text-generation-webui/server.py

import re
import time
import glob
from sys import exit
import torch
import argparse
import json
from pathlib import Path
import gradio as gr
import transformers
from html_generator import *
from transformers import AutoTokenizer, AutoModelForCausalLM
import warnings
import gc
from tqdm import tqdm


transformers.logging.set_verbosity_error()

parser = argparse.ArgumentParser()
parser.add_argument('--model', type=str, help='Name of the model to load by default.')
parser.add_argument('--notebook', action='store_true', help='Launch the web UI in notebook mode, where the output is written to the same text box as the input.')
parser.add_argument('--chat', action='store_true', help='Launch the web UI in chat mode.')
parser.add_argument('--cai-chat', action='store_true', help='Launch the web UI in chat mode with a style similar to Character.AI\'s. If the file profile.png or profile.jpg exists in the same folder as server.py, this image will be used as the bot\'s profile picture.')
parser.add_argument('--cpu', action='store_true', help='Use the CPU to generate text.')
parser.add_argument('--load-in-8bit', action='store_true', help='Load the model with 8-bit precision.')
parser.add_argument('--auto-devices', action='store_true', help='Automatically split the model across the available GPU(s) and CPU.')
parser.add_argument('--disk', action='store_true', help='If the model is too large for your GPU(s) and CPU combined, send the remaining layers to the disk.')
parser.add_argument('--disk-cache-dir', type=str, help='Directory to save the disk cache to. Defaults to "cache/".')
parser.add_argument('--gpu-memory', type=int, help='Maximum GPU memory in GiB to allocate. This is useful if you get out of memory errors while trying to generate text. Must be an integer number.')
parser.add_argument('--cpu-memory', type=int, help='Maximum CPU memory in GiB to allocate for offloaded weights. Must be an integer number. Defaults to 99 GiB.')
parser.add_argument('--no-stream', action='store_true', help='Don\'t stream the text output in real time. This slightly improves the text generation performance.')
parser.add_argument('--settings', type=str, help='Load the default interface settings from this json file. See settings-template.json for an example.')
parser.add_argument('--listen', action='store_true', help='Make the web UI reachable from your local network.')
parser.add_argument('--share', action='store_true', help='Create a public URL. This is useful for running the web UI on Google Colab or similar.')
args = parser.parse_args()

loaded_preset = None
available_models = sorted(set([item.replace('.pt', '') for item in map(lambda x : str(x.name), list(Path('models/').glob('*'))+list(Path('torch-dumps/').glob('*'))) if not item.endswith('.txt')]), key=str.lower)
available_presets = sorted(set(map(lambda x : '.'.join(str(x.name).split('.')[:-1]), Path('presets').glob('*.txt'))), key=str.lower)
available_characters = sorted(set(map(lambda x : '.'.join(str(x.name).split('.')[:-1]), Path('characters').glob('*.json'))), key=str.lower)

settings = {
    'max_new_tokens': 200,
    'max_new_tokens_min': 1,
    'max_new_tokens_max': 2000,
    'preset': 'NovelAI-Sphinx Moth',
    'name1': 'Person 1',
    'name2': 'Person 2',
    'context': 'This is a conversation between two people.',
    'prompt': 'Common sense questions and answers\n\nQuestion: \nFactual answer:',
    'prompt_gpt4chan': '-----\n--- 865467536\nInput text\n--- 865467537\n',
    'stop_at_newline': True,
    'history_size': 8,
    'history_size_min': 0,
    'history_size_max': 64,
    'preset_pygmalion': 'Pygmalion',
    'name1_pygmalion': 'You',
    'name2_pygmalion': 'Kawaii',
    'context_pygmalion': 'This is a conversation between two people.\n<START>',
    'stop_at_newline_pygmalion': False,
}

if args.settings is not None and Path(args.settings).exists():
    with open(Path(args.settings), 'r') as f:
        new_settings = json.load(f)
    for item in new_settings:
        if item in settings:
            settings[item] = new_settings[item]

def load_model(model_name):
    print(f"Loading {model_name}...")
    t0 = time.time()

    # Default settings
    if not (args.cpu or args.load_in_8bit or args.auto_devices or args.disk or args.gpu_memory is not None):
        if Path(f"torch-dumps/{model_name}.pt").exists():
            print("Loading in .pt format...")
            model = torch.load(Path(f"torch-dumps/{model_name}.pt"))
        elif model_name.lower().startswith(('gpt-neo', 'opt-', 'galactica')) and any(size in model_name.lower() for size in ('13b', '20b', '30b')):
            model = AutoModelForCausalLM.from_pretrained(Path(f"models/{model_name}"), device_map='auto', load_in_8bit=True)
        else:
            model = AutoModelForCausalLM.from_pretrained(Path(f"models/{model_name}"), low_cpu_mem_usage=True, torch_dtype=torch.float16).cuda()

    # Custom
    else:
        settings = ["low_cpu_mem_usage=True"]
        command = "AutoModelForCausalLM.from_pretrained"

        if args.cpu:
            settings.append("torch_dtype=torch.float32")
        else:
            settings.append("device_map='auto'")
            if args.gpu_memory is not None:
                if args.cpu_memory is not None:
                    settings.append(f"max_memory={{0: '{args.gpu_memory}GiB', 'cpu': '{args.cpu_memory}GiB'}}")
                else:
                    settings.append(f"max_memory={{0: '{args.gpu_memory}GiB', 'cpu': '99GiB'}}")
            if args.disk:
                if args.disk_cache_dir is not None:
                    settings.append("offload_folder='"+args.disk_cache_dir+"'")
                else:
                    settings.append("offload_folder='cache'")
            if args.load_in_8bit:
                settings.append("load_in_8bit=True")
            else:
                settings.append("torch_dtype=torch.float16")

        settings = ', '.join(set(settings))
        command = f"{command}(Path(f'models/{model_name}'), {settings})"
        model = eval(command)

    # Loading the tokenizer
    if model_name.lower().startswith(('gpt4chan', 'gpt-4chan', '4chan')) and Path(f"models/gpt-j-6B/").exists():
        tokenizer = AutoTokenizer.from_pretrained(Path("models/gpt-j-6B/"))
    else:
        tokenizer = AutoTokenizer.from_pretrained(Path(f"models/{model_name}/"))
    tokenizer.truncation_side = 'left'

    print(f"Loaded the model in {(time.time()-t0):.2f} seconds.")
    return model, tokenizer

# Removes empty replies from gpt4chan outputs
def fix_gpt4chan(s):
    for i in range(10):
        s = re.sub("--- [0-9]*\n>>[0-9]*\n---", "---", s)
        s = re.sub("--- [0-9]*\n *\n---", "---", s)
        s = re.sub("--- [0-9]*\n\n\n---", "---", s)
    return s

# Fix the LaTeX equations in galactica
def fix_galactica(s):
    s = s.replace(r'\[', r'$')
    s = s.replace(r'\]', r'$')
    s = s.replace(r'\(', r'$')
    s = s.replace(r'\)', r'$')
    s = s.replace(r'$$', r'$')
    return s

def encode(prompt, tokens):
    if not args.cpu:
        torch.cuda.empty_cache()
        input_ids = tokenizer.encode(str(prompt), return_tensors='pt', truncation=True, max_length=2048-tokens).cuda()
    else:
        input_ids = tokenizer.encode(str(prompt), return_tensors='pt', truncation=True, max_length=2048-tokens)
    return input_ids

def decode(output_ids):
    reply = tokenizer.decode(output_ids, skip_special_tokens=True)
    reply = reply.replace(r'<|endoftext|>', '')
    return reply

def formatted_outputs(reply, model_name):
    if not (args.chat or args.cai_chat):
        if model_name.lower().startswith('galactica'):
            reply = fix_galactica(reply)
            return reply, reply, generate_basic_html(reply)
        elif model_name.lower().startswith('gpt4chan'):
            reply = fix_gpt4chan(reply)
            return reply, 'Only applicable for GALACTICA models.', generate_4chan_html(reply)
        else:
            return reply, 'Only applicable for GALACTICA models.', generate_basic_html(reply)
    else:
        return reply

def generate_reply(question, tokens, inference_settings, selected_model, eos_token=None):
    global model, tokenizer, model_name, loaded_preset, preset

    if selected_model != model_name:
        model_name = selected_model
        model = tokenizer = None
        if not args.cpu:
            gc.collect()
            torch.cuda.empty_cache()
        model, tokenizer = load_model(model_name)
    if inference_settings != loaded_preset:
        with open(Path(f'presets/{inference_settings}.txt'), 'r') as infile:
            preset = infile.read()
        loaded_preset = inference_settings

    cuda = "" if args.cpu else ".cuda()"
    n = None if eos_token is None else tokenizer.encode(eos_token, return_tensors='pt')[0][-1]
    input_ids = encode(question, tokens)

    # Generate the entire reply at once
    if args.no_stream:
        output = eval(f"model.generate(input_ids, eos_token_id={n}, {preset}){cuda}")
        reply = decode(output[0])
        yield formatted_outputs(reply, model_name)

    # Generate the reply 1 token at a time
    else:
        yield formatted_outputs(question, model_name)
        preset = preset.replace('max_new_tokens=tokens', 'max_new_tokens=1')
        for i in tqdm(range(tokens)):
            output = eval(f"model.generate(input_ids, {preset}){cuda}")
            reply = decode(output[0])
            if eos_token is not None and reply[-1] == eos_token:
                break

            yield formatted_outputs(reply, model_name)
            input_ids = output

# Choosing the default model
if args.model is not None:
    model_name = args.model
else:
    if len(available_models) == 0:
        print("No models are available! Please download at least one.")
        exit(0)
    elif len(available_models) == 1:
        i = 0
    else:
        print("The following models are available:\n")
        for i,model in enumerate(available_models):
            print(f"{i+1}. {model}")
        print(f"\nWhich one do you want to load? 1-{len(available_models)}\n")
        i = int(input())-1
        print()
    model_name = available_models[i]
model, tokenizer = load_model(model_name)

# UI settings
if model_name.lower().startswith('gpt4chan'):
    default_text = settings['prompt_gpt4chan']
else:
    default_text = settings['prompt']

description = f"\n\n# Text generation lab\nGenerate text using Large Language Models.\n"
css = ".my-4 {margin-top: 0} .py-6 {padding-top: 2.5rem}"
if args.chat or args.cai_chat:
    history = []
    character = None

    # This gets the new line characters right.
    def clean_chat_message(text):
        text = text.replace('\n', '\n\n')
        text = re.sub(r"\n{3,}", "\n\n", text)
        text = text.strip()
        return text

    def generate_chat_prompt(text, tokens, name1, name2, context, history_size):
        text = clean_chat_message(text)

        rows = [f"{context.strip()}\n"]
        i = len(history)-1
        count = 0
        while i >= 0 and len(encode(''.join(rows), tokens)[0]) < 2048-tokens:
            rows.insert(1, f"{name2}: {history[i][1].strip()}\n")
            count += 1
            if not (i == 0 and len(history[i][0]) == 0):
                rows.insert(1, f"{name1}: {history[i][0].strip()}\n")
                count += 1
            i -= 1
            if history_size != 0 and count >= history_size:
                break
        rows.append(f"{name1}: {text}\n")
        rows.append(f"{name2}:")

        while len(rows) > 3 and len(encode(''.join(rows), tokens)[0]) >= 2048-tokens:
            rows.pop(1)
            rows.pop(1)

        question = ''.join(rows)
        question = question.replace('<|BEGIN-VISIBLE-CHAT|>', '')
        return question

    def chatbot_wrapper(text, tokens, inference_settings, selected_model, name1, name2, context, check, history_size):
        question = generate_chat_prompt(text, tokens, name1, name2, context, history_size)
        history.append(['', ''])
        eos_token = '\n' if check else None
        for reply in generate_reply(question, tokens, inference_settings, selected_model, eos_token=eos_token):
            next_character_found = False

            previous_idx = [m.start() for m in re.finditer(f"(^|\n){name2}:", question)]
            idx = [m.start() for m in re.finditer(f"(^|\n){name2}:", reply)]
            idx = idx[len(previous_idx)-1]

            reply = reply[idx + len(f"\n{name2}:"):]

            if check:
                reply = reply.split('\n')[0].strip()
            else:
                idx = reply.find(f"\n{name1}:")
                if idx != -1:
                    reply = reply[:idx]
                    next_character_found = True
                reply = clean_chat_message(reply)

            history[-1] = [text, reply]
            if next_character_found:
                break

            # Prevent the chat log from flashing if something like "\nYo" is generated just
            # before "\nYou:" is completed
            tmp = f"\n{name1}:"
            next_character_substring_found = False
            for j in range(1, len(tmp)):
                if reply[-j:] == tmp[:j]:
                    next_character_substring_found = True

            if not next_character_substring_found:
                yield history

        yield history

    def cai_chatbot_wrapper(text, tokens, inference_settings, selected_model, name1, name2, context, check, history_size):
        for history in chatbot_wrapper(text, tokens, inference_settings, selected_model, name1, name2, context, check, history_size):
            yield generate_chat_html(history, name1, name2, character)

    def remove_last_message(name1, name2):
        history.pop()
        if args.cai_chat:
            return generate_chat_html(history, name1, name2, character)
        else:
            return history

    def clear():
        global history
        history = []

    def clear_html():
        return generate_chat_html([], "", "", character)

    def redraw_html(name1, name2):
        global history
        return generate_chat_html(history, name1, name2, character)

    def save_history():
        _history = copy.deepcopy(history)
        for i in range(len(_history)):
            if '<|BEGIN-VISIBLE-CHAT|>' in history[i][0]:
                _history[i][0] = _history[i][0].replace('<|BEGIN-VISIBLE-CHAT|>', '')
                _history = _history[i:]
                break

        if not Path('logs').exists():
            Path('logs').mkdir()
        with open(Path('logs/conversation.json'), 'w') as f:
            f.write(json.dumps({'data': _history}))
        return Path('logs/conversation.json')

    def load_history(file):
        global history
        history = json.loads(file.decode('utf-8'))['data']

    def tokenize_example_dialogue(dialogue, name1, name2):
        dialogue = re.sub('<START>', '', dialogue)
        dialogue = re.sub('(\n|^)[Aa]non:', '\\1You:', dialogue)

        idx = [m.start() for m in re.finditer(f"(^|\n)({name1}|{name2}):", dialogue)]
        messages = []
        for i in range(len(idx)-1):
            messages.append(dialogue[idx[i]:idx[i+1]].strip())
        history = []
        entry = ['', '']
        for i in messages:
            if i.startswith(f'{name1}:'):
                entry[0] = i[len(f'{name1}:'):].strip()
            elif i.startswith(f'{name2}:'):
                entry[1] = i[len(f'{name2}:'):].strip()
                if not (len(entry[0]) == 0 and len(entry[1]) == 0):
                    history.append(entry)
                entry = ['', '']
        return history

    def load_character(_character, name1, name2):
        global history, character
        context = ""
        history = []
        if _character != 'None':
            character = _character
            with open(Path(f'characters/{_character}.json'), 'r') as f:
                data = json.loads(f.read())
            name2 = data['char_name']
            if 'char_persona' in data and data['char_persona'] != '':
                context += f"{data['char_name']}'s Persona: {data['char_persona']}\n"
            if 'world_scenario' in data and data['world_scenario'] != '':
                context += f"Scenario: {data['world_scenario']}\n"
            context = f"{context.strip()}\n<START>\n"
            if 'example_dialogue' in data and data['example_dialogue'] != '':
                history = tokenize_example_dialogue(data['example_dialogue'], name1, name2)
            if 'char_greeting' in data and len(data['char_greeting'].strip()) > 0:
                history += [['<|BEGIN-VISIBLE-CHAT|>', data['char_greeting']]]
            else:
                history += [['<|BEGIN-VISIBLE-CHAT|>', "Hello there!"]]
        else:
            character = None
            context = settings['context_pygmalion']
            name2 = settings['name2_pygmalion']

        if args.cai_chat:
            return name2, context, generate_chat_html(history, name1, name2, character)
        else:
            return name2, context, history

    suffix = '_pygmalion' if 'pygmalion' in model_name.lower() else ''
    with gr.Blocks(css=css+".h-\[40vh\] {height: 66.67vh} .gradio-container {max-width: 800px; margin-left: auto; margin-right: auto}", analytics_enabled=False) as interface:
        if args.cai_chat:
            display1 = gr.HTML(value=generate_chat_html([], "", "", character))
        else:
            display1 = gr.Chatbot()
        textbox = gr.Textbox(lines=2, label='Input')
        btn = gr.Button("Generate")
        with gr.Row():
            btn2 = gr.Button("Clear history")
            stop = gr.Button("Stop")
            btn3 = gr.Button("Remove last message")

        with gr.Row():
            with gr.Column():
                length_slider = gr.Slider(minimum=settings['max_new_tokens_min'], maximum=settings['max_new_tokens_max'], step=1, label='max_new_tokens', value=settings['max_new_tokens'])
                model_menu = gr.Dropdown(choices=available_models, value=model_name, label='Model')
            with gr.Column():
                history_size_slider = gr.Slider(minimum=settings['history_size_min'], maximum=settings['history_size_max'], step=1, label='Chat history size (0 for no limit)', value=settings['history_size'])
                preset_menu = gr.Dropdown(choices=available_presets, value=settings[f'preset{suffix}'], label='Settings preset')

        name1 = gr.Textbox(value=settings[f'name1{suffix}'], lines=1, label='Your name')
        name2 = gr.Textbox(value=settings[f'name2{suffix}'], lines=1, label='Bot\'s name')
        context = gr.Textbox(value=settings[f'context{suffix}'], lines=2, label='Context')
        with gr.Row():
            character_menu = gr.Dropdown(choices=["None"]+available_characters, value="None", label='Character')
        with gr.Row():
            check = gr.Checkbox(value=settings[f'stop_at_newline{suffix}'], label='Stop generating at new line character?')
        with gr.Row():
            with gr.Column():
                gr.Markdown("Upload chat history")
                upload = gr.File(type='binary')
            with gr.Column():
                gr.Markdown("Download chat history")
                save_btn = gr.Button(value="Click me")
                download = gr.File()

        input_params = [textbox, length_slider, preset_menu, model_menu, name1, name2, context, check, history_size_slider]
        if args.cai_chat:
            gen_event = btn.click(cai_chatbot_wrapper, input_params, display1, show_progress=args.no_stream, api_name="textgen")
            gen_event2 = textbox.submit(cai_chatbot_wrapper, input_params, display1, show_progress=args.no_stream)
            btn2.click(clear_html, [], display1, show_progress=False)
        else:
            gen_event = btn.click(chatbot_wrapper, input_params, display1, show_progress=args.no_stream, api_name="textgen")
            gen_event2 = textbox.submit(chatbot_wrapper, input_params, display1, show_progress=args.no_stream)
            btn2.click(lambda x: "", display1, display1, show_progress=False)

        btn2.click(clear)
        btn3.click(remove_last_message, [name1, name2], display1, show_progress=False)
        btn.click(lambda x: "", textbox, textbox, show_progress=False)
        textbox.submit(lambda x: "", textbox, textbox, show_progress=False)
        stop.click(None, None, None, cancels=[gen_event, gen_event2])
        save_btn.click(save_history, inputs=[], outputs=[download])
        upload.upload(load_history, [upload], [])
        character_menu.change(load_character, [character_menu, name1, name2], [name2, context, display1])

        if args.cai_chat:
            upload.upload(redraw_html, [name1, name2], [display1])
        else:
            upload.upload(lambda : history, [], [display1])

elif args.notebook:
    with gr.Blocks(css=css, analytics_enabled=False) as interface:
        gr.Markdown(description)
        with gr.Tab('Raw'):
            textbox = gr.Textbox(value=default_text, lines=23)
        with gr.Tab('Markdown'):
            markdown = gr.Markdown()
        with gr.Tab('HTML'):
            html = gr.HTML()
        btn = gr.Button("Generate")
        stop = gr.Button("Stop")

        length_slider = gr.Slider(minimum=settings['max_new_tokens_min'], maximum=settings['max_new_tokens_max'], step=1, label='max_new_tokens', value=settings['max_new_tokens'])
        with gr.Row():
            with gr.Column():
                model_menu = gr.Dropdown(choices=available_models, value=model_name, label='Model')
            with gr.Column():
                preset_menu = gr.Dropdown(choices=available_presets, value=settings['preset'], label='Settings preset')

        gen_event = btn.click(generate_reply, [textbox, length_slider, preset_menu, model_menu], [textbox, markdown, html], show_progress=args.no_stream, api_name="textgen")
        gen_event2 = textbox.submit(generate_reply, [textbox, length_slider, preset_menu, model_menu], [textbox, markdown, html], show_progress=args.no_stream)
        stop.click(None, None, None, cancels=[gen_event, gen_event2])

else:
    with gr.Blocks(css=css, analytics_enabled=False) as interface:
        gr.Markdown(description)
        with gr.Row():
            with gr.Column():
                textbox = gr.Textbox(value=default_text, lines=15, label='Input')
                length_slider = gr.Slider(minimum=settings['max_new_tokens_min'], maximum=settings['max_new_tokens_max'], step=1, label='max_new_tokens', value=settings['max_new_tokens'])
                preset_menu = gr.Dropdown(choices=available_presets, value=settings['preset'], label='Settings preset')
                model_menu = gr.Dropdown(choices=available_models, value=model_name, label='Model')
                btn = gr.Button("Generate")
                with gr.Row():
                    with gr.Column():
                        cont = gr.Button("Continue")
                    with gr.Column():
                        stop = gr.Button("Stop")
            with gr.Column():
                with gr.Tab('Raw'):
                    output_textbox = gr.Textbox(lines=15, label='Output')
                with gr.Tab('Markdown'):
                    markdown = gr.Markdown()
                with gr.Tab('HTML'):
                    html = gr.HTML()

        gen_event = btn.click(generate_reply, [textbox, length_slider, preset_menu, model_menu], [output_textbox, markdown, html], show_progress=args.no_stream, api_name="textgen")
        gen_event2 = textbox.submit(generate_reply, [textbox, length_slider, preset_menu, model_menu], [output_textbox, markdown, html], show_progress=args.no_stream)
        cont_event = cont.click(generate_reply, [output_textbox, length_slider, preset_menu, model_menu], [output_textbox, markdown, html], show_progress=args.no_stream)
        stop.click(None, None, None, cancels=[gen_event, gen_event2, cont_event])

interface.queue()
if args.listen:
    interface.launch(share=args.share, server_name="0.0.0.0")
else:
    interface.launch(share=args.share)