From 05dbed86bfe3fe8cf9329eaba88d59f07d325afe Mon Sep 17 00:00:00 2001 From: lilacheden Date: Mon, 20 Jan 2025 09:45:57 +0200 Subject: [PATCH] fix external_rag example Signed-off-by: lilacheden --- ...te_external_rag_results_with_binary_llm_as_judge.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/examples/evaluate_external_rag_results_with_binary_llm_as_judge.py b/examples/evaluate_external_rag_results_with_binary_llm_as_judge.py index e06d553fb..823c6ff4e 100644 --- a/examples/evaluate_external_rag_results_with_binary_llm_as_judge.py +++ b/examples/evaluate_external_rag_results_with_binary_llm_as_judge.py @@ -54,16 +54,16 @@ # Select the desired metric(s). # Each metric measures a certain aspect of the generated answer (answer_correctness, faithfulness, # answer_relevance and context_relevance). -# All available metrics are under "catalog.metrics.rag.autorag.", ending with "judge" +# All available metrics are under "catalog.metrics.rag.external_rag.", ending with "judge" # By default, all judges use llama_3_3_70b_instruct. We will soon see how to change this. metric_names = [ - "metrics.rag.autorag.answer_correctness.llama_3_3_70b_instruct_wml_judge", - "metrics.rag.autorag.faithfulness.llama_3_3_70b_instruct_wml_judge", + "metrics.rag.external_rag.answer_correctness.llama_3_3_70b_instruct_watsonx_judge", + "metrics.rag.external_rag.faithfulness.llama_3_3_70b_instruct_watsonx_judge", ] -# select the desired model. +# select the desired model, including provider. # all available models are under "catalog.engines.classification" -model_names = ["engines.classification.mixtral_8x7b_instruct_v01_wml"] +model_names = ["engines.classification.llama_3_3_70b_instruct_rits"] if __name__ == "__main__": multi_stream = MultiStream.from_iterables({"test": test_examples}, copying=True)