[Bug] Unable to download models

Maarten-buelens commented 1 year ago

Describe the bug

When I run the example the model is unable to download.

when I try to manually download the model from

https://coqui.gateway.scarf.sh/v0.10.1_models/tts_models--multilingual--multi-dataset--your_tts.zip

I get redirected to

https://huggingface.co/erogol/v0.10.1_models/resolve/main/tts_models--multilingual--multi-dataset--your_tts.zip

which says "Repository not found"

To Reproduce

Run the basic example for python

code:

from TTS.api import TTS

# Running a multi-speaker and multi-lingual model

# List available 🐸TTS models and choose the first one
model_name = TTS.list_models()[0]
# Init TTS
tts = TTS(model_name)

# Run TTS

# ❗ Since this model is multi-speaker and multi-lingual, we must set the target speaker and the language
# Text to speech with a numpy output
wav = tts.tts("This is a test! This is also a test!!", speaker=tts.speakers[0], language=tts.languages[0])

output:

zipfile.BadZipFile: File is not a zip file

### Expected behavior

Being able to download the pretrained models

### Logs

```shell
> Downloading model to /home/mb/.local/share/tts/tts_models--multilingual--multi-dataset--your_tts
  0%|                                                                                                                                         | 0.00/29.0 [00:00<?, ?iB/s] > Error: Bad zip file - https://coqui.gateway.scarf.sh/v0.10.1_models/tts_models--multilingual--multi-dataset--your_tts.zip
Traceback (most recent call last):
  File "/home/mb/Documents/ai/TTS/TTS/utils/manage.py", line 434, in _download_zip_file
    with zipfile.ZipFile(temp_zip_name) as z:
  File "/usr/lib/python3.10/zipfile.py", line 1267, in __init__
    self._RealGetContents()
  File "/usr/lib/python3.10/zipfile.py", line 1334, in _RealGetContents
    raise BadZipFile("File is not a zip file")
zipfile.BadZipFile: File is not a zip file

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/mb/Documents/ai/tts/main.py", line 8, in <module>
    tts = TTS(model_name)
  File "/home/mb/Documents/ai/TTS/TTS/api.py", line 289, in __init__
    self.load_tts_model_by_name(model_name, gpu)
  File "/home/mb/Documents/ai/TTS/TTS/api.py", line 385, in load_tts_model_by_name
    model_path, config_path, vocoder_path, vocoder_config_path, model_dir = self.download_model_by_name(
  File "/home/mb/Documents/ai/TTS/TTS/api.py", line 348, in download_model_by_name
    model_path, config_path, model_item = self.manager.download_model(model_name)
  File "/home/mb/Documents/ai/TTS/TTS/utils/manage.py", line 303, in download_model
    self._download_zip_file(model_item["github_rls_url"], output_path, self.progress_bar)
  File "/home/mb/Documents/ai/TTS/TTS/utils/manage.py", line 439, in _download_zip_file
    raise zipfile.BadZipFile  # pylint: disable=raise-missing-from
zipfile.BadZipFile
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 29.0/29.0 [00:00<00:00, 185iB/s]

Environment

{
    "CUDA": {
        "GPU": [],
        "available": false,
        "version": "11.7"
    },
    "Packages": {
        "PyTorch_debug": false,
        "PyTorch_version": "2.0.1+cu117",
        "TTS": "0.14.3",
        "numpy": "1.23.5"
    },
    "System": {
        "OS": "Linux",
        "architecture": [
            "64bit",
            "ELF"
        ],
        "processor": "x86_64",
        "python": "3.10.6",
        "version": "#202303130630~1685473338~22.04~995127e SMP PREEMPT_DYNAMIC Tue M"
    }
}

Additional context

No response

imaginashaun commented 1 year ago

Any solution for this?

Maarten-buelens commented 1 year ago

No not yet.

ACDwyser commented 1 year ago

Same problem when using the docker images. CPU and GPU.

libTorrentUser commented 1 year ago

same error seems to happen when using the command line utility (tts) tts --text "hello world." --out_path audio/hello.wav

damozhang commented 1 year ago

Same error

wilyarti commented 1 year ago

Same error. It stopped working over night.

rdlaitila commented 1 year ago

Here is my workaround for the docker usage. Seems their proxy urls in .model.json are broken.

If you've attempted to download any broken model links you probably need to clear them from /root/.local/share/tts/

Replace the proxy url's with github release links:

cp /root/TTS/.models.json /root/TTS/.models.json.bak
sed -i 's|https://coqui.gateway.scarf.sh|https://github.com/coqui-ai/TTS/releases/download|g' /root/TTS/.models.json
tts-server --model_name tts_models/en/vctk/vits
> Downloading model to /root/.local/share/tts/tts_models--en--vctk--vits
 > Model's license - apache 2.0
 > Check https://choosealicense.com/licenses/apache-2.0/ for more info.
 > Using model: vits
 > Setting up Audio Processor...
 | > sample_rate:22050
 | > resample:False
 | > num_mels:80
 | > log_func:np.log10
 | > min_level_db:0
 | > frame_shift_ms:None
 | > frame_length_ms:None
 | > ref_level_db:None
 | > fft_size:1024
 | > power:None
 | > preemphasis:0.0
 | > griffin_lim_iters:None
 | > signal_norm:None
 | > symmetric_norm:None
 | > mel_fmin:0
 | > mel_fmax:None
 | > pitch_fmin:None
 | > pitch_fmax:None
 | > spec_gain:20.0
 | > stft_pad_mode:reflect
 | > max_norm:1.0
 | > clip_norm:True
 | > do_trim_silence:False
 | > trim_db:60
 | > do_sound_norm:False
 | > do_amp_to_db_linear:True
 | > do_amp_to_db_mel:True
 | > do_rms_norm:False
 | > db_level:None
 | > stats_path:None
 | > base:10
 | > hop_length:256
 | > win_length:1024
 > initialization of speaker-embedding layers.
 * Serving Flask app 'TTS.server.server'
 * Debug mode: off
INFO:werkzeug:WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
 * Running on all addresses (::)
 * Running on http://[::1]:5002
 * Running on http://[::1]:5002
INFO:werkzeug:Press CTRL+C to quit

Curious of the need to retarget these model downloads away from github.

Bonk1971 commented 1 year ago

Just following so I know when this is fixed.

hkbb2014 commented 1 year ago

same here

MansiDS98 commented 1 year ago

I am also facing the same since yesterday, please provide a fix asap

pridi commented 1 year ago

Same for me since yesterday. Following to see when fixed...

imaginashaun commented 1 year ago

I just saw his tweet about "GitHub changed the max artifact size in releases and made it 25MB" This could be why it's not reachable anymore

AlexTs10 commented 1 year ago

in the TTS/models.json file in the repo the model links work fine (you can download them). but even if i download them manually i can't get it to work. anyone try this ?

carloscdias commented 1 year ago

Since the link for the model download is back I got it to work again by simply deleting the directory containing the corrupted zip files. In my case: rm -rf ~/.local/share/tts/tts_models--en--ljspeech--tacotron2-DDC

Maarten-buelens commented 1 year ago

in the TTS/models.json file in the repo the model links work fine (you can download them). but even if i download them manually i can't get it to work. anyone try this ?

If you manually download the zip you need to unpack it to the ~/.local/share/tts/{model name} directory or you could try the automatic download again.

Edresson commented 1 year ago

Sorry, guys, we had a problem with our proxy. Now all is good.

ihmily commented 9 months ago

Error: raise BadZipFile("File is not a zip file") zipfile.BadZipFile: File is not a zip file

I think the following is the root cause of this problem：

It is indeed a network issue that caused the download of resources to fail or be incomplete.

Reason:

This error is likely caused by the absence of the averaged_perceptron_tagger package in NLTK, which is used for part-of-speech tagging. It contains a part-of-speech tagger based on the averaged perceptron algorithm. If you are using this tagger in your code but haven't downloaded the corresponding data package beforehand, you will encounter an error indicating that the averaged_perceptron_tagger.zip file is missing. It is also possible that you are missing the cmudict CMU pronunciation dictionary data package file.

Normally, when you run the program for the first time, NLTK automatically downloads the relevant data packages it needs. In debug mode, you should see the following information:

[nltk_data] Downloading package averaged_perceptron_tagger to
[nltk_data]     /root/nltk_data...
[nltk_data]   Unzipping taggers/averaged_perceptron_tagger.zip.
[nltk_data] Downloading package cmudict to /root/nltk_data...
[nltk_data]   Unzipping corpora/cmudict.zip.

However, due to reasons such as network issues , the automatic download might fail, resulting in missing files and causing the loading error.

Solution: Redownload the missing data package files.

Method 1:

Create a download.py file and write the following code in it:

import nltk
print(nltk.data.path)
nltk.download('averaged_perceptron_tagger')
nltk.download('cmudict')

Save and run the file:

python download.py

This will display the file index location and automatically download the missing averaged_perceptron_tagger.zip and cmudict.zip files to a subdirectory under the /root/nltk_data directory. After the download is complete, check if you have an nltk_data folder in the root directory and extract the contents of the downloaded zip files there.

. Method 2:

If the above code still fails to download the data packages, you can manually search and download the compressed zip files by opening the following address :

https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/index.xml

The download links for the averaged_perceptron_tagger.zip and cmudict.zip data packages are:

https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/packages/taggers/averaged_perceptron_tagger.zip https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/packages/corpora/cmudict.zip

Then, upload the downloaded zip files to the file index location that was printed when running python download.py, such as /root/nltk_data or /root/miniconda3/envs/EmotiVoice/nltk_data. If the directory doesn't exist, create one and extract the zip files there.

After extraction, the directory structure of nltk_data should look like this:

├── nltk_data
│   ├── corpora
│   │   ├── cmudict
│   │   │   ├── README
│   │   │   └── cmudict
│   │   └── cmudict.zip
│   └── taggers
│       ├── averaged_perceptron_tagger
│       │   └── averaged_perceptron_tagger.pickle
│       └── averaged_perceptron_tagger.zip

Reference souce：运行报错 raise BadZipFile("File is not a zip file")