
"# CONFIG.windowing = False\n", "# CONFIG.windowing = False\n",
"# CONFIG.prenet_dropout = False\n", "# CONFIG.prenet_dropout = False\n",
"# CONFIG.separate_stopnet = True\n", "# CONFIG.separate_stopnet = True\n",
"CONFIG.use_forward_attn = True\n",
"# CONFIG.forward_attn_mask = True\n",
"# CONFIG.stopnet = True\n", "# CONFIG.stopnet = True\n",
"\n", "\n",
"# Set the vocoder\n", "# Set the vocoder\n",
"use_gl = True # use GL if True\n", "use_gl = False # use GL if True\n",
"batched_wavernn = True # use batched wavernn inference if True" "batched_wavernn = True # use batched wavernn inference if True"
] ]
}, },
@ -165,9 +140,17 @@
"# LOAD TTS MODEL\n", "# LOAD TTS MODEL\n",
"from utils.text.symbols import symbols, phonemes\n", "from utils.text.symbols import symbols, phonemes\n",
"\n", "\n",
"# multi speaker \n",
"if CONFIG.use_speaker_embedding:\n",
" speakers = json.load(open(f\"{ROOT_PATH}/speakers.json\", 'r'))\n",
" speakers_idx_to_id = {v: k for k, v in speakers.items()}\n",
"else:\n",
" speakers = []\n",
" speaker_id = None\n",
"\n",
"# load the model\n", "# load the model\n",
"num_chars = len(phonemes) if CONFIG.use_phonemes else len(symbols)\n", "num_chars = len(phonemes) if CONFIG.use_phonemes else len(symbols)\n",
"model = setup_model(num_chars, CONFIG)\n", "model = setup_model(num_chars, len(speakers), CONFIG)\n",
"\n", "\n",
"# load the audio processor\n", "# load the audio processor\n",
"ap = AudioProcessor(**CONFIG.audio) \n", "ap = AudioProcessor(**CONFIG.audio) \n",
@ -184,7 +167,12 @@
"if use_cuda:\n", "if use_cuda:\n",
" model.cuda()\n", " model.cuda()\n",
"model.eval()\n", "model.eval()\n",
"print(cp['step'])" "print(cp['step'])\n",
"print(cp['r'])\n",
"\n",
"# set model stepsize\n",
"if 'r' in cp:\n",
" model.decoder.set_r(cp['r'])"
] ]
}, },
{ {
@ -196,25 +184,28 @@
"# LOAD WAVERNN\n", "# LOAD WAVERNN\n",
"if use_gl == False:\n", "if use_gl == False:\n",
" from WaveRNN.models.wavernn import Model\n", " from WaveRNN.models.wavernn import Model\n",
" from WaveRNN.utils.audio import AudioProcessor as AudioProcessorVocoder\n",
" bits = 10\n", " bits = 10\n",
"\n", " ap_vocoder = AudioProcessorVocoder(**VOCODER_CONFIG.audio) \n",
" wavernn = Model(\n", " wavernn = Model(\n",
" rnn_dims=512,\n", " rnn_dims=512,\n",
" fc_dims=512,\n", " fc_dims=512,\n",
" mode=\"mold\",\n", " mode=VOCODER_CONFIG.mode,\n",
" pad=2,\n", " mulaw=VOCODER_CONFIG.mulaw,\n",
" upsample_factors=VOCODER_CONFIG.upsample_factors, # set this depending on dataset\n", " pad=VOCODER_CONFIG.pad,\n",
" upsample_factors=VOCODER_CONFIG.upsample_factors,\n",
" feat_dims=VOCODER_CONFIG.audio[\"num_mels\"],\n", " feat_dims=VOCODER_CONFIG.audio[\"num_mels\"],\n",
" compute_dims=128,\n", " compute_dims=128,\n",
" res_out_dims=128,\n", " res_out_dims=128,\n",
" res_blocks=10,\n", " res_blocks=10,\n",
" hop_length=ap.hop_length,\n", " hop_length=ap_vocoder.hop_length,\n",
" sample_rate=ap.sample_rate,\n", " sample_rate=ap_vocoder.sample_rate,\n",
" use_upsample_net = True,\n",
" use_aux_net = True\n",
" ).cuda()\n", " ).cuda()\n",
"\n", "\n",
"\n",
" check = torch.load(VOCODER_MODEL_PATH)\n", " check = torch.load(VOCODER_MODEL_PATH)\n",
" wavernn.load_state_dict(check['model'])\n", " wavernn.load_state_dict(check['model'], strict=False)\n",
" if use_cuda:\n", " if use_cuda:\n",
" wavernn.cuda()\n", " wavernn.cuda()\n",
" wavernn.eval();\n", " wavernn.eval();\n",
@ -230,111 +221,67 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 5, "execution_count": null,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"ename": "NameError",
"evalue": "name 'model' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-5-e285d5bde9fb>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meval\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdecoder\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax_decoder_steps\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m2000\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0mspeaker_id\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 4\u001b[0m \u001b[0msentence\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"Bill got in the habit of asking himself “Is that thought true?” And if he wasn’t absolutely certain it was, he just let it go.\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0malign\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspec\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstop_tokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwav\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtts\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msentence\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mCONFIG\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_cuda\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0map\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspeaker_id\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mspeaker_id\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_gl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0muse_gl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfigures\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"
]
}
],
"source": [ "source": [
"model.eval()\n", "model.eval()\n",
"model.decoder.max_decoder_steps = 2000\n", "model.decoder.max_decoder_steps = 2000\n",
"speaker_id = 0\n", "speaker_id = None\n",
"sentence = \"Bill got in the habit of asking himself “Is that thought true?” And if he wasn’t absolutely certain it was, he just let it go.\"\n", "sentence = \"Bill got in the habit of asking himself “Is that thought true?” and if he wasn’t absolutely certain it was, he just let it go.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"scrolled": true
},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'model' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-6-621056ffa667>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0msentence\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"Be a voice, not an echo.\"\u001b[0m \u001b[0;31m# 'echo' is not in training set.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0malign\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspec\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstop_tokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwav\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtts\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msentence\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mCONFIG\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_cuda\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0map\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspeaker_id\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mspeaker_id\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_gl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0muse_gl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfigures\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"
]
}
],
"source": [
"sentence = \"Be a voice, not an echo.\" # 'echo' is not in training set. \n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'model' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-7-26967668a1a1>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0msentence\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"The human voice is the most perfect instrument of all.\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0malign\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspec\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstop_tokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwav\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtts\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msentence\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mCONFIG\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_cuda\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0map\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspeaker_id\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mspeaker_id\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_gl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0muse_gl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfigures\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"
]
}
],
"source": [
"sentence = \"The human voice is the most perfect instrument of all.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'model' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m<ipython-input-8-28cb5023e353>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0msentence\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"I'm sorry Dave. I'm afraid I can't do that.\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0malign\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspec\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstop_tokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwav\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtts\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msentence\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mCONFIG\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_cuda\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0map\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mspeaker_id\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mspeaker_id\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muse_gl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0muse_gl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfigures\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
"\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"
]
}
],
"source": [
"sentence = \"I'm sorry Dave. I'm afraid I can't do that.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true, "outputs": [],
"jupyter": { "source": [
"outputs_hidden": true "model.eval()\n",
}, "model.decoder.max_decoder_steps = 2000\n",
"scrolled": true "sentence = \"Bill got in the habit of asking himself “Is that thought true?” and if he wasn’t absolutely certain it was, he just let it go.\"\n",
}, "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"sentence = \"Be a voice, not an echo.\" # 'echo' is not in training set. \n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"sentence = \"The human voice is the most perfect instrument of all.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"sentence = \"I'm sorry Dave. I'm afraid I can't do that.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"This cake is great. It's so delicious and moist.\"\n", "sentence = \"This cake is great. It's so delicious and moist.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
@ -347,76 +294,51 @@
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Generative adversarial network or variational auto-encoder.\"\n", "sentence = \"Generative adversarial network or variational auto-encoder.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Scientists at the CERN laboratory say they have discovered a new particle.\"\n", "sentence = \"Scientists at the CERN laboratory say they have discovered a new particle.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Here’s a way to measure the acute emotional intelligence that has never gone out of style.\"\n", "sentence = \"Here’s a way to measure the acute emotional intelligence that has never gone out of style.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"President Trump met with other leaders at the Group of 20 conference.\"\n", "sentence = \"President Trump met with other leaders at the Group of 20 conference.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"The buses aren't the problem, they actually provide a solution.\"\n", "sentence = \"The buses aren't the problem, they actually provide a solution.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
@ -429,136 +351,91 @@
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Generative adversarial network or variational auto-encoder.\"\n", "sentence = \"Generative adversarial network or variational auto-encoder.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Basilar membrane and otolaryngology are not auto-correlations.\"\n", "sentence = \"Basilar membrane and otolaryngology are not auto-correlations.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \" He has read the whole thing.\"\n", "sentence = \" He has read the whole thing.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"He reads books.\"\n", "sentence = \"He reads books.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Thisss isrealy awhsome.\"\n", "sentence = \"Thisss isrealy awhsome.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"This is your internet browser, Firefox.\"\n", "sentence = \"This is your internet browser, Firefox.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"This is your internet browser Firefox.\"\n", "sentence = \"This is your internet browser Firefox.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"The quick brown fox jumps over the lazy dog.\"\n", "sentence = \"The quick brown fox jumps over the lazy dog.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Does the quick brown fox jump over the lazy dog?\"\n", "sentence = \"Does the quick brown fox jump over the lazy dog?\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
@ -568,7 +445,7 @@
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Eren, how are you?\"\n", "sentence = \"Eren, how are you?\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
@ -581,107 +458,62 @@
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Encouraged, he started with a minute a day.\"\n", "sentence = \"Encouraged, he started with a minute a day.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"His meditation consisted of “body scanning” which involved focusing his mind and energy on each section of the body from head to toe .\"\n", "sentence = \"His meditation consisted of “body scanning” which involved focusing his mind and energy on each section of the body from head to toe .\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Recent research at Harvard has shown meditating for as little as 8 weeks can actually increase the grey matter in the parts of the brain responsible for emotional regulation and learning . \"\n", "sentence = \"Recent research at Harvard has shown meditating for as little as 8 weeks can actually increase the grey matter in the parts of the brain responsible for emotional regulation and learning . \"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"If he decided to watch TV he really watched it.\"\n", "sentence = \"If he decided to watch TV he really watched it.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
},
"scrolled": true
},
"outputs": [], "outputs": [],
"source": [ "source": [
"sentence = \"Often we try to bring about change through sheer effort and we put all of our energy into a new initiative .\"\n", "sentence = \"Often we try to bring about change through sheer effort and we put all of our energy into a new initiative .\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,
"metadata": { "metadata": {},
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [], "outputs": [],
"source": [ "source": [
"# for twb dataset\n", "# for twb dataset\n",
"sentence = \"In our preparation for Easter, God in his providence offers us each year the season of Lent as a sacramental sign of our conversion.\"\n", "sentence = \"In our preparation for Easter, God in his providence offers us each year the season of Lent as a sacramental sign of our conversion.\"\n",
"align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, speaker_id=speaker_id, use_gl=use_gl, figures=True)" "align, spec, stop_tokens, wav = tts(model, sentence, CONFIG, use_cuda, ap, use_gl=use_gl, figures=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true,
"jupyter": {
"outputs_hidden": true
}
},
"outputs": [],
"source": [
"# !zip benchmark_samples/samples.zip benchmark_samples/*"
] ]
} }
], ],