src/display/utils.py

from dataclasses import dataclass, make_dataclass

from src.display.about import create_task_list

def fields(raw_class):
    return [v for k, v in raw_class.__dict__.items() if k[:2] != "__" and k[-2:] != "__"]


# These classes are for user facing column names,
# to avoid having to change them all around the code
# when a modif is needed
@dataclass
class ColumnContent:
    name: str
    type: str
    displayed_by_default: bool
    hidden: bool = False
    never_hidden: bool = False
    dummy: bool = False

Tasks, Groups = create_task_list()

## Leaderboard columns
auto_eval_column_dict = []
# Init
auto_eval_column_dict.append(["model_submission_date", ColumnContent, ColumnContent("Submission Date", "str", True, never_hidden=True)])
auto_eval_column_dict.append(["model", ColumnContent, ColumnContent("Model", "markdown", True, never_hidden=True)])
#Scores
auto_eval_column_dict.append(["average", ColumnContent, ColumnContent("Average ⬆️", "number", True)])


for task in Tasks:
    auto_eval_column_dict.append([task.benchmark, ColumnContent, ColumnContent(task.col_name, "number", True)])
# Dummy column for the search bar (hidden by the custom CSS)
auto_eval_column_dict.append(["dummy", ColumnContent, ColumnContent("model_name_for_query", "str", False, dummy=True)])

# We use make dataclass to dynamically fill the scores from Tasks
AutoEvalColumn = make_dataclass("AutoEvalColumn", auto_eval_column_dict, frozen=True)

## For the queue columns in the submission tab
@dataclass(frozen=True)
class EvalQueueColumn:  # Queue column
    model = ColumnContent("model", "markdown", True)
    submitted_time = ColumnContent("submitted_time", "str", True)
    status = ColumnContent("status", "str", True)
    
# Column selection
COLS = [c.name for c in fields(AutoEvalColumn) if not c.hidden]
TYPES = [c.type for c in fields(AutoEvalColumn) if not c.hidden]

EVAL_COLS = [c.name for c in fields(EvalQueueColumn)]
EVAL_TYPES = [c.type for c in fields(EvalQueueColumn)]

BENCHMARK_COLS = [t.col_name for t in Tasks]





#for grouping


## Leaderboard columns
auto_eval_group_dict = []
# Init
auto_eval_group_dict.append(["model_submission_date", ColumnContent, ColumnContent("Submission Date", "str", True, never_hidden=True)])
auto_eval_group_dict.append(["model", ColumnContent, ColumnContent("Model", "markdown", True, never_hidden=True)])
#Scores
auto_eval_group_dict.append(["average", ColumnContent, ColumnContent("Average ⬆️", "number", True)])


for task in Groups:
    auto_eval_group_dict.append([task.benchmark, ColumnContent, ColumnContent(task.col_name, "number", True)])
# Dummy column for the search bar (hidden by the custom CSS)
auto_eval_group_dict.append(["dummy", ColumnContent, ColumnContent("model_name_for_query", "str", False, dummy=True)])

# We use make dataclass to dynamically fill the scores from Tasks
AutoEvalColumnGroup = make_dataclass("AutoEvalColumnGroup", auto_eval_group_dict, frozen=True)

## For the queue columns in the submission tab
@dataclass(frozen=True)
class EvalQueueColumnGroup:  # Queue column
    model = ColumnContent("model", "markdown", True)
    submitted_time = ColumnContent("submitted_time", "str", True)
    status = ColumnContent("status", "str", True)
    
# Column selection
COLS_GROUP = [c.name for c in fields(AutoEvalColumnGroup) if not c.hidden]
TYPES_GROUP = [c.type for c in fields(AutoEvalColumnGroup) if not c.hidden]

EVAL_COLS_GROUP = [c.name for c in fields(EvalQueueColumnGroup)]
EVAL_TYPES_GROUP = [c.type for c in fields(EvalQueueColumnGroup)]

BENCHMARK_COLS_GROUP = [t.col_name for t in Groups]