' in inputs:
+ inputs = inputs.split('
')[1]
+ inputs = inputs.split("
")[0]
+ elif len(self.history) == 1:
+ inputs = self.history[-1]["content"]
+ del self.history[-1]
else:
yield chatbot, f"{STANDARD_ERROR_MSG}上下文是空的"
return
@@ -613,13 +674,15 @@ class BaseLLMModel:
def set_single_turn(self, new_single_turn):
self.single_turn = new_single_turn
- def reset(self):
+ def reset(self, remain_system_prompt=False):
self.history = []
self.all_token_counts = []
self.interrupted = False
- pathlib.Path(os.path.join(HISTORY_DIR, self.user_identifier, new_auto_history_filename(
- os.path.join(HISTORY_DIR, self.user_identifier)))).touch()
- return [], self.token_message([0])
+ self.history_file_path = new_auto_history_filename(self.user_identifier)
+ history_name = self.history_file_path[:-5]
+ choices = [history_name] + get_history_names(self.user_identifier)
+ system_prompt = self.system_prompt if remain_system_prompt else ""
+ return [], self.token_message([0]), gr.Radio.update(choices=choices, value=history_name), system_prompt
def delete_first_conversation(self):
if self.history:
@@ -630,18 +693,18 @@ class BaseLLMModel:
def delete_last_conversation(self, chatbot):
if len(chatbot) > 0 and STANDARD_ERROR_MSG in chatbot[-1][1]:
msg = "由于包含报错信息，只删除chatbot记录"
- chatbot.pop()
+ chatbot = chatbot[:-1]
return chatbot, self.history
if len(self.history) > 0:
- self.history.pop()
- self.history.pop()
+ self.history = self.history[:-2]
if len(chatbot) > 0:
msg = "删除了一组chatbot对话"
- chatbot.pop()
+ chatbot = chatbot[:-1]
if len(self.all_token_counts) > 0:
msg = "删除了一组对话的token计数记录"
self.all_token_counts.pop()
msg = "删除了一组对话"
+ self.auto_save(chatbot)
return chatbot, msg
def token_message(self, token_lst=None):
@@ -652,16 +715,36 @@ class BaseLLMModel:
token_sum += sum(token_lst[: i + 1])
return i18n("Token 计数: ") + f"{sum(token_lst)}" + i18n("，本次对话累计消耗了 ") + f"{token_sum} tokens"
- def save_chat_history(self, filename, chatbot, user_name):
+ def rename_chat_history(self, filename, chatbot, user_name):
if filename == "":
- return
+ return gr.update()
if not filename.endswith(".json"):
filename += ".json"
- return save_file(filename, self.system_prompt, self.history, chatbot, user_name)
+ self.delete_chat_history(self.history_file_path, user_name)
+ # 命名重复检测
+ repeat_file_index = 2
+ full_path = os.path.join(HISTORY_DIR, user_name, filename)
+ while os.path.exists(full_path):
+ full_path = os.path.join(HISTORY_DIR, user_name, f"{repeat_file_index}_{filename}")
+ repeat_file_index += 1
+ filename = os.path.basename(full_path)
+
+ self.history_file_path = filename
+ save_file(filename, self.system_prompt, self.history, chatbot, user_name)
+ return init_history_list(user_name)
+
+ def auto_name_chat_history(self, name_chat_method, user_question, chatbot, user_name, single_turn_checkbox):
+ if len(self.history) == 2 and not single_turn_checkbox:
+ user_question = self.history[0]["content"]
+ if type(user_question) == list:
+ user_question = user_question[0]["text"]
+ filename = replace_special_symbols(user_question)[:16] + ".json"
+ return self.rename_chat_history(filename, chatbot, user_name)
+ else:
+ return gr.update()
def auto_save(self, chatbot):
- history_file_path = get_history_filepath(self.user_identifier)
- save_file(history_file_path, self.system_prompt,
+ save_file(self.history_file_path, self.system_prompt,
self.history, chatbot, self.user_identifier)
def export_markdown(self, filename, chatbot, user_name):
@@ -669,19 +752,27 @@ class BaseLLMModel:
return
if not filename.endswith(".md"):
filename += ".md"
- return save_file(filename, self.system_prompt, self.history, chatbot, user_name)
-
- def load_chat_history(self, filename, user_name):
- logging.debug(f"{user_name} 加载对话历史中……")
- logging.info(f"filename: {filename}")
- if type(filename) != str and filename is not None:
- filename = filename.name
+ save_file(filename, self.system_prompt, self.history, chatbot, user_name)
+
+ def load_chat_history(self, new_history_file_path=None, username=None):
+ logging.debug(f"{self.user_identifier} 加载对话历史中……")
+ if new_history_file_path is not None:
+ if type(new_history_file_path) != str:
+ # copy file from new_history_file_path.name to os.path.join(HISTORY_DIR, self.user_identifier)
+ new_history_file_path = new_history_file_path.name
+ shutil.copyfile(new_history_file_path, os.path.join(
+ HISTORY_DIR, self.user_identifier, os.path.basename(new_history_file_path)))
+ self.history_file_path = os.path.basename(new_history_file_path)
+ else:
+ self.history_file_path = new_history_file_path
try:
- if "/" not in filename:
+ if self.history_file_path == os.path.basename(self.history_file_path):
history_file_path = os.path.join(
- HISTORY_DIR, user_name, filename)
+ HISTORY_DIR, self.user_identifier, self.history_file_path)
else:
- history_file_path = filename
+ history_file_path = self.history_file_path
+ if not self.history_file_path.endswith(".json"):
+ history_file_path += ".json"
with open(history_file_path, "r", encoding="utf-8") as f:
json_s = json.load(f)
try:
@@ -697,13 +788,17 @@ class BaseLLMModel:
logging.info(new_history)
except:
pass
- logging.debug(f"{user_name} 加载对话历史完毕")
+ if len(json_s["chatbot"]) < len(json_s["history"])//2:
+ logging.info("Trimming corrupted history...")
+ json_s["history"] = json_s["history"][-len(json_s["chatbot"]):]
+ logging.info(f"Trimmed history: {json_s['history']}")
+ logging.debug(f"{self.user_identifier} 加载对话历史完毕")
self.history = json_s["history"]
- return os.path.basename(filename), json_s["system"], json_s["chatbot"]
+ return os.path.basename(self.history_file_path), json_s["system"], json_s["chatbot"]
except:
# 没有对话历史或者对话历史解析失败
- logging.info(f"没有找到对话历史记录 {filename}")
- return gr.update(), self.system_prompt, gr.update()
+ logging.info(f"没有找到对话历史记录 {self.history_file_path}")
+ return self.history_file_path, "", []
def delete_chat_history(self, filename, user_name):
if filename == "CANCELED":
@@ -712,25 +807,29 @@ class BaseLLMModel:
return i18n("你没有选择任何对话历史"), gr.update(), gr.update()
if not filename.endswith(".json"):
filename += ".json"
- if "/" not in filename:
+ if filename == os.path.basename(filename):
history_file_path = os.path.join(HISTORY_DIR, user_name, filename)
else:
history_file_path = filename
+ md_history_file_path = history_file_path[:-5] + ".md"
try:
os.remove(history_file_path)
- return i18n("删除对话历史成功"), get_history_names(False, user_name), []
+ os.remove(md_history_file_path)
+ return i18n("删除对话历史成功"), get_history_list(user_name), []
except:
logging.info(f"删除对话历史失败 {history_file_path}")
- return i18n("对话历史")+filename+i18n("已经被删除啦"), gr.update(), gr.update()
+ return i18n("对话历史")+filename+i18n("已经被删除啦"), get_history_list(user_name), []
def auto_load(self):
- if self.user_identifier == "":
- self.reset()
- return self.system_prompt, gr.update()
- history_file_path = get_history_filepath(self.user_identifier)
- filename, system_prompt, chatbot = self.load_chat_history(
- history_file_path, self.user_identifier)
- return system_prompt, chatbot
+ filepath = get_history_filepath(self.user_identifier)
+ if not filepath:
+ self.history_file_path = new_auto_history_filename(
+ self.user_identifier)
+ else:
+ self.history_file_path = filepath
+ filename, system_prompt, chatbot = self.load_chat_history()
+ filename = filename[:-5]
+ return filename, system_prompt, chatbot
def like(self):
"""like the last response, implement if needed
diff --git a/modules/models/midjourney.py b/modules/models/midjourney.py
index 65a560fc2427aad735d227d4d25b61b72f3ace5a..a54bde5d60ebd853d23182bd4a64b4fba3ee4ca1 100644
--- a/modules/models/midjourney.py
+++ b/modules/models/midjourney.py
@@ -2,11 +2,10 @@ import base64
import io
import json
import logging
+import os
import pathlib
-import time
import tempfile
-import os
-
+import time
from datetime import datetime
import requests
@@ -14,7 +13,7 @@ import tiktoken
from PIL import Image
from modules.config import retrieve_proxy
-from modules.models.models import XMChat
+from modules.models.XMChat import XMChat
mj_proxy_api_base = os.getenv("MIDJOURNEY_PROXY_API_BASE")
mj_discord_proxy_url = os.getenv("MIDJOURNEY_DISCORD_PROXY_URL")
@@ -218,10 +217,10 @@ class Midjourney_Client(XMChat):
logging.info("使用图片作为输入")
return None, chatbot, None
- def reset(self):
+ def reset(self, remain_system_prompt=False):
self.image_bytes = None
self.image_path = None
- return [], "已重置"
+ return super().reset()
def get_answer_at_once(self):
content = self.history[-1]['content']
@@ -367,7 +366,7 @@ UPSCALE - 确认后放大图片，第一个数值为需要放大的图片（1~4
请使用SD进行UPSCALE
VARIATION - 图片变体，第一个数值为需要放大的图片（1~4），第二参数为任务ID
/mj VARIATION::1::123456789
-
+
【绘图参数】
所有命令默认会带上参数--v 5.2
其他参数参照 https://docs.midjourney.com/docs/parameter-list
diff --git a/modules/models/models.py b/modules/models/models.py
index fe1edbc91fc611881f6dff34affd23dc93596699..1f360f3552326ccb09b4844ed7c754ca2c1adeae 100644
--- a/modules/models/models.py
+++ b/modules/models/models.py
@@ -1,546 +1,19 @@
from __future__ import annotations
-from typing import TYPE_CHECKING, List
import logging
-import json
-import commentjson as cjson
import os
-import sys
-import requests
-import urllib3
-import platform
-import base64
-from io import BytesIO
-from PIL import Image
-from tqdm import tqdm
import colorama
-import asyncio
-import aiohttp
-from enum import Enum
-import uuid
+import commentjson as cjson
+
+from modules import config
-from ..presets import *
from ..index_func import *
+from ..presets import *
from ..utils import *
-from .. import shared
-from ..config import retrieve_proxy, usage_limit, sensitive_id
-from modules import config
from .base_model import BaseLLMModel, ModelType
-class OpenAIClient(BaseLLMModel):
- def __init__(
- self,
- model_name,
- api_key,
- system_prompt=INITIAL_SYSTEM_PROMPT,
- temperature=1.0,
- top_p=1.0,
- user_name=""
- ) -> None:
- super().__init__(
- model_name=model_name,
- temperature=temperature,
- top_p=top_p,
- system_prompt=system_prompt,
- user=user_name
- )
- self.api_key = api_key
- self.need_api_key = True
- self._refresh_header()
-
- def get_answer_stream_iter(self):
- response = self._get_response(stream=True)
- if response is not None:
- iter = self._decode_chat_response(response)
- partial_text = ""
- for i in iter:
- partial_text += i
- yield partial_text
- else:
- yield STANDARD_ERROR_MSG + GENERAL_ERROR_MSG
-
- def get_answer_at_once(self):
- response = self._get_response()
- response = json.loads(response.text)
- content = response["choices"][0]["message"]["content"]
- total_token_count = response["usage"]["total_tokens"]
- return content, total_token_count
-
- def count_token(self, user_input):
- input_token_count = count_token(construct_user(user_input))
- if self.system_prompt is not None and len(self.all_token_counts) == 0:
- system_prompt_token_count = count_token(
- construct_system(self.system_prompt)
- )
- return input_token_count + system_prompt_token_count
- return input_token_count
-
- def billing_info(self):
- try:
- curr_time = datetime.datetime.now()
- last_day_of_month = get_last_day_of_month(
- curr_time).strftime("%Y-%m-%d")
- first_day_of_month = curr_time.replace(day=1).strftime("%Y-%m-%d")
- usage_url = f"{shared.state.usage_api_url}?start_date={first_day_of_month}&end_date={last_day_of_month}"
- try:
- usage_data = self._get_billing_data(usage_url)
- except Exception as e:
- # logging.error(f"获取API使用情况失败: " + str(e))
- if "Invalid authorization header" in str(e):
- return i18n("**获取API使用情况失败**，需在填写`config.json`中正确填写sensitive_id")
- elif "Incorrect API key provided: sess" in str(e):
- return i18n("**获取API使用情况失败**，sensitive_id错误或已过期")
- return i18n("**获取API使用情况失败**")
- # rounded_usage = "{:.5f}".format(usage_data["total_usage"] / 100)
- rounded_usage = round(usage_data["total_usage"] / 100, 5)
- usage_percent = round(usage_data["total_usage"] / usage_limit, 2)
- from ..webui import get_html
- # return i18n("**本月使用金额** ") + f"\u3000 ${rounded_usage}"
- return get_html("billing_info.html").format(
- label = i18n("本月使用金额"),
- usage_percent = usage_percent,
- rounded_usage = rounded_usage,
- usage_limit = usage_limit
- )
- except requests.exceptions.ConnectTimeout:
- status_text = (
- STANDARD_ERROR_MSG + CONNECTION_TIMEOUT_MSG + ERROR_RETRIEVE_MSG
- )
- return status_text
- except requests.exceptions.ReadTimeout:
- status_text = STANDARD_ERROR_MSG + READ_TIMEOUT_MSG + ERROR_RETRIEVE_MSG
- return status_text
- except Exception as e:
- import traceback
- traceback.print_exc()
- logging.error(i18n("获取API使用情况失败:") + str(e))
- return STANDARD_ERROR_MSG + ERROR_RETRIEVE_MSG
-
- def set_token_upper_limit(self, new_upper_limit):
- pass
-
- @shared.state.switching_api_key # 在不开启多账号模式的时候，这个装饰器不会起作用
- def _get_response(self, stream=False):
- openai_api_key = self.api_key
- system_prompt = self.system_prompt
- history = self.history
- logging.debug(colorama.Fore.YELLOW +
- f"{history}" + colorama.Fore.RESET)
- headers = {
- "Content-Type": "application/json",
- "Authorization": f"Bearer {openai_api_key}",
- }
-
- if system_prompt is not None:
- history = [construct_system(system_prompt), *history]
-
- payload = {
- "model": self.model_name,
- "messages": history,
- "temperature": self.temperature,
- "top_p": self.top_p,
- "n": self.n_choices,
- "stream": stream,
- "presence_penalty": self.presence_penalty,
- "frequency_penalty": self.frequency_penalty,
- }
-
- if self.max_generation_token is not None:
- payload["max_tokens"] = self.max_generation_token
- if self.stop_sequence is not None:
- payload["stop"] = self.stop_sequence
- if self.logit_bias is not None:
- payload["logit_bias"] = self.logit_bias
- if self.user_identifier:
- payload["user"] = self.user_identifier
-
- if stream:
- timeout = TIMEOUT_STREAMING
- else:
- timeout = TIMEOUT_ALL
-
- # 如果有自定义的api-host，使用自定义host发送请求，否则使用默认设置发送请求
- if shared.state.completion_url != COMPLETION_URL:
- logging.debug(f"使用自定义API URL: {shared.state.completion_url}")
-
- with retrieve_proxy():
- try:
- response = requests.post(
- shared.state.completion_url,
- headers=headers,
- json=payload,
- stream=stream,
- timeout=timeout,
- )
- except:
- return None
- return response
-
- def _refresh_header(self):
- self.headers = {
- "Content-Type": "application/json",
- "Authorization": f"Bearer {sensitive_id}",
- }
-
-
- def _get_billing_data(self, billing_url):
- with retrieve_proxy():
- response = requests.get(
- billing_url,
- headers=self.headers,
- timeout=TIMEOUT_ALL,
- )
-
- if response.status_code == 200:
- data = response.json()
- return data
- else:
- raise Exception(
- f"API request failed with status code {response.status_code}: {response.text}"
- )
-
- def _decode_chat_response(self, response):
- error_msg = ""
- for chunk in response.iter_lines():
- if chunk:
- chunk = chunk.decode()
- chunk_length = len(chunk)
- try:
- chunk = json.loads(chunk[6:])
- except:
- print(i18n("JSON解析错误,收到的内容: ") + f"{chunk}")
- error_msg += chunk
- continue
- if chunk_length > 6 and "delta" in chunk["choices"][0]:
- if chunk["choices"][0]["finish_reason"] == "stop":
- break
- try:
- yield chunk["choices"][0]["delta"]["content"]
- except Exception as e:
- # logging.error(f"Error: {e}")
- continue
- if error_msg:
- raise Exception(error_msg)
-
- def set_key(self, new_access_key):
- ret = super().set_key(new_access_key)
- self._refresh_header()
- return ret
-
-
-class ChatGLM_Client(BaseLLMModel):
- def __init__(self, model_name, user_name="") -> None:
- super().__init__(model_name=model_name, user=user_name)
- from transformers import AutoTokenizer, AutoModel
- import torch
- global CHATGLM_TOKENIZER, CHATGLM_MODEL
- if CHATGLM_TOKENIZER is None or CHATGLM_MODEL is None:
- system_name = platform.system()
- model_path = None
- if os.path.exists("models"):
- model_dirs = os.listdir("models")
- if model_name in model_dirs:
- model_path = f"models/{model_name}"
- if model_path is not None:
- model_source = model_path
- else:
- model_source = f"THUDM/{model_name}"
- CHATGLM_TOKENIZER = AutoTokenizer.from_pretrained(
- model_source, trust_remote_code=True
- )
- quantified = False
- if "int4" in model_name:
- quantified = True
- model = AutoModel.from_pretrained(
- model_source, trust_remote_code=True
- )
- if torch.cuda.is_available():
- # run on CUDA
- logging.info("CUDA is available, using CUDA")
- model = model.half().cuda()
- # mps加速还存在一些问题，暂时不使用
- elif system_name == "Darwin" and model_path is not None and not quantified:
- logging.info("Running on macOS, using MPS")
- # running on macOS and model already downloaded
- model = model.half().to("mps")
- else:
- logging.info("GPU is not available, using CPU")
- model = model.float()
- model = model.eval()
- CHATGLM_MODEL = model
-
- def _get_glm_style_input(self):
- history = [x["content"] for x in self.history]
- query = history.pop()
- logging.debug(colorama.Fore.YELLOW +
- f"{history}" + colorama.Fore.RESET)
- assert (
- len(history) % 2 == 0
- ), f"History should be even length. current history is: {history}"
- history = [[history[i], history[i + 1]]
- for i in range(0, len(history), 2)]
- return history, query
-
- def get_answer_at_once(self):
- history, query = self._get_glm_style_input()
- response, _ = CHATGLM_MODEL.chat(
- CHATGLM_TOKENIZER, query, history=history)
- return response, len(response)
-
- def get_answer_stream_iter(self):
- history, query = self._get_glm_style_input()
- for response, history in CHATGLM_MODEL.stream_chat(
- CHATGLM_TOKENIZER,
- query,
- history,
- max_length=self.token_upper_limit,
- top_p=self.top_p,
- temperature=self.temperature,
- ):
- yield response
-
-
-class LLaMA_Client(BaseLLMModel):
- def __init__(
- self,
- model_name,
- lora_path=None,
- user_name=""
- ) -> None:
- super().__init__(model_name=model_name, user=user_name)
- from lmflow.datasets.dataset import Dataset
- from lmflow.pipeline.auto_pipeline import AutoPipeline
- from lmflow.models.auto_model import AutoModel
- from lmflow.args import ModelArguments, DatasetArguments, InferencerArguments
-
- self.max_generation_token = 1000
- self.end_string = "\n\n"
- # We don't need input data
- data_args = DatasetArguments(dataset_path=None)
- self.dataset = Dataset(data_args)
- self.system_prompt = ""
-
- global LLAMA_MODEL, LLAMA_INFERENCER
- if LLAMA_MODEL is None or LLAMA_INFERENCER is None:
- model_path = None
- if os.path.exists("models"):
- model_dirs = os.listdir("models")
- if model_name in model_dirs:
- model_path = f"models/{model_name}"
- if model_path is not None:
- model_source = model_path
- else:
- model_source = f"decapoda-research/{model_name}"
- # raise Exception(f"models目录下没有这个模型: {model_name}")
- if lora_path is not None:
- lora_path = f"lora/{lora_path}"
- model_args = ModelArguments(model_name_or_path=model_source, lora_model_path=lora_path, model_type=None, config_overrides=None, config_name=None, tokenizer_name=None, cache_dir=None,
- use_fast_tokenizer=True, model_revision='main', use_auth_token=False, torch_dtype=None, use_lora=False, lora_r=8, lora_alpha=32, lora_dropout=0.1, use_ram_optimized_load=True)
- pipeline_args = InferencerArguments(
- local_rank=0, random_seed=1, deepspeed='configs/ds_config_chatbot.json', mixed_precision='bf16')
-
- with open(pipeline_args.deepspeed, "r", encoding="utf-8") as f:
- ds_config = json.load(f)
- LLAMA_MODEL = AutoModel.get_model(
- model_args,
- tune_strategy="none",
- ds_config=ds_config,
- )
- LLAMA_INFERENCER = AutoPipeline.get_pipeline(
- pipeline_name="inferencer",
- model_args=model_args,
- data_args=data_args,
- pipeline_args=pipeline_args,
- )
-
- def _get_llama_style_input(self):
- history = []
- instruction = ""
- if self.system_prompt:
- instruction = (f"Instruction: {self.system_prompt}\n")
- for x in self.history:
- if x["role"] == "user":
- history.append(f"{instruction}Input: {x['content']}")
- else:
- history.append(f"Output: {x['content']}")
- context = "\n\n".join(history)
- context += "\n\nOutput: "
- return context
-
- def get_answer_at_once(self):
- context = self._get_llama_style_input()
-
- input_dataset = self.dataset.from_dict(
- {"type": "text_only", "instances": [{"text": context}]}
- )
-
- output_dataset = LLAMA_INFERENCER.inference(
- model=LLAMA_MODEL,
- dataset=input_dataset,
- max_new_tokens=self.max_generation_token,
- temperature=self.temperature,
- )
-
- response = output_dataset.to_dict()["instances"][0]["text"]
- return response, len(response)
-
- def get_answer_stream_iter(self):
- context = self._get_llama_style_input()
- partial_text = ""
- step = 1
- for _ in range(0, self.max_generation_token, step):
- input_dataset = self.dataset.from_dict(
- {"type": "text_only", "instances": [
- {"text": context + partial_text}]}
- )
- output_dataset = LLAMA_INFERENCER.inference(
- model=LLAMA_MODEL,
- dataset=input_dataset,
- max_new_tokens=step,
- temperature=self.temperature,
- )
- response = output_dataset.to_dict()["instances"][0]["text"]
- if response == "" or response == self.end_string:
- break
- partial_text += response
- yield partial_text
-
-
-class XMChat(BaseLLMModel):
- def __init__(self, api_key, user_name=""):
- super().__init__(model_name="xmchat", user=user_name)
- self.api_key = api_key
- self.session_id = None
- self.reset()
- self.image_bytes = None
- self.image_path = None
- self.xm_history = []
- self.url = "https://xmbot.net/web"
- self.last_conv_id = None
-
- def reset(self):
- self.session_id = str(uuid.uuid4())
- self.last_conv_id = None
- return [], "已重置"
-
- def image_to_base64(self, image_path):
- # 打开并加载图片
- img = Image.open(image_path)
-
- # 获取图片的宽度和高度
- width, height = img.size
-
- # 计算压缩比例，以确保最长边小于4096像素
- max_dimension = 2048
- scale_ratio = min(max_dimension / width, max_dimension / height)
-
- if scale_ratio < 1:
- # 按压缩比例调整图片大小
- new_width = int(width * scale_ratio)
- new_height = int(height * scale_ratio)
- img = img.resize((new_width, new_height), Image.ANTIALIAS)
-
- # 将图片转换为jpg格式的二进制数据
- buffer = BytesIO()
- if img.mode == "RGBA":
- img = img.convert("RGB")
- img.save(buffer, format='JPEG')
- binary_image = buffer.getvalue()
-
- # 对二进制数据进行Base64编码
- base64_image = base64.b64encode(binary_image).decode('utf-8')
-
- return base64_image
-
- def try_read_image(self, filepath):
- def is_image_file(filepath):
- # 判断文件是否为图片
- valid_image_extensions = [
- ".jpg", ".jpeg", ".png", ".bmp", ".gif", ".tiff"]
- file_extension = os.path.splitext(filepath)[1].lower()
- return file_extension in valid_image_extensions
-
- if is_image_file(filepath):
- logging.info(f"读取图片文件: {filepath}")
- self.image_bytes = self.image_to_base64(filepath)
- self.image_path = filepath
- else:
- self.image_bytes = None
- self.image_path = None
-
- def like(self):
- if self.last_conv_id is None:
- return "点赞失败，你还没发送过消息"
- data = {
- "uuid": self.last_conv_id,
- "appraise": "good"
- }
- requests.post(self.url, json=data)
- return "👍点赞成功，感谢反馈～"
-
- def dislike(self):
- if self.last_conv_id is None:
- return "点踩失败，你还没发送过消息"
- data = {
- "uuid": self.last_conv_id,
- "appraise": "bad"
- }
- requests.post(self.url, json=data)
- return "👎点踩成功，感谢反馈～"
-
- def prepare_inputs(self, real_inputs, use_websearch, files, reply_language, chatbot):
- fake_inputs = real_inputs
- display_append = ""
- limited_context = False
- return limited_context, fake_inputs, display_append, real_inputs, chatbot
-
- def handle_file_upload(self, files, chatbot, language):
- """if the model accepts multi modal input, implement this function"""
- if files:
- for file in files:
- if file.name:
- logging.info(f"尝试读取图像: {file.name}")
- self.try_read_image(file.name)
- if self.image_path is not None:
- chatbot = chatbot + [((self.image_path,), None)]
- if self.image_bytes is not None:
- logging.info("使用图片作为输入")
- # XMChat的一轮对话中实际上只能处理一张图片
- self.reset()
- conv_id = str(uuid.uuid4())
- data = {
- "user_id": self.api_key,
- "session_id": self.session_id,
- "uuid": conv_id,
- "data_type": "imgbase64",
- "data": self.image_bytes
- }
- response = requests.post(self.url, json=data)
- response = json.loads(response.text)
- logging.info(f"图片回复: {response['data']}")
- return None, chatbot, None
-
- def get_answer_at_once(self):
- question = self.history[-1]["content"]
- conv_id = str(uuid.uuid4())
- self.last_conv_id = conv_id
- data = {
- "user_id": self.api_key,
- "session_id": self.session_id,
- "uuid": conv_id,
- "data_type": "text",
- "data": question
- }
- response = requests.post(self.url, json=data)
- try:
- response = json.loads(response.text)
- return response["data"], len(response["data"])
- except Exception as e:
- return response.text, len(response.text)
-
-
def get_model(
model_name,
lora_model_path=None,
@@ -548,21 +21,23 @@ def get_model(
temperature=None,
top_p=None,
system_prompt=None,
- user_name=""
+ user_name="",
+ original_model = None
) -> BaseLLMModel:
msg = i18n("模型设置为了：") + f" {model_name}"
model_type = ModelType.get_type(model_name)
lora_selector_visibility = False
- lora_choices = []
+ lora_choices = ["No LoRA"]
dont_change_lora_selector = False
if model_type != ModelType.OpenAI:
config.local_embedding = True
# del current_model.model
- model = None
+ model = original_model
chatbot = gr.Chatbot.update(label=model_name)
try:
if model_type == ModelType.OpenAI:
logging.info(f"正在加载OpenAI模型: {model_name}")
+ from .OpenAI import OpenAIClient
access_key = os.environ.get("OPENAI_API_KEY", access_key)
model = OpenAIClient(
model_name=model_name,
@@ -572,19 +47,31 @@ def get_model(
top_p=top_p,
user_name=user_name,
)
+ elif model_type == ModelType.OpenAIInstruct:
+ logging.info(f"正在加载OpenAI Instruct模型: {model_name}")
+ from .OpenAIInstruct import OpenAI_Instruct_Client
+ access_key = os.environ.get("OPENAI_API_KEY", access_key)
+ model = OpenAI_Instruct_Client(
+ model_name, api_key=access_key, user_name=user_name)
+ elif model_type == ModelType.OpenAIVision:
+ logging.info(f"正在加载OpenAI Vision模型: {model_name}")
+ from .OpenAIVision import OpenAIVisionClient
+ access_key = os.environ.get("OPENAI_API_KEY", access_key)
+ model = OpenAIVisionClient(
+ model_name, api_key=access_key, user_name=user_name)
elif model_type == ModelType.ChatGLM:
logging.info(f"正在加载ChatGLM模型: {model_name}")
+ from .ChatGLM import ChatGLM_Client
model = ChatGLM_Client(model_name, user_name=user_name)
elif model_type == ModelType.LLaMA and lora_model_path == "":
msg = f"现在请为 {model_name} 选择LoRA模型"
logging.info(msg)
lora_selector_visibility = True
if os.path.isdir("lora"):
- lora_choices = get_file_names(
- "lora", plain=True, filetypes=[""])
- lora_choices = ["No LoRA"] + lora_choices
+ lora_choices = ["No LoRA"] + get_file_names_by_pinyin("lora", filetypes=[""])
elif model_type == ModelType.LLaMA and lora_model_path != "":
logging.info(f"正在加载LLaMA模型: {model_name} + {lora_model_path}")
+ from .LLaMA import LLaMA_Client
dont_change_lora_selector = True
if lora_model_path == "No LoRA":
lora_model_path = None
@@ -594,6 +81,7 @@ def get_model(
model = LLaMA_Client(
model_name, lora_model_path, user_name=user_name)
elif model_type == ModelType.XMChat:
+ from .XMChat import XMChat
if os.environ.get("XMCHAT_API_KEY") != "":
access_key = os.environ.get("XMCHAT_API_KEY")
model = XMChat(api_key=access_key, user_name=user_name)
@@ -605,26 +93,41 @@ def get_model(
model = MOSS_Client(model_name, user_name=user_name)
elif model_type == ModelType.YuanAI:
from .inspurai import Yuan_Client
- model = Yuan_Client(model_name, api_key=access_key, user_name=user_name, system_prompt=system_prompt)
+ model = Yuan_Client(model_name, api_key=access_key,
+ user_name=user_name, system_prompt=system_prompt)
elif model_type == ModelType.Minimax:
from .minimax import MiniMax_Client
if os.environ.get("MINIMAX_API_KEY") != "":
access_key = os.environ.get("MINIMAX_API_KEY")
- model = MiniMax_Client(model_name, api_key=access_key, user_name=user_name, system_prompt=system_prompt)
+ model = MiniMax_Client(
+ model_name, api_key=access_key, user_name=user_name, system_prompt=system_prompt)
elif model_type == ModelType.ChuanhuAgent:
from .ChuanhuAgent import ChuanhuAgent_Client
model = ChuanhuAgent_Client(model_name, access_key, user_name=user_name)
+ msg = i18n("启用的工具：") + ", ".join([i.name for i in model.tools])
elif model_type == ModelType.GooglePaLM:
- from .Google_PaLM import Google_PaLM_Client
+ from .GooglePaLM import Google_PaLM_Client
access_key = os.environ.get("GOOGLE_PALM_API_KEY", access_key)
- model = Google_PaLM_Client(model_name, access_key, user_name=user_name)
+ model = Google_PaLM_Client(
+ model_name, access_key, user_name=user_name)
elif model_type == ModelType.LangchainChat:
- from .azure import Azure_OpenAI_Client
+ from .Azure import Azure_OpenAI_Client
model = Azure_OpenAI_Client(model_name, user_name=user_name)
elif model_type == ModelType.Midjourney:
from .midjourney import Midjourney_Client
mj_proxy_api_secret = os.getenv("MIDJOURNEY_PROXY_API_SECRET")
- model = Midjourney_Client(model_name, mj_proxy_api_secret, user_name=user_name)
+ model = Midjourney_Client(
+ model_name, mj_proxy_api_secret, user_name=user_name)
+ elif model_type == ModelType.Spark:
+ from .spark import Spark_Client
+ model = Spark_Client(model_name, os.getenv("SPARK_APPID"), os.getenv(
+ "SPARK_API_KEY"), os.getenv("SPARK_API_SECRET"), user_name=user_name)
+ elif model_type == ModelType.Claude:
+ from .Claude import Claude_Client
+ model = Claude_Client(model_name="claude-2", api_secret=os.getenv("CLAUDE_API_SECRET"))
+ elif model_type == ModelType.Qwen:
+ from .Qwen import Qwen_Client
+ model = Qwen_Client(model_name, user_name=user_name)
elif model_type == ModelType.Unknown:
raise ValueError(f"未知模型: {model_name}")
logging.info(msg)
@@ -633,6 +136,9 @@ def get_model(
traceback.print_exc()
msg = f"{STANDARD_ERROR_MSG}: {e}"
presudo_key = hide_middle_chars(access_key)
+ if original_model is not None and model is not None:
+ model.history = original_model.history
+ model.history_file_path = original_model.history_file_path
if dont_change_lora_selector:
return model, msg, chatbot, gr.update(), access_key, presudo_key
else:
diff --git a/modules/models/spark.py b/modules/models/spark.py
new file mode 100644
index 0000000000000000000000000000000000000000..1c5790f5ad3802301d788623326fb21afcb030aa
--- /dev/null
+++ b/modules/models/spark.py
@@ -0,0 +1,166 @@
+import _thread as thread
+import base64
+import datetime
+import hashlib
+import hmac
+import json
+from collections import deque
+from urllib.parse import urlparse
+import ssl
+from datetime import datetime
+from time import mktime
+from urllib.parse import urlencode
+from wsgiref.handlers import format_date_time
+from threading import Condition
+import websocket
+import logging
+
+from .base_model import BaseLLMModel, CallbackToIterator
+
+
+class Ws_Param(object):
+ # 来自官方 Demo
+ # 初始化
+ def __init__(self, APPID, APIKey, APISecret, Spark_url):
+ self.APPID = APPID
+ self.APIKey = APIKey
+ self.APISecret = APISecret
+ self.host = urlparse(Spark_url).netloc
+ self.path = urlparse(Spark_url).path
+ self.Spark_url = Spark_url
+
+ # 生成url
+ def create_url(self):
+ # 生成RFC1123格式的时间戳
+ now = datetime.now()
+ date = format_date_time(mktime(now.timetuple()))
+
+ # 拼接字符串
+ signature_origin = "host: " + self.host + "\n"
+ signature_origin += "date: " + date + "\n"
+ signature_origin += "GET " + self.path + " HTTP/1.1"
+
+ # 进行hmac-sha256进行加密
+ signature_sha = hmac.new(
+ self.APISecret.encode("utf-8"),
+ signature_origin.encode("utf-8"),
+ digestmod=hashlib.sha256,
+ ).digest()
+
+ signature_sha_base64 = base64.b64encode(
+ signature_sha).decode(encoding="utf-8")
+
+ authorization_origin = f'api_key="{self.APIKey}", algorithm="hmac-sha256", headers="host date request-line", signature="{signature_sha_base64}"'
+
+ authorization = base64.b64encode(authorization_origin.encode("utf-8")).decode(
+ encoding="utf-8"
+ )
+
+ # 将请求的鉴权参数组合为字典
+ v = {"authorization": authorization, "date": date, "host": self.host}
+ # 拼接鉴权参数，生成url
+ url = self.Spark_url + "?" + urlencode(v)
+ # 此处打印出建立连接时候的url,参考本demo的时候可取消上方打印的注释，比对相同参数时生成的url与自己代码生成的url是否一致
+ return url
+
+
+class Spark_Client(BaseLLMModel):
+ def __init__(self, model_name, appid, api_key, api_secret, user_name="") -> None:
+ super().__init__(model_name=model_name, user=user_name)
+ self.api_key = api_key
+ self.appid = appid
+ self.api_secret = api_secret
+ if None in [self.api_key, self.appid, self.api_secret]:
+ raise Exception("请在配置文件或者环境变量中设置讯飞的API Key、APP ID和API Secret")
+ if "2.0" in self.model_name:
+ self.spark_url = "wss://spark-api.xf-yun.com/v2.1/chat"
+ self.domain = "generalv2"
+ if "3.0" in self.model_name:
+ self.spark_url = "wss://spark-api.xf-yun.com/v3.1/chat"
+ self.domain = "generalv3"
+ else:
+ self.spark_url = "wss://spark-api.xf-yun.com/v1.1/chat"
+ self.domain = "general"
+
+ # 收到websocket错误的处理
+ def on_error(self, ws, error):
+ ws.iterator.callback("出现了错误:" + error)
+
+ # 收到websocket关闭的处理
+ def on_close(self, ws, one, two):
+ pass
+
+ # 收到websocket连接建立的处理
+ def on_open(self, ws):
+ thread.start_new_thread(self.run, (ws,))
+
+ def run(self, ws, *args):
+ data = json.dumps(
+ self.gen_params()
+ )
+ ws.send(data)
+
+ # 收到websocket消息的处理
+ def on_message(self, ws, message):
+ ws.iterator.callback(message)
+
+ def gen_params(self):
+ """
+ 通过appid和用户的提问来生成请参数
+ """
+ data = {
+ "header": {"app_id": self.appid, "uid": "1234"},
+ "parameter": {
+ "chat": {
+ "domain": self.domain,
+ "random_threshold": self.temperature,
+ "max_tokens": 4096,
+ "auditing": "default",
+ }
+ },
+ "payload": {"message": {"text": self.history}},
+ }
+ return data
+
+ def get_answer_stream_iter(self):
+ wsParam = Ws_Param(self.appid, self.api_key, self.api_secret, self.spark_url)
+ websocket.enableTrace(False)
+ wsUrl = wsParam.create_url()
+ ws = websocket.WebSocketApp(
+ wsUrl,
+ on_message=self.on_message,
+ on_error=self.on_error,
+ on_close=self.on_close,
+ on_open=self.on_open,
+ )
+ ws.appid = self.appid
+ ws.domain = self.domain
+
+ # Initialize the CallbackToIterator
+ ws.iterator = CallbackToIterator()
+
+ # Start the WebSocket connection in a separate thread
+ thread.start_new_thread(
+ ws.run_forever, (), {"sslopt": {"cert_reqs": ssl.CERT_NONE}}
+ )
+
+ # Iterate over the CallbackToIterator instance
+ answer = ""
+ total_tokens = 0
+ for message in ws.iterator:
+ data = json.loads(message)
+ code = data["header"]["code"]
+ if code != 0:
+ ws.close()
+ raise Exception(f"请求错误: {code}, {data}")
+ else:
+ choices = data["payload"]["choices"]
+ status = choices["status"]
+ content = choices["text"][0]["content"]
+ if "usage" in data["payload"]:
+ total_tokens = data["payload"]["usage"]["text"]["total_tokens"]
+ answer += content
+ if status == 2:
+ ws.iterator.finish() # Finish the iterator when the status is 2
+ ws.close()
+ yield answer, total_tokens
diff --git a/modules/overwrites.py b/modules/overwrites.py
index a4ef6167eb7ce75ed8b88024ad1187b24f2fc191..971c342b2467b179bb7bfebe54dfaca044231b9b 100644
--- a/modules/overwrites.py
+++ b/modules/overwrites.py
@@ -44,32 +44,36 @@ def postprocess_chat_messages(
) -> str | dict | None:
if chat_message is None:
return None
- elif isinstance(chat_message, (tuple, list)):
- file_uri = chat_message[0]
- if utils.validate_url(file_uri):
- filepath = file_uri
- else:
- filepath = self.make_temp_copy_if_needed(file_uri)
-
- mime_type = client_utils.get_mimetype(filepath)
- return {
- "name": filepath,
- "mime_type": mime_type,
- "alt_text": chat_message[1] if len(chat_message) > 1 else None,
- "data": None, # These last two fields are filled in by the frontend
- "is_file": True,
- }
- elif isinstance(chat_message, str):
- # chat_message = inspect.cleandoc(chat_message)
- # escape html spaces
- # chat_message = chat_message.replace(" ", " ")
- if role == "bot":
- chat_message = convert_bot_before_marked(chat_message)
- elif role == "user":
- chat_message = convert_user_before_marked(chat_message)
- return chat_message
else:
- raise ValueError(f"Invalid message for Chatbot component: {chat_message}")
+ if isinstance(chat_message, (tuple, list)):
+ if len(chat_message) > 0 and "text" in chat_message[0]:
+ chat_message = chat_message[0]["text"]
+ else:
+ file_uri = chat_message[0]
+ if utils.validate_url(file_uri):
+ filepath = file_uri
+ else:
+ filepath = self.make_temp_copy_if_needed(file_uri)
+
+ mime_type = client_utils.get_mimetype(filepath)
+ return {
+ "name": filepath,
+ "mime_type": mime_type,
+ "alt_text": chat_message[1] if len(chat_message) > 1 else None,
+ "data": None, # These last two fields are filled in by the frontend
+ "is_file": True,
+ }
+ if isinstance(chat_message, str):
+ # chat_message = inspect.cleandoc(chat_message)
+ # escape html spaces
+ # chat_message = chat_message.replace(" ", " ")
+ if role == "bot":
+ chat_message = convert_bot_before_marked(chat_message)
+ elif role == "user":
+ chat_message = convert_user_before_marked(chat_message)
+ return chat_message
+ else:
+ raise ValueError(f"Invalid message for Chatbot component: {chat_message}")
@@ -103,4 +107,3 @@ def BlockContext_init(self, *args, **kwargs):
original_BlockContext_init = gr.blocks.BlockContext.__init__
gr.blocks.BlockContext.__init__ = BlockContext_init
-
diff --git a/modules/presets.py b/modules/presets.py
index a56d50e1c7aefae37b3252b983d445ea327471a4..e0a2fbb6096d8f4a00fe6f948378ee8001bb3329 100644
--- a/modules/presets.py
+++ b/modules/presets.py
@@ -14,7 +14,9 @@ LLAMA_INFERENCER = None
# ChatGPT 设置
INITIAL_SYSTEM_PROMPT = "You are a helpful assistant."
API_HOST = "api.openai.com"
-COMPLETION_URL = "https://api.openai.com/v1/chat/completions"
+OPENAI_API_BASE = "https://api.openai.com/v1"
+CHAT_COMPLETION_URL = "https://api.openai.com/v1/chat/completions"
+COMPLETION_URL = "https://api.openai.com/v1/completions"
BALANCE_API_URL="https://api.openai.com/dashboard/billing/credit_grants"
USAGE_API_URL="https://api.openai.com/dashboard/billing/usage"
HISTORY_DIR = Path("history")
@@ -36,6 +38,7 @@ BILLING_NOT_APPLICABLE_MSG = i18n("账单信息不适用") # 本地运行的模
TIMEOUT_STREAMING = 60 # 流式对话时的超时时间
TIMEOUT_ALL = 200 # 非流式对话时的超时时间
ENABLE_STREAMING_OPTION = True # 是否启用选择选择是否实时显示回答的勾选框
+ENABLE_LLM_NAME_CHAT_OPTION = True # 是否启用选择是否使用LLM模型的勾选框
HIDE_MY_KEY = False # 如果你想在UI中隐藏你的 API 密钥，将此值设置为 True
CONCURRENT_COUNT = 100 # 允许同时使用的用户数量
@@ -48,16 +51,15 @@ CHUANHU_DESCRIPTION = i18n("由Bilibili [土川虎虎虎](https://space.bilibili
ONLINE_MODELS = [
- "gpt-3.5-turbo",
- "gpt-3.5-turbo-16k",
- "gpt-3.5-turbo-0301",
- "gpt-3.5-turbo-0613",
- "gpt-4",
- "gpt-4-0314",
- "gpt-4-0613",
- "gpt-4-32k",
- "gpt-4-32k-0314",
- "gpt-4-32k-0613",
+ "GPT3.5 Turbo",
+ "GPT3.5 Turbo Instruct",
+ "GPT3.5 Turbo 16K",
+ "GPT3.5 Turbo 0301",
+ "GPT3.5 Turbo 0613",
+ "GPT4",
+ "GPT4 32K",
+ "GPT4 Turbo",
+ "GPT4 Vision",
"川虎助理",
"川虎助理 Pro",
"GooglePaLM",
@@ -67,9 +69,12 @@ ONLINE_MODELS = [
"yuanai-1.0-translate",
"yuanai-1.0-dialog",
"yuanai-1.0-rhythm_poems",
- "minimax-abab4-chat",
"minimax-abab5-chat",
- "midjourney"
+ "midjourney",
+ "讯飞星火大模型V3.0",
+ "讯飞星火大模型V2.0",
+ "讯飞星火大模型V1.5",
+ "Claude"
]
LOCAL_MODELS = [
@@ -80,12 +85,69 @@ LOCAL_MODELS = [
"chatglm2-6b-int4",
"StableLM",
"MOSS",
- "llama-7b-hf",
- "llama-13b-hf",
- "llama-30b-hf",
- "llama-65b-hf",
+ "Llama-2-7B-Chat",
+ "Qwen 7B",
+ "Qwen 14B"
]
+# Additional metadata for online and local models
+MODEL_METADATA = {
+ "Llama-2-7B":{
+ "repo_id": "TheBloke/Llama-2-7B-GGUF",
+ "filelist": ["llama-2-7b.Q6_K.gguf"],
+ },
+ "Llama-2-7B-Chat":{
+ "repo_id": "TheBloke/Llama-2-7b-Chat-GGUF",
+ "filelist": ["llama-2-7b-chat.Q6_K.gguf"],
+ },
+ "Qwen 7B": {
+ "repo_id": "Qwen/Qwen-7B-Chat-Int4",
+ },
+ "Qwen 14B": {
+ "repo_id": "Qwen/Qwen-14B-Chat-Int4",
+ },
+ "GPT3.5 Turbo": {
+ "model_name": "gpt-3.5-turbo",
+ "token_limit": 4096,
+ },
+ "GPT3.5 Turbo Instruct": {
+ "model_name": "gpt-3.5-turbo-instruct",
+ "token_limit": 4096,
+ },
+ "GPT3.5 Turbo 16K": {
+ "model_name": "gpt-3.5-turbo-16k",
+ "token_limit": 16384,
+ },
+ "GPT3.5 Turbo 0301": {
+ "model_name": "gpt-3.5-turbo-0301",
+ "token_limit": 4096,
+ },
+ "GPT3.5 Turbo 0613": {
+ "model_name": "gpt-3.5-turbo-0613",
+ "token_limit": 4096,
+ },
+ "GPT4": {
+ "model_name": "gpt-4",
+ "token_limit": 8192,
+ },
+ "GPT4 32K": {
+ "model_name": "gpt-4-32k",
+ "token_limit": 32768,
+ },
+ "GPT4 Turbo": {
+ "model_name": "gpt-4-1106-preview",
+ "token_limit": 128000,
+ },
+ "GPT4 Vision": {
+ "model_name": "gpt-4-vision-preview",
+ "token_limit": 128000,
+ },
+ "Claude": {
+ "model_name": "Claude",
+ "token_limit": 4096,
+ },
+}
+
if os.environ.get('HIDE_LOCAL_MODELS', 'false') == 'true':
MODELS = ONLINE_MODELS
else:
@@ -101,19 +163,6 @@ for dir_name in os.listdir("models"):
if dir_name not in MODELS:
MODELS.append(dir_name)
-MODEL_TOKEN_LIMIT = {
- "gpt-3.5-turbo": 4096,
- "gpt-3.5-turbo-16k": 16384,
- "gpt-3.5-turbo-0301": 4096,
- "gpt-3.5-turbo-0613": 4096,
- "gpt-4": 8192,
- "gpt-4-0314": 8192,
- "gpt-4-0613": 8192,
- "gpt-4-32k": 32768,
- "gpt-4-32k-0314": 32768,
- "gpt-4-32k-0613": 32768
-}
-
TOKEN_OFFSET = 1000 # 模型的token上限减去这个值，得到软上限。到达软上限之后，自动尝试减少token占用。
DEFAULT_TOKEN_LIMIT = 3000 # 默认的token上限
REDUCE_TOKEN_FACTOR = 0.5 # 与模型token上限想乘，得到目标token数。减少token占用时，将token占用减少到目标token数以下。
@@ -125,11 +174,18 @@ REPLY_LANGUAGES = [
"日本語",
"Español",
"Français",
+ "Russian",
"Deutsch",
"한국어",
"跟随问题语言（不稳定）"
]
+HISTORY_NAME_METHODS = [
+ i18n("根据日期时间"),
+ i18n("第一条提问"),
+ i18n("模型自动总结（消耗tokens）"),
+]
+
WEBSEARCH_PTOMPT_TEMPLATE = """\
Web search results:
@@ -175,6 +231,15 @@ SUMMARIZE_PROMPT = """Write a concise summary of the following:
CONCISE SUMMARY IN 中文:"""
+SUMMARY_CHAT_SYSTEM_PROMPT = """\
+Please summarize the following conversation for a chat topic.
+No more than 16 characters.
+No special characters.
+Punctuation mark is banned.
+Not including '.' ':' '?' '!' '“' '*' '<' '>'.
+Reply in user's language.
+"""
+
ALREADY_CONVERTED_MARK = ""
START_OF_OUTPUT_MARK = ""
END_OF_OUTPUT_MARK = ""
@@ -243,6 +308,7 @@ small_and_beautiful_theme = gr.themes.Soft(
block_title_background_fill_dark="*primary_900",
block_label_background_fill_dark="*primary_900",
input_background_fill="#F6F6F6",
- chatbot_code_background_color="*neutral_950",
+ # chatbot_code_background_color="*neutral_950",
+ # gradio 会把这个几个chatbot打头的变量应用到其他md渲染的地方，鬼晓得怎么想的。。。
chatbot_code_background_color_dark="*neutral_950",
)
diff --git a/modules/repo.py b/modules/repo.py
index 2788de5b06a744bc436df677a973d89c26489a8a..cb69adeed2bd17f089ed233f53fe9cf4c121ca1c 100644
--- a/modules/repo.py
+++ b/modules/repo.py
@@ -6,30 +6,33 @@ from functools import lru_cache
import logging
import gradio as gr
import datetime
+import platform
# This file is mainly used to describe repo version info, execute the git command, python pip command, shell command, etc.
# Part of the code in this file is referenced from stable-diffusion-webui/modules/launch_utils.py
python = sys.executable
-pip = os.environ.get('PIP', "pip")
-git = os.environ.get('GIT', "git")
+pip = os.environ.get("PIP", "pip")
+git = os.environ.get("GIT", "git")
# Pypi index url
-index_url = os.environ.get('INDEX_URL', "")
+index_url = os.environ.get("INDEX_URL", "")
# Whether to default to printing command output
default_command_live = True
-def run(command, desc=None, errdesc=None, custom_env=None, live: bool = default_command_live) -> str:
+def run(
+ command, desc=None, errdesc=None, custom_env=None, live: bool = default_command_live
+) -> str:
if desc is not None:
print(desc)
run_kwargs = {
"args": command,
"shell": True,
"env": os.environ if custom_env is None else custom_env,
- "encoding": 'utf8',
- "errors": 'ignore',
+ "encoding": "utf8",
+ "errors": "ignore",
}
if not live:
@@ -48,29 +51,32 @@ def run(command, desc=None, errdesc=None, custom_env=None, live: bool = default_
error_bits.append(f"stderr: {result.stderr}")
raise RuntimeError("\n".join(error_bits))
- return (result.stdout or "")
+ return result.stdout or ""
def run_pip(command, desc=None, pref=None, live=default_command_live):
# if args.skip_install:
# return
- index_url_line = f' --index-url {index_url}' if index_url != '' else ''
+ index_url_line = f" --index-url {index_url}" if index_url != "" else ""
return run(
- f'"{python}" -m pip {command} --prefer-binary{index_url_line}',
- desc=f"{pref} Installing {desc}...",
- errdesc=f"Couldn't install {desc}",
- live=live
+ f'"{python}" -m pip {command} --prefer-binary{index_url_line}',
+ desc=f"{pref} Installing {desc}...",
+ errdesc=f"Couldn't install {desc}",
+ live=live,
)
@lru_cache()
def commit_hash():
try:
- return subprocess.check_output([git, "rev-parse", "HEAD"], shell=False, encoding='utf8').strip()
+ return subprocess.check_output(
+ [git, "rev-parse", "HEAD"], shell=False, encoding="utf8"
+ ).strip()
except Exception:
return "
"
+
def commit_html():
commit = commit_hash()
if commit != "":
@@ -80,6 +86,7 @@ def commit_html():
commit_info = "unknown \U0001F615"
return commit_info
+
@lru_cache()
def tag_html():
try:
@@ -87,7 +94,7 @@ def tag_html():
try:
# tag = subprocess.check_output([git, "describe", "--tags", "--exact-match"], shell=False, encoding='utf8').strip()
tag = run(f"{git} describe --tags --exact-match", live=False).strip()
- except Exception:
+ except Exception:
tag = ""
except Exception:
tag = ""
@@ -98,14 +105,16 @@ def tag_html():
tag_info = f'{latest_tag} * '
else:
tag_info = f'{tag} '
-
+
return tag_info
+
def repo_tag_html():
commit_version = commit_html()
tag_version = tag_html()
return tag_version if tag_version != "unknown \U0001F615" else commit_version
+
def versions_html():
python_version = ".".join([str(x) for x in sys.version_info[0:3]])
repo_version = repo_tag_html()
@@ -117,20 +126,44 @@ def versions_html():
ChuanhuChat : {repo_version}
"""
+
def version_time():
+ git = "git"
+ cmd = f"{git} log -1 --format=%cd --date=iso-strict"
+ commit_time = "unknown"
try:
- commit_time = subprocess.check_output(f"TZ=UTC {git} log -1 --format=%cd --date='format-local:%Y-%m-%dT%H:%M:%SZ'", shell=True, encoding='utf8').strip()
- # commit_time = run(f"TZ=UTC {git} log -1 --format=%cd --date='format-local:%Y-%m-%dT%H:%M:%SZ'").strip()
+ if platform.system() == "Windows":
+ # For Windows
+ env = dict(os.environ) # copy the current environment
+ env["TZ"] = "UTC" # set timezone to UTC
+ raw_commit_time = subprocess.check_output(
+ cmd, shell=True, encoding="utf8", env=env
+ ).strip()
+ else:
+ # For Unix systems
+ cmd = f"TZ=UTC {cmd}"
+ raw_commit_time = subprocess.check_output(
+ cmd, shell=True, encoding="utf8"
+ ).strip()
+
+ # Convert the date-time to the desired format
+ commit_datetime = datetime.datetime.strptime(
+ raw_commit_time, "%Y-%m-%dT%H:%M:%S%z"
+ )
+ commit_time = commit_datetime.strftime("%Y-%m-%dT%H:%M:%SZ")
+
+ # logging.info(f"commit time: {commit_time}")
except Exception:
commit_time = "unknown"
return commit_time
-
def get_current_branch():
try:
# branch = run(f"{git} rev-parse --abbrev-ref HEAD").strip()
- branch = subprocess.check_output([git, "rev-parse", "--abbrev-ref", "HEAD"], shell=False, encoding='utf8').strip()
+ branch = subprocess.check_output(
+ [git, "rev-parse", "--abbrev-ref", "HEAD"], shell=False, encoding="utf8"
+ ).strip()
except Exception:
branch = ""
return branch
@@ -139,7 +172,10 @@ def get_current_branch():
def get_latest_release():
try:
import requests
- release = requests.get("https://api.github.com/repos/GaiZhenbiao/ChuanhuChatGPT/releases/latest").json()
+
+ release = requests.get(
+ "https://api.github.com/repos/GaiZhenbiao/ChuanhuChatGPT/releases/latest"
+ ).json()
tag = release["tag_name"]
release_note = release["body"]
need_pip = release_note.find("requirements reinstall needed") != -1
@@ -149,21 +185,34 @@ def get_latest_release():
need_pip = False
return {"tag": tag, "release_note": release_note, "need_pip": need_pip}
+
def get_tag_commit_hash(tag):
try:
import requests
- tags = requests.get("https://api.github.com/repos/GaiZhenbiao/ChuanhuChatGPT/tags").json()
+
+ tags = requests.get(
+ "https://api.github.com/repos/GaiZhenbiao/ChuanhuChatGPT/tags"
+ ).json()
commit_hash = [x["commit"]["sha"] for x in tags if x["name"] == tag][0]
except Exception:
commit_hash = ""
return commit_hash
+
def repo_need_stash():
try:
- return subprocess.check_output([git, "diff-index", "--quiet", "HEAD", "--"], shell=False, encoding='utf8').strip() != ""
+ return (
+ subprocess.check_output(
+ [git, "diff-index", "--quiet", "HEAD", "--"],
+ shell=False,
+ encoding="utf8",
+ ).strip()
+ != ""
+ )
except Exception:
return True
+
def background_update():
# {git} fetch --all && ({git} pull https://github.com/GaiZhenbiao/ChuanhuChatGPT.git main -f || ({git} stash && {git} pull https://github.com/GaiZhenbiao/ChuanhuChatGPT.git main -f && {git} stash pop)) && {pip} install -r requirements.txt")
try:
@@ -175,50 +224,83 @@ def background_update():
timestamp = datetime.datetime.now().strftime("%Y%m%d%H%M%S")
current_branch = get_current_branch()
- updater_branch = f'tmp_{timestamp}'
- backup_branch = f'backup_{timestamp}'
+ updater_branch = f"tmp_{timestamp}"
+ backup_branch = f"backup_{timestamp}"
track_repo = "https://github.com/GaiZhenbiao/ChuanhuChatGPT.git"
try:
try:
- run(f"{git} fetch {track_repo}", desc="[Updater] Fetching from github...", live=False)
+ run(
+ f"{git} fetch {track_repo}",
+ desc="[Updater] Fetching from github...",
+ live=False,
+ )
except Exception:
- logging.error(f"Update failed in fetching, check your network connection")
+ logging.error(
+ f"Update failed in fetching, check your network connection"
+ )
return "failed"
-
- run(f'{git} stash push --include-untracked -m "updater-{timestamp}"',
- desc=f'[Updater] Restoring you local changes on stash updater-{timestamp}', live=False) if need_stash else None
-
+
+ run(
+ f'{git} stash push --include-untracked -m "updater-{timestamp}"',
+ desc=f"[Updater] Restoring you local changes on stash updater-{timestamp}",
+ live=False,
+ ) if need_stash else None
+
run(f"{git} checkout -b {backup_branch}", live=False)
run(f"{git} checkout -b {updater_branch}", live=False)
run(f"{git} reset --hard FETCH_HEAD", live=False)
- run(f"{git} reset --hard {latest_release_hash}", desc=f'[Updater] Checking out {latest_release_tag}...', live=False)
+ run(
+ f"{git} reset --hard {latest_release_hash}",
+ desc=f"[Updater] Checking out {latest_release_tag}...",
+ live=False,
+ )
run(f"{git} checkout {current_branch}", live=False)
-
+
try:
- run(f"{git} merge --no-edit {updater_branch} -q", desc=f"[Updater] Trying to apply latest update on version {latest_release_tag}...")
+ run(
+ f"{git} merge --no-edit {updater_branch} -q",
+ desc=f"[Updater] Trying to apply latest update on version {latest_release_tag}...",
+ )
except Exception:
logging.error(f"Update failed in merging")
try:
- run(f"{git} merge --abort", desc="[Updater] Conflict detected, canceling update...")
+ run(
+ f"{git} merge --abort",
+ desc="[Updater] Conflict detected, canceling update...",
+ )
run(f"{git} reset --hard {backup_branch}", live=False)
run(f"{git} branch -D -f {updater_branch}", live=False)
run(f"{git} branch -D -f {backup_branch}", live=False)
run(f"{git} stash pop", live=False) if need_stash else None
- logging.error(f"Update failed, but your file was safely reset to the state before the update.")
+ logging.error(
+ f"Update failed, but your file was safely reset to the state before the update."
+ )
return "failed"
except Exception as e:
- logging.error(f"!!!Update failed in resetting, try to reset your files manually. {e}")
+ logging.error(
+ f"!!!Update failed in resetting, try to reset your files manually. {e}"
+ )
return "failed"
if need_stash:
try:
- run(f"{git} stash apply", desc="[Updater] Trying to restore your local modifications...", live=False)
+ run(
+ f"{git} stash apply",
+ desc="[Updater] Trying to restore your local modifications...",
+ live=False,
+ )
except Exception:
- run(f"{git} reset --hard {backup_branch}", desc="[Updater] Conflict detected, canceling update...", live=False)
+ run(
+ f"{git} reset --hard {backup_branch}",
+ desc="[Updater] Conflict detected, canceling update...",
+ live=False,
+ )
run(f"{git} branch -D -f {updater_branch}", live=False)
run(f"{git} branch -D -f {backup_branch}", live=False)
run(f"{git} stash pop", live=False)
- logging.error(f"Update failed in applying your local changes, but your file was safely reset to the state before the update.")
+ logging.error(
+ f"Update failed in applying your local changes, but your file was safely reset to the state before the update."
+ )
return "failed"
run(f"{git} stash drop", live=False)
@@ -228,12 +310,17 @@ def background_update():
logging.error(f"Update failed: {e}")
return "failed"
if need_pip:
- try:
- run_pip(f"install -r requirements.txt", pref="[Updater]", desc="requirements", live=False)
+ try:
+ run_pip(
+ f"install -r requirements.txt",
+ pref="[Updater]",
+ desc="requirements",
+ live=False,
+ )
except Exception:
logging.error(f"Update failed in pip install")
return "failed"
return "success"
except Exception as e:
logging.error(f"Update failed: {e}")
- return "failed"
+ return "failed"
diff --git a/modules/shared.py b/modules/shared.py
index 89f0779459225957c13865ef7f7448efae6d1998..381d7374a18a80f55a148b37115980f0aabf066d 100644
--- a/modules/shared.py
+++ b/modules/shared.py
@@ -1,4 +1,4 @@
-from modules.presets import COMPLETION_URL, BALANCE_API_URL, USAGE_API_URL, API_HOST
+from modules.presets import CHAT_COMPLETION_URL, BALANCE_API_URL, USAGE_API_URL, API_HOST, OPENAI_API_BASE
import os
import queue
import openai
@@ -6,9 +6,10 @@ import openai
class State:
interrupted = False
multi_api_key = False
- completion_url = COMPLETION_URL
+ chat_completion_url = CHAT_COMPLETION_URL
balance_api_url = BALANCE_API_URL
usage_api_url = USAGE_API_URL
+ openai_api_base = OPENAI_API_BASE
def interrupt(self):
self.interrupted = True
@@ -22,13 +23,14 @@ class State:
api_host = f"https://{api_host}"
if api_host.endswith("/v1"):
api_host = api_host[:-3]
- self.completion_url = f"{api_host}/v1/chat/completions"
+ self.chat_completion_url = f"{api_host}/v1/chat/completions"
+ self.openai_api_base = f"{api_host}/v1"
self.balance_api_url = f"{api_host}/dashboard/billing/credit_grants"
self.usage_api_url = f"{api_host}/dashboard/billing/usage"
os.environ["OPENAI_API_BASE"] = api_host
def reset_api_host(self):
- self.completion_url = COMPLETION_URL
+ self.chat_completion_url = CHAT_COMPLETION_URL
self.balance_api_url = BALANCE_API_URL
self.usage_api_url = USAGE_API_URL
os.environ["OPENAI_API_BASE"] = f"https://{API_HOST}"
@@ -36,7 +38,7 @@ class State:
def reset_all(self):
self.interrupted = False
- self.completion_url = COMPLETION_URL
+ self.chat_completion_url = CHAT_COMPLETION_URL
def set_api_key_queue(self, api_key_list):
self.multi_api_key = True
diff --git a/modules/utils.py b/modules/utils.py
index fcc7d4b198a8e796d3ef5016c8eb0226ca4d6f9a..8eaf63edf23a2b19975011fef0451eceaedc2963 100644
--- a/modules/utils.py
+++ b/modules/utils.py
@@ -68,15 +68,15 @@ def delete_last_conversation(current_model, *args):
def set_system_prompt(current_model, *args):
return current_model.set_system_prompt(*args)
-def save_chat_history(current_model, *args):
- return current_model.save_chat_history(*args)
+def rename_chat_history(current_model, *args):
+ return current_model.rename_chat_history(*args)
+
+def auto_name_chat_history(current_model, *args):
+ return current_model.auto_name_chat_history(*args)
def export_markdown(current_model, *args):
return current_model.export_markdown(*args)
-def load_chat_history(current_model, *args):
- return current_model.load_chat_history(*args)
-
def upload_chat_history(current_model, *args):
return current_model.load_chat_history(*args)
@@ -331,55 +331,97 @@ def construct_assistant(text):
def save_file(filename, system, history, chatbot, user_name):
- logging.debug(f"{user_name} 保存对话历史中……")
os.makedirs(os.path.join(HISTORY_DIR, user_name), exist_ok=True)
- if filename.endswith(".json"):
- json_s = {"system": system, "history": history, "chatbot": chatbot}
- if "/" in filename or "\\" in filename:
- history_file_path = filename
- else:
- history_file_path = os.path.join(HISTORY_DIR, user_name, filename)
- with open(history_file_path, "w", encoding='utf-8') as f:
- json.dump(json_s, f, ensure_ascii=False)
- elif filename.endswith(".md"):
- md_s = f"system: \n- {system} \n"
- for data in history:
- md_s += f"\n{data['role']}: \n- {data['content']} \n"
- with open(os.path.join(HISTORY_DIR, user_name, filename), "w", encoding="utf8") as f:
- f.write(md_s)
- logging.debug(f"{user_name} 保存对话历史完毕")
+ if filename is None:
+ filename = new_auto_history_filename(user_name)
+ if filename.endswith(".md"):
+ filename = filename[:-3]
+ if not filename.endswith(".json") and not filename.endswith(".md"):
+ filename += ".json"
+ if filename == ".json":
+ raise Exception("文件名不能为空")
+
+ json_s = {"system": system, "history": history, "chatbot": chatbot}
+ repeat_file_index = 2
+ if not filename == os.path.basename(filename):
+ history_file_path = filename
+ else:
+ history_file_path = os.path.join(HISTORY_DIR, user_name, filename)
+
+ with open(history_file_path, "w", encoding='utf-8') as f:
+ json.dump(json_s, f, ensure_ascii=False)
+
+ filename = os.path.basename(filename)
+ filename_md = filename[:-5] + ".md"
+ md_s = f"system: \n- {system} \n"
+ for data in history:
+ md_s += f"\n{data['role']}: \n- {data['content']} \n"
+ with open(os.path.join(HISTORY_DIR, user_name, filename_md), "w", encoding="utf8") as f:
+ f.write(md_s)
return os.path.join(HISTORY_DIR, user_name, filename)
def sorted_by_pinyin(list):
return sorted(list, key=lambda char: lazy_pinyin(char)[0][0])
+def sorted_by_last_modified_time(list, dir):
+ return sorted(list, key=lambda char: os.path.getctime(os.path.join(dir, char)), reverse=True)
-def get_file_names(dir, plain=False, filetypes=[".json"]):
- logging.debug(f"获取文件名列表，目录为{dir}，文件类型为{filetypes}，是否为纯文本列表{plain}")
+def get_file_names_by_type(dir, filetypes=[".json"]):
+ logging.debug(f"获取文件名列表，目录为{dir}，文件类型为{filetypes}")
files = []
- try:
- for type in filetypes:
- files += [f for f in os.listdir(dir) if f.endswith(type)]
- except FileNotFoundError:
- files = []
- files = sorted_by_pinyin(files)
- if files == []:
- files = [""]
+ for type in filetypes:
+ files += [f for f in os.listdir(dir) if f.endswith(type)]
logging.debug(f"files are:{files}")
- if plain:
- return files
- else:
- return gr.Dropdown.update(choices=files)
+ return files
+
+def get_file_names_by_pinyin(dir, filetypes=[".json"]):
+ files = get_file_names_by_type(dir, filetypes)
+ if files != [""]:
+ files = sorted_by_pinyin(files)
+ logging.debug(f"files are:{files}")
+ return files
+
+def get_file_names_dropdown_by_pinyin(dir, filetypes=[".json"]):
+ files = get_file_names_by_pinyin(dir, filetypes)
+ return gr.Dropdown.update(choices=files)
+
+def get_file_names_by_last_modified_time(dir, filetypes=[".json"]):
+ files = get_file_names_by_type(dir, filetypes)
+ if files != [""]:
+ files = sorted_by_last_modified_time(files, dir)
+ logging.debug(f"files are:{files}")
+ return files
-def get_history_names(plain=False, user_name=""):
+def get_history_names(user_name=""):
logging.debug(f"从用户 {user_name} 中获取历史记录文件名列表")
if user_name == "" and hide_history_when_not_logged_in:
- return ""
+ return []
else:
- return get_file_names(os.path.join(HISTORY_DIR, user_name), plain)
+ history_files = get_file_names_by_last_modified_time(os.path.join(HISTORY_DIR, user_name))
+ history_files = [f[:f.rfind(".")] for f in history_files]
+ return history_files
+
+def get_first_history_name(user_name=""):
+ history_names = get_history_names(user_name)
+ return history_names[0] if history_names else None
+def get_history_list(user_name=""):
+ history_names = get_history_names(user_name)
+ return gr.Radio.update(choices=history_names)
+
+def init_history_list(user_name=""):
+ history_names = get_history_names(user_name)
+ return gr.Radio.update(choices=history_names, value=history_names[0] if history_names else "")
+
+def filter_history(user_name, keyword):
+ history_names = get_history_names(user_name)
+ try:
+ history_names = [name for name in history_names if re.search(keyword, name)]
+ return gr.update(choices=history_names)
+ except:
+ return gr.update(choices=history_names)
def load_template(filename, mode=0):
logging.debug(f"加载模板文件{filename}，模式为{mode}（0为返回字典和下拉菜单，1为返回下拉菜单，2为返回字典）")
@@ -406,9 +448,14 @@ def load_template(filename, mode=0):
)
-def get_template_names(plain=False):
+def get_template_names():
logging.debug("获取模板文件名列表")
- return get_file_names(TEMPLATES_DIR, plain, filetypes=[".csv", "json"])
+ return get_file_names_by_pinyin(TEMPLATES_DIR, filetypes=[".csv", "json"])
+
+def get_template_dropdown():
+ logging.debug("获取模板下拉菜单")
+ template_names = get_template_names()
+ return gr.Dropdown.update(choices=template_names)
def get_template_content(templates, selection, original_system_prompt):
@@ -614,36 +661,21 @@ def toggle_like_btn_visibility(selected_model_name):
else:
return gr.update(visible=False)
-def new_auto_history_filename(dirname):
- latest_file = get_latest_filepath(dirname)
+def new_auto_history_filename(username):
+ latest_file = get_first_history_name(username)
if latest_file:
- with open(os.path.join(dirname, latest_file), 'r', encoding="utf-8") as f:
+ with open(os.path.join(HISTORY_DIR, username, latest_file + ".json"), 'r', encoding="utf-8") as f:
if len(f.read()) == 0:
return latest_file
- now = datetime.datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
+ now = i18n("新对话 ") + datetime.datetime.now().strftime('%m-%d %H-%M')
return f'{now}.json'
-def get_latest_filepath(dirname):
- pattern = re.compile(r'\d{4}-\d{2}-\d{2}_\d{2}-\d{2}-\d{2}')
- latest_time = None
- latest_file = None
- for filename in os.listdir(dirname):
- if os.path.isfile(os.path.join(dirname, filename)):
- match = pattern.search(filename)
- if match and match.group(0) == filename[:19]:
- time_str = filename[:19]
- filetime = datetime.datetime.strptime(time_str, '%Y-%m-%d_%H-%M-%S')
- if not latest_time or filetime > latest_time:
- latest_time = filetime
- latest_file = filename
- return latest_file
-
def get_history_filepath(username):
dirname = os.path.join(HISTORY_DIR, username)
os.makedirs(dirname, exist_ok=True)
- latest_file = get_latest_filepath(dirname)
+ latest_file = get_first_history_name(username)
if not latest_file:
- latest_file = new_auto_history_filename(dirname)
+ latest_file = new_auto_history_filename(username)
latest_file = os.path.join(dirname, latest_file)
return latest_file
@@ -651,7 +683,7 @@ def get_history_filepath(username):
def beautify_err_msg(err_msg):
if "insufficient_quota" in err_msg:
return i18n("剩余配额不足，[进一步了解](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/%E5%B8%B8%E8%A7%81%E9%97%AE%E9%A2%98#you-exceeded-your-current-quota-please-check-your-plan-and-billing-details)")
- if "The model: gpt-4 does not exist" in err_msg:
+ if "The model `gpt-4` does not exist" in err_msg:
return i18n("你没有权限访问 GPT4，[进一步了解](https://github.com/GaiZhenbiao/ChuanhuChatGPT/issues/843)")
if "Resource not found" in err_msg:
return i18n("请查看 config_example.json，配置 Azure OpenAI")
@@ -681,3 +713,14 @@ def get_file_hash(file_src=None, file_paths=None):
md5_hash.update(chunk)
return md5_hash.hexdigest()
+
+def myprint(**args):
+ print(args)
+
+def replace_special_symbols(string, replace_string=" "):
+ # 定义正则表达式，匹配所有特殊符号
+ pattern = r'[!@#$%^&*()<>?/\|}{~:]'
+
+ new_string = re.sub(pattern, replace_string, string)
+
+ return new_string
diff --git a/modules/webui.py b/modules/webui.py
index 61f863d7ca3b8975222b90d4f66a2c6cdc9d2e0d..a9308d394ca706b1fe6cb4d9e86f3a38b0a23a6f 100644
--- a/modules/webui.py
+++ b/modules/webui.py
@@ -56,11 +56,24 @@ def reload_javascript():
js += ''
js += ' 
')
+ gr.Markdown("# "+i18n("川虎Chat"))
+ gr.HTML(get_html("footer.html").format(
+ versions=versions_html()), elem_id="footer")
+ gr.Markdown(CHUANHU_DESCRIPTION, elem_id="description")
+
+ with gr.Box(elem_id="chuanhu-training"):
+ with gr.Row():
+ gr.Markdown("## "+i18n("训练"))
+ gr.HTML(get_html("close_btn.html").format(
+ obj="box"), elem_classes="close-btn")
+ with gr.Tabs(elem_id="chuanhu-training-tabs"):
+ with gr.Tab(label="OpenAI "+i18n("微调")):
+ openai_train_status = gr.Markdown(label=i18n("训练状态"), value=i18n(
+ "查看[使用介绍](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#微调-gpt-35)"))
+
+ with gr.Tab(label=i18n("准备数据集")):
+ dataset_preview_json = gr.JSON(
+ label=i18n("数据集预览"))
+ dataset_selection = gr.Files(label=i18n("选择数据集"), file_types=[
+ ".xlsx", ".jsonl"], file_count="single")
+ upload_to_openai_btn = gr.Button(
+ i18n("上传到OpenAI"), variant="primary", interactive=False)
+
+ with gr.Tab(label=i18n("训练")):
+ openai_ft_file_id = gr.Textbox(label=i18n(
+ "文件ID"), value="", lines=1, placeholder=i18n("上传到 OpenAI 后自动填充"))
+ openai_ft_suffix = gr.Textbox(label=i18n(
+ "模型名称后缀"), value="", lines=1, placeholder=i18n("可选，用于区分不同的模型"))
+ openai_train_epoch_slider = gr.Slider(label=i18n(
+ "训练轮数（Epochs）"), minimum=1, maximum=100, value=3, step=1, interactive=True)
+ openai_start_train_btn = gr.Button(
+ i18n("开始训练"), variant="primary", interactive=False)
+
+ with gr.Tab(label=i18n("状态")):
+ openai_status_refresh_btn = gr.Button(i18n("刷新状态"))
+ openai_cancel_all_jobs_btn = gr.Button(
+ i18n("取消所有任务"))
+ add_to_models_btn = gr.Button(
+ i18n("添加训练好的模型到模型列表"), interactive=False)
+
+ with gr.Box(elem_id="web-config", visible=False):
+ gr.HTML(get_html('web_config.html').format(
+ enableCheckUpdate_config=check_update,
+ hideHistoryWhenNotLoggedIn_config=hide_history_when_not_logged_in,
+ forView_i18n=i18n("仅供查看"),
+ deleteConfirm_i18n_pref=i18n("你真的要删除 "),
+ deleteConfirm_i18n_suff=i18n(" 吗？"),
+ usingLatest_i18n=i18n("您使用的就是最新版！"),
+ updatingMsg_i18n=i18n("正在尝试更新..."),
+ updateSuccess_i18n=i18n("更新成功，请重启本程序"),
+ updateFailure_i18n=i18n(
+ "更新失败，请尝试[手动更新](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#手动更新)"),
+ regenerate_i18n=i18n("重新生成"),
+ deleteRound_i18n=i18n("删除这轮问答"),
+ renameChat_i18n=i18n("重命名该对话"),
+ validFileName_i18n=i18n("请输入有效的文件名，不要包含以下特殊字符："),
+ ))
+ with gr.Box(elem_id="fake-gradio-components", visible=False):
+ updateChuanhuBtn = gr.Button(
+ visible=False, elem_classes="invisible-btn", elem_id="update-chuanhu-btn")
+ changeSingleSessionBtn = gr.Button(
+ visible=False, elem_classes="invisible-btn", elem_id="change-single-session-btn")
+ changeOnlineSearchBtn = gr.Button(
+ visible=False, elem_classes="invisible-btn", elem_id="change-online-search-btn")
+ historySelectBtn = gr.Button(
+ visible=False, elem_classes="invisible-btn", elem_id="history-select-btn") # Not used
# https://github.com/gradio-app/gradio/pull/3296
+
def create_greeting(request: gr.Request):
- if hasattr(request, "username") and request.username: # is not None or is not ""
+ if hasattr(request, "username") and request.username: # is not None or is not ""
logging.info(f"Get User Name: {request.username}")
- user_info, user_name = gr.Markdown.update(value=f"User: {request.username}"), request.username
+ user_info, user_name = gr.Markdown.update(
+ value=f"User: {request.username}"), request.username
else:
- user_info, user_name = gr.Markdown.update(value=f"", visible=False), ""
- current_model = get_model(model_name = MODELS[DEFAULT_MODEL], access_key = my_api_key)[0]
+ user_info, user_name = gr.Markdown.update(
+ value=f"", visible=False), ""
+ current_model = get_model(
+ model_name=MODELS[DEFAULT_MODEL], access_key=my_api_key)[0]
current_model.set_user_identifier(user_name)
- chatbot = gr.Chatbot.update(label=MODELS[DEFAULT_MODEL])
- return user_info, user_name, current_model, toggle_like_btn_visibility(DEFAULT_MODEL), *current_model.auto_load(), get_history_names(False, user_name), chatbot
- demo.load(create_greeting, inputs=None, outputs=[user_info, user_name, current_model, like_dislike_area, systemPromptTxt, chatbot, historyFileSelectDropdown, chatbot], api_name="load")
+ if not hide_history_when_not_logged_in or user_name:
+ filename, system_prompt, chatbot = current_model.auto_load()
+ else:
+ system_prompt = gr.update()
+ filename = gr.update()
+ chatbot = gr.Chatbot.update(label=MODELS[DEFAULT_MODEL])
+ return user_info, user_name, current_model, toggle_like_btn_visibility(DEFAULT_MODEL), filename, system_prompt, chatbot, init_history_list(user_name)
+ demo.load(create_greeting, inputs=None, outputs=[
+ user_info, user_name, current_model, like_dislike_area, saveFileName, systemPromptTxt, chatbot, historySelectList], api_name="load")
chatgpt_predict_args = dict(
fn=predict,
inputs=[
@@ -369,42 +537,58 @@ with gr.Blocks(theme=small_and_beautiful_theme) as demo:
)
transfer_input_args = dict(
- fn=transfer_input, inputs=[user_input], outputs=[user_question, user_input, submitBtn, cancelBtn], show_progress=True
+ fn=transfer_input, inputs=[user_input], outputs=[
+ user_question, user_input, submitBtn, cancelBtn], show_progress=True
)
get_usage_args = dict(
- fn=billing_info, inputs=[current_model], outputs=[usageTxt], show_progress=False
+ fn=billing_info, inputs=[current_model], outputs=[
+ usageTxt], show_progress=False
)
load_history_from_file_args = dict(
fn=load_chat_history,
- inputs=[current_model, historyFileSelectDropdown, user_name],
+ inputs=[current_model, historySelectList, user_name],
outputs=[saveFileName, systemPromptTxt, chatbot]
)
refresh_history_args = dict(
- fn=get_history_names, inputs=[gr.State(False), user_name], outputs=[historyFileSelectDropdown]
+ fn=get_history_list, inputs=[user_name], outputs=[historySelectList]
)
+ auto_name_chat_history_args = dict(
+ fn=auto_name_chat_history,
+ inputs=[current_model, name_chat_method, user_question, chatbot, user_name, single_turn_checkbox],
+ outputs=[historySelectList],
+ show_progress=False,
+ )
# Chatbot
cancelBtn.click(interrupt, [current_model], [])
- user_input.submit(**transfer_input_args).then(**chatgpt_predict_args).then(**end_outputing_args)
+ user_input.submit(**transfer_input_args).then(**
+ chatgpt_predict_args).then(**end_outputing_args).then(**auto_name_chat_history_args)
user_input.submit(**get_usage_args)
- submitBtn.click(**transfer_input_args).then(**chatgpt_predict_args, api_name="predict").then(**end_outputing_args)
+ # user_input.submit(auto_name_chat_history, [current_model, user_question, chatbot, user_name], [historySelectList], show_progress=False)
+
+ submitBtn.click(**transfer_input_args).then(**chatgpt_predict_args,
+ api_name="predict").then(**end_outputing_args).then(**auto_name_chat_history_args)
submitBtn.click(**get_usage_args)
- index_files.change(handle_file_upload, [current_model, index_files, chatbot, language_select_dropdown], [index_files, chatbot, status_display])
- summarize_btn.click(handle_summarize_index, [current_model, index_files, chatbot, language_select_dropdown], [chatbot, status_display])
+ # submitBtn.click(auto_name_chat_history, [current_model, user_question, chatbot, user_name], [historySelectList], show_progress=False)
+
+ index_files.upload(handle_file_upload, [current_model, index_files, chatbot, language_select_dropdown], [
+ index_files, chatbot, status_display])
+ summarize_btn.click(handle_summarize_index, [
+ current_model, index_files, chatbot, language_select_dropdown], [chatbot, status_display])
emptyBtn.click(
reset,
- inputs=[current_model],
- outputs=[chatbot, status_display],
+ inputs=[current_model, retain_system_prompt_checkbox],
+ outputs=[chatbot, status_display, historySelectList, systemPromptTxt],
show_progress=True,
- _js='clearChatbot',
+ _js='(a,b)=>{return clearChatbot(a,b);}',
)
retryBtn.click(**start_outputing_args).then(
@@ -452,17 +636,24 @@ with gr.Blocks(theme=small_and_beautiful_theme) as demo:
two_column.change(update_doc_config, [two_column], None)
# LLM Models
- keyTxt.change(set_key, [current_model, keyTxt], [user_api_key, status_display], api_name="set_key").then(**get_usage_args)
+ keyTxt.change(set_key, [current_model, keyTxt], [
+ user_api_key, status_display], api_name="set_key").then(**get_usage_args)
keyTxt.submit(**get_usage_args)
- single_turn_checkbox.change(set_single_turn, [current_model, single_turn_checkbox], None)
- model_select_dropdown.change(get_model, [model_select_dropdown, lora_select_dropdown, user_api_key, temperature_slider, top_p_slider, systemPromptTxt, user_name], [current_model, status_display, chatbot, lora_select_dropdown, user_api_key, keyTxt], show_progress=True, api_name="get_model")
- model_select_dropdown.change(toggle_like_btn_visibility, [model_select_dropdown], [like_dislike_area], show_progress=False)
- lora_select_dropdown.change(get_model, [model_select_dropdown, lora_select_dropdown, user_api_key, temperature_slider, top_p_slider, systemPromptTxt, user_name], [current_model, status_display, chatbot], show_progress=True)
+ single_turn_checkbox.change(
+ set_single_turn, [current_model, single_turn_checkbox], None)
+ model_select_dropdown.change(get_model, [model_select_dropdown, lora_select_dropdown, user_api_key, temperature_slider, top_p_slider, systemPromptTxt, user_name, current_model], [
+ current_model, status_display, chatbot, lora_select_dropdown, user_api_key, keyTxt], show_progress=True, api_name="get_model")
+ model_select_dropdown.change(toggle_like_btn_visibility, [model_select_dropdown], [
+ like_dislike_area], show_progress=False)
+ lora_select_dropdown.change(get_model, [model_select_dropdown, lora_select_dropdown, user_api_key, temperature_slider,
+ top_p_slider, systemPromptTxt, user_name, current_model], [current_model, status_display, chatbot], show_progress=True)
# Template
- systemPromptTxt.change(set_system_prompt, [current_model, systemPromptTxt], None)
- templateRefreshBtn.click(get_template_names, None, [templateFileSelectDropdown])
- templateFileSelectDropdown.change(
+ systemPromptTxt.change(set_system_prompt, [
+ current_model, systemPromptTxt], None)
+ templateRefreshBtn.click(get_template_dropdown, None, [
+ templateFileSelectDropdown])
+ templateFileSelectDropdown.input(
load_template,
[templateFileSelectDropdown],
[promptTemplates, templateSelectDropdown],
@@ -476,47 +667,80 @@ with gr.Blocks(theme=small_and_beautiful_theme) as demo:
)
# S&L
- saveHistoryBtn.click(
- save_chat_history,
+ renameHistoryBtn.click(
+ rename_chat_history,
[current_model, saveFileName, chatbot, user_name],
- downloadFile,
+ [historySelectList],
show_progress=True,
+ _js='(a,b,c,d)=>{return saveChatHistory(a,b,c,d);}'
)
- saveHistoryBtn.click(get_history_names, [gr.State(False), user_name], [historyFileSelectDropdown])
exportMarkdownBtn.click(
export_markdown,
[current_model, saveFileName, chatbot, user_name],
- downloadFile,
+ [],
show_progress=True,
)
historyRefreshBtn.click(**refresh_history_args)
- historyDeleteBtn.click(delete_chat_history, [current_model, historyFileSelectDropdown, user_name], [status_display, historyFileSelectDropdown, chatbot], _js='(a,b,c)=>{return showConfirmationDialog(a, b, c);}')
- historyFileSelectDropdown.change(**load_history_from_file_args)
- downloadFile.change(upload_chat_history, [current_model, downloadFile, user_name], [saveFileName, systemPromptTxt, chatbot])
+ historyDeleteBtn.click(delete_chat_history, [current_model, historySelectList, user_name], [status_display, historySelectList, chatbot], _js='(a,b,c)=>{return showConfirmationDialog(a, b, c);}').then(
+ reset,
+ inputs=[current_model, retain_system_prompt_checkbox],
+ outputs=[chatbot, status_display, historySelectList, systemPromptTxt],
+ show_progress=True,
+ _js='(a,b)=>{return clearChatbot(a,b);}',
+ )
+ historySelectList.input(**load_history_from_file_args)
+ uploadFileBtn.upload(upload_chat_history, [current_model, uploadFileBtn, user_name], [
+ saveFileName, systemPromptTxt, chatbot]).then(**refresh_history_args)
+ historyDownloadBtn.click(None, [
+ user_name, historySelectList], None, _js='(a,b)=>{return downloadHistory(a,b,".json");}')
+ historyMarkdownDownloadBtn.click(None, [
+ user_name, historySelectList], None, _js='(a,b)=>{return downloadHistory(a,b,".md");}')
+ historySearchTextbox.input(
+ filter_history,
+ [user_name, historySearchTextbox],
+ [historySelectList]
+ )
# Train
- dataset_selection.upload(handle_dataset_selection, dataset_selection, [dataset_preview_json, upload_to_openai_btn, openai_train_status])
- dataset_selection.clear(handle_dataset_clear, [], [dataset_preview_json, upload_to_openai_btn])
- upload_to_openai_btn.click(upload_to_openai, [dataset_selection], [openai_ft_file_id, openai_train_status], show_progress=True)
-
- openai_ft_file_id.change(lambda x: gr.update(interactive=True) if len(x) > 0 else gr.update(interactive=False), [openai_ft_file_id], [openai_start_train_btn])
- openai_start_train_btn.click(start_training, [openai_ft_file_id, openai_ft_suffix, openai_train_epoch_slider], [openai_train_status])
-
- openai_status_refresh_btn.click(get_training_status, [], [openai_train_status, add_to_models_btn])
- add_to_models_btn.click(add_to_models, [], [model_select_dropdown, openai_train_status], show_progress=True)
- openai_cancel_all_jobs_btn.click(cancel_all_jobs, [], [openai_train_status], show_progress=True)
+ dataset_selection.upload(handle_dataset_selection, dataset_selection, [
+ dataset_preview_json, upload_to_openai_btn, openai_train_status])
+ dataset_selection.clear(handle_dataset_clear, [], [
+ dataset_preview_json, upload_to_openai_btn])
+ upload_to_openai_btn.click(upload_to_openai, [dataset_selection], [
+ openai_ft_file_id, openai_train_status], show_progress=True)
+
+ openai_ft_file_id.change(lambda x: gr.update(interactive=True) if len(
+ x) > 0 else gr.update(interactive=False), [openai_ft_file_id], [openai_start_train_btn])
+ openai_start_train_btn.click(start_training, [
+ openai_ft_file_id, openai_ft_suffix, openai_train_epoch_slider], [openai_train_status])
+
+ openai_status_refresh_btn.click(get_training_status, [], [
+ openai_train_status, add_to_models_btn])
+ add_to_models_btn.click(add_to_models, [], [
+ model_select_dropdown, openai_train_status], show_progress=True)
+ openai_cancel_all_jobs_btn.click(
+ cancel_all_jobs, [], [openai_train_status], show_progress=True)
# Advanced
- max_context_length_slider.change(set_token_upper_limit, [current_model, max_context_length_slider], None)
- temperature_slider.change(set_temperature, [current_model, temperature_slider], None)
+ max_context_length_slider.change(
+ set_token_upper_limit, [current_model, max_context_length_slider], None)
+ temperature_slider.change(
+ set_temperature, [current_model, temperature_slider], None)
top_p_slider.change(set_top_p, [current_model, top_p_slider], None)
- n_choices_slider.change(set_n_choices, [current_model, n_choices_slider], None)
- stop_sequence_txt.change(set_stop_sequence, [current_model, stop_sequence_txt], None)
- max_generation_slider.change(set_max_tokens, [current_model, max_generation_slider], None)
- presence_penalty_slider.change(set_presence_penalty, [current_model, presence_penalty_slider], None)
- frequency_penalty_slider.change(set_frequency_penalty, [current_model, frequency_penalty_slider], None)
- logit_bias_txt.change(set_logit_bias, [current_model, logit_bias_txt], None)
- user_identifier_txt.change(set_user_identifier, [current_model, user_identifier_txt], None)
+ n_choices_slider.change(
+ set_n_choices, [current_model, n_choices_slider], None)
+ stop_sequence_txt.change(
+ set_stop_sequence, [current_model, stop_sequence_txt], None)
+ max_generation_slider.change(
+ set_max_tokens, [current_model, max_generation_slider], None)
+ presence_penalty_slider.change(
+ set_presence_penalty, [current_model, presence_penalty_slider], None)
+ frequency_penalty_slider.change(
+ set_frequency_penalty, [current_model, frequency_penalty_slider], None)
+ logit_bias_txt.change(
+ set_logit_bias, [current_model, logit_bias_txt], None)
+ user_identifier_txt.change(set_user_identifier, [
+ current_model, user_identifier_txt], None)
default_btn.click(
reset_default, [], [apihostTxt, proxyTxt, status_display], show_progress=True
@@ -533,7 +757,7 @@ with gr.Blocks(theme=small_and_beautiful_theme) as demo:
# [status_display],
# show_progress=True,
# )
- checkUpdateBtn.click(fn=None, _js='manualCheckUpdate')
+ # checkUpdateBtn.click(fn=None, _js='manualCheckUpdate')
# Invisible elements
updateChuanhuBtn.click(
@@ -542,6 +766,25 @@ with gr.Blocks(theme=small_and_beautiful_theme) as demo:
[status_display],
show_progress=True,
)
+ changeSingleSessionBtn.click(
+ fn=lambda value: gr.Checkbox.update(value=value),
+ inputs=[single_turn_checkbox],
+ outputs=[single_turn_checkbox],
+ _js='(a)=>{return bgChangeSingleSession(a);}'
+ )
+ changeOnlineSearchBtn.click(
+ fn=lambda value: gr.Checkbox.update(value=value),
+ inputs=[use_websearch_checkbox],
+ outputs=[use_websearch_checkbox],
+ _js='(a)=>{return bgChangeOnlineSearch(a);}'
+ )
+ historySelectBtn.click( # This is an experimental feature... Not actually used.
+ fn=load_chat_history,
+ inputs=[current_model, historySelectList],
+ outputs=[saveFileName, systemPromptTxt, chatbot],
+ _js='(a,b)=>{return bgSelectHistory(a,b);}'
+ )
+
logging.info(
colorama.Back.GREEN
@@ -554,6 +797,8 @@ demo.title = i18n("川虎Chat 🚀")
if __name__ == "__main__":
reload_javascript()
demo.queue(concurrency_count=CONCURRENT_COUNT).launch(
- blocked_paths=["config.json"],
+ allowed_paths=["history", "web_assets"],
+ auth=auth_from_conf if authflag else None,
favicon_path="./web_assets/favicon.ico",
+ inbrowser=not dockerflag, # 禁止在docker下开启inbrowser
)
diff --git a/README.md b/README.md
index 820a9a57349cfbf6d565c797ed822c398347e682..f22239cee71784ded3c9c459db9b57bda9a92ff3 100644
--- a/README.md
+++ b/README.md
@@ -1,13 +1,15 @@
+```
---
title: ChuanhuChatGPT
emoji: 🐯
colorFrom: yellow
colorTo: yellow
sdk: gradio
-sdk_version: 3.40.0
+sdk_version: 3.43.2
app_file: ChuanhuChatbot.py
pinned: false
license: gpl-3.0
---
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
\ No newline at end of file
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+```
diff --git a/config_example.json b/config_example.json
index 0b77caefbb39ef08d6a53b3b40ee67bb8a3b1576..8558c0229b9ba3ef7194a318bbbd8cfc2fe46cde 100644
--- a/config_example.json
+++ b/config_example.json
@@ -11,6 +11,10 @@
"midjourney_proxy_api_secret": "", // 你的 MidJourney Proxy API Secret，用于鉴权访问 api，可选
"midjourney_discord_proxy_url": "", // 你的 MidJourney Discord Proxy URL，用于对生成对图进行反代，可选
"midjourney_temp_folder": "./tmp", // 你的 MidJourney 临时文件夹，用于存放生成的图片，填空则关闭自动下载切图（直接显示MJ的四宫格图）
+ "spark_appid": "", // 你的 讯飞星火大模型 API AppID，用于讯飞星火大模型对话模型
+ "spark_api_key": "", // 你的 讯飞星火大模型 API Key，用于讯飞星火大模型对话模型
+ "spark_api_secret": "", // 你的 讯飞星火大模型 API Secret，用于讯飞星火大模型对话模型
+ "claude_api_secret":"",// 你的 Claude API Secret，用于 Claude 对话模型
//== Azure ==
@@ -23,14 +27,15 @@
"azure_embedding_model_name": "text-embedding-ada-002", // 你的 Azure OpenAI Embedding 模型名称
//== 基础配置 ==
- "language": "auto", // 界面语言，可选"auto", "zh-CN", "en-US", "ja-JP", "ko-KR", "sv-SE"
+ "language": "auto", // 界面语言，可选"auto", "zh_CN", "en_US", "ja_JP", "ko_KR", "sv_SE", "ru_RU", "vi_VN"
"users": [], // 用户列表，[[用户名1, 密码1], [用户名2, 密码2], ...]
"local_embedding": false, //是否在本地编制索引
"hide_history_when_not_logged_in": false, //未登录情况下是否不展示对话历史
"check_update": true, //是否启用检查更新
"default_model": "gpt-3.5-turbo", // 默认模型
- "bot_avatar": "default", // 机器人头像，可填写图片链接、Data URL (base64)，或者"none"（不显示头像）
- "user_avatar": "default", // 用户头像，可填写图片链接、Data URL (base64)，或者"none"（不显示头像）
+ "chat_name_method_index": 2, // 选择对话名称的方法。0: 使用日期时间命名；1: 使用第一条提问命名，2: 使用模型自动总结
+ "bot_avatar": "default", // 机器人头像，可填写本地或网络图片链接，或者"none"（不显示头像）
+ "user_avatar": "default", // 用户头像，可填写本地或网络图片链接，或者"none"（不显示头像）
//== API 用量 ==
"show_api_billing": false, //是否显示OpenAI API用量（启用需要填写sensitive_id）
@@ -57,11 +62,13 @@
//== 高级配置 ==
// 是否多个API Key轮换使用
"multi_api_key": false,
- "api_key_list": [
- "sk-xxxxxxxxxxxxxxxxxxxxxxxx1",
- "sk-xxxxxxxxxxxxxxxxxxxxxxxx2",
- "sk-xxxxxxxxxxxxxxxxxxxxxxxx3"
- ],
+ // "available_models": ["GPT3.5 Turbo", "GPT4 Turbo", "GPT4 Vision"], // 可用的模型列表，将覆盖默认的可用模型列表
+ // "extra_models": ["模型名称3", "模型名称4", ...], // 额外的模型，将添加到可用的模型列表之后
+ // "api_key_list": [
+ // "sk-xxxxxxxxxxxxxxxxxxxxxxxx1",
+ // "sk-xxxxxxxxxxxxxxxxxxxxxxxx2",
+ // "sk-xxxxxxxxxxxxxxxxxxxxxxxx3"
+ // ],
// 自定义OpenAI API Base
// "openai_api_base": "https://api.openai.com",
// 自定义使用代理（请替换代理URL）
diff --git a/locale/en_US.json b/locale/en_US.json
index 17a5aa618ee8e1c4425a7ce69e1d86adfbd24b6c..9ceaa244f8d2b468ca02520bc1cbefe477a9fdaa 100644
--- a/locale/en_US.json
+++ b/locale/en_US.json
@@ -1,87 +1,141 @@
{
- "未命名对话历史记录": "Unnamed Dialog History",
- "在这里输入": "Type in here",
- "🧹 新的对话": "🧹 New Dialogue",
- "🔄 重新生成": "🔄 Regeneration",
- "🗑️ 删除最旧对话": "🗑️ Delete oldest dialog",
- "🗑️ 删除最新对话": "🗑️ Delete latest dialog",
- "模型": "Model",
- "多账号模式已开启，无需输入key，可直接开始对话": "Multi-account mode is enabled, no need to enter key, you can start the dialogue directly",
+ " 吗？": " ?",
+ "# ⚠️ 务必谨慎更改 ⚠️": "# ⚠️ Caution: Changes require care. ⚠️",
"**发送消息** 或 **提交key** 以显示额度": "**Send message** or **Submit key** to display credit",
- "选择模型": "Select Model",
- "选择LoRA模型": "Select LoRA Model",
- "实时传输回答": "Stream output",
- "单轮对话": "Single-turn dialogue",
- "使用在线搜索": "Use online search",
- "选择回复语言（针对搜索&索引功能）": "Select reply language (for search & index)",
- "上传索引文件": "Upload",
- "双栏pdf": "Two-column pdf",
- "识别公式": "formula OCR",
- "在这里输入System Prompt...": "Type in System Prompt here...",
- "加载Prompt模板": "Load Prompt Template",
- "选择Prompt模板集合文件": "Select Prompt Template Collection File",
- "🔄 刷新": "🔄 Refresh",
+ "**本月使用金额** ": "**Monthly usage** ",
+ "**获取API使用情况失败**": "**Failed to get API usage**",
+ "**获取API使用情况失败**，sensitive_id错误或已过期": "**Failed to get API usage**, wrong or expired sensitive_id",
+ "**获取API使用情况失败**，需在填写`config.json`中正确填写sensitive_id": "**Failed to get API usage**, correct sensitive_id needed in `config.json`",
+ "API key为空，请检查是否输入正确。": "API key is empty, check whether it is entered correctly.",
+ "API密钥更改为了": "The API key is changed to",
+ "JSON解析错误,收到的内容: ": "JSON parsing error, received content: ",
+ "SSL错误，无法获取对话。": "SSL error, unable to get dialogue.",
+ "Token 计数: ": "Token Count: ",
+ "☹️发生了错误：": "☹️Error: ",
+ "⚠️ 为保证API-Key安全，请在配置文件`config.json`中修改网络设置": "⚠️ To ensure the security of API-Key, please modify the network settings in the configuration file `config.json`.",
+ "。你仍然可以使用聊天功能。": ". You can still use the chat function.",
+ "上传": "Upload",
+ "上传了": "Uploaded",
+ "上传到 OpenAI 后自动填充": "Automatically filled after uploading to OpenAI",
+ "上传到OpenAI": "Upload to OpenAI",
+ "上传文件": "Upload files",
+ "仅供查看": "For viewing only",
"从Prompt模板中加载": "Load from Prompt Template",
- "保存/加载": "Save/Load",
- "保存/加载对话历史记录": "Save/Load Dialog History",
"从列表中加载对话": "Load dialog from list",
- "设置文件名: 默认为.json，可选为.md": "Set file name: default is .json, optional is .md",
- "设置保存文件名": "Set save file name",
- "对话历史记录": "Dialog History",
- "💾 保存对话": "💾 Save Dialog",
- "📝 导出为Markdown": "📝 Export as Markdown",
- "默认保存于history文件夹": "Default save in history folder",
- "高级": "Advanced",
- "# ⚠️ 务必谨慎更改 ⚠️": "# ⚠️ Caution: Changes require care. ⚠️",
- "参数": "Parameters",
- "停止符，用英文逗号隔开...": "Type in stop token here, separated by comma...",
- "用于定位滥用行为": "Used to locate abuse",
- "用户名": "Username",
- "在这里输入API-Host...": "Type in API-Host here...",
- "🔄 切换API地址": "🔄 Switch API Address",
- "未设置代理...": "No proxy...",
"代理地址": "Proxy address",
- "🔄 设置代理地址": "🔄 Set Proxy Address",
- "🔙 恢复默认网络设置": "🔙 Reset Network Settings",
- "🔄 检查更新...": "🔄 Check for Update...",
+ "代理错误，无法获取对话。": "Proxy error, unable to get dialogue.",
+ "你没有权限访问 GPT4，[进一步了解](https://github.com/GaiZhenbiao/ChuanhuChatGPT/issues/843)": "You do not have permission to access GPT-4, [learn more](https://github.com/GaiZhenbiao/ChuanhuChatGPT/issues/843)",
+ "你没有选择任何对话历史": "You have not selected any conversation history.",
+ "你真的要删除 ": "Are you sure you want to delete ",
+ "使用在线搜索": "Use online search",
+ "停止符，用英文逗号隔开...": "Type in stop token here, separated by comma...",
+ "关于": "About",
+ "准备数据集": "Prepare Dataset",
+ "切换亮暗色主题": "Switch light/dark theme",
+ "删除对话历史成功": "Successfully deleted conversation history.",
+ "删除这轮问答": "Delete this round of Q&A",
+ "刷新状态": "Refresh Status",
+ "剩余配额不足，[进一步了解](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/%E5%B8%B8%E8%A7%81%E9%97%AE%E9%A2%98#you-exceeded-your-current-quota-please-check-your-plan-and-billing-details)": "Insufficient remaining quota, [learn more](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/%E5%B8%B8%E8%A7%81%E9%97%AE%E9%A2%98#you-exceeded-your-current-quota-please-check-your-plan-and-billing-details)",
+ "加载Prompt模板": "Load Prompt Template",
+ "单轮对话": "Single-turn",
+ "历史记录（JSON）": "History file (JSON)",
+ "参数": "Parameters",
+ "双栏pdf": "Two-column pdf",
"取消": "Cancel",
- "更新": "Update",
- "详情": "Details",
+ "取消所有任务": "Cancel All Tasks",
+ "可选，用于区分不同的模型": "Optional, used to distinguish different models",
+ "启用的工具：": "Enabled tools: ",
+ "在工具箱中管理知识库文件": "Manage knowledge base files in the toolbox",
+ "在线搜索": "Web search",
+ "在这里输入": "Type in here",
+ "在这里输入System Prompt...": "Type in System Prompt here...",
+ "多账号模式已开启，无需输入key，可直接开始对话": "Multi-account mode is enabled, no need to enter key, you can start the dialogue directly",
"好": "OK",
- "更新成功，请重启本程序": "Updated successfully, please restart this program",
- "更新失败，请尝试[手动更新](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#手动更新)": "Update failed, please try [manually updating](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#手动更新)",
+ "实时传输回答": "Stream output",
+ "对话": "Dialogue",
+ "对话历史": "Conversation history",
+ "对话历史记录": "Dialog History",
+ "对话命名方式": "History naming method",
+ "导出为 Markdown": "Export as Markdown",
+ "川虎Chat": "Chuanhu Chat",
"川虎Chat 🚀": "Chuanhu Chat 🚀",
+ "工具箱": "Toolbox",
+ "已经被删除啦": "It has been deleted.",
"开始实时传输回答……": "Start streaming output...",
- "Token 计数: ": "Token Count: ",
- "，本次对话累计消耗了 ": ", Total cost for this dialogue is ",
- "**获取API使用情况失败**": "**Failed to get API usage**",
- "**获取API使用情况失败**，需在填写`config.json`中正确填写sensitive_id": "**Failed to get API usage**, correct sensitive_id needed in `config.json`",
- "**获取API使用情况失败**，sensitive_id错误或已过期": "**Failed to get API usage**, wrong or expired sensitive_id",
- "**本月使用金额** ": "**Monthly usage** ",
+ "开始训练": "Start Training",
+ "微调": "Fine-tuning",
+ "总结": "Summarize",
+ "总结完成": "Summary completed.",
+ "您使用的就是最新版！": "You are using the latest version!",
+ "您的IP区域：": "Your IP region: ",
+ "您的IP区域：未知。": "Your IP region: Unknown.",
+ "拓展": "Extensions",
+ "搜索（支持正则）...": "Search (supports regex)...",
+ "数据集预览": "Dataset Preview",
+ "文件ID": "File ID",
+ "新对话 ": "New Chat ",
+ "新建对话保留Prompt": "Retain Prompt For New Chat",
+ "暂时未知": "Unknown",
+ "更新": "Update",
+ "更新失败，请尝试[手动更新](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#手动更新)": "Update failed, please try [manually updating](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#手动更新)",
+ "更新成功，请重启本程序": "Updated successfully, please restart this program",
+ "未命名对话历史记录": "Unnamed Dialog History",
+ "未设置代理...": "No proxy...",
"本月使用金额": "Monthly usage",
- "获取API使用情况失败:": "Failed to get API usage:",
- "API密钥更改为了": "The API key is changed to",
- "JSON解析错误,收到的内容: ": "JSON parsing error, received content: ",
+ "查看[使用介绍](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#微调-gpt-35)": "View the [usage guide](https://github.com/GaiZhenbiao/ChuanhuChatGPT/wiki/使用教程#微调-gpt-35) for more details",
+ "根据日期时间": "By date and time",
+ "模型": "Model",
+ "模型名称后缀": "Model Name Suffix",
+ "模型自动总结（消耗tokens）": "Auto summary by LLM (Consume tokens)",
"模型设置为了：": "Model is set to: ",
- "☹️发生了错误：": "☹️Error: ",
+ "正在尝试更新...": "Trying to update...",
+ "添加训练好的模型到模型列表": "Add trained model to the model list",
+ "状态": "Status",
+ "生成内容总结中……": "Generating content summary...",
+ "用于定位滥用行为": "Used to locate abuse",
+ "用户名": "Username",
+ "由Bilibili [土川虎虎虎](https://space.bilibili.com/29125536)、[明昭MZhao](https://space.bilibili.com/24807452) 和 [Keldos](https://github.com/Keldos-Li) 开发