Tested on windows?

[blakejrobinson]

Was keen to give this a try as it looks very cool, but ran into issues on windows.

Install gives hardlink error

F:\Software\AI\GlaDOS>python scripts\install.py
UV is already installed
Using CPython 3.12.8
Creating virtual environment at: .venv
Activate with: .venv\Scripts\activate
Resolved 44 packages in 32ms
      Built glados @ file:///F:/Software/AI/GlaDOS
Prepared 1 package in 551ms
░░░░░░░░░░░░░░░░░░░░ [0/44] Installing wheels...                                                                                                                                                                                    warning: Failed to hardlink files; falling back to full copy. This may lead to degraded performance.
         If the cache and target directories are on different filesystems, hardlinking may not be supported.
         If this is intentional, set `export UV_LINK_MODE=copy` or use `--link-mode=copy` to suppress this warning.
Installed 44 packages in 674ms
 + annotated-types==0.7.0
 + certifi==2025.1.31
 + cffi==1.17.1
 + charset-normalizer==3.4.1
 + colorama==0.4.6
 + coloredlogs==15.0.1
 + flatbuffers==25.2.10
 + glados==0.1.0 (from file:///F:/Software/AI/GlaDOS)
 + humanfriendly==10.0
 + idna==3.10
 + jinja2==3.1.6
 + levenshtein==0.27.1
 + linkify-it-py==2.0.3
 + llvmlite==0.44.0
 + loguru==0.7.3
 + markdown-it-py==3.0.0
 + markupsafe==3.0.2
 + mdit-py-plugins==0.4.2
 + mdurl==0.1.2
 + mpmath==1.3.0
 + numba==0.61.0
 + numpy==2.1.3
 + onnxruntime-gpu==1.21.0
 + packaging==24.2
 + platformdirs==4.3.7
 + protobuf==6.30.2
 + pycparser==2.22
 + pydantic==2.11.1
 + pydantic-core==2.33.0
 + pygments==2.19.1
 + pyreadline3==3.5.4
 + pyyaml==6.0.2
 + rapidfuzz==3.12.2
 + requests==2.32.3
 + rich==13.9.4
 + sounddevice==0.5.1
 + soundfile==0.13.1
 + sympy==1.13.3
 + textual==3.0.0
 + typing-extensions==4.13.0
 + typing-inspection==0.4.0
 + uc-micro-py==1.0.3
 + urllib3==2.3.0
 + win32-setctime==1.2.0

uv run glados tui fails to run

╭─────────────────────────────────────────────────────────────────────────────────────────────── Traceback (most recent call last) ────────────────────────────────────────────────────────────────────────────────────────────────╮
│ F:\Software\AI\GlaDOS\src\glados\tui.py:114 in _display_next_char                                                                                                                                                                │
│                                                                                                                                                                                                                                  │
│   111 │   │   try:                                                                             ╭────── locals ───────╮                                                                                                           │
│   112 │   │   │   if not self._vertical_scroll.is_vertical_scroll_end:                         │ self = Typewriter() │                                                                                                           │
│   113 │   │   │   │   self._vertical_scroll.scroll_down()                                      ╰─────────────────────╯                                                                                                           │
│ ❱ 114 │   │   │   self._static.update(next(self._iter_text))                                                                                                                                                                     │
│   115 │   │   except StopIteration:                                                                                                                                                                                              │
│   116 │   │   │   if self._repeat:                                                                                                                                                                                               │
│   117 │   │   │   │   self._iter_text = self._get_iterator()                                                                                                                                                                     │
│                                                                                                                                                                                                                                  │
│ F:\Software\AI\GlaDOS\.venv\Lib\site-packages\textual\widgets\_static.py:106 in update                                                                                                                                           │
│                                                                                                                                                                                                                                  │
│   103 │   │   """                                                                              ╭───────── locals ─────────╮                                                                                                      │
│   104 │   │                                                                                    │ content = '[[blink]▃[/]' │                                                                                                      │
│   105 │   │   self._content = content                                                          │    self = Static()       │                                                                                                      │
│ ❱ 106 │   │   self._visual = visualize(self, content, markup=self._render_markup)              ╰──────────────────────────╯                                                                                                      │
│   107 │   │   self.refresh(layout=True)                                                                                                                                                                                          │
│   108                                                                                                                                                                                                                            │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
MarkupError: Expected markup style value (found '[blink]▃[/]').

NOTE: 1 of 2 errors shown. Run with textual run --dev to see all errors.

and running normal (uv run glados) does not seem to work. I hear it say 'All n-' and then it cuts off. I can see speech being detected but GLaDOS doesn't seem to be talking or responding.

I imagine it's going to be some kind of dependency issue somewhere.

I'm on Windows 10 Pro 19045, Python 3.12.8

[dnhkng]

Is Ollama installed and running correctly?

edit configs/glados_config.yaml, and make this change:

Glados:
  completion_url: "http://localhost:11434/api/chat"
  model: "llama3.2"
  api_key: "Bearer your_api_key_here"
  interruptible: false
  ...

[blakejrobinson]

Ah good shout on ollama - I did have it installed but I'd changed the port it listens on - updating the config fixed that.

I still can't run tui mode, but basic mode runs for me now.

[scarecrow180]

Find the app.py file at ... \GLaDOS\src\glados\api\app.py

replace the app.py with the following:

from dataclasses import dataclass
import io
from typing import Literal

from litestar import Litestar, post, get
from litestar.response import Stream

from .log import structlog_plugin
from .tts import write_glados_audio_file

# Define supported voice and response formats                                                                                     
Voice = Literal["glados"]
ResponseFormat = Literal["mp3", "wav", "ogg"]

@dataclass
class RequestData:
    input: str
    model: str = "glados"
    voice: Voice = "glados"
    response_format: ResponseFormat = "mp3"
    speed: float = 1.0

CONTENT_TYPES: dict[ResponseFormat, str] = {
    "mp3": "audio/mpeg",
    "wav": "audio/wav",
    "ogg": "audio/ogg",
}

@post("/v1/audio/speech")
async def create_speech(data: RequestData) -> Stream:
    """
    Generate speech audio from input text.

    Parameters:
        data: The request data containing input text and speech parameters

    Returns:
        Stream: Stream of bytes data containing the generated speech
    """
    # TODO: Handle other voices and speed variations
    buffer = io.BytesIO()
    write_glados_audio_file(buffer, data.input, format=data.response_format)
    buffer.seek(0)
    return Stream(
        buffer,
        headers={
            "content-type": CONTENT_TYPES[data.response_format],
            "content-disposition": f'attachment; filename="speech.{data.response_format}"',
        },
    )

# NEW ENDPOINT: GET /v1/audio/voices
@get("/v1/audio/voices")
async def get_voices() -> dict:
    """
    Return a list of available voices.
    """
    return {
        "data": [
            {
                "voice_id": "glados",
                "name": "GLaDOS AI",
                "description": "The default GLaDOS voice.",
            }
        ]
    }

# NEW ENDPOINT: GET /v1/audio/models                                                                                                                    the cake is a lie
@get("/v1/audio/models")
async def get_models() -> dict:
    """
    Return a list of available models.
    """
    return {
        "data": [
            {
                "model_id": "glados",
                "name": "GLaDOS Model",
                "description": "The default TTS model.",
            }
        ]
    }

# Register all endpoints with the Litestar app
app = Litestar(
    route_handlers=[create_speech, get_voices, get_models],
    plugins=[structlog_plugin]
)

In your open-webui or otherwise open ai compatible, use http://localhost:5050/v1 for the url. Api key doesn't matter, but most importantly you have to type glados in each selection box for voice and model. just: glados

it took me a year to figure this out...

[dnhkng]

@scarecrow180 If you find a bug, please make a pull request! I need all the help I can get!

[Anadeprado]

Hi cake-lovers 🍰

I have tested it on Windows 11 (GPU NVIDIA GeForce RTX 2070 8Gb) and it works perfectly, fast as a rocket! 🚀
But I had 2 problems when installing. I comment them in case it helps anyone:

1. onnxruntime-gpu libs issue:

 ImportError: DLL load failed while importing onnxruntime_pybind11_state

Solved by installing the latest Microsoft Visual C++ Redistributable

2. I wasn't sure whether I needed to create my own Ollama server or if one was already set up internally. I also wasn’t sure how to create it. It would be helpful to add a third step in the 'Set up a local LLM server' section, stating: 'You can start your Ollama server by running ollama serve.

I will open a pull request with everything I find.
Thanks for the great project!! 🤖❤

[dnhkng]

Yes, please write a PR on Windows installation. I don't use Windows so often.

[dnhkng]

@Anadeprado Hows the PR going?