dosubot[bot]
commented
10 months ago 🤖
Hello,
The error you're encountering is due to the way the Bedrock class in LangChain is designed. It checks the alternation between 'Human:' and 'Assistant:' in the input text. If the alternation is not correct, it raises a ValueError with the message "Error: Prompt must alternate between '\n\nHuman:' and '\n\nAssistant:'."
Here is the relevant code:
def _human_assistant_format(input_text: str) -> str:
...
count = 0
# track alternation
for i in range(len(input_text)):
if input_text[i : i + len(HUMAN_PROMPT)] == HUMAN_PROMPT:
if count % 2 == 0:
count += 1
else:
raise ValueError(ALTERNATION_ERROR)
if input_text[i : i + len(ASSISTANT_PROMPT)] == ASSISTANT_PROMPT:
if count % 2 == 1:
count += 1
else:
raise ValueError(ALTERNATION_ERROR)
...In this code, count is used to track the alternation between 'Human:' and 'Assistant:'. If 'Human:' is found when count is even, count is incremented. If 'Assistant:' is found when count is odd, count is also incremented. If these conditions are not met, a ValueError is raised.
Therefore, to avoid this error, you need to ensure that your input text alternates correctly between 'Human:' and 'Assistant:'.
As for the PromptTemplate class, it is used to format the prompts that are fed into the language models in the SmartLLMChain class. The PromptTemplate class formats the prompts based on the examples and the input variables provided. The formatted prompts are then used by the SmartLLMChain class to generate responses from the language models.
Here is an example of how the PromptTemplate class might be used:
prompt = FewShotPromptWithTemplates(
examples=[{"input": "Translate this English text to French: {text}", "output": "Traduisez ce texte anglais en français : {text}"}],
example_prompt=PromptTemplate(template="{input}\n{output}\n"),
suffix=StringPromptTemplate(template="Translate this English text to French: {text}"),
input_variables=["text"],
template_format="f-string"
)
formatted_prompt = prompt.format(text="Hello, world!")In this example, the PromptTemplate class is used to format a prompt for a translation task. The formatted prompt is then used by the SmartLLMChain class to generate a French translation of the text "Hello, world!".
I hope this helps! Let me know if you have any other questions.
Sources
- libs/langchain/langchain/llms/bedrock.py
- libs/langchain/tests/unit_tests/llms/test_bedrock.py
- libs/langchain/langchain/prompts/few_shot_with_templates.py
- libs/langchain/langchain/prompts/base.py
- libs/langchain/langchain/prompts/pipeline.py
This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.
Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.
If you want to continue the conversation, start your reply with @dosu-bot.
austinmw
seabasshn
priestleyn
DamianJot
rizblie
russell-dot-js
System Info
langchain version v0.0.305
Python 3.11
Who can help?
@hwchase17 @agola11
Information
Related Components
Reproduction
Results in error:
Expected behavior
No error