Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
from src.display.formatting import model_hyperlink | |
from src.display.utils import AutoEvalColumn | |
# Models which have been flagged by users as being problematic for a reason or another | |
# (Model name to forum discussion link) | |
FLAGGED_MODELS = { | |
"Voicelab/trurl-2-13b": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/202", | |
"deepnight-research/llama-2-70B-inst": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/207", | |
"Aspik101/trurl-2-13b-pl-instruct_unload": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/213", | |
"Fredithefish/ReasonixPajama-3B-HF": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/236", | |
"TigerResearch/tigerbot-7b-sft-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/237", | |
"gaodrew/gaodrew-gorgonzola-13b": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/215", | |
"AIDC-ai-business/Marcoroni-70B": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/287", | |
"AIDC-ai-business/Marcoroni-13B": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/287", | |
"AIDC-ai-business/Marcoroni-7B": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/287", | |
"fblgit/una-xaberius-34b-v1beta": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/444", | |
"jan-hq/trinity-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"rwitz2/go-bruins-v2.1.1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"rwitz2/go-bruins-v2.1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"GreenNode/GreenNodeLM-v3olet-7B": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"GreenNode/GreenNodeLM-7B-v4leo": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"GreenNode/LeoScorpius-GreenNode-7B-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"viethq188/LeoScorpius-7B-Chat-DPO": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"GreenNode/GreenNodeLM-7B-v2leo": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"janai-hq/trinity-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"ignos/LeoScorpius-GreenNode-Alpaca-7B-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"fblgit/una-cybertron-7b-v3-OMA": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"mncai/mistral-7b-dpo-merge-v1.1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"mncai/mistral-7b-dpo-v6": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"Toten5/LeoScorpius-GreenNode-7B-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"GreenNode/GreenNodeLM-7B-v1olet": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"quantumaikr/quantum-dpo-v0.1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"quantumaikr/quantum-v0.01": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"quantumaikr/quantum-trinity-v0.1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"mncai/mistral-7b-dpo-v5": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"cookinai/BruinHermes": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"jan-ai/Pandora-10.7B-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"v1olet/v1olet_marcoroni-go-bruins-merge-7B": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"v1olet/v1olet_merged_dpo_7B_v3": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"rwitz2/pee": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
"zyh3826 / GML-Mistral-merged-v1": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/503", | |
"dillfrescott/trinity-medium": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/474", | |
} | |
# Models which have been requested by orgs to not be submitted on the leaderboard | |
DO_NOT_SUBMIT_MODELS = [ | |
"Voicelab/trurl-2-13b", # trained on MMLU | |
"TigerResearch/tigerbot-70b-chat", # per authors request | |
"TigerResearch/tigerbot-70b-chat-v2", # per authors request | |
"TigerResearch/tigerbot-70b-chat-v4-4k", # per authors request | |
] | |
def flag_models(leaderboard_data: list[dict]): | |
for model_data in leaderboard_data: | |
if model_data["model_name_for_query"] in FLAGGED_MODELS: | |
issue_num = FLAGGED_MODELS[model_data["model_name_for_query"]].split("/")[-1] | |
issue_link = model_hyperlink( | |
FLAGGED_MODELS[model_data["model_name_for_query"]], | |
f"See discussion #{issue_num}", | |
) | |
model_data[ | |
AutoEvalColumn.model.name | |
] = f"{model_data[AutoEvalColumn.model.name]} has been flagged! {issue_link}" | |
model_data[AutoEvalColumn.flagged.name] = True | |
else: | |
model_data[AutoEvalColumn.flagged.name] = False | |
def remove_forbidden_models(leaderboard_data: list[dict]): | |
indices_to_remove = [] | |
for ix, model in enumerate(leaderboard_data): | |
if model["model_name_for_query"] in DO_NOT_SUBMIT_MODELS: | |
indices_to_remove.append(ix) | |
for ix in reversed(indices_to_remove): | |
leaderboard_data.pop(ix) | |
return leaderboard_data | |
def filter_models(leaderboard_data: list[dict]): | |
leaderboard_data = remove_forbidden_models(leaderboard_data) | |
flag_models(leaderboard_data) | |