{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "3abd2681",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:06:08.344969Z",
"start_time": "2023-09-15T00:06:04.392233Z"
}
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/tony/anaconda3/envs/suno_env/lib/python3.10/site-packages/whisper/timing.py:57: NumbaDeprecationWarning: \u001b[1mThe 'nopython' keyword argument was not supplied to the 'numba.jit' decorator. The implicit default value for this argument is currently False, but it will be changed to True in Numba 0.59.0. See https://numba.readthedocs.io/en/stable/reference/deprecation.html#deprecation-of-object-mode-fall-back-behaviour-when-using-jit for details.\u001b[0m\n",
" def backtrace(trace: np.ndarray):\n"
]
}
],
"source": [
"import os\n",
"os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"1\"\n",
"\n",
"from suno_utils.audio import Audio\n",
"from suno_utils.tasks.mert_v2 import (\n",
" encode as mert_encode,\n",
" preload_models as preload_mert_models\n",
")\n",
"from suno_utils.tasks.dac import (\n",
" encode as codec_encode,\n",
" decode as codec_decode,\n",
" preload_models as preload_codec_models,\n",
")\n",
"\n",
"from suno_utils.tasks.gpt_v2.chirp_v1 import preload_models as preload_chirp_models\n",
"from suno_utils.tasks.gpt_v2.chirp_v1 import generate_audio\n",
"from suno_utils.tasks.gpt_v2.generation import generate\n",
"\n",
"%load_ext autoreload\n",
"%autoreload 2"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "f97f7e06",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:06:41.199611Z",
"start_time": "2023-09-15T00:06:08.347408Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"model loaded: 1720.9M params, 1.411 loss\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Some weights of the model checkpoint at m-a-p/MERT-v1-95M were not used when initializing MERTModel: ['encoder.pos_conv_embed.conv.weight_g', 'encoder.pos_conv_embed.conv.weight_v']\n",
"- This IS expected if you are initializing MERTModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
"- This IS NOT expected if you are initializing MERTModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
"Some weights of MERTModel were not initialized from the model checkpoint at m-a-p/MERT-v1-95M and are newly initialized: ['encoder.pos_conv_embed.conv.parametrizations.weight.original0', 'encoder.pos_conv_embed.conv.parametrizations.weight.original1']\n",
"You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n",
"/home/tony/anaconda3/envs/suno_env/lib/python3.10/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.\n",
" warnings.warn(\"torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.\")\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Downloading s3://suno-data/georg/trained_models/chirp_v1/lid.176.bin to /tmp/tmpppe5rhkw\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Warning : `load_model` does not return WordVectorModel or SupervisedModel any more, but a `FastText` object which is very similar.\n"
]
}
],
"source": [
"preload_chirp_models(\n",
" centroids_filepath=\"/home/mikeys/bundle/2x1k_centroids_mert.npy\",\n",
" gpt_ckpt_path=\"/home/tony/Data/Chirp/xl_2.pt\",\n",
" codec_ckpt_path=\"/home/mikeys/bundle/d_codec_25x8.pt\",\n",
" common_genre_tags_path=\"/home/mikeys/bundle/common_tags.json\",\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "cceb2397",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:07:12.144770Z",
"start_time": "2023-09-15T00:06:41.201353Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"hacking text with intro for non-English\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1310/1310 [00:29<00:00, 44.04it/s]\n"
]
}
],
"source": [
"text = \"\"\"\n",
"[Verse]\n",
"花褪残红\n",
"青杏小\n",
"燕子飞时\n",
"绿水\n",
"人家绕\n",
"枝上柳绵\n",
"吹又少\n",
"\n",
"[Chorus]\n",
"天涯何处无芳草\n",
"无芳草\n",
"\"\"\"\n",
"out = generate_audio(\n",
" text,\n",
" text_tags=\"[mandopop]\", \n",
" n_batch=2,\n",
" history_audio=None,\n",
" history_text=None,\n",
" max_gen_duration_s=15,\n",
" max_history_duration_s=10.01,\n",
" temp_semantic=0.8,\n",
" temp_coarse=0.8,\n",
" top_k_semantic=250,\n",
" top_k_coarse=500,\n",
" cfg_coef=2.25,\n",
" cfg_coef_tags=1.75,\n",
" min_eos_p=0.5,\n",
" rep_penality=0.2,\n",
" allow_genre_randomization=False,\n",
" silent=False,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "f1cbb567",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:07:12.647489Z",
"start_time": "2023-09-15T00:07:12.146063Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"#codec_decode(out[1][1][0]).play()\n",
"for o in out:\n",
" o.play()\n",
"# or out[0][0].play()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "e3c7d909",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:07:12.677010Z",
"start_time": "2023-09-15T00:07:12.648851Z"
}
},
"outputs": [],
"source": [
"my_audio = out[1]"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "dd1076d4",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:21:30.151353Z",
"start_time": "2023-09-15T00:21:11.496315Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"False, 恰你永止于惩而垂眠过而小冬,\"还街时ニ儿竟養殷其惩而辽悲're in the dark night sky, 2.889\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 560/560 [00:13<00:00, 41.25it/s]\n"
]
}
],
"source": [
"text_2 = \"\"\"\n",
"[Chorus]\n",
"墙里秋千\n",
"墙外道\n",
"墙外行人\n",
"墙里\n",
"佳人笑\n",
"笑渐不闻\n",
"声渐悄\n",
"\n",
"[Beat]\n",
"多情却被无情恼\n",
"无情恼\n",
"[Beat]\n",
"\"\"\"\n",
"out_2 = generate_audio(\n",
" text_2,\n",
" text_tags=\"[kpop]\", \n",
" n_batch=2,\n",
" history_audio=my_audio,\n",
" history_text=None,\n",
" max_gen_duration_s=5,\n",
" max_history_duration_s=10.01,\n",
" temp_semantic=0.8,\n",
" temp_coarse=0.8,\n",
" top_k_semantic=250,\n",
" top_k_coarse=500,\n",
" cfg_coef=2.25,\n",
" cfg_coef_tags=1.75,\n",
" min_eos_p=0.8,\n",
" rep_penality=0.2,\n",
" allow_genre_randomization=False,\n",
" silent=False,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "c8705ad1",
"metadata": {
"ExecuteTime": {
"end_time": "2023-09-15T00:21:30.279214Z",
"start_time": "2023-09-15T00:21:30.152742Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"for o2 in out_2:\n",
" o2.play()"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "ab458636",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.12"
},
"toc": {
"base_numbering": 1,
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": false,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 5
}