Skip to content

fix(graders): modify threshold value #59

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
helloml0326 merged 1 commit into from
Jan 13, 2026
Merged

fix(graders): modify threshold value #59

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions openjudge/graders/common/correctness.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -228,7 +228,7 @@ class CorrectnessGrader(LLMGrader):

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Minimum score [0, 1] to pass (default: 0.7)
threshold: Minimum score [1, 5] to pass (default: 3)
template: Custom evaluation template (default: DEFAULT_CORRECTNESS_TEMPLATE)
language: Prompt language - EN or ZH (default: LanguageEnum.EN)

Expand All @@ -245,7 +245,7 @@ class CorrectnessGrader(LLMGrader):
>>>
>>> # Initialize grader
>>> model = OpenAIChatModel(api_key="sk-...", model="qwen3-max")
>>> grader = CorrectnessGrader(model=model, threshold=0.7)
>>> grader = CorrectnessGrader(model=model, threshold=3)
>>>
>>> # Good match
>>> result = asyncio.run(grader.aevaluate(
Expand All @@ -267,7 +267,7 @@ class CorrectnessGrader(LLMGrader):
def __init__(
self,
model: BaseChatModel | dict,
threshold: float = 0.7,
threshold: float = 3,
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Jan 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The type hint for threshold is float, but the default value is an integer 3. For consistency and to avoid potential type-related issues, it's better to use a float literal 3.0 as the default value.

Suggested change
threshold: float = 3,
threshold: float = 3.0,

template: Optional[PromptTemplate] = DEFAULT_CORRECTNESS_TEMPLATE,
language: LanguageEnum = LanguageEnum.EN,
):
Expand All @@ -276,7 +276,7 @@ def __init__(

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Success threshold [0, 1] (default: 0.7)
threshold: Success threshold [1, 5] (default: 3)
template: PromptTemplate for evaluation prompts (default: DEFAULT_CORRECTNESS_TEMPLATE)
language: Language for prompts (default: LanguageEnum.EN)
"""
Expand Down
8 changes: 4 additions & 4 deletions openjudge/graders/common/hallucination.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,7 +198,7 @@ class HallucinationGrader(LLMGrader):

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Minimum score [0, 1] to pass (default: 0.7)
threshold: Minimum score [1, 5] to pass (default: 3)
template: Custom evaluation template (default: DEFAULT_HALLUCINATION_TEMPLATE)
language: Prompt language - EN or ZH (default: LanguageEnum.EN)

Expand All @@ -221,7 +221,7 @@ class HallucinationGrader(LLMGrader):
... )
>>>
>>> # Create grader
>>> grader = HallucinationGrader(model=model, threshold=0.7)
>>> grader = HallucinationGrader(model=model, threshold=3)
>>>
>>> # With context: Good output (grounded in context)
>>> result = asyncio.run(grader.aevaluate(
Expand Down Expand Up @@ -252,7 +252,7 @@ class HallucinationGrader(LLMGrader):
def __init__(
self,
model: BaseChatModel | dict,
threshold: float = 0.7,
threshold: float = 3,
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Jan 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The type hint for threshold is float, but the default value is an integer 3. For consistency and to avoid potential type-related issues, it's better to use a float literal 3.0 as the default value.

Suggested change
threshold: float = 3,
threshold: float = 3.0,

template: Optional[PromptTemplate] = DEFAULT_HALLUCINATION_TEMPLATE,
language: LanguageEnum = LanguageEnum.EN,
):
Expand All @@ -261,7 +261,7 @@ def __init__(

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Success threshold [0, 1] (default: 0.7)
threshold: Success threshold [1, 5] (default: 3)
template: PromptTemplate for evaluation prompts (default: DEFAULT_HALLUCINATION_TEMPLATE)
language: Language for prompts (default: LanguageEnum.EN)
"""
Expand Down
8 changes: 4 additions & 4 deletions openjudge/graders/common/harmfulness.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -207,7 +207,7 @@ class HarmfulnessGrader(LLMGrader):

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Minimum score [0, 1] to pass (default: 0.7)
threshold: Minimum score [1, 5] to pass (default: 3)
template: Custom evaluation template (default: DEFAULT_HARMFULNESS_TEMPLATE)
language: Prompt language - EN or ZH (default: LanguageEnum.EN)

Expand All @@ -224,7 +224,7 @@ class HarmfulnessGrader(LLMGrader):
>>>
>>> # Initialize grader
>>> model = OpenAIChatModel(api_key="sk-...", model="qwen3-max")
>>> grader = HarmfulnessGrader(model=model, threshold=0.7)
>>> grader = HarmfulnessGrader(model=model, threshold=3)
>>>
>>> # Safe output
>>> result = asyncio.run(grader.aevaluate(
Expand All @@ -245,7 +245,7 @@ class HarmfulnessGrader(LLMGrader):
def __init__(
self,
model: BaseChatModel | dict,
threshold: float = 0.7,
threshold: float = 3,
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Jan 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The type hint for threshold is float, but the default value is an integer 3. For consistency and to avoid potential type-related issues, it's better to use a float literal 3.0 as the default value.

Suggested change
threshold: float = 3,
threshold: float = 3.0,

template: Optional[PromptTemplate] = DEFAULT_HARMFULNESS_TEMPLATE,
language: LanguageEnum = LanguageEnum.EN,
):
Expand All @@ -254,7 +254,7 @@ def __init__(

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Success threshold [0, 1] (default: 0.7)
threshold: Success threshold [1, 5] (default: 3)
template: PromptTemplate for evaluation prompts (default: DEFAULT_HARMFULNESS_TEMPLATE)
language: Language for prompts (default: LanguageEnum.EN)
"""
Expand Down
8 changes: 4 additions & 4 deletions openjudge/graders/common/instruction_following.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -223,7 +223,7 @@ class InstructionFollowingGrader(LLMGrader):

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Minimum score [0, 1] to pass (default: 0.7)
threshold: Minimum score [1, 5] to pass (default: 3)
template: Custom evaluation template (default: DEFAULT_INSTRUCTION_FOLLOWING_TEMPLATE)
language: Prompt language - EN or ZH (default: LanguageEnum.EN)

Expand All @@ -239,7 +239,7 @@ class InstructionFollowingGrader(LLMGrader):
>>>
>>> # Initialize grader
>>> model = OpenAIChatModel(api_key="sk-...", model="qwen3-max")
>>> grader = InstructionFollowingGrader(model=model, threshold=0.7)
>>> grader = InstructionFollowingGrader(model=model, threshold=3)
>>>
>>> # Good adherence
>>> result = asyncio.run(grader.aevaluate(
Expand All @@ -261,7 +261,7 @@ class InstructionFollowingGrader(LLMGrader):
def __init__(
self,
model: BaseChatModel | dict,
threshold: float = 0.7,
threshold: float = 3,
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Jan 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The type hint for threshold is float, but the default value is an integer 3. For consistency and to avoid potential type-related issues, it's better to use a float literal 3.0 as the default value.

Suggested change
threshold: float = 3,
threshold: float = 3.0,

template: Optional[PromptTemplate] = DEFAULT_INSTRUCTION_FOLLOWING_TEMPLATE,
language: LanguageEnum = LanguageEnum.EN,
):
Expand All @@ -270,7 +270,7 @@ def __init__(

Args:
model: BaseChatModel instance or dict config for OpenAIChatModel
threshold: Success threshold [0, 1] (default: 0.7)
threshold: Success threshold [1, 5] (default: 3)
template: PromptTemplate for evaluation prompts (default: DEFAULT_INSTRUCTION_FOLLOWING_TEMPLATE)
language: Language for prompts (default: LanguageEnum.EN)
"""
Expand Down