[BUG]In the Jupyter notebook, define feedback function error UnicodeEncodeError

[hamster1111]

Bug Description
I encountered a UnicodeEncodeError while defining the feedback function

To Reproduce

from trulens.core import TruSession
from trulens.core import Feedback
from trulens.apps.llamaindex import TruLlama
from trulens.providers.langchain import Langchain
from trulens.feedback.v2.feedback import Groundedness
from langchain_community.llms import SparkLLM

tru = TruSession()
tru.reset_database()
langc = Langchain(chain=spark_llm)

qa_relevance = (
    Feedback(langc.relevance_with_cot_reasons, name="Answer Relevance")
    .on_input_output()
)

qs_relevance = (
    Feedback(langc.relevance_with_cot_reasons, name="Context Relevance")
    .on_input()
    .on(TruLlama.select_source_nodes().node.text)
    .aggregate(np.mean)
)


groundedness = (
    Feedback(langc.groundedness_measure_with_cot_reasons, name="Groundedness")
    .on(TruLlama.select_source_nodes().node.text)
    .on_output()
)
feedbacks = [qa_relevance, qs_relevance, groundedness]
tru_recorder = TruLlama(
        query_engine,
        app_id=app_id,
        feedbacks=feedbacks
    )

Expected behavior
It should not report an error, but instead obtain a result similar to the following:

✅ In Groundedness, input source will be set to __record__.app.first.steps__.context.first.invoke.rets[:].page_content.collect() .
✅ In Groundedness, input statement will be set to __record__.main_output or `Select.RecordOutput` .
✅ In Answer Relevance, input prompt will be set to __record__.main_input or `Select.RecordInput` .
✅ In Answer Relevance, input response will be set to __record__.main_output or `Select.RecordOutput` .
✅ In Context Relevance, input question will be set to __record__.main_input or `Select.RecordInput` .
✅ In Context Relevance, input context will be set to __record__.app.first.steps__.context.first.invoke.rets[:].page_content .

Relevant Logs/Tracebacks
The error message is as follows:

UnicodeEncodeError                        Traceback (most recent call last)
Cell In[24], line 4
      1 
      2 qa_relevance = (
      3     Feedback(langc.relevance_with_cot_reasons, name="Answer Relevance")
----> 4     .on_input_output()
      5 )
      7 qs_relevance = (
      8     Feedback(langc.relevance_with_cot_reasons, name="Context Relevance")
      9     .on_input()
     10     .on(TruLlama.select_source_nodes().node.text)
     11     .aggregate(np.mean)
     12 )
     15 groundedness = (
     16     Feedback(langc.groundedness_measure_with_cot_reasons, name="Groundedness")
     17     .on(TruLlama.select_source_nodes().node.text)
     18     .on_output()
     19 )

File D:\work\hamster\pythonProject\reinforce_interview\RAG_reinforce\venv\Lib\site-packages\trulens\core\feedback\feedback.py:296, in Feedback.on_input_output(self)
    289 def on_input_output(self) -> Feedback:
    290     """
    291     Specifies that the feedback implementation arguments are to be the main
    292     app input and output in that order.
    293 
    294     Returns a new Feedback object with the specification.
    295     """
--> 296     return self.on_input().on_output()

File D:\work\hamster\pythonProject\reinforce_interview\RAG_reinforce\venv\Lib\site-packages\trulens\core\feedback\feedback.py:591, in Feedback.on_prompt(self, arg)
    589 if arg is None:
    590     arg = self._next_unselected_arg_name()
--> 591     self._print_guessed_selector(arg, select_schema.Select.RecordInput)
    593 new_selectors[arg] = select_schema.Select.RecordInput
    595 ret = self.model_copy()

File D:\work\hamster\pythonProject\reinforce_interview\RAG_reinforce\venv\Lib\site-packages\trulens\core\feedback\feedback.py:323, in Feedback._print_guessed_selector(self, par_name, par_path)
    320 else:
    321     alias_info = ""
--> 323 print(
    324     f"{text_utils.UNICODE_CHECK} In {self.supplied_name if self.supplied_name is not None else self.name}, "
    325     f"input {par_name} will be set to {par_path}{alias_info} ."
    326 )

File D:\work\hamster\pythonProject\reinforce_interview\RAG_reinforce\venv\Scripts\pywin32_postinstall.py:27, in Tee.write(self, what)
     25     except OSError:
     26         pass
---> 27 tee_f.write(what)

UnicodeEncodeError: 'gbk' codec can't encode character '\u2705' in position 0: illegal multibyte sequence

Environment:

• OS: Windows
• Python Version: 3.12
• TruLens version:1.3.3

[auyil]

Not sure about your case, but I also had this error. I fixed it by adding encoding="utf-8" when reading the pdf for RAG.

from llama_index.core import SimpleDirectoryReader

documents = SimpleDirectoryReader(
    input_files=["./eBook-How-to-Build-a-Career-in-AI.pdf"], encoding="utf-8"
).load_data()