{% include 'pages/navbar.html' %}
← LLM campaigns

New LLM {% if mode == 'llm_gen' %}generation{% else %}evaluation{% endif %} campaign

Preset
{% if mode == 'llm_eval' %} {% include 'include/annotation_span_categories.html' %}
{% endif %}
Prompt template
Prompt for the model. See the instructions for configuring the prompt at factgenie wiki.
System message
Role of the model (optional).
{% if mode == 'llm_gen' %}
Start with
The model will need to start the generated text with this string (optional).
{% endif %}
LLM evaluator
Select one of the available LLM evaluators (or add your own class to factgenie/models.py).
API URL
For Ollama API, this is the API URL, e.g. http://my-server.com:11434/api/. The parameter is ignored for OpenAI API.
Model
The identifier of the model you are querying.
Model arguments
API arguments to be passed to the model. Internally, we use ast.literal_eval() for determining the argument type. Use floating point format (e.g. 0.0) for the numbers that should be interpreted as floats.
Extra arguments
Any other parameters used by the metric.
{% include 'include/data_selection.html' %}
{% include 'include/example_annotation_modal.html' %}