From db9157fbcc2d1a9e97bf3d728a0a47111af78a76 Mon Sep 17 00:00:00 2001 From: Tyrin Todd Date: Thu, 23 Apr 2026 23:09:20 -0700 Subject: [PATCH] server-inference push halueval_summarization --- README.md | 5 ++++- config.json | 48 ++++++++++++++++++++++++++++++++++++++++++++++++ program.json | 44 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 96 insertions(+), 1 deletion(-) create mode 100644 config.json create mode 100644 program.json diff --git a/README.md b/README.md index 0580404..3d28582 100644 --- a/README.md +++ b/README.md @@ -1,2 +1,5 @@ -# halueval_summarization +--- +{} +--- +# halueval_summarization \ No newline at end of file diff --git a/config.json b/config.json new file mode 100644 index 0000000..6d8fd39 --- /dev/null +++ b/config.json @@ -0,0 +1,48 @@ +{ + "model": null, + "signature": { + "description": "Determine whether the summary is hallucinated given the source document.As you reason, note any uncertainties, ambiguities, and gaps \u2014 both in how the instructions apply to the task at hand and in whether you have the knowledge needed to solve it. Be honest in your reasoning when you are unsure about your answer.", + "properties": { + "document": { + "__dspy_field_type": "input", + "desc": "Source document that the summary should reflect.", + "prefix": "Document:", + "title": "Document", + "type": "string" + }, + "summary": { + "__dspy_field_type": "input", + "desc": "Summary to evaluate for hallucination.", + "prefix": "Summary:", + "title": "Summary", + "type": "string" + }, + "reasoning": { + "__dspy_field_type": "output", + "desc": "Step-by-step reasoning, including uncertainty, ambiguity, or uncertainty in your thought process when relevant.", + "prefix": "Reasoning:", + "title": "Reasoning", + "type": "string" + }, + "hallucination": { + "__dspy_field_type": "output", + "desc": "Final hallucination label: 'yes' if the summary is hallucinated, otherwise 'no'.", + "enum": [ + "yes", + "no" + ], + "prefix": "Hallucination:", + "title": "Hallucination", + "type": "string" + } + }, + "required": [ + "document", + "summary", + "reasoning", + "hallucination" + ], + "title": "HaluEvalSummarizationJudgeSig", + "type": "object" + } +} \ No newline at end of file diff --git a/program.json b/program.json new file mode 100644 index 0000000..65ef2dc --- /dev/null +++ b/program.json @@ -0,0 +1,44 @@ +{ + "traces": [], + "train": [], + "demos": [], + "signature": { + "instructions": "Determine whether the summary is hallucinated given the source document.As you reason, note any uncertainties, ambiguities, and gaps \u2014 both in how the instructions apply to the task at hand and in whether you have the knowledge needed to solve it. Be honest in your reasoning when you are unsure about your answer.", + "fields": [ + { + "prefix": "Document:", + "description": "Source document that the summary should reflect." + }, + { + "prefix": "Summary:", + "description": "Summary to evaluate for hallucination." + }, + { + "prefix": "Reasoning:", + "description": "Step-by-step reasoning, including uncertainty, ambiguity, or uncertainty in your thought process when relevant." + }, + { + "prefix": "Hallucination:", + "description": "Final hallucination label: 'yes' if the summary is hallucinated, otherwise 'no'." + } + ] + }, + "lm": { + "model": "together_ai/Qwen/Qwen2.5-7B-Instruct-Turbo", + "model_type": "chat", + "cache": true, + "num_retries": 3, + "finetuning_model": null, + "launch_kwargs": {}, + "train_kwargs": {}, + "temperature": null, + "max_tokens": null + }, + "metadata": { + "dependency_versions": { + "python": "3.11", + "dspy": "3.1.3", + "cloudpickle": "3.1" + } + } +} \ No newline at end of file