ValueError: No valid option generated in #select!

[Sam1320]

Error was raised running the following:

prompt = guidance('''Is the following sentence offensive? Please answer with a single word, either "Yes", "Nein", or "Vielleicht".
Sentence: {{example}}
Answer:{{#select "answer" logprobs='logprobs'}} Ja{{or}} Nein{{or}} Vielleicht{{/select}}''')
prompt = prompt(example='I hate tacos.')
prompt

I tried multiple variations. When the answer format specified at the beginning of the prompt and the options inside the select tag vary then very often this error happens.

[slundberg]

Can you share what version of guidance you are using? I can't get similar errors on the latest 0.0.47

[amanape]

Maybe try importing differently:

from guidance import Guidance
from guidance.llms import OpenAI

# example llm setup
# There is also a `token` parameter if necessary.
davinci = OpenAI(model='text-davinci-003')

guidance = Guidance('guidance-name')
prompt = guidance('''
Is the following sentence offensive? Please answer with a single word, either "Yes", "Nein", or "Vielleicht".
Sentence: {{example}}
Answer:{{#select "answer" logprobs='logprobs'}} Ja{{or}} Nein{{or}} Vielleicht{{/select}}
''', llm=davinci)

executed_prompt = prompt(example='I hate tacos.')

[Sam1320]

@slundberg my version is indeed 0.0.47. I tried the way you suggested @amanape but it also raised the same error.

This is another example that causes the error and is the closest to the example given in the readme:

import guidance
guidance.llm = guidance.llms.OpenAI('text-davinci-003')
guidance.llms.OpenAI.cache.clear()

prompt = guidance('''Is the following sentence offensive? Please answer with a single word, either "Yes", "No", or "Maybe".
Sentence: {{example}}
Answer:{{#select "answer" logprobs='logprobs'}} Yes{{or}} Nein{{or}} Maybe{{/select}}''')
prompt = prompt(example='I hate tacos.')
prompt

[amanape]

@Sam1320 This seems to have worked for me https://github.com/microsoft/guidance/issues/40#issuecomment-1554084184

[Sam1320]

@Sam1320 This seems to have worked for me #40 (comment)

Indeed also for me. Thanks @amanape!

[slundberg]

Thanks for reporting this! This should now be fixed in 0.0.49 :)