model

finetune_configs/internlm_chat_7b_qlora_alpace_e3.py

@@ -0,0 +1,219 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import torch
+from datasets import load_dataset
+from mmengine.dataset import DefaultSampler
+from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
+                            LoggerHook, ParamSchedulerHook)
+from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
+from peft import LoraConfig
+from torch.optim import AdamW
+from transformers import (AutoModelForCausalLM, AutoTokenizer,
+                          BitsAndBytesConfig)
+
+from xtuner.dataset import process_hf_dataset
+from xtuner.dataset.collate_fns import default_collate_fn
+from xtuner.dataset.map_fns import alpaca_map_fn, template_map_fn_factory
+from xtuner.engine.hooks import (DatasetInfoHook, EvaluateChatHook,
+                                 VarlenAttnArgsToMessageHubHook)
+from xtuner.engine.runner import TrainLoop
+from xtuner.model import SupervisedFinetune
+from xtuner.parallel.sequence import SequenceParallelSampler
+from xtuner.utils import PROMPT_TEMPLATE, SYSTEM_TEMPLATE
+
+#######################################################################
+#                          PART 1  Settings                           #
+#######################################################################
+# Model
+pretrained_model_name_or_path = '/group_share/lntelligent-Medical-Guidance-Large-Model/InternLM/XTuner/model/internlm2-chat-7b'
+use_varlen_attn = False
+
+# Data
+alpaca_en_path = 'dataset/gen_dataset/train_dataset/90_train.jsonl'
+prompt_template = PROMPT_TEMPLATE.internlm2_chat
+max_length = 2048
+pack_to_max_length = True
+
+# parallel
+sequence_parallel_size = 1
+
+# Scheduler & Optimizer
+batch_size = 1  # per_device
+accumulative_counts = 16
+accumulative_counts *= sequence_parallel_size
+dataloader_num_workers = 0
+max_epochs = 3
+optim_type = AdamW
+lr = 2e-4
+betas = (0.9, 0.999)
+weight_decay = 0
+max_norm = 1  # grad clip
+warmup_ratio = 0.03
+
+# Save
+save_steps = 500
+save_total_limit = 2  # Maximum checkpoints to keep (-1 means unlimited)
+
+# Evaluate the generation performance during the training
+evaluation_freq = 500
+SYSTEM = SYSTEM_TEMPLATE.alpaca
+evaluation_inputs = [
+    '请介绍一下你自己', 'Please introduce yourself'
+]
+
+#######################################################################
+#                      PART 2  Model & Tokenizer                      #
+#######################################################################
+tokenizer = dict(
+    type=AutoTokenizer.from_pretrained,
+    pretrained_model_name_or_path=pretrained_model_name_or_path,
+    trust_remote_code=True,
+    padding_side='right')
+
+model = dict(
+    type=SupervisedFinetune,
+    use_varlen_attn=use_varlen_attn,
+    llm=dict(
+        type=AutoModelForCausalLM.from_pretrained,
+        pretrained_model_name_or_path=pretrained_model_name_or_path,
+        trust_remote_code=True,
+        torch_dtype=torch.float16,
+        quantization_config=dict(
+            type=BitsAndBytesConfig,
+            load_in_4bit=True,
+            load_in_8bit=False,
+            llm_int8_threshold=6.0,
+            llm_int8_has_fp16_weight=False,
+            bnb_4bit_compute_dtype=torch.float16,
+            bnb_4bit_use_double_quant=True,
+            bnb_4bit_quant_type='nf4')),
+    lora=dict(
+        type=LoraConfig,
+        r=64,
+        lora_alpha=16,
+        lora_dropout=0.1,
+        bias='none',
+        task_type='CAUSAL_LM'))
+
+#######################################################################
+#                      PART 3  Dataset & Dataloader                   #
+#######################################################################
+alpaca_en = dict(
+    type=process_hf_dataset,
+    dataset=dict(type=load_dataset, path='json', data_files=dict(train=alpaca_en_path)),
+    tokenizer=tokenizer,
+    max_length=max_length,
+    dataset_map_fn=None,
+    template_map_fn=dict(
+        type=template_map_fn_factory, template=prompt_template),
+    remove_unused_columns=True,
+    shuffle_before_pack=True,
+    pack_to_max_length=pack_to_max_length,
+    use_varlen_attn=use_varlen_attn)
+
+sampler = SequenceParallelSampler \
+    if sequence_parallel_size > 1 else DefaultSampler
+train_dataloader = dict(
+    batch_size=batch_size,
+    num_workers=dataloader_num_workers,
+    dataset=alpaca_en,
+    sampler=dict(type=sampler, shuffle=True),
+    collate_fn=dict(type=default_collate_fn, use_varlen_attn=use_varlen_attn))
+
+#######################################################################
+#                    PART 4  Scheduler & Optimizer                    #
+#######################################################################
+# optimizer
+optim_wrapper = dict(
+    type=AmpOptimWrapper,
+    optimizer=dict(
+        type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
+    clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
+    accumulative_counts=accumulative_counts,
+    loss_scale='dynamic',
+    dtype='float16')
+
+# learning policy
+# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md  # noqa: E501
+param_scheduler = [
+    dict(
+        type=LinearLR,
+        start_factor=1e-5,
+        by_epoch=True,
+        begin=0,
+        end=warmup_ratio * max_epochs,
+        convert_to_iter_based=True),
+    dict(
+        type=CosineAnnealingLR,
+        eta_min=0.0,
+        by_epoch=True,
+        begin=warmup_ratio * max_epochs,
+        end=max_epochs,
+        convert_to_iter_based=True)
+]
+
+# train, val, test setting
+train_cfg = dict(type=TrainLoop, max_epochs=max_epochs)
+
+#######################################################################
+#                           PART 5  Runtime                           #
+#######################################################################
+# Log the dialogue periodically during the training process, optional
+custom_hooks = [
+    dict(type=DatasetInfoHook, tokenizer=tokenizer),
+    dict(
+        type=EvaluateChatHook,
+        tokenizer=tokenizer,
+        every_n_iters=evaluation_freq,
+        evaluation_inputs=evaluation_inputs,
+        system=SYSTEM,
+        prompt_template=prompt_template)
+]
+
+if use_varlen_attn:
+    custom_hooks += [dict(type=VarlenAttnArgsToMessageHubHook)]
+
+# configure default hooks
+default_hooks = dict(
+    # record the time of every iteration.
+    timer=dict(type=IterTimerHook),
+    # print log every 10 iterations.
+    logger=dict(type=LoggerHook, log_metric_by_epoch=False, interval=10),
+    # enable the parameter scheduler.
+    param_scheduler=dict(type=ParamSchedulerHook),
+    # save checkpoint per `save_steps`.
+    checkpoint=dict(
+        type=CheckpointHook,
+        by_epoch=False,
+        interval=save_steps,
+        max_keep_ckpts=save_total_limit),
+    # set sampler seed in distributed evrionment.
+    sampler_seed=dict(type=DistSamplerSeedHook),
+)
+
+# configure environment
+env_cfg = dict(
+    # whether to enable cudnn benchmark
+    cudnn_benchmark=False,
+    # set multi process parameters
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
+    # set distributed parameters
+    dist_cfg=dict(backend='nccl'),
+)
+
+# set visualizer
+visualizer = None
+
+# set log level
+log_level = 'INFO'
+
+# load from which checkpoint
+load_from = None
+
+# whether to resume training from the loaded checkpoint
+resume = False
+
+# Defaults to use random seed and disable `deterministic`
+randomness = dict(seed=None, deterministic=False)
+
+# set log processor
+log_processor = dict(by_epoch=False)

pages/selling_page.py

@@ -0,0 +1,282 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# @Time    : 2024.4.16
+# @Author  : HinGwenWong
+
+import random
+from datetime import datetime
+from pathlib import Path
+
+import streamlit as st
+
+from utils.web_configs import WEB_CONFIGS
+
+# 设置页面配置，包括标题、图标、布局和菜单项
+st.set_page_config(
+    page_title="智能医导大模型",
+    page_icon="🛒",
+    layout="wide",
+    initial_sidebar_state="expanded",
+    menu_items={
+        "Get Help": "https://github.com/nhbdgtgefr/Intelligent-Medical-Guidance-Large-Model/tree/main",
+        "About": "# 智能医导大模型",
+    },
+)
+
+from audiorecorder import audiorecorder
+
+from utils.asr.asr_worker import process_asr
+from utils.digital_human.digital_human_worker import show_video
+from utils.infer.lmdeploy_infer import get_turbomind_response
+from utils.model_loader import ASR_HANDLER, LLM_MODEL, RAG_RETRIEVER
+from utils.tools import resize_image
+
+
+def on_btn_click(*args, **kwargs):
+    """
+    处理按钮点击事件的函数。
+    """
+    if kwargs["info"] == "清除对话历史":
+        st.session_state.messages = []
+    elif kwargs["info"] == "返回科室页":
+        st.session_state.page_switch = "app.py"
+    else:
+        st.session_state.button_msg = kwargs["info"]
+
+
+def init_sidebar():
+    """
+    初始化侧边栏界面，展示商品信息，并提供操作按钮。
+    """
+    asr_text = ""
+    with st.sidebar:
+        # 标题
+        st.markdown("## 智能医导大模型")
+        st.markdown("[智能医导大模型](https://github.com/nhbdgtgefr/Intelligent-Medical-Guidance-Large-Model)")
+        st.subheader("功能点：", divider="grey")
+        # st.markdown(
+        #     "1. 📜 **主播文案一键生成**\n2. 🚀 KV cache + Turbomind **推理加速**\n3. 📚 RAG **检索增强生成**\n4. 🔊 TTS **文字转语音**\n5. 🦸 **数字人生成**\n6. 🌐 **Agent 网络查询**\n7. 🎙️ **ASR 语音转文字**"
+        # )
+
+        st.subheader("目前讲解")
+        with st.container(height=400, border=True):
+            st.subheader(st.session_state.product_name)
+
+            image = resize_image(st.session_state.image_path, max_height=100)
+            st.image(image, channels="bgr")
+
+            st.subheader("科室特点", divider="grey")
+            st.markdown(st.session_state.hightlight)
+
+            want_to_buy_list = [
+                "我打算买了。",
+                "我准备入手了。",
+                "我决定要买了。",
+                "我准备下单了。",
+                "我将要购买这款产品。",
+                "我准备买下来了。",
+                "我准备将这个买下。",
+                "我准备要购买了。",
+                "我决定买下它。",
+                "我准备将其买下。",
+            ]
+            buy_flag = st.button("加入信息🛒", on_click=on_btn_click, kwargs={"info": random.choice(want_to_buy_list)})
+
+        # TODO 加入卖货信息
+        # 卖出 xxx 个
+        # 成交额
+
+        if WEB_CONFIGS.ENABLE_ASR:
+            Path(WEB_CONFIGS.ASR_WAV_SAVE_PATH).mkdir(parents=True, exist_ok=True)
+
+            st.subheader(f"语音输入", divider="grey")
+            audio = audiorecorder(
+                start_prompt="开始录音", stop_prompt="停止录音", pause_prompt="", show_visualizer=True, key=None
+            )
+
+            if len(audio) > 0:
+
+                # 将录音保存 wav 文件
+                save_tag = datetime.now().strftime("%Y-%m-%d-%H-%M-%S") + ".wav"
+                wav_path = str(Path(WEB_CONFIGS.ASR_WAV_SAVE_PATH).joinpath(save_tag).absolute())
+
+                # st.audio(audio.export().read()) # 前端显示
+                audio.export(wav_path, format="wav")  # 使用 pydub 保存到 wav 文件
+
+                # To get audio properties, use pydub AudioSegment properties:
+                # st.write(
+                #     f"Frame rate: {audio.frame_rate}, Frame width: {audio.frame_width}, Duration: {audio.duration_seconds} seconds"
+                # )
+
+                # 语音识别
+                asr_text = process_asr(ASR_HANDLER, wav_path)
+
+                # 删除过程文件
+                # Path(wav_path).unlink()
+
+        # 是否生成 TTS
+        if WEB_CONFIGS.ENABLE_TTS:
+            st.subheader("TTS 配置", divider="grey")
+            st.session_state.gen_tts_checkbox = st.toggle("生成语音", value=st.session_state.gen_tts_checkbox)
+
+        if WEB_CONFIGS.ENABLE_DIGITAL_HUMAN:
+            # 是否生成 数字人
+            st.subheader(f"数字人 配置", divider="grey")
+            st.session_state.gen_digital_human_checkbox = st.toggle(
+                "生成数字人视频", value=st.session_state.gen_digital_human_checkbox
+            )
+
+        if WEB_CONFIGS.ENABLE_AGENT:
+            # 是否使用 agent
+            st.subheader(f"Agent 配置", divider="grey")
+            with st.container(border=True):
+                st.markdown("**插件列表**")
+                st.button("结合天气查询到货时间", type="primary")
+            st.session_state.enable_agent_checkbox = st.toggle("使用 Agent 能力", value=st.session_state.enable_agent_checkbox)
+
+        st.subheader("页面切换", divider="grey")
+        st.button("返回科室页", on_click=on_btn_click, kwargs={"info": "返回科室页"})
+
+        st.subheader("对话设置", divider="grey")
+        st.button("清除对话历史", on_click=on_btn_click, kwargs={"info": "清除对话历史"})
+
+        # 模型配置
+        # st.markdown("## 模型配置")
+        # max_length = st.slider("Max Length", min_value=8, max_value=32768, value=32768)
+        # top_p = st.slider("Top P", 0.0, 1.0, 0.8, step=0.01)
+        # temperature = st.slider("Temperature", 0.0, 1.0, 0.7, step=0.01)
+
+    return asr_text
+
+
+def init_message_block(meta_instruction, user_avator, robot_avator):
+
+    # 在应用重新运行时显示聊天历史消息
+    for message in st.session_state.messages:
+        with st.chat_message(message["role"], avatar=message.get("avatar")):
+            st.markdown(message["content"])
+
+            if message.get("wav") is not None:
+                # 展示语音
+                print(f"Load wav {message['wav']}")
+                with open(message["wav"], "rb") as f_wav:
+                    audio_bytes = f_wav.read()
+                st.audio(audio_bytes, format="audio/wav")
+
+    # 如果聊天历史为空，则显示产品介绍
+    if len(st.session_state.messages) == 0:
+        # 直接产品介绍
+        get_turbomind_response(
+            st.session_state.first_input,
+            meta_instruction,
+            user_avator,
+            robot_avator,
+            LLM_MODEL,
+            session_messages=st.session_state.messages,
+            add_session_msg=False,
+            first_input_str="",
+            enable_agent=False,
+        )
+
+    # 初始化按钮消息状态
+    if "button_msg" not in st.session_state:
+        st.session_state.button_msg = "x-x"
+
+
+def process_message(user_avator, prompt, meta_instruction, robot_avator):
+    # Display user message in chat message container
+    with st.chat_message("user", avatar=user_avator):
+        st.markdown(prompt)
+
+    get_turbomind_response(
+        prompt,
+        meta_instruction,
+        user_avator,
+        robot_avator,
+        LLM_MODEL,
+        session_messages=st.session_state.messages,
+        add_session_msg=True,
+        first_input_str=st.session_state.first_input,
+        rag_retriever=RAG_RETRIEVER,
+        product_name=st.session_state.product_name,
+        enable_agent=st.session_state.enable_agent_checkbox,
+        # departure_place=st.session_state.departure_place,
+        # delivery_company_name=st.session_state.delivery_company_name,
+    )
+
+
+def main(meta_instruction):
+
+    # 检查页面切换状态并进行切换
+    if st.session_state.page_switch != st.session_state.current_page:
+        st.switch_page(st.session_state.page_switch)
+
+    # 页面标题
+    st.title("智能医导大模型")
+
+    # 说明
+    st.info(
+        "本项目是基于人工智能的文字、语音、视频生成领域搭建的智能医导大模型。用户被授予使用此工具创建文字、语音、视频的自由，但用户在使用过程中应该遵守当地法律，并负责任地使用。开发人员不对用户可能的不当使用承担任何责任。",
+        icon="❗",
+    )
+
+    # 初始化侧边栏
+    asr_text = init_sidebar()
+
+    # 初始化聊天历史记录
+    if "messages" not in st.session_state:
+        st.session_state.messages = []
+
+    message_col = None
+    if st.session_state.gen_digital_human_checkbox and WEB_CONFIGS.ENABLE_DIGITAL_HUMAN:
+
+        with st.container():
+            message_col, video_col = st.columns([0.6, 0.4])
+
+            with video_col:
+                # 创建 empty 控件
+                st.session_state.video_placeholder = st.empty()
+                with st.session_state.video_placeholder.container():
+                    show_video(st.session_state.digital_human_video_path, autoplay=True, loop=True, muted=True)
+
+            with message_col:
+                init_message_block(meta_instruction, WEB_CONFIGS.USER_AVATOR, WEB_CONFIGS.ROBOT_AVATOR)
+    else:
+        init_message_block(meta_instruction, WEB_CONFIGS.USER_AVATOR, WEB_CONFIGS.ROBOT_AVATOR)
+
+    # 输入框显示提示信息
+    hint_msg = "你好，你可以向我提出任何关于就诊的问题，我将竭诚为您服务"
+    if st.session_state.button_msg != "x-x":
+        prompt = st.session_state.button_msg
+        st.session_state.button_msg = "x-x"
+        st.chat_input(hint_msg)
+    elif asr_text != "" and st.session_state.asr_text_cache != asr_text:
+        prompt = asr_text
+        st.chat_input(hint_msg)
+        st.session_state.asr_text_cache = asr_text
+    else:
+        prompt = st.chat_input(hint_msg)
+
+    # 接收用户输入
+    if prompt:
+
+        if message_col is None:
+            process_message(WEB_CONFIGS.USER_AVATOR, prompt, meta_instruction, WEB_CONFIGS.ROBOT_AVATOR)
+        else:
+            # 数字人启动，页面会分块，放入信息块中
+            with message_col:
+                process_message(WEB_CONFIGS.USER_AVATOR, prompt, meta_instruction, WEB_CONFIGS.ROBOT_AVATOR)
+
+
+# st.sidebar.page_link("app.py", label="商品页")
+# st.sidebar.page_link("./pages/selling_page.py", label="主播卖货", disabled=True)
+
+# META_INSTRUCTION 
+print("into sales page")
+st.session_state.current_page = "pages/selling_page.py"
+
+if "sales_info" not in st.session_state or st.session_state.sales_info == "":
+    st.session_state.page_switch = "app.py"
+    st.switch_page("app.py")
+
+main((st.session_state.sales_info))