Closed Innomen closed 5 months ago
Hi @Innomen
Youve not given me much to go on as to how you got to this point. e.g. have you updated? did you git pull? Did you re-run the requirements with atsetup (Standalone > Option 1)?
Either way, "PytorchStreamReader failed reading zip archive: failed finding central directory" suggests your model files in your models folder is corrupted in some way.
You may wish to re-name the xttsv2_2.0.2
folder to a xttsv2_2.0.2.old
and let it download a fresh copy of the model files and confirm it works at that.
You should have the following files and they should be the sizes shown:
Thanks
I couldn't make update work at all so I started from scratch. This is the behavior of a fresh install.
Hi @Innomen
Could you please:
1) Confirm the model files exist and match the size shown in the folder alltalk_tts\models\xttsv2_2.0.2
2) if not, could you confirm how did the setup. Did you git clone
or did you pull down the zip file?
3) Could you please provide me a diagnostics log file, which you can generate by using atsetup.
Thanks
You were right about file sizes. The model.pth was 22 megs, I copied the 1gb one from my old install folder, and now it will launch.
I couldn't ever generate a diagnostics file until after I copied over the model file:
8) Generate a diagnostics file
9) Exit/Quit
Enter your choice: 8 The input line is too long.
E:\alltalk_tts>
Also I'm using the zip, not git clone. diagnostics.log
Seems like it's working now.
I think for some reason it wasn't downloading the model file? If I encounter any other issues I'll pester you. Thanks for pushing the update.
Hi @Innomen
Great glad its sorted. The only other time I've seen this behaviour is when Huggingface, where the model is downloaded from, has outages or problems. So its likely to be related somehow to internet connectivity type behaviour. Its a rare occurrence, but probably something I should add an entry in the help section for.
Thanks
[AllTalk Startup] DeepSpeed Detected [AllTalk Startup] Activate DeepSpeed in AllTalk settings [AllTalk Model] XTTSv2 Local Loading xttsv2_2.0.2 into cpu ERROR: Traceback (most recent call last): File "E:\alltalk_tts\alltalk_environment\env\Lib\site-packages\starlette\routing.py", line 734, in lifespan async with self.lifespan_context(app) as maybe_state: File "E:\alltalk_tts\alltalk_environment\env\Lib\contextlib.py", line 210, in aenter return await anext(self.gen) ^^^^^^^^^^^^^^^^^^^^^ File "E:\alltalk_tts\tts_server.py", line 132, in startup_shutdown await setup() File "E:\alltalk_tts\tts_server.py", line 177, in setup model = await xtts_manual_load_model() ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\alltalk_tts\tts_server.py", line 248, in xtts_manual_load_model model.load_checkpoint( File "E:\alltalk_tts\alltalk_environment\env\Lib\site-packages\TTS\tts\models\xtts.py", line 771, in load_checkpoint checkpoint = self.get_compatible_checkpoint_state_dict(model_path) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\alltalk_tts\alltalk_environment\env\Lib\site-packages\TTS\tts\models\xtts.py", line 714, in get_compatible_checkpoint_state_dict checkpoint = load_fsspec(model_path, map_location=torch.device("cpu"))["model"] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\alltalk_tts\alltalk_environment\env\Lib\site-packages\TTS\utils\io.py", line 54, in load_fsspec return torch.load(f, map_location=map_location, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\alltalk_tts\alltalk_environment\env\Lib\site-packages\torch\serialization.py", line 1005, in load with _open_zipfile_reader(opened_file) as opened_zipfile: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\alltalk_tts\alltalk_environment\env\Lib\site-packages\torch\serialization.py", line 457, in init super().init(torch._C.PyTorchFileReader(name_or_buffer)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory
ERROR: Application startup failed. Exiting. [AllTalk Startup] Warning TTS Subprocess has NOT started up yet, Will keep trying for 120 seconds maximum. Please wait.