long-code-arena / src /get_results_for_task.py
saridormi's picture
Add metrics computation for CMG task
cdf268e
raw
history blame
967 Bytes
import pandas as pd # type: ignore[import]
def get_results_for_task_stub(task: str) -> pd.DataFrame:
stub_df = pd.DataFrame(
[
{
"Model Name": "GPT-4",
"Availability": "Proprietary",
"Context Size": "16k",
"BLEU": "X",
"ROUGE": "X",
"ChrF": "X",
"BERTScore": "X",
"BERTScore (Normalized)": "X",
"Submitted By": "🏟 Long Code Arena Team",
},
{
"Model Name": "CodeLlama-7b (instruct)",
"Availability": "Llama 2 license",
"Context Size": "16k",
"BLEU": "X",
"ROUGE": "X",
"ChrF": "X",
"BERTScore": "X",
"BERTScore (Normalized)": "X",
"Submitted By": "🏟 Long Code Arena Team",
},
]
)
return stub_df