###### # Project : GPT4ALL-UI # File : api.py # Author : ParisNeo with the help of the community # Supported by Nomic-AI # Licence : Apache 2.0 # Description : # A simple api to communicate with gpt4all-ui and its models. ###### import gc import sys from datetime import datetime from gpt4all_api.db import DiscussionsDB from pathlib import Path import importlib from pyaipersonality import AIPersonality import multiprocessing as mp import threading import time import requests import urllib.request from tqdm import tqdm __author__ = "parisneo" __github__ = "https://github.com/nomic-ai/gpt4all-ui" __copyright__ = "Copyright 2023, " __license__ = "Apache 2.0" import subprocess import pkg_resources # =========================================================== # Manage automatic install scripts def is_package_installed(package_name): try: dist = pkg_resources.get_distribution(package_name) return True except pkg_resources.DistributionNotFound: return False def install_package(package_name): try: # Check if the package is already installed __import__(package_name) print(f"{package_name} is already installed.") except ImportError: print(f"{package_name} is not installed. Installing...") # Install the package using pip subprocess.check_call(["pip", "install", package_name]) print(f"{package_name} has been successfully installed.") def parse_requirements_file(requirements_path): with open(requirements_path, 'r') as f: for line in f: line = line.strip() if not line or line.startswith('#'): # Skip empty and commented lines continue package_name, _, version_specifier = line.partition('==') package_name, _, version_specifier = line.partition('>=') if is_package_installed(package_name): # The package is already installed print(f"{package_name} is already installed.") else: # The package is not installed, install it if version_specifier: install_package(f"{package_name}{version_specifier}") else: install_package(package_name) # =========================================================== class ModelProcess: def __init__(self, config=None): self.config = config self.generate_queue = mp.Queue() self.generation_queue = mp.Queue() self.cancel_queue = mp.Queue(maxsize=1) self.clear_queue_queue = mp.Queue(maxsize=1) self.set_config_queue = mp.Queue(maxsize=1) self.set_config_result_queue = mp.Queue(maxsize=1) self.started_queue = mp.Queue() self.process = None self.is_generating = mp.Value('i', 0) self.model_ready = mp.Value('i', 0) self.ready = False self.id=0 self.n_predict=2048 self.reset_config_result() def reset_config_result(self): self._set_config_result = { 'status': 'succeeded', 'backend_status':'ok', 'model_status':'ok', 'personality_status':'ok', 'errors':[] } def load_backend(self, backend_name:str): backend_path = Path("backends")/backend_name # first find out if there is a requirements.txt file requirements_file = backend_path/"requirements.txt" if requirements_file.exists(): parse_requirements_file(requirements_file) # define the full absolute path to the module absolute_path = backend_path.resolve() # infer the module name from the file path module_name = backend_path.stem # use importlib to load the module from the file path loader = importlib.machinery.SourceFileLoader(module_name, str(absolute_path/"__init__.py")) backend_module = loader.load_module() backend_class = getattr(backend_module, backend_module.backend_name) return backend_class def start(self): if self.process is None: self.process = mp.Process(target=self._run) self.process.start() def stop(self): if self.process is not None: self.generate_queue.put(None) self.process.join() self.process = None def set_backend(self, backend_path): self.backend = backend_path def set_model(self, model_path): self.model = model_path def set_config(self, config): self.set_config_queue.put(config) # Wait for it t o be consumed while self.set_config_result_queue.empty(): time.sleep(0.5) return self.set_config_result_queue.get() def generate(self, full_prompt, prompt, id, n_predict): self.generate_queue.put((full_prompt, prompt, id, n_predict)) def cancel_generation(self): self.cancel_queue.put(('cancel',)) def clear_queue(self): self.clear_queue_queue.put(('clear_queue',)) def rebuild_backend(self, config): try: backend = self.load_backend(config["backend"]) print("Backend loaded successfully") except Exception as ex: print("Couldn't build backend") print(ex) backend = None self._set_config_result['backend_status'] ='failed' self._set_config_result['errors'].append(f"couldn't build backend:{ex}") return backend def _rebuild_model(self): try: print("Rebuilding model") self.backend = self.load_backend(self.config["backend"]) print("Backend loaded successfully") try: model_file = Path("models")/self.config["backend"]/self.config["model"] print(f"Loading model : {model_file}") self.model = self.backend(self.config) self.model_ready.value = 1 print("Model created successfully\n") except Exception as ex: print("Couldn't build model") print(ex) self.model = None self._set_config_result['model_status'] ='failed' self._set_config_result['errors'].append(f"couldn't build model:{ex}") except Exception as ex: print("Couldn't build backend") print(ex) self.backend = None self.model = None def rebuild_personality(self): try: personality_path = f"personalities/{self.config['personality_language']}/{self.config['personality_category']}/{self.config['personality']}" personality = AIPersonality(personality_path) except Exception as ex: print(f"Personality file not found or is corrupted ({personality_path}).\nPlease verify that the personality you have selected exists or select another personality. Some updates may lead to change in personality name or category, so check the personality selection in settings to be sure.") if self.config["debug"]: print(ex) personality = AIPersonality() return personality def _rebuild_personality(self): try: personality_path = f"personalities/{self.config['personality_language']}/{self.config['personality_category']}/{self.config['personality']}" self.personality = AIPersonality(personality_path) except Exception as ex: print(f"Personality file not found or is corrupted ({personality_path}).\nPlease verify that the personality you have selected exists or select another personality. Some updates may lead to change in personality name or category, so check the personality selection in settings to be sure.") if self.config["debug"]: print(ex) self.personality = AIPersonality() self._set_config_result['personality_status'] ='failed' self._set_config_result['errors'].append(f"couldn't load personality:{ex}") def step_callback(self, text, message_type): self.generation_queue.put((text,self.id, message_type)) def _run(self): self._rebuild_model() self._rebuild_personality() if self.model_ready.value == 1: self.n_predict = 1 self._generate("I",1) print() print("Ready to receive data") else: print("No model loaded. Waiting for new configuration instructions") self.ready = True print(f"Listening on :http://{self.config['host']}:{self.config['port']}") while True: try: self._check_set_config_queue() self._check_cancel_queue() self._check_clear_queue() if not self.generate_queue.empty(): command = self.generate_queue.get() if command is None: break if self.cancel_queue.empty() and self.clear_queue_queue.empty(): self.is_generating.value = 1 self.started_queue.put(1) self.id=command[2] self.n_predict=command[3] if self.personality.processor is not None: if self.personality.processor_cfg is not None: if "custom_workflow" in self.personality.processor_cfg: if self.personality.processor_cfg["custom_workflow"]: output = self.personality.processor.run_workflow(self._generate, command[1], command[0], self.step_callback) self._callback(output) self.is_generating.value = 0 continue self._generate(command[0], self.n_predict, self._callback) while not self.generation_queue.empty(): time.sleep(1) self.is_generating.value = 0 time.sleep(1) except Exception as ex: time.sleep(1) print(ex) def _generate(self, prompt, n_predict=50, callback=None): if self.model is not None: self.id = self.id if self.config["override_personality_model_parameters"]: output = self.model.generate( prompt, new_text_callback=callback, n_predict=n_predict, temp=self.config['temperature'], top_k=self.config['top_k'], top_p=self.config['top_p'], repeat_penalty=self.config['repeat_penalty'], repeat_last_n = self.config['repeat_last_n'], seed=self.config['seed'], n_threads=self.config['n_threads'] ) else: output = self.model.generate( prompt, new_text_callback=callback, n_predict=self.n_predict, temp=self.personality.model_temperature, top_k=self.personality.model_top_k, top_p=self.personality.model_top_p, repeat_penalty=self.personality.model_repeat_penalty, repeat_last_n = self.personality.model_repeat_last_n, #seed=self.config['seed'], n_threads=self.config['n_threads'] ) else: print("No model is installed or selected. Please make sure to install a model and select it inside your configuration before attempting to communicate with the model.") print("To do this: Install the model to your models/ folder.") print("Then set your model information in your local configuration file that you can find in configs/local_default.yaml") print("You can also use the ui to set your model in the settings page.") output = "" return output def _callback(self, text): if not self.ready: print(".",end="") sys.stdout.flush() return True else: # Stream the generated text to the main process self.generation_queue.put((text,self.id)) self._check_set_config_queue() self._check_cancel_queue() self._check_clear_queue() # if stop generation is detected then stop if self.is_generating.value==1: return True else: return False def _check_cancel_queue(self): while not self.cancel_queue.empty(): command = self.cancel_queue.get() if command is not None: self._cancel_generation() def _check_clear_queue(self): while not self.clear_queue_queue.empty(): command = self.clear_queue_queue.get() if command is not None: self._clear_queue() def _check_set_config_queue(self): while not self.set_config_queue.empty(): config = self.set_config_queue.get() if config is not None: print("Inference process : Setting configuration") self.reset_config_result() self._set_config(config) self.set_config_result_queue.put(self._set_config_result) def _cancel_generation(self): self.is_generating.value = 0 def _clear_queue(self): while not self.generate_queue.empty(): self.generate_queue.get() def _set_config(self, config): bk_cfg = self.config self.config = config print("Changing configuration") # verify that the backend is the same if self.config["backend"]!=bk_cfg["backend"] or self.config["model"]!=bk_cfg["model"]: self._rebuild_model() # verify that the personality is the same if self.config["personality"]!=bk_cfg["personality"] or self.config["personality_category"]!=bk_cfg["personality_category"] or self.config["personality_language"]!=bk_cfg["personality_language"]: self._rebuild_personality() class GPT4AllAPI(): def __init__(self, config:dict, socketio, config_file_path:str) -> None: self.socketio = socketio #Create and launch the process self.process = ModelProcess(config) self.process.start() self.config = config self.backend = self.process.rebuild_backend(self.config) self.personality = self.process.rebuild_personality() if config["debug"]: print(print(f"{self.personality}")) self.config_file_path = config_file_path self.cancel_gen = False # Keeping track of current discussion and message self.current_discussion = None self._current_user_message_id = 0 self._current_ai_message_id = 0 self._message_id = 0 self.db_path = config["db_path"] # Create database object self.db = DiscussionsDB(self.db_path) # If the database is empty, populate it with tables self.db.populate() # This is used to keep track of messages self.full_message_list = [] # ========================================================================================= # Socket IO stuff # ========================================================================================= @socketio.on('connect') def connect(): print('Client connected') @socketio.on('disconnect') def disconnect(): print('Client disconnected') @socketio.on('install_model') def install_model(data): def install_model_(): print("Install model triggered") model_path = data["path"] progress = 0 installation_dir = Path(f'./models/{self.config["backend"]}/') filename = Path(model_path).name installation_path = installation_dir / filename print("Model install requested") print(f"Model path : {model_path}") if installation_path.exists(): print("Error: Model already exists") socketio.emit('install_progress',{'status': 'failed', 'error': 'model already exists'}) socketio.emit('install_progress',{'status': 'progress', 'progress': progress}) def callback(progress): socketio.emit('install_progress',{'status': 'progress', 'progress': progress}) self.download_file(model_path, installation_path, callback) socketio.emit('install_progress',{'status': 'succeeded', 'error': ''}) tpe = threading.Thread(target=install_model_, args=()) tpe.start() @socketio.on('uninstall_model') def uninstall_model(data): model_path = data['path'] installation_dir = Path(f'./models/{self.config["backend"]}/') filename = Path(model_path).name installation_path = installation_dir / filename if not installation_path.exists(): socketio.emit('install_progress',{'status': 'failed', 'error': 'The model does not exist'}) installation_path.unlink() socketio.emit('install_progress',{'status': 'succeeded', 'error': ''}) @socketio.on('generate_msg') def generate_msg(data): if self.process.model_ready.value==1: if self.current_discussion is None: if self.db.does_last_discussion_have_messages(): self.current_discussion = self.db.create_discussion() else: self.current_discussion = self.db.load_last_discussion() message = data["prompt"] message_id = self.current_discussion.add_message( "user", message, parent=self.message_id ) self.current_user_message_id = message_id tpe = threading.Thread(target=self.start_message_generation, args=(message, message_id)) tpe.start() else: self.socketio.emit('infos', { "status":'model_not_ready', "type": "input_message_infos", 'logo': self.personality.logo "bot": self.personality.name, "user": self.personality.user_name, "message":"", "user_message_id": self.current_user_message_id, "ai_message_id": self.current_ai_message_id, } ) @socketio.on('generate_msg_from') def handle_connection(data): message_id = int(data['id']) message = data["prompt"] self.current_user_message_id = message_id tpe = threading.Thread(target=self.start_message_generation, args=(message, message_id)) tpe.start() # generation status self.generating=False #properties @property def message_id(self): return self._message_id @property def current_user_message_id(self): return self._current_user_message_id @current_user_message_id.setter def current_user_message_id(self, id): self._current_user_message_id=id self._message_id = id @property def current_ai_message_id(self): return self._current_ai_message_id @current_ai_message_id.setter def current_ai_message_id(self, id): self._current_ai_message_id=id self._message_id = id def download_file(self, url, installation_path, callback=None): """ Downloads a file from a URL, reports the download progress using a callback function, and displays a progress bar. Args: url (str): The URL of the file to download. installation_path (str): The path where the file should be saved. callback (function, optional): A callback function to be called during the download with the progress percentage as an argument. Defaults to None. """ try: response = requests.get(url, stream=True) # Get the file size from the response headers total_size = int(response.headers.get('content-length', 0)) with open(installation_path, 'wb') as file: downloaded_size = 0 with tqdm(total=total_size, unit='B', unit_scale=True, ncols=80) as progress_bar: for chunk in response.iter_content(chunk_size=8192): if chunk: file.write(chunk) downloaded_size += len(chunk) if callback is not None: percentage = (downloaded_size / total_size) * 100 callback(percentage) progress_bar.update(len(chunk)) if callback is not None: callback(100.0) print("File downloaded successfully") except Exception as e: print("Couldn't download file:", str(e)) def load_backend(self, backend_name): backend_path = Path("backends")/backend_name # define the full absolute path to the module absolute_path = backend_path.resolve() # infer the module name from the file path module_name = backend_path.stem # use importlib to load the module from the file path loader = importlib.machinery.SourceFileLoader(module_name, str(absolute_path/"__init__.py")) backend_module = loader.load_module() backend_class = getattr(backend_module, backend_module.backend_name) return backend_class def condition_chatbot(self): if self.current_discussion is None: self.current_discussion = self.db.load_last_discussion() if self.personality.welcome_message!="": message_id = self.current_discussion.add_message( self.personality.name, self.personality.welcome_message, DiscussionsDB.MSG_TYPE_NORMAL, 0, -1 ) self.current_ai_message_id = message_id else: message_id = 0 return message_id def prepare_reception(self): self.bot_says = "" self.full_text = "" self.is_bot_text_started = False def create_new_discussion(self, title): self.current_discussion = self.db.create_discussion(title) # Get the current timestamp timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S") # Chatbot conditionning self.condition_chatbot() return timestamp def prepare_query(self, message_id=-1): messages = self.current_discussion.get_messages() self.full_message_list = [] for message in messages: if message["id"]< message_id or message_id==-1: if message["type"]==self.db.MSG_TYPE_NORMAL: if message["sender"]==self.personality.name: self.full_message_list.append(self.personality.ai_message_prefix+message["content"]) else: self.full_message_list.append(self.personality.user_message_prefix + message["content"]) else: break if self.personality.processor is not None: preprocessed_prompt = self.personality.processor.process_model_input(message["content"]) else: preprocessed_prompt = message["content"] if preprocessed_prompt is not None: self.full_message_list.append(self.personality.user_message_prefix+preprocessed_prompt+self.personality.link_text+self.personality.ai_message_prefix) else: self.full_message_list.append(self.personality.user_message_prefix+message["content"]+self.personality.link_text+self.personality.ai_message_prefix) link_text = self.personality.link_text if len(self.full_message_list) > self.config["nb_messages_to_remember"]: discussion_messages = self.personality.personality_conditioning+ link_text.join(self.full_message_list[-self.config["nb_messages_to_remember"]:]) else: discussion_messages = self.personality.personality_conditioning+ link_text.join(self.full_message_list) return discussion_messages, message["content"] def get_discussion_to(self, message_id=-1): messages = self.current_discussion.get_messages() self.full_message_list = [] for message in messages: if message["id"]<= message_id or message_id==-1: if message["type"]!=self.db.MSG_TYPE_CONDITIONNING: if message["sender"]==self.personality.name: self.full_message_list.append(self.personality.ai_message_prefix+message["content"]) else: self.full_message_list.append(self.personality.user_message_prefix + message["content"]) link_text = self.personality.link_text if len(self.full_message_list) > self.config["nb_messages_to_remember"]: discussion_messages = self.personality.personality_conditioning+ link_text.join(self.full_message_list[-self.config["nb_messages_to_remember"]:]) else: discussion_messages = self.personality.personality_conditioning+ link_text.join(self.full_message_list) return discussion_messages # Removes the last return def remove_text_from_string(self, string, text_to_find): """ Removes everything from the first occurrence of the specified text in the string (case-insensitive). Parameters: string (str): The original string. text_to_find (str): The text to find in the string. Returns: str: The updated string. """ index = string.lower().find(text_to_find.lower()) if index != -1: string = string[:index] return string def process_chunk(self, chunk): print(chunk,end="") sys.stdout.flush() self.bot_says += chunk if not self.personality.detect_antiprompt(self.bot_says): self.socketio.emit('message', { 'data': self.bot_says, 'user_message_id':self.current_user_message_id, 'ai_message_id':self.current_ai_message_id, 'discussion_id':self.current_discussion.discussion_id } ) if self.cancel_gen: print("Generation canceled") self.process.cancel_generation() self.cancel_gen = False else: self.bot_says = self.remove_text_from_string(self.bot_says, self.personality.user_message_prefix.strip()) self.process.cancel_generation() print("The model is halucinating") def start_message_generation(self, message, message_id): bot_says = "" # send the message to the bot print(f"Received message : {message}") if self.current_discussion: # First we need to send the new message ID to the client self.current_ai_message_id = self.current_discussion.add_message( self.personality.name, "", parent = self.current_user_message_id ) # first the content is empty, but we'll fill it at the end self.socketio.emit('infos', { "status":'generation_started', "type": "input_message_infos", "bot": self.personality.name, "user": self.personality.user_name, "message":message,#markdown.markdown(message), "user_message_id": self.current_user_message_id, "ai_message_id": self.current_ai_message_id, } ) # prepare query and reception self.discussion_messages, self.current_message = self.prepare_query(message_id) self.prepare_reception() self.generating = True print(">Generating message") self.process.generate(self.discussion_messages, self.current_message, message_id, n_predict = self.config['n_predict']) self.process.started_queue.get() while(self.process.is_generating.value): # Simulating other commands being issued chunk = "" while not self.process.generation_queue.empty(): chk, tok = self.process.generation_queue.get() chunk += chk if chunk!="": self.process_chunk(chunk) print() print("## Done ##") print() # Send final message self.socketio.emit('final', { 'data': self.bot_says, 'ai_message_id':self.current_ai_message_id, 'parent':self.current_user_message_id, 'discussion_id':self.current_discussion.discussion_id } ) self.current_discussion.update_message(self.current_ai_message_id, self.bot_says) self.full_message_list.append(self.bot_says) self.cancel_gen = False return bot_says else: #No discussion available print("No discussion selected!!!") print("## Done ##") print() self.cancel_gen = False return ""