Anyone have an idea why truberics collector works in the first instance, but not in the second? The only difference in the code is that the second collector.st_feedback is inside the if statement.
Clue #1: From a UX standpoint it appears to work. The feedback icons can be clicked and they trigger a text input field. But no response is generated.
Clue #2: Trubrics collector.log_prompt works fine in both, which rules out any issues with the feedback collector.
I prefer to have the feedback widget only show once the response results are returned.
Snippet #1
query = st.chat_input("Type your question or task here", max_chars=200)
if query:
response = ASK.rag(query,retriever)
collector.log_prompt(
config_model={"model": "gpt-3.5-turbo"},
prompt=query,
generation=response['result'],
)
user_feedback = collector.st_feedback(
component="default",
feedback_type="thumbs",
model="gpt-3.5-turbo",
open_feedback_label="[Optional] Provide additional feedback",
align="flex-end",
prompt_id=None, # checkout collector.log_prompt() to log your user prompts
)
Snippet #2
query = st.chat_input("Type your question or task here", max_chars=200)
if query:
response = ASK.rag(query,retriever)
collector.log_prompt(
config_model={"model": "gpt-3.5-turbo"},
prompt=query,
generation=response['result'],
)
user_feedback = collector.st_feedback(
component="default",
feedback_type="thumbs",
model="gpt-3.5-turbo",
open_feedback_label="[Optional] Provide additional feedback",
align="flex-end",
prompt_id=None, # checkout collector.log_prompt() to log your user prompts
)
-
Share the link to the public deployed app.
https://ask-test.streamlit.app/ -
Share the link to your app’s public GitHub repository (including a [requirements file]
GitHub - dvvilkins/ASK at test -
Share the full text of the error message (not a screenshot).
-
Share the Streamlit and Python versions.
lastest of Streamlit
Python 3.9