matrix-pygmalion-bot/matrix_pygmalion_bot/bot/wrappers/runpod.py

import asyncio
import requests
import json
import os, tempfile
import io
import base64
from PIL import Image, PngImagePlugin
import logging

logger = logging.getLogger(__name__)

class RunpodWrapper(object):
    """Base Class for runpod"""

    def __init__(self, api_key: str, endpoint_name: str, model_name: str):
        self.api_key = api_key
        self.endpoint_name = endpoint_name
        self.model_name = model_name

    async def generate(self, input_data: str, typing_fn, timeout=180):
        # Set the API endpoint URL
        endpoint = f"https://api.runpod.ai/v2/{self.endpoint_name}/run"

        # Set the headers for the request
        headers = {
            "Content-Type": "application/json",
            "Authorization": f"Bearer {self.api_key}"
        }

        logger.info(f"sending request to runpod.io. endpoint=\"{self.endpoint_name}\"")

        # Make the request
        try:
            r = requests.post(endpoint, json=input_data, headers=headers, timeout=timeout)
        except requests.exceptions.RequestException as e:
            raise ValueError(f"<HTTP ERROR>")
        r_json = r.json()
        logger.debug(r_json)

        if r.status_code == 200:
            status = r_json["status"]
            job_id = r_json["id"]
            TIMEOUT = 360
            DELAY = 5
            for i in range(TIMEOUT//DELAY):
                endpoint = f"https://api.runpod.ai/v2/{self.endpoint_name}/status/{job_id}"
                r = requests.get(endpoint, headers=headers)
                r_json = r.json()
                logger.info(r_json)
                status = r_json["status"]
                if status == 'IN_PROGRESS':
                    await typing_fn()
                    await asyncio.sleep(DELAY)
                elif status == 'IN_QUEUE':
                    await asyncio.sleep(DELAY)
                elif status == 'COMPLETED':
                    output = r_json["output"]
                    return output
                else:
                    err_msg = r_json["error"] if "error" in r_json else ""
                    err_msg = err_msg.replace("\\n", "\n")
                    raise ValueError(f"<ERROR> RETURN CODE {status}: {err_msg}")
            raise ValueError(f"<ERROR> TIMEOUT")
        else:
            raise ValueError(f"<ERROR>")


class RunpodTextWrapper(RunpodWrapper):

    async def generate(self, prompt, typing_fn, temperature=0.72, max_new_tokens=200, timeout=180):

        # Define your inputs
        input_data = {
            "input": {
                "prompt": prompt,
                "max_length": min(max_new_tokens, 2048),
                "temperature": bot.temperature,
                "do_sample": True,
            }
        }
        output = await super().generate(input_data, api_key, typing_fn, timeout)
        output = output.removeprefix(prompt)
        return(output)

    async def generate2(self, prompt, typing_fn, temperature=0.72, max_new_tokens=200, timeout=180):
        generate(prompt, typing_fn, temperature, nax_new_tokens, timeout)


class RunpodImageWrapper(RunpodWrapper):
    async def download_image(self, url, path):
        r = requests.get(url, stream=True)
        if r.status_code == 200:
            with open(path, 'wb') as f:
                for chunk in r:
                    f.write(chunk)

    async def generate(self, input_prompt: str, negative_prompt: str, typing_fn, timeout=180):

        # Define your inputs
        input_data = {
            "input": {
                "prompt": input_prompt,
                "negative_prompt": negative_prompt,
                "width": 512,
                "height": 768,
                "num_outputs": 3,
    #            "nsfw": True
            },
        }

        output = await super().generate(input_data, typing_fn, timeout)

        os.makedirs("./.data/images", exist_ok=True)
        files = []
        for image in output:
            temp_name = next(tempfile._get_candidate_names())
            filename = "./.data/images/" + temp_name + ".jpg"
            await self.download_image(image["image"], filename)
            files.append(filename)

        return files


class RunpodImageWrapper2(RunpodWrapper):
    async def download_image(self, url, path):
        r = requests.get(url, stream=True)
        if r.status_code == 200:
            with open(path, 'wb') as f:
                for chunk in r:
                    f.write(chunk)

    async def generate(self, input_prompt: str, negative_prompt: str, typing_fn, timeout=180):

        # Define your inputs
        input_data = {
            "input": {
                "prompt": input_prompt,
                "negative_prompt": negative_prompt,
                "h": 768,
                "w": 768,
                "num_images": 3,
                "seed": -1
            },
        }

        output = await super().generate(input_data, typing_fn, timeout)

        os.makedirs("./.data/images", exist_ok=True)
        files = []
        for image in output:
            temp_name = next(tempfile._get_candidate_names())
            filename = "./.data/images/" + temp_name + ".jpg"
            await self.download_image(image["image"], filename)
            files.append(filename)

        return files


class RunpodImageAutomaticWrapper(RunpodWrapper):

    async def generate(self, input_prompt: str, negative_prompt: str, typing_fn, timeout=180):

        # Define your inputs
        input_data = {
            "input": {
                "prompt": input_prompt,
                "nagative_prompt": negative_prompt,
                "steps": 25,
                "cfg_scale": 7,
                "seed": -1,
                "width": 512,
                "height": 768,
                "batch_size": 3,
        #        "sampler_index": "DPM++ 2M Karras",
        #        "enable_hr": True,
        #        "hr_scale": 2,
        #        "hr_upscaler": "ESRGAN_4x", # "Latent"
        #        "denoising_strength": 0.5,
        #        "hr_second_pass_steps": 15,
                "restore_faces": True,
        #        "gfpgan_visibility": 0.5,
        #        "codeformer_visibility": 0.5,
        #        "codeformer_weight": 0.5,
        ##        "override_settings": {
        ##            "filter_nsfw": False,
        ##        },
                "api_endpoint": "txt2img",
            },
            "cmd": "txt2img"
        }

        output = await super().generate(input_data, typing_fn, timeout)

        upscale = False
        if upscale:
            count = 0
            for i in output['images']:
                payload = {
                    "init_images": [i],
                    "prompt": input_prompt,
                    "nagative_prompt": negative_prompt,
                    "steps": 20,
                    "seed": -1,
                    #"sampler_index": "Euler",
    #                                  tile_width, tile_height, mask_blur, padding, seams_fix_width, seams_fix_denoise, seams_fix_padding, upscaler_index, save_upscaled_image, redraw_mode, save_seams_fix_image, seams_fix_mask_blur, seams_fix_type, target_size_type, custom_width, custom_height, custom_scale
    #                "script_args": ["",512,0,8,32,64,0.275,32,3,False,0,True,8,3,2,1080,1440,1.875],
    #                "script_name": "Ultimate SD upscale",
                }
                upscaled_output = await serverless_automatic_request(payload, "img2img", api_url, api_key, typing_fn)
                output['images'][count] = upscaled_output['images'][count]


        os.makedirs("./.data/images", exist_ok=True)
        files = []
        for i in output['images']:
            temp_name = next(tempfile._get_candidate_names())
            filename = "./.data/images/" + temp_name + ".png"
            image = Image.open(io.BytesIO(base64.b64decode(i.split(",",1)[0])))
            info = output['info']
            parameters = output['parameters']
            pnginfo = PngImagePlugin.PngInfo()
            pnginfo.add_text("parameters", info)
            image.save(filename, pnginfo=pnginfo)
            files.append(filename)

        return files