src/about.py

from dataclasses import dataclass
from enum import Enum


@dataclass
class Task:
    benchmark: str
    metric: str
    col_name: str
    higher_is_better: bool = True
    scale_by_100: bool = True


# Select your tasks here
# ---------------------------------------------------
class Tasks(Enum):
    # task_key in the json file, metric_key in the json file, name to display in the leaderboard
    task1 = Task("ami_2020_aggressiveness", "f1,none", "AMI 2020 Agg")
    task2 = Task("ami_2020_misogyny", "f1,none", "AMI 2020 Miso")
    task0 = Task("arc_challenge_ita", "acc_norm,none", "ARC-C")
    task4 = Task("belebele_ita", "acc_norm,none", "Belebele")
    task3 = Task("gente_rephrasing", "acc,none", "GeNTE Neutralizing")
    task12 = Task("haspeede2_hs", "f1,none", "HaSpeeDe2 HS")
    task13 = Task("haspeede2_stereo", "f1,none", "HaSpeeDe2 Stereo")
    task5 = Task("hatecheck_ita", "f1,none", "HateCheck")
    task6 = Task("honest_ita", "acc,none", "HONEST", higher_is_better=False)
    task14 = Task("ironita_irony", "f1,none", "IronITA Irony")
    task15 = Task("ironita_sarcasm", "f1,none", "IronITA Sarcasm")
    task7 = Task("itacola", "mcc,none", "ItaCoLA", scale_by_100=False)
    task8 = Task("news_sum", "bertscore,none", "News Sum")
    task16 = Task("sentipolc", "f1,none", "SENTIPOLC")
    task9 = Task("squad_it", "squad_f1,get-answer", "SQuAD it")
    task10 = Task("truthfulqa_mc2_ita", "acc,none", "TruthfulQA")
    task11 = Task("xcopa_it", "acc,none", "XCOPA")


NUM_FEWSHOT = 0  # Change with your few shot
# ---------------------------------------------------


# Your leaderboard name
TITLE = """<h1 align="center" id="space-title">ItaEval leaderboard</h1>"""

# What does your leaderboard evaluate?
INTRODUCTION_TEXT = """
This leaderboard evaluates language models on <b>ItaEval</b>, a new unified benchmark for Italian.

Some information:
- compared to other leaderboard you may found online, we do not support automatic evaluation for new model submissions
"""

ITA_EVAL_REPO = "https://github.com/g8a9/ita-eval"

# Which evaluations are you running? how can people reproduce what you have?
LLM_BENCHMARKS_TEXT = f"""
## How it works

## Reproducibility
To reproduce our results, head to {ITA_EVAL_REPO} for all the instructions.

If all the setup goes smoothly, you can run 'MODEL' on ItaEval with:
```bash
MODEL="..."
lm_eval -mixed_precision=bf16 --model hf \
    --model_args pretrained=$MODEL,dtype=bfloat16 \
    --tasks ita_eval \
    --device cuda:0 \
    --batch_size "auto" \
    --log_samples \
    --output_path $FAST/ita_eval_v1/$MODEL \
    --use_cache $FAST/ita_eval_v1/$MODEL \
    --cache_requests "true"
```
"""

EVALUATION_QUEUE_TEXT = """
We do not plan to accept autonomous submissions, yet. 
"""

CITATION_BUTTON_LABEL = "Copy the following snippet to cite these results"
CITATION_BUTTON_TEXT = r"""
We are working on it! :)
"""