Spaces:
Runtime error
Runtime error
Download app.py from ananya1011/tinymonkeystage: direct link, hf CLI and curl.
- Browser
- Download file 3.51 kB
-
https://huggingface.co/spaces/ananya1011/tinymonkeystage/resolve/main/app.py
- Command line
-
hf download hf://spaces/ananya1011/tinymonkeystage/app.py
-
curl -L -o app.py https://huggingface.co/spaces/ananya1011/tinymonkeystage/resolve/main/app.py
3.51 kB
| import openai | |
| import json | |
| from flask import Flask, request, jsonify | |
| from flask_cors import CORS, cross_origin | |
| app = Flask(__name__) | |
| cors = CORS(app) | |
| app.config['CORS_HEADERS'] = 'Content-Type' | |
| openai.api_key = "" | |
| db = {} | |
| def get_continue_story(): | |
| args = request.args | |
| if args['id'] not in db.keys() and not len(args['decision']): | |
| db[args['id']] = [ | |
| {"role": "system", "content": f"""You're a story writer. The mood of the story: {args['mood']}. | |
| The main character: {args['mainCharacter']}. | |
| Setting of the story: {args['settingStory']}."""}] | |
| if args['id'] not in db.keys() and len(args['decision']): | |
| db[args['id']] = [ | |
| {"role": "system", "content": f"""You're a story writer. | |
| The beginning of the story is as follows: {args['decision']}"""}] | |
| args['decision'] = "" | |
| if len(args['decision']): | |
| db[args['id']].append({"role": "assistant", "content": args['decision']}) | |
| db[args['id']].append({"role": "user", "content": args['storyPoint']}) | |
| try: | |
| completion = openai.ChatCompletion.create( | |
| model="gpt-3.5-turbo", | |
| messages=db[args['id']], | |
| temperature=0.8, | |
| max_tokens=340 | |
| ) | |
| db[args['id']].append(completion.choices[0].message) | |
| except Exception as e: | |
| print(e) | |
| return jsonify({ "ok": False, "error": str(e)}), 500 | |
| resp = args['decision'] + completion.choices[0].message.content | |
| if resp[-1] != '.': | |
| resp = crop_text(resp) | |
| if 'happy ending' in args['storyPoint']: | |
| with open('./bd.txt', 'a') as fl: | |
| story = '\n'.join([i['content'] for i in db[args['id']]]) | |
| fl.write(f"""{args['id']} | |
| {story}""" + "\n") | |
| return jsonify({ "ok": True, "new_part": resp }) | |
| def get_story_decisions(): | |
| args = request.args | |
| while True: | |
| completion = openai.ChatCompletion.create( | |
| model="gpt-3.5-turbo", | |
| messages=get_decision(db[args['id']]), | |
| temperature=0.5, | |
| max_tokens=340 | |
| ) | |
| resp = completion.choices[0].message.content | |
| print(resp) | |
| try: | |
| res = response_processing(resp) | |
| if len(res.keys()) != 3: | |
| continue | |
| except Exception as e: | |
| print(e) | |
| continue | |
| else: | |
| break | |
| return jsonify({ "ok": True, "data": res }) | |
| def response_processing(resp): | |
| result = {} | |
| mas = [i for i in resp.split('{')][1:] | |
| mas = [i.split('}') for i in mas] | |
| mas = [i[0].replace('\n', '') for i in mas] | |
| for m in mas: | |
| c = m.split(':') | |
| c = [i.split('"') for i in c] | |
| result[c[0][1]] = c[1][1] | |
| return json.loads(json.dumps(result)) | |
| def get_decision(db): | |
| return db + [{ | |
| "role": "user", "content": """Come up with three short different story continuations. | |
| Write them to me in json format, where the key is the name of the continuation, and the value is the text. | |
| Output example: | |
| { | |
| 'name_story': 'value', | |
| 'name_story': 'value', | |
| 'name_story': 'value' | |
| }""" | |
| }] | |
| def crop_text(text): | |
| text_arr = text.split(".") | |
| res = ".".join(text_arr[:-1]) | |
| return f'{res}.' | |
| if __name__ == '__main__': | |
| app.run(debug=True,port=5623) | |