OpenAI throttling breaks summarization of large files

[dbillor]

this allows model to summarize large documents.

[ryanmac]

I've seen this too. OpenAI responds too many tokens after summary text.

[dbillor]

I have a fix can i contribute? https://github.com/Torantulino/Auto-GPT/pull/214 here is the PR

[SvenMeyer]

crash reproducible when processing larger file in chunks - is this fixed with with #210 ?

NEXT ACTION:  COMMAND = browse_website ARGUMENTS = {'url': 'https://docs.centrifuge.io/build/tinlake/', 'question': 'What is the overall architecture of the Tinlake system?'}
Text length: 50733 characters
Adding chunk 1 / 5 to memory
Summarizing chunk 1 / 5 of length 12070 characters, or 2990 tokens
Added chunk 1 summary to memory, of length 846 characters
Adding chunk 2 / 5 to memory
Summarizing chunk 2 / 5 of length 12788 characters, or 2930 tokens
Added chunk 2 summary to memory, of length 692 characters
Adding chunk 3 / 5 to memory
Summarizing chunk 3 / 5 of length 11341 characters, or 2974 tokens
SYSTEM:  Command browse_website returned: Error: The server is overloaded or not ready yet.
Traceback (most recent call last):
  File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/usr/lib/python3.10/runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "/home/sum/DEV/AI/Auto-GPT/Auto-GPT-0.2.2/autogpt/__main__.py", line 5, in <module>
    autogpt.cli.main()
  File "/usr/lib/python3.10/site-packages/click/core.py", line 1130, in __call__
    return self.main(*args, **kwargs)
  File "/usr/lib/python3.10/site-packages/click/core.py", line 1055, in main
    rv = self.invoke(ctx)
  File "/usr/lib/python3.10/site-packages/click/core.py", line 1635, in invoke
    rv = super().invoke(ctx)
  File "/usr/lib/python3.10/site-packages/click/core.py", line 1404, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/usr/lib/python3.10/site-packages/click/core.py", line 760, in invoke
    return __callback(*args, **kwargs)
  File "/usr/lib/python3.10/site-packages/click/decorators.py", line 26, in new_func
    return f(get_current_context(), *args, **kwargs)
  File "/home/sum/DEV/AI/Auto-GPT/Auto-GPT-0.2.2/autogpt/cli.py", line 151, in main
    agent.start_interaction_loop()
  File "/home/sum/DEV/AI/Auto-GPT/Auto-GPT-0.2.2/autogpt/agent/agent.py", line 75, in start_interaction_loop
    assistant_reply = chat_with_ai(
  File "/home/sum/DEV/AI/Auto-GPT/Auto-GPT-0.2.2/autogpt/chat.py", line 159, in chat_with_ai
    assistant_reply = create_chat_completion(
  File "/home/sum/DEV/AI/Auto-GPT/Auto-GPT-0.2.2/autogpt/llm_utils.py", line 93, in create_chat_completion
    response = openai.ChatCompletion.create(
  File "/home/sum/.local/lib/python3.10/site-packages/openai/api_resources/chat_completion.py", line 25, in create
    return super().create(*args, **kwargs)
  File "/home/sum/.local/lib/python3.10/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 153, in create
    response, _, api_key = requestor.request(
  File "/home/sum/.local/lib/python3.10/site-packages/openai/api_requestor.py", line 226, in request
    resp, got_stream = self._interpret_response(result, stream)
  File "/home/sum/.local/lib/python3.10/site-packages/openai/api_requestor.py", line 619, in _interpret_response
    self._interpret_response_line(
  File "/home/sum/.local/lib/python3.10/site-packages/openai/api_requestor.py", line 662, in _interpret_response_line
    raise error.ServiceUnavailableError(
openai.error.ServiceUnavailableError: The server is overloaded or not ready yet.

[Pwuts]

I strongly suspect this is a result of using Auto-GPT with a free account. The instructions clearly state to set up billing as there are strict rate limits for free accounts (max 3 requests/minute).

[maxluc1980]

I strongly suspect this is a result of using Auto-GPT with a free account. The instructions clearly state to set up billing as there are strict rate limits for free accounts (max 3 requests/minute).

@Pwuts I use the latest stable build of Auto-GPT with a paid account and run into the same issue on every task I give Auto-GPT. I also:

• re-installed it fresh multiple times
• re-installed the required files
• switched to gpt3only
• reduced the limit of tokens used to 1K

all to no avail.

I haven't been able to fully complete a single task up to this date.

[ZylerKade]

I strongly suspect this is a result of using Auto-GPT with a free account. The instructions clearly state to set up billing as there are strict rate limits for free accounts (max 3 requests/minute).

@Pwuts I use the latest stable build of Auto-GPT with a paid account and run into the same issue on every task I give Auto-GPT. I also:

• re-installed it fresh multiple times
• re-installed the required files
• switched to gpt3only
• reduced the limit of tokens used to 1K

all to no avail.

I haven't been able to fully complete a single task up to this date.

Same situation. Using Stable AutoGPT v0.3.0 and Paid OpenAI account that was working perfectly fine with the stable AutoGPT 0.2.2. Can't get it to run with CMD for v0.3.0 though.

[Pwuts]

@maxluc1980 @ZylerKade can either of you post a log?