Hi,
The following code snippet from Chapter 3 :
from langchain.llms import HuggingFaceHub
llm = HuggingFaceHub(
model_kwargs={"temperature": 0.5, "max_length": 64},
repo_id="google/flan-t5-xxl"
)
prompt = "In which country is Tokyo?"
completion = llm(prompt)
print(completion)
Is giving:
ValueError Traceback (most recent call last)
Cell In[6], line 2
1 prompt = "In which country is Tokyo?"
----> 2 completion = llm(prompt)
3 print(completion)
File D:\generative_ai_with_langchain\pyenv\Lib\site-packages\langchain\llms\base.py:825, in BaseLLM.call(self, prompt, stop, callbacks, tags, metadata, **kwargs)
818 if not isinstance(prompt, str):
819 raise ValueError(
820 "Argument prompt
is expected to be a string. Instead found "
821 f"{type(prompt)}. If you want to run the LLM on multiple prompts, use "
822 "generate
instead."
823 )
824 return (
--> 825 self.generate(
826 [prompt],
827 stop=stop,
828 callbacks=callbacks,
829 tags=tags,
830 metadata=metadata,
831 **kwargs,
832 )
833 .generations[0][0]
834 .text
835 )
File D:\generative_ai_with_langchain\pyenv\Lib\site-packages\langchain\llms\base.py:621, in BaseLLM.generate(self, prompts, stop, callbacks, tags, metadata, **kwargs)
612 raise ValueError(
613 "Asked to cache, but no cache found at langchain.cache
."
614 )
615 run_managers = [
616 callback_manager.on_llm_start(
617 dumpd(self), [prompt], invocation_params=params, options=options
618 )[0]
619 for callback_manager, prompt in zip(callback_managers, prompts)
620 ]
--> 621 output = self._generate_helper(
622 prompts, stop, run_managers, bool(new_arg_supported), **kwargs
623 )
624 return output
625 if len(missing_prompts) > 0:
File D:\generative_ai_with_langchain\pyenv\Lib\site-packages\langchain\llms\base.py:523, in BaseLLM._generate_helper(self, prompts, stop, run_managers, new_arg_supported, **kwargs)
521 for run_manager in run_managers:
522 run_manager.on_llm_error(e)
--> 523 raise e
524 flattened_outputs = output.flatten()
525 for manager, flattened_output in zip(run_managers, flattened_outputs):
File D:\generative_ai_with_langchain\pyenv\Lib\site-packages\langchain\llms\base.py:510, in BaseLLM._generate_helper(self, prompts, stop, run_managers, new_arg_supported, **kwargs)
500 def _generate_helper(
501 self,
502 prompts: List[str],
(...)
506 **kwargs: Any,
507 ) -> LLMResult:
508 try:
509 output = (
--> 510 self._generate(
511 prompts,
512 stop=stop,
513 # TODO: support multiple run managers
514 run_manager=run_managers[0] if run_managers else None,
515 **kwargs,
516 )
517 if new_arg_supported
518 else self._generate(prompts, stop=stop)
519 )
520 except (KeyboardInterrupt, Exception) as e:
521 for run_manager in run_managers:
File D:\generative_ai_with_langchain\pyenv\Lib\site-packages\langchain\llms\base.py:1000, in LLM._generate(self, prompts, stop, run_manager, **kwargs)
997 new_arg_supported = inspect.signature(self._call).parameters.get("run_manager")
998 for prompt in prompts:
999 text = (
-> 1000 self._call(prompt, stop=stop, run_manager=run_manager, **kwargs)
1001 if new_arg_supported
1002 else self._call(prompt, stop=stop, **kwargs)
1003 )
1004 generations.append([Generation(text=text)])
1005 return LLMResult(generations=generations)
File D:\generative_ai_with_langchain\pyenv\Lib\site-packages\langchain\llms\huggingface_hub.py:112, in HuggingFaceHub._call(self, prompt, stop, run_manager, **kwargs)
110 response = self.client(inputs=prompt, params=params)
111 if "error" in response:
--> 112 raise ValueError(f"Error raised by inference API: {response['error']}")
113 if self.client.task == "text-generation":
114 # Text generation return includes the starter text.
115 text = response[0]["generated_text"][len(prompt) :]
ValueError: Error raised by inference API: Service Unavailable