aiwaves-cn/recurrentgpt

Official Code for Paper: RecurrentGPT: Interactive Generation of (Arbitrarily) Long Text

recurrentgpt's Issues

Use the 'gpt-3.5-turbo-16k-0613' model - Much better outputs

I am using the gpt-3.5-turbo-16k-0613 model and I am getting much better, longer outputs.

I'm getting an issue with the 'cookie'

Trying to run this locally.

I got the cookie issue, and I tried editing the python file to skip the cookie if it wasn't there for the first time. I was able to generate the text one time, but after clicking next I get another cookie error.

I think this is something wrong with my local setup, or is this the project?

Syntax Error

python: 3.8

Code is not modified, and error reporting is executed.

I found some bug

human.step_with_edit() maybe is human.step()

may be a stupid question

I want to ask a question, does this interactive novel have an ending?

Cookie Problem

How to deploy interactive novels locally?

How can I deploy the interactive fiction writing project locally? I didn't find this part in the github code, but I'm very interested in it. Can you upload it? Thanks~

Unable to continue, an error will be reported

local gradio demo has a bug

File "xxx/RecurrentGPT/gradio_server.py", line 65, in init
init_paragraphs = get_init(chatbot, init_prompt(novel_type, description))
File "xxx/RecurrentGPT/utils.py", line 38, in get_init
with open(init_text,'r',encoding='utf-8') as f:
TypeError: expected str, bytes or os.PathLike object, not Chatbot

I guess this demo use web session for GPT4.
How about call gpt-3.5-turbo api for the demo ?
Thank you for your interesting work.

Something went wrong Expecting value: line 1 column 1 (char 0)

没有见到类似的问题

我运行gradio_server.py，后端警告如下：
UserWarning: Failed to initialize NumPy: module compiled against API version 0x10 but this version of numpy is 0xf . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem . (Triggered internally at ..\torch\csrc\utils\tensor_numpy.cpp:84.)
t = torch.tensor([], dtype=storage.dtype, device=storage._untyped_storage.device)

打开本地网页应用后，当我尝试开始运行就出现了以下问题导致无法运行
Something went wrong
Expecting value: line 1 column 1 (char 0)

猜测可能是环境中依赖版本不匹配导致的，我想可能需要用conda导出一个完整的可运行环境的requirements.txt

RecurrentGPT中文版本来啦~~

我把文章的那个功能改成了中文版，也发在GitHub上了。特来官方这里报告一声，链接：https://github.com/femnn/RecurrentGPT-zh

How can I use it locally?

This is an exciting extension of ChatGPT.
Can you write a detailed teaching for local use?
Thank.

Streamlit version

Hi,

I've created a Streamlit version of RecurrentGPT, called StoryGPT.

https://github.com/tractorjuice/storygpt

No Chinese version

I see "第一人称互动式小说生成器" while opening the DEMO website (https://www.aiwaves.org/interactivefiction)
But it disappears while I run it on Hugging Face (https://imliguojie-recurrent.hf.space)

Could you put it as an option in the configuration file?

fine tune

How can we fine-tune this recurrent gpt model so that it can output some other type of information but not only the novels?

How to deploy interactive novels locally?

How to deploy interactive novels locally? At the same time, I also encountered the problem of cookies when deploying locally

what is sentence_transformers, cannot find in code

Encountering cookie error during local deployment with third-party ChatGPT library and latest code

I am having trouble deploying RecurrentGPT locally and encountering a cookie error when using a third-party ChatGPT library for continuous generation. I am using the latest code and I am not sure if the official API endpoint has this issue. However, I believe this issue is not related to the API endpoint, as I am occasionally able to continue generating text. Can someone please help me with this issue?

Thank you.

Good Project！

If I can， I want to divide on this project。 but i m no good at ml or llm..haha. I will focus this project

Add a License to the Repository

I noticed that this repository does not have a license. Would you consider adding a license to clarify the terms under which others can use the code?

Can I use the language model for local deployment?

Can I use the language model for local deployment?