diff --git a/config.py b/config.py index e668ac90..ac8d2fdf 100644 --- a/config.py +++ b/config.py @@ -45,6 +45,7 @@ AVAIL_LLM_MODELS = ["qwen-max", "o1-mini", "o1-mini-2024-09-12", "o1", "o1-2024- "gemini-1.5-pro", "chatglm3", "chatglm4", "deepseek-chat", "deepseek-coder", "deepseek-reasoner", "volcengine-deepseek-r1-250120", "volcengine-deepseek-v3-241226", + "dashscope-deepseek-r1", "dashscope-deepseek-v3", ] EMBEDDING_MODEL = "text-embedding-3-small" diff --git a/main.py b/main.py index 73a3e2b8..062cba5f 100644 --- a/main.py +++ b/main.py @@ -184,7 +184,7 @@ def main(): from themes.gui_floating_menu import define_gui_floating_menu area_input_secondary, txt2, area_customize, _, resetBtn2, clearBtn2, stopBtn2 = \ define_gui_floating_menu(customize_btns, functional, predefined_btns, cookies, web_cookie_cache) - + # 浮动时间线定义 gr.Spark() diff --git a/request_llms/bridge_all.py b/request_llms/bridge_all.py index e911a765..52ed2fa7 100644 --- a/request_llms/bridge_all.py +++ b/request_llms/bridge_all.py @@ -530,6 +530,15 @@ model_info = { "tokenizer": tokenizer_gpt35, "token_cnt": get_token_num_gpt35, }, + "gemini-2.0-flash": { + "fn_with_ui": genai_ui, + "fn_without_ui": genai_noui, + "endpoint": gemini_endpoint, + "has_multimodal_capacity": True, + "max_token": 1024 * 204800, + "tokenizer": tokenizer_gpt35, + "token_cnt": get_token_num_gpt35, + }, # cohere "cohere-command-r-plus": { @@ -813,8 +822,9 @@ if "qwen-local" in AVAIL_LLM_MODELS: }) except: logger.error(trimmed_format_exc()) -# -=-=-=-=-=-=- 通义-在线模型 -=-=-=-=-=-=- -qwen_models = ["qwen-max-latest", "qwen-max-2025-01-25","qwen-max","qwen-turbo","qwen-plus"] + +# -=-=-=-=-=-=- 阿里云百炼（通义）-在线模型 -=-=-=-=-=-=- +qwen_models = ["qwen-max-latest", "qwen-max-2025-01-25","qwen-max","qwen-turbo","qwen-plus","dashscope-deepseek-r1","dashscope-deepseek-v3"] if any(item in qwen_models for item in AVAIL_LLM_MODELS): try: from .bridge_qwen import predict_no_ui_long_connection as qwen_noui @@ -864,10 +874,30 @@ if any(item in qwen_models for item in AVAIL_LLM_MODELS): "max_token": 30720, "tokenizer": tokenizer_gpt35, "token_cnt": get_token_num_gpt35, + }, + "dashscope-deepseek-r1": { + "fn_with_ui": qwen_ui, + "fn_without_ui": qwen_noui, + "enable_reasoning": True, + "can_multi_thread": True, + "endpoint": None, + "max_token": 57344, + "tokenizer": tokenizer_gpt35, + "token_cnt": get_token_num_gpt35, + }, + "dashscope-deepseek-v3": { + "fn_with_ui": qwen_ui, + "fn_without_ui": qwen_noui, + "can_multi_thread": True, + "endpoint": None, + "max_token": 57344, + "tokenizer": tokenizer_gpt35, + "token_cnt": get_token_num_gpt35, } }) except: logger.error(trimmed_format_exc()) + # -=-=-=-=-=-=- 零一万物模型 -=-=-=-=-=-=- yi_models = ["yi-34b-chat-0205","yi-34b-chat-200k","yi-large","yi-medium","yi-spark","yi-large-turbo","yi-large-preview"] if any(item in yi_models for item in AVAIL_LLM_MODELS): diff --git a/request_llms/bridge_chatgpt.py b/request_llms/bridge_chatgpt.py index 7d5cbe64..b1eddb9e 100644 --- a/request_llms/bridge_chatgpt.py +++ b/request_llms/bridge_chatgpt.py @@ -368,12 +368,12 @@ def predict(inputs:str, llm_kwargs:dict, plugin_kwargs:dict, chatbot:ChatBotWith log_chat(llm_model=llm_kwargs["llm_model"], input_str=inputs, output_str=gpt_replying_buffer) break # 对于符合规范的接口，这里可以break else: - continue # 对于不符合规范的狗屎接口，这里需要继续 + continue # 对于不符合规范的接口，这里需要继续 # 到这里，我们已经可以假定必须包含choice了 try: status_text = f"finish_reason: {chunkjson['choices'][0].get('finish_reason', 'null')}" except: - logger.error(f"一些垃圾第三方接口出现这样的错误，兼容一下吧: {chunk_decoded}") + logger.error(f"一些第三方接口出现这样的错误，兼容一下吧: {chunk_decoded}") # 处理数据流的主体 if has_content: # 正常情况 @@ -382,9 +382,9 @@ def predict(inputs:str, llm_kwargs:dict, plugin_kwargs:dict, chatbot:ChatBotWith # 一些第三方接口的出现这样的错误，兼容一下吧 continue else: - # 至此已经超出了正常接口应该进入的范围，一些垃圾第三方接口会出现这样的错误 + # 至此已经超出了正常接口应该进入的范围，一些第三方接口会出现这样的错误 if chunkjson['choices'][0]["delta"].get("content", None) is None: - logger.error(f"一些垃圾第三方接口出现这样的错误，兼容一下吧: {chunk_decoded}") + logger.error(f"一些第三方接口出现这样的错误，兼容一下吧: {chunk_decoded}") continue gpt_replying_buffer = gpt_replying_buffer + chunkjson['choices'][0]["delta"]["content"] diff --git a/request_llms/com_qwenapi.py b/request_llms/com_qwenapi.py index 70872e16..edbf9d7b 100644 --- a/request_llms/com_qwenapi.py +++ b/request_llms/com_qwenapi.py @@ -3,6 +3,7 @@ from toolbox import get_conf import threading timeout_bot_msg = '[Local Message] Request timeout. Network error.' +model_prefix_to_remove = 'dashscope-' class QwenRequestInstance(): def __init__(self): @@ -20,6 +21,13 @@ class QwenRequestInstance(): raise RuntimeError('请配置 DASHSCOPE_API_KEY') dashscope.api_key = get_conf("DASHSCOPE_API_KEY") + def format_reasoning(self, reasoning_content:str, main_content:str): + if reasoning_content: + reasoning_content_paragraphs = ''.join([f'

{line}

' for line in reasoning_content.split('\n')]) + formatted_reasoning_content = f'
{reasoning_content_paragraphs}
\n\n---\n\n' + return formatted_reasoning_content + main_content + else: + return main_content def generate(self, inputs, llm_kwargs, history, system_prompt): # import _thread as thread @@ -28,9 +36,13 @@ class QwenRequestInstance(): if top_p == 0: top_p += 1e-5 if top_p == 1: top_p -= 1e-5 + model_name = llm_kwargs['llm_model'] + if model_name.startswith(model_prefix_to_remove): model_name = model_name[len(model_prefix_to_remove):] + + self.reasoning_buf = "" self.result_buf = "" responses = Generation.call( - model=llm_kwargs['llm_model'], + model=model_name, messages=generate_message_payload(inputs, llm_kwargs, history, system_prompt), top_p=top_p, temperature=llm_kwargs.get('temperature', 1.0), @@ -46,18 +58,24 @@ class QwenRequestInstance(): self.result_buf += response.output.choices[0].message.content except: pass - yield self.result_buf + yield self.format_reasoning(self.reasoning_buf, self.result_buf) break elif response.output.choices[0].finish_reason == 'length': self.result_buf += "[Local Message] 生成长度过长，后续输出被截断" - yield self.result_buf + yield self.format_reasoning(self.reasoning_buf, self.result_buf) break else: + try: + contain_reasoning = hasattr(response.output.choices[0].message, 'reasoning_content') + except: + contain_reasoning = False + if contain_reasoning: + self.reasoning_buf += response.output.choices[0].message.reasoning_content self.result_buf += response.output.choices[0].message.content - yield self.result_buf + yield self.format_reasoning(self.reasoning_buf, self.result_buf) else: self.result_buf += f"[Local Message] 请求错误：状态码：{response.status_code}，错误码:{response.code}，消息：{response.message}" - yield self.result_buf + yield self.format_reasoning(self.reasoning_buf, self.result_buf) break # 耗尽generator避免报错 diff --git a/shared_utils/advanced_markdown_format.py b/shared_utils/advanced_markdown_format.py index 7be1db20..a8a10411 100644 --- a/shared_utils/advanced_markdown_format.py +++ b/shared_utils/advanced_markdown_format.py @@ -347,7 +347,7 @@ def markdown_convertion(txt): # 在文本中插入一个base64编码的原始文本，以便在复制时能够获得原始文本 raw_text_encoded = compress_string(txt) - raw_text_node = f'' + raw_text_node = f'" # 用于查找数学公式的正则表达式 diff --git a/themes/common.css b/themes/common.css index 2f7dda2f..610d46c2 100644 --- a/themes/common.css +++ b/themes/common.css @@ -335,4 +335,74 @@ .raw_text { display: none; +} + +.message_tail { + justify-content: center; + align-items: center; +} +.message_tail_stop { + border: dotted !important; + margin-top: 5px !important; + padding-left: 8px !important; + padding-top: 1px !important; + padding-bottom: 1px !important; + padding-right: 12px !important; +} + +/* ant 开始 */ + +/* 通用按钮样式 */ +.ant-btn { + display: inline-flex; + align-items: center; + justify-content: center; + padding: 8px 16px; + font-size: 14px; + font-weight: bold; + border-radius: 4px; + cursor: pointer; + transition: background-color 0.3s; +} + +/* 按钮颜色和状态 */ +.ant-btn-primary { + background-color: #1890ff; + color: white; + border: none; +} + +.ant-btn-primary:hover { + background-color: #40a9ff; +} + +.ant-btn-loading { + pointer-events: none; + opacity: 0.7; +} + +/* 图标样式 */ +.ant-btn-icon { + display: inline-flex; + margin-right: 8px; +} + +.anticon { + width: 1em; + height: 1em; + fill: currentColor; +} + +.anticon-spin { + animation: spin 1s linear infinite; +} + +/* 动画效果 */ +@keyframes spin { + from { + transform: rotate(0deg); + } + to { + transform: rotate(360deg); + } } \ No newline at end of file diff --git a/themes/common.js b/themes/common.js index f8510624..f05b9b71 100644 --- a/themes/common.js +++ b/themes/common.js @@ -289,7 +289,7 @@ function addCopyButton(botElement, index, is_last_in_arr) { const audioIcon = ''; // const cancelAudioIcon = ''; - // 此功能没准备好 + // audio functionality if (allow_auto_read_continously && is_last_in_arr && allow_auto_read_tts_flag) { process_latest_text_output(botElement.innerText, index); } @@ -300,41 +300,6 @@ function addCopyButton(botElement, index, is_last_in_arr) { return; } - // var copyButton = document.createElement('button'); - // copyButton.classList.add('copy-bot-btn'); - // copyButton.setAttribute('aria-label', 'Copy'); - // copyButton.innerHTML = copyIcon; - // copyButton.addEventListener('click', async () => { - // const textToCopy = botElement.innerText; - // try { - // // push_text_to_audio(textToCopy).catch(console.error); - // if ("clipboard" in navigator) { - // await navigator.clipboard.writeText(textToCopy); - // copyButton.innerHTML = copiedIcon; - // setTimeout(() => { - // copyButton.innerHTML = copyIcon; - // }, 1500); - // } else { - // const textArea = document.createElement("textarea"); - // textArea.value = textToCopy; - // document.body.appendChild(textArea); - // textArea.select(); - // try { - // document.execCommand('copy'); - // copyButton.innerHTML = copiedIcon; - // setTimeout(() => { - // copyButton.innerHTML = copyIcon; - // }, 1500); - // } catch (error) { - // console.error("Copy failed: ", error); - // } - // document.body.removeChild(textArea); - // } - // } catch (error) { - // console.error("Copy failed: ", error); - // } - // }); - // 原始文本拷贝 var copyButtonOrig = document.createElement('button'); copyButtonOrig.classList.add('copy-bot-btn'); @@ -436,7 +401,6 @@ function do_something_but_not_too_frequently(min_interval, func) { function chatbotContentChanged(attempt = 1, force = false) { - // https://github.com/GaiZhenbiao/ChuanhuChatGPT/tree/main/web_assets/javascript for (var i = 0; i < attempt; i++) { setTimeout(() => { const messages = gradioApp().querySelectorAll('#gpt-chatbot .message-wrap .message.bot'); @@ -447,6 +411,9 @@ function chatbotContentChanged(attempt = 1, force = false) { // Now pass both the message element and the is_last_in_arr boolean to addCopyButton addCopyButton(message, index, is_last_in_arr); + // if last message, add stop btn link + addStopButton(message, index, is_last_in_arr); + // save_conversation_history save_conversation_history_slow_down(); }); @@ -457,6 +424,79 @@ function chatbotContentChanged(attempt = 1, force = false) { } +function addStopButton(botElement, index, is_last_in_arr) { + function is_generating() { + var statePanelElement = document.getElementById("state-panel"); + var generatingElement = statePanelElement.querySelector(".generating"); + if (generatingElement) { + return true; + } else { + return false; + } + } + function on_stop_btn_click() { + let stopButton = document.getElementById("elem_stop"); + stopButton.click(); + } + function remove_stop_generate_btn(messageTailElement) { + // remove all child elements of messageTailElement + while (messageTailElement.firstChild) { + messageTailElement.removeChild(messageTailElement.firstChild); + } + messageTailElement.style.display = 'none'; + messageTailElement.classList.add('removed'); + } + function add_stop_generate_btn() { + // write here: add a beautiful stop btn `bottomElement` as child, when clicked execute on_stop_btn_click + console.log("get_gradio_component") + const bottomElement = document.createElement('button'); + bottomElement.className = 'ant-btn ant-btn-primary'; + bottomElement.innerHTML = ` + + + + 终止 + `; + bottomElement.classList.add('message_tail_stop'); + bottomElement.addEventListener('click', on_stop_btn_click); + messageTailElement.appendChild(bottomElement); + } + + // find a sub element of class `message_tail` + const messageTailElement = botElement.querySelector('.message_tail'); + // if not is_last_in_arr, hide this elem (display none) + if (!messageTailElement) { + return; + } + if (messageTailElement.classList.contains('removed')) { + return; + } + if (!is_last_in_arr) { + remove_stop_generate_btn(messageTailElement); + return; + } + messageTailElement.style.display = 'flex'; + const messageTailStopElem = messageTailElement.querySelector('.message_tail_stop'); + if(!is_generating()) { + setTimeout(() => { + if(!is_generating()) { + remove_stop_generate_btn(messageTailElement); + return; + } else { + if (!messageTailStopElem) { + add_stop_generate_btn() + } + } + }, 500); + } else { + if (!messageTailStopElem) { + add_stop_generate_btn() + } + } + +} // -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=