updates

Dockerfile

@@ -37,7 +37,7 @@ COPY --chown=user . $HOME/app
 RUN pip install gdown
 RUN --mount=type=secret,id=truepic_sign_gdoc_id,mode=0444,required=true \
    gdown --id $(cat /run/secrets/truepic_sign_gdoc_id)
-RUN tar -xf truepic-sign-v0.2.0-ubuntu-20.04.tar.gz
+RUN tar -xf truepic-sign-v0.2.0-ubuntu-20.04.tar.gz 
 RUN chmod +x truepic-sign
 
 WORKDIR $HOME/app

app.py

@@ -1,15 +0,0 @@
-from text_to_image import TextToImageTool
-import gradio as gr
-
-tool = TextToImageTool()
-
-def fn(*args, **kwargs):
-    return tool(*args, **kwargs)
-
-gr.Interface(
-    fn=fn,
-    inputs=tool.inputs,
-    outputs=tool.outputs,
-    title="TextToImageTool",
-    article=tool.description,
-).queue(concurrency_count=5).launch()

main.py

@@ -13,7 +13,7 @@ import uuid
 import torch
 from diffusers import StableDiffusionPipeline, DPMSolverMultistepScheduler, EulerDiscreteScheduler
 
-app = FastAPI()
+app = FastAPI() 
 
 @app.get("/generate")
 def generate_image(prompt, inference_steps, model):

text_to_image.py

@@ -1,50 +0,0 @@
-from transformers.tools.base import Tool, get_default_device
-from transformers.utils import is_accelerate_available
-import torch
-
-from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler
-
-
-TEXT_TO_IMAGE_DESCRIPTION = (
-    "This is a tool that creates an image according to a prompt, which is a text description. It takes an input named `prompt` which "
-    "contains the image description and outputs an image."
-)
-
-
-class TextToImageTool(Tool):
-    default_checkpoint = "runwayml/stable-diffusion-v1-5"
-    description = TEXT_TO_IMAGE_DESCRIPTION
-    inputs = ['text']
-    outputs = ['image']
-
-    def __init__(self, device=None, **hub_kwargs) -> None:
-        if not is_accelerate_available():
-            raise ImportError("Accelerate should be installed in order to use tools.")
-
-        super().__init__()
-
-        self.device = device
-        self.pipeline = None
-        self.hub_kwargs = hub_kwargs
-
-    def setup(self):
-        if self.device is None:
-            self.device = get_default_device()
-
-        self.pipeline = DiffusionPipeline.from_pretrained(self.default_checkpoint)
-        self.pipeline.scheduler = DPMSolverMultistepScheduler.from_config(self.pipeline.scheduler.config)
-        self.pipeline.to(self.device)
-
-        if self.device.type == "cuda":
-            self.pipeline.to(torch_dtype=torch.float16)
-
-        self.is_initialized = True
-
-    def __call__(self, prompt):
-        if not self.is_initialized:
-            self.setup()
-
-        negative_prompt = "low quality, bad quality, deformed, low resolution"
-        added_prompt = " , highest quality, highly realistic, very high resolution"
-
-        return self.pipeline(prompt + added_prompt, negative_prompt=negative_prompt, num_inference_steps=25).images[0]