[BUG] TTS Server starting for OuteTTS not possible

[mindreframer]

Issue Description

I do not understand, how I should start OuteTTS model server in TTS mode. If I just provide the OuteTTS-0.2-500M:q8_0 name, CLI ignores the model_type argument. If I provide model_path + local_path (with no value, to indicate, that we deal with a local path), it fails with Model /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf not found in Model Hub error.

Steps to Reproduce

$ nexa server --model_type TTS  --model_path OuteTTS-0.2-500M:q8_0
INFO:     Started server process [5006]
INFO:     Waiting for application startup.
Model OuteTTS-0.2-500M:q8_0 already exists at /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf
model_type: NLP
INFO:     Application startup complete.
INFO:     Uvicorn running on http://localhost:8000 (Press CTRL+C to quit)
^CINFO:     Shutting down

Now with model_path + local_path + model_type

nexa server  --model_path /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf --model_type TTS -lp
INFO:     Started server process [5653]
INFO:     Waiting for application startup.
model_type: TTS
ERROR:    Traceback (most recent call last):
  File "starlette/routing.py", line 693, in lifespan
  File "starlette/routing.py", line 569, in __aenter__
  File "starlette/routing.py", line 670, in startup
  File "nexa/gguf/server/nexa_service.py", line 704, in startup_event
  File "nexa/gguf/server/nexa_service.py", line 467, in load_model
ValueError: Model /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf not found in Model Hub. If you are using local path, be sure to add --local_path and --model_type flags.

OS

Mac Os Sonoma (14.7 (23H124))

Python Version

3.12.2

Nexa SDK Version

nexa -V 0.0.9.7

GPU (if using one)

Apple M1 Air

[Davidqian123]

Now we have already supported TTS server

[keithem]

OK, give me a minute Get Outlook for iOS <https://aka.ms/o0ukef> ------------------------------ *From:* Roman Heinrich ***@***.***> *Sent:* Tuesday, January 7, 2025 10:21:42 AM *To:* NexaAI/nexa-sdk ***@***.***> *Cc:* Subscribed ***@***.***> *Subject:* [NexaAI/nexa-sdk] [BUG] TTS Server starting for OuteTTS not possible (Issue #346) Issue Description I do not understand, how I should start OuteTTS model server in TTS mode. If I just provide the OuteTTS-0.2-500M:q8_0 name, CLI ignores the model_type argument. If I provide model_path + local_path (with no value, to indicate, that we deal with a local path), it fails with Model /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf not found in Model Hub error. Steps to Reproduce $ nexa server --model_type TTS --model_path OuteTTS-0.2-500M:q8_0 INFO: Started server process [5006] INFO: Waiting for application startup. Model OuteTTS-0.2-500M:q8_0 already exists at /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf model_type: NLP INFO: Application startup complete. INFO: Uvicorn running on http://localhost:8000 (Press CTRL+C to quit) ^CINFO: Shutting down Now with model_path + local_path + model_type nexa server --model_path /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf --model_type TTS -lp INFO: Started server process [5653] INFO: Waiting for application startup. model_type: TTS ERROR: Traceback (most recent call last): File "starlette/routing.py", line 693, in lifespan File "starlette/routing.py", line 569, in __aenter__ File "starlette/routing.py", line 670, in startup File "nexa/gguf/server/nexa_service.py", line 704, in startup_event File "nexa/gguf/server/nexa_service.py", line 467, in load_model ValueError: Model /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf not found in Model Hub. If you are using local path, be sure to add --local_path and --model_type flags. OS Mac Os Sonoma (14.7 (23H124)) Python Version 3.12.2 Nexa SDK Version nexa -V 0.0.9.7 GPU (if using one) Apple M1 Air — Reply to this email directly, view it on GitHub <#346>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BN5OQI6YJLLZH33PXLP77632JPWINAVCNFSM6AAAAABUX5RCV2VHI2DSMVQWIX3LMV43ASLTON2WKOZSG43TGMJQGMYDENQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[keithem]

Subject: Re: Issue with OuteTTS Model Server Hi [Recipient], I hope this email finds you well. I wanted to follow up on the issue you are experiencing with the OuteTTS model server. I recommend reviewing the documentation for the OuteTTS model server to ensure you are using the correct arguments. If you still encounter issues, feel free to reach out for further assistance. Please let me know if you need any additional help. Best regards, [Your Name]

…

On Wed, Feb 5, 2025 at 10:52 AM Keith Cox ***@***.***> wrote: OK, give me a minute Get Outlook for iOS <https://aka.ms/o0ukef> ------------------------------ *From:* Roman Heinrich ***@***.***> *Sent:* Tuesday, January 7, 2025 10:21:42 AM *To:* NexaAI/nexa-sdk ***@***.***> *Cc:* Subscribed ***@***.***> *Subject:* [NexaAI/nexa-sdk] [BUG] TTS Server starting for OuteTTS not possible (Issue #346) Issue Description I do not understand, how I should start OuteTTS model server in TTS mode. If I just provide the OuteTTS-0.2-500M:q8_0 name, CLI ignores the model_type argument. If I provide model_path + local_path (with no value, to indicate, that we deal with a local path), it fails with Model /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf not found in Model Hub error. Steps to Reproduce $ nexa server --model_type TTS --model_path OuteTTS-0.2-500M:q8_0 INFO: Started server process [5006] INFO: Waiting for application startup. Model OuteTTS-0.2-500M:q8_0 already exists at /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf model_type: NLP INFO: Application startup complete. INFO: Uvicorn running on http://localhost:8000 (Press CTRL+C to quit) ^CINFO: Shutting down Now with model_path + local_path + model_type nexa server --model_path /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf --model_type TTS -lp INFO: Started server process [5653] INFO: Waiting for application startup. model_type: TTS ERROR: Traceback (most recent call last): File "starlette/routing.py", line 693, in lifespan File "starlette/routing.py", line 569, in __aenter__ File "starlette/routing.py", line 670, in startup File "nexa/gguf/server/nexa_service.py", line 704, in startup_event File "nexa/gguf/server/nexa_service.py", line 467, in load_model ValueError: Model /Users/roman/.cache/nexa/hub/official/OuteTTS-0.2-500M/q8_0.gguf not found in Model Hub. If you are using local path, be sure to add --local_path and --model_type flags. OS Mac Os Sonoma (14.7 (23H124)) Python Version 3.12.2 Nexa SDK Version nexa -V 0.0.9.7 GPU (if using one) Apple M1 Air — Reply to this email directly, view it on GitHub <#346>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BN5OQI6YJLLZH33PXLP77632JPWINAVCNFSM6AAAAABUX5RCV2VHI2DSMVQWIX3LMV43ASLTON2WKOZSG43TGMJQGMYDENQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>