chavinlo/rvc-runpod

Serverless RVC endpoint for Runpod

📻 What I am listening to: https://www.last.fm/user/hololens

What I worked on previously:

• 🎞️ TempoFunk, a Text-To-Video Model: https://huggingface.co/TempoFunk/makeavid-sd-jax
• 🏃💨 SDA, a simplified Stable Diffusion TensorRT pipeline: https://github.com/chavinlo/sda-node
• A few more projects (decentralized training, riffusion control, etc...)

HuggingFace: https://huggingface.co/chavinlo

Discord: @qasb

Bye!

rvc-runpod's Issues

Endpoint fails to execute on Runpod with the docker image provided (AttributeError: 'NoneType' object has no attribute 'tobytes')

I followed the howto, pulled the current docker image from "depositame/rvc_runpod_serverless/general" and deployed the image on my runpod console. Deplyoment ran smoothly on runpod.

When creating a request for the endpoint on runpod with an adjusted input_json (the audio_url and model_name given in the howto don not exist any longer), i run into an python error which is visible in the runpod logs. For input paramters i used an mp3 file (small 3min song) and another model from "sail-rvc".

Here is the error i am getting:
AttributeError: 'NoneType' object has no attribute 'tobytes'
temp_dir = Path(dir) / self.hash_bytes(data.tobytes())

More context to the log/trace:
2024-06-28 19:18:26.620
[ocmtv88uo73xas]
[info]
INFO | 68a8cde3-c7ec-4ba3-80e3-3a8cf59b1c44-e1 | Finished
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[info]
}
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[info]
"pod_id": "ocmtv88uo73xas"
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[info]
"host_name": "ocmtv88uo73xas-64411128",
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[info]
"error_traceback": "Traceback (most recent call last):\n File "/opt/conda/lib/python3.10/site-packages/runpod/serverless/modules/job.py", line 85, in run_job\n job_output = handler(job)\n File "/rvc_serverless/main.py", line 283, in handler\n return self.infer(request)\n File "/rvc_serverless/main.py", line 229, in infer\n result = client.predict(\n File "/opt/conda/lib/python3.10/site-packages/gradio_client/client.py", line 285, in predict\n return self.submit(*args, api_name=api_name, fn_index=fn_index).result()\n File "/opt/conda/lib/python3.10/site-packages/gradio_client/client.py", line 982, in result\n raise self.future._exception # type: ignore\n File "/opt/conda/lib/python3.10/concurrent/futures/thread.py", line 58, in run\n result = self.fn(*self.args, **self.kwargs)\n File "/opt/conda/lib/python3.10/site-packages/gradio_client/client.py", line 645, in _inner\n predictions = _predict(*data)\n File "/opt/conda/lib/python3.10/site-packages/gradio_client/client.py", line 676, in _predict\n raise ValueError(result["error"])\nValueError: None\n",
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[info]
"error_message": "None",
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[info]
"error_type": "<class 'ValueError'>",
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[error]
ERROR | {
2024-06-28 19:18:26.466
[ocmtv88uo73xas]
[error]
ERROR | 68a8cde3-c7ec-4ba3-80e3-3a8cf59b1c44-e1 | Captured Handler Exception
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
AttributeError: 'NoneType' object has no attribute 'tobytes'
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
temp_dir = Path(dir) / self.hash_bytes(data.tobytes())
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
File "/opt/conda/lib/python3.10/site-packages/gradio/components/base.py", line 321, in audio_to_temp_file
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
file_path = self.audio_to_temp_file(
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
prediction_value = block.postprocess(prediction_value)
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
File "/opt/conda/lib/python3.10/site-packages/gradio/blocks.py", line 1321, in postprocess_data
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
data = self.postprocess_data(fn_index, result["prediction"], state)
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
File "/opt/conda/lib/python3.10/site-packages/gradio/blocks.py", line 1387, in process_api
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
output = await app.get_blocks().process_api(
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
File "/opt/conda/lib/python3.10/site-packages/gradio/routes.py", line 439, in run_predict
2024-06-28 19:18:26.457
[ocmtv88uo73xas]
[info]
Traceback (most recent call last):
2024-06-28 19:18:25.221
[ocmtv88uo73xas]
[info]
Downloading model.index: 0%| | 0.00/373M [00:00<?, ?B/s] Downloading model.index: 3%|▎ | 10.5M/373M [00:00<00:04, 76.9MB/s] Downloading model.index: 6%|▌ | 21.0M/373M [00:00<00:04, 78.7MB/s] Downloading model.index: 8%|▊ | 31.5M/373M [00:00<00:04, 79.2MB/s] Downloading model.index: 11%|█ | 41.9M/373M [00:00<00:04, 79.5MB/s] Downloading model.index: 14%|█▍ | 52.4M/373M [00:00<00:04, 79.6MB/s] Downloading model.index: 17%|█▋ | 62.9M/373M [00:00<00:03, 79.7MB/s] Downloading model.index: 20%|█▉ | 73.4M/373M [00:00<00:03, 79.8MB/s] Downloading model.index: 22%|██▏ | 83.9M/373M [00:01<00:03, 79.7MB/s] Downloading model.index: 25%|██▌ | 94.4M/373M [00:01<00:03, 79.8MB/s] Downloading model.index: 28%|██▊ | 105M/373M [00:01<00:03, 79.8MB/s] Downloading model.index: 31%|███ | 115M/373M [00:01<00:03, 79.7MB/s] Downloading model.index: 34%|███▎ | 126M/373M [00:01<00:03, 79.8MB/s] Downloading model.index: 37%|███▋ | 136M/373M [00:01<00:03, 65.0MB/s] Downloading model.index: 39%|███▉ | 147M/373M [00:01<00:03, 67.3MB/s] Downloading model.index: 42%|████▏ | 157M/373M [00:02<00:03, 70.6MB/s] Downloading model.index: 45%|████▍ | 168M/373M [00:02<00:02, 70.5MB/s] Downloading model.index: 48%|████▊ | 178M/373M [00:02<00:02, 68.4MB/s] Downloading model.index: 51%|█████ | 189M/373M [00:02<00:02, 71.0MB/s] Downloading model.index: 53%|█████▎ | 199M/373M [00:02<00:02, 67.2MB/s] Downloading model.index: 56%|█████▌ | 210M/373M [00:02<00:02, 69.7MB/s] Downloading model.index: 59%|█████▉ | 220M/373M [00:03<00:02, 67.2MB/s] Downloading model.index: 62%|██████▏ | 231M/373M [00:03<00:02, 68.6MB/s] Downloading model.index: 65%|██████▍ | 241M/373M [00:03<00:01, 71.2MB/s] Downloading model.index: 67%|██████▋ | 252M/373M [00:03<00:01, 73.5MB/s] Downloading model.index: 70%|███████ | 262M/373M [00:03<00:01, 75.2MB/s] Downloading model.index: 73%|███████▎ | 273M/373M [00:03<00:01, 76.6MB/s] Downloading model.index: 76%|███████▌ | 283M/373M [00:03<00:01, 77.5MB/s] Downloading model.index: 79%|███████▊ | 294M/373M [00:04<00:01, 70.2MB/s] Downloading model.index: 82%|████████▏ | 304M/373M [00:04<00:00, 72.4MB/s] Downloading model.index: 84%|████████▍ | 315M/373M [00:04<00:00, 74.3MB/s] Downloading model.index: 87%|████████▋ | 325M/373M [00:04<00:00, 75.9MB/s] Downloading model.index: 90%|████████▉ | 336M/373M [00:04<00:00, 76.9MB/s] Downloading model.index: 93%|█████████▎| 346M/373M [00:04<00:00, 77.7MB/s] Downloading model.index: 96%|█████████▌| 357M/373M [00:04<00:00, 78.3MB/s] Downloading model.index: 98%|█████████▊| 367M/373M [00:04<00:00, 78.6MB/s] Downloading model.index: 100%|██████████| 373M/373M [00:05<00:00, 74.5MB/s]
2024-06-28 19:18:22.713
[ocmtv88uo73xas]
[info]
Downloading model.pth: 0%| | 0.00/55.2M [00:00<?, ?B/s] Downloading model.pth: 19%|█▉ | 10.5M/55.2M [00:00<00:01, 30.1MB/s] Downloading model.pth: 38%|███▊ | 21.0M/55.2M [00:00<00:00, 47.1MB/s] Downloading model.pth: 57%|█████▋ | 31.5M/55.2M [00:00<00:00, 57.3MB/s] Downloading model.pth: 76%|███████▌ | 41.9M/55.2M [00:00<00:00, 64.0MB/s] Downloading model.pth: 95%|█████████▍| 52.4M/55.2M [00:00<00:00, 68.9MB/s] Downloading model.pth: 100%|██████████| 55.2M/55.2M [00:00<00:00, 59.9MB/s]
2024-06-28 19:18:22.713
[ocmtv88uo73xas]
[info]
Downloading config.json: 0%| | 0.00/130 [00:00<?, ?B/s] Downloading config.json: 100%|██████████| 130/130 [00:00<00:00, 688kB/s]
2024-06-28 19:18:22.713
[ocmtv88uo73xas]
[info]
INFO | 68a8cde3-c7ec-4ba3-80e3-3a8cf59b1c44-e1 | Started
2024-06-28 19:18:22.713
[ocmtv88uo73xas]

what is the location for placing the pretained models in dockerfile ? or is it not possible in rvc ?

what would be the default location for placing the pretained models ? where i can find the existing models in my container ? like if i want to put models directly in my docker image so where should i place them ? not only for "sail-rvc" but if i need different model so whats the default location in docker image / container ?

ModuleNotFoundError: No module named 'faiss'

Getting something like this

2024-07-22T02:28:48.567516413Z The token has not been saved to the git credentials helper. Pass `add_to_git_credential=True` in this function directly or `--add-to-git-credential` if using via `huggingface-cli` if you want to set the git credential as well.
2024-07-22T02:28:48.567550868Z Token is valid (permission: fineGrained).
2024-07-22T02:28:48.567557250Z Your token has been saved to /root/.cache/huggingface/token
2024-07-22T02:28:48.567563282Z Login successful
2024-07-22T02:28:48.567569003Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567573972Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567578791Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567583590Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567588419Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567593228Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567598047Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567602846Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567607716Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.567612575Z Loaded as API: http://127.0.0.1:7860/ ✔
2024-07-22T02:28:48.569759813Z Traceback (most recent call last):
2024-07-22T02:28:48.569782936Z   File "/rvc_serverless/main.py", line 295, in <module>
2024-07-22T02:28:48.569789138Z     main()
2024-07-22T02:28:48.569794969Z   File "/rvc_serverless/main.py", line 288, in main
2024-07-22T02:28:48.569821700Z     pipeline.prepare()
2024-07-22T02:28:48.569828753Z   File "/rvc_serverless/main.py", line 102, in prepare
2024-07-22T02:28:48.569833822Z     raise Exception("Cannot connect to gradio API")
2024-07-22T02:28:48.569838942Z Exception: Cannot connect to gradio API
2024-07-22T02:28:54.809514940Z python infer-web.py --paperspace --pycmd python
2024-07-22T02:28:56.082303869Z Traceback (most recent call last):
2024-07-22T02:28:56.082381235Z   File "/rvc_repo/infer-web.py", line 22, in <module>
2024-07-22T02:28:56.082386916Z     import faiss
2024-07-22T02:28:56.082392186Z ModuleNotFoundError: No module named 'faiss'
2024-07-22T02:28:56.360716508Z make: *** [Makefile:59: run-ui] Error 1

Transfer.sh

Transfer.sh down.

ModuleNotFoundError

hello, thx for the project , i get errors when following the how to,

: No module named 'faiss' 
: No module named 'infer_pack' 

do i need to install something else ?