关于自部署时文件索引功能的使用。

[Olimiya]

首先感谢作者大大的开源项目。
我在本地部署时，之前在hugging face上正常使用的文件索引功能无法正常使用了。

1. 首先是部署后上传文件出现一直闪烁的现象，找到相同的issue：[Bug]: 本地部署上传索引文件后页面一直在闪烁 #644 (comment)
2. 更新后，再次上传，出现Error错误，报错日志为：KeyError: 'type'
   看了这个issue：[Bug]: config文件把本地索引改成true，然后上传文件后会报错 #676 (comment)
3. 以后改了config为本地索引编制。后续就提示缺少torch。然后再安装了requirements_advanced.
4. 后续发现我自己小服务器装不下torch。
5. 问题：请问自部署使用文件索引功能，是必须要求安装requirements_advanced是吗，这里面torch做的事情是什么呢。如果有这个要求的话，能否补充一下使用说明，感谢。

[GaiZhenbiao]

torch是运行本地embedding模型所需要的。本地embedding必须安装requirements_advanced.txt。本地embedding对服务器的性能要求没本地LLM那么高，但也需要至少2GB内存。

[GaiZhenbiao]

是的。

[Olimiya]

目前完成的requirements_advanced依赖的安装，然后还是Error，提示报错如下。请问一下是什么原因呢
2023-04-21 11:28:03,656 [INFO] [models.py:548] 正在加载OpenAI模型: gpt-3.5-turbo
2023-04-21 11:28:03,656 [INFO] [models.py:585] 模型设置为了： gpt-3.5-turbo
2023-04-21 11:28:16,115 [INFO] [llama_func.py:131] 找到了缓存的索引文件，加载中……
Traceback (most recent call last):
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/gradio/routes.py", line 395, in run_predict
output = await app.get_blocks().process_api(
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/gradio/blocks.py", line 1193, in process_api
result = await self.call_function(
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/gradio/blocks.py", line 916, in call_function
prediction = await anyio.to_thread.run_sync(
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/anyio/to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/anyio/_backends/_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/anyio/_backends/_asyncio.py", line 867, in run
result = context.run(func, *args)
File "/home/ubuntu/ChuanhuChatGPT/modules/utils.py", line 114, in handle_file_upload
return current_model.handle_file_upload(*args)
File "/home/ubuntu/ChuanhuChatGPT/modules/models/base_model.py", line 178, in handle_file_upload
construct_index(self.api_key, file_src=files)
File "/home/ubuntu/ChuanhuChatGPT/modules/llama_func.py", line 132, in construct_index
return GPTSimpleVectorIndex.load_from_disk(f"./index/{index_name}.json")
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/llama_index/indices/base.py", line 364, in load_from_disk
return cls.load_from_string(file_contents, **kwargs)
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/llama_index/indices/base.py", line 340, in load_from_string
return cls.load_from_dict(result_dict, **kwargs)
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/llama_index/indices/vector_store/base.py", line 260, in load_from_dict
vector_store = load_vector_store_from_dict(
File "/home/ubuntu/.virtualenvs/GPT/lib/python3.8/site-packages/llama_index/vector_stores/registry.py", line 52, in load_vector_store_from_dict
type = vector_store_dict[TYPE_KEY]
KeyError: 'type'
2023-04-21 11:29:02,052 [INFO] [models.py:548] 正在加载OpenAI模型: gpt-3.5-turbo
2023-04-21 11:29:02,052 [INFO] [models.py:585] 模型设置为了： gpt-3.5-turbo

[GaiZhenbiao]

把索引文件缓存（在index文件夹里）全部删掉再试试。本地embedding的索引文件与OpenAI的不兼容。

[Olimiya]

尝试以后上传文件一段时间后（应该是在解析）
软件崩了，Something went wrong
Connection errored out。
可能是系统内存不够被杀掉了。

[Olimiya]

日志内容。
2023-04-21 11:33:33,089 [INFO] [SentenceTransformer.py:66] Load pretrained SentenceTransformer: sentence-transformers/distiluse-base-multilingual-cased-v2
Downloading (…)6015c/.gitattributes: 100%|██████████| 690/690 [00:00<00:00, 81.8kB/s]
Downloading (…)_Pooling/config.json: 100%|██████████| 190/190 [00:00<00:00, 24.2kB/s]
Downloading (…)/2_Dense/config.json: 100%|██████████| 114/114 [00:00<00:00, 40.3kB/s]
Downloading pytorch_model.bin: 100%|██████████| 1.58M/1.58M [00:01<00:00, 1.43MB/s]
Downloading (…)ff6066015c/README.md: 100%|██████████| 2.38k/2.38k [00:00<00:00, 794kB/s]
Downloading (…)6066015c/config.json: 100%|██████████| 610/610 [00:00<00:00, 202kB/s]
Downloading (…)ce_transformers.json: 100%|██████████| 122/122 [00:00<00:00, 41.7kB/s]
Downloading pytorch_model.bin: 100%|██████████| 539M/539M [00:43<00:00, 12.4MB/s]
Downloading (…)nce_bert_config.json: 100%|██████████| 53.0/53.0 [00:00<00:00, 17.5kB/s]
Downloading (…)cial_tokens_map.json: 100%|██████████| 112/112 [00:00<00:00, 38.3kB/s]
Downloading (…)6015c/tokenizer.json: 100%|██████████| 1.96M/1.96M [00:05<00:00, 347kB/s]
Downloading (…)okenizer_config.json: 100%|██████████| 531/531 [00:00<00:00, 163kB/s]
Downloading (…)ff6066015c/vocab.txt: 100%|██████████| 996k/996k [00:00<00:00, 1.57MB/s]
Downloading (…)066015c/modules.json: 100%|██████████| 341/341 [00:00<00:00, 117kB/s]%