from functools import partial
import json
import numpy as np
from datasets import load_dataset
import gradio as gr
from huggingface_hub import get_hf_file_metadata, HfApi, hf_hub_download, hf_hub_url
from huggingface_hub.repocard import metadata_load
import pandas as pd
from tqdm.autonotebook import tqdm
TASKS = [
"BitextMining",
"Classification",
"Clustering",
"PairClassification",
"Reranking",
"Retrieval",
"STS",
"Summarization",
]
TASK_LIST_BITEXT_MINING = ['BUCC (de-en)', 'BUCC (fr-en)', 'BUCC (ru-en)', 'BUCC (zh-en)', 'Tatoeba (afr-eng)', 'Tatoeba (amh-eng)', 'Tatoeba (ang-eng)', 'Tatoeba (ara-eng)', 'Tatoeba (arq-eng)', 'Tatoeba (arz-eng)', 'Tatoeba (ast-eng)', 'Tatoeba (awa-eng)', 'Tatoeba (aze-eng)', 'Tatoeba (bel-eng)', 'Tatoeba (ben-eng)', 'Tatoeba (ber-eng)', 'Tatoeba (bos-eng)', 'Tatoeba (bre-eng)', 'Tatoeba (bul-eng)', 'Tatoeba (cat-eng)', 'Tatoeba (cbk-eng)', 'Tatoeba (ceb-eng)', 'Tatoeba (ces-eng)', 'Tatoeba (cha-eng)', 'Tatoeba (cmn-eng)', 'Tatoeba (cor-eng)', 'Tatoeba (csb-eng)', 'Tatoeba (cym-eng)', 'Tatoeba (dan-eng)', 'Tatoeba (deu-eng)', 'Tatoeba (dsb-eng)', 'Tatoeba (dtp-eng)', 'Tatoeba (ell-eng)', 'Tatoeba (epo-eng)', 'Tatoeba (est-eng)', 'Tatoeba (eus-eng)', 'Tatoeba (fao-eng)', 'Tatoeba (fin-eng)', 'Tatoeba (fra-eng)', 'Tatoeba (fry-eng)', 'Tatoeba (gla-eng)', 'Tatoeba (gle-eng)', 'Tatoeba (glg-eng)', 'Tatoeba (gsw-eng)', 'Tatoeba (heb-eng)', 'Tatoeba (hin-eng)', 'Tatoeba (hrv-eng)', 'Tatoeba (hsb-eng)', 'Tatoeba (hun-eng)', 'Tatoeba (hye-eng)', 'Tatoeba (ido-eng)', 'Tatoeba (ile-eng)', 'Tatoeba (ina-eng)', 'Tatoeba (ind-eng)', 'Tatoeba (isl-eng)', 'Tatoeba (ita-eng)', 'Tatoeba (jav-eng)', 'Tatoeba (jpn-eng)', 'Tatoeba (kab-eng)', 'Tatoeba (kat-eng)', 'Tatoeba (kaz-eng)', 'Tatoeba (khm-eng)', 'Tatoeba (kor-eng)', 'Tatoeba (kur-eng)', 'Tatoeba (kzj-eng)', 'Tatoeba (lat-eng)', 'Tatoeba (lfn-eng)', 'Tatoeba (lit-eng)', 'Tatoeba (lvs-eng)', 'Tatoeba (mal-eng)', 'Tatoeba (mar-eng)', 'Tatoeba (max-eng)', 'Tatoeba (mhr-eng)', 'Tatoeba (mkd-eng)', 'Tatoeba (mon-eng)', 'Tatoeba (nds-eng)', 'Tatoeba (nld-eng)', 'Tatoeba (nno-eng)', 'Tatoeba (nob-eng)', 'Tatoeba (nov-eng)', 'Tatoeba (oci-eng)', 'Tatoeba (orv-eng)', 'Tatoeba (pam-eng)', 'Tatoeba (pes-eng)', 'Tatoeba (pms-eng)', 'Tatoeba (pol-eng)', 'Tatoeba (por-eng)', 'Tatoeba (ron-eng)', 'Tatoeba (rus-eng)', 'Tatoeba (slk-eng)', 'Tatoeba (slv-eng)', 'Tatoeba (spa-eng)', 'Tatoeba (sqi-eng)', 'Tatoeba (srp-eng)', 'Tatoeba (swe-eng)', 'Tatoeba (swg-eng)', 'Tatoeba (swh-eng)', 'Tatoeba (tam-eng)', 'Tatoeba (tat-eng)', 'Tatoeba (tel-eng)', 'Tatoeba (tgl-eng)', 'Tatoeba (tha-eng)', 'Tatoeba (tuk-eng)', 'Tatoeba (tur-eng)', 'Tatoeba (tzl-eng)', 'Tatoeba (uig-eng)', 'Tatoeba (ukr-eng)', 'Tatoeba (urd-eng)', 'Tatoeba (uzb-eng)', 'Tatoeba (vie-eng)', 'Tatoeba (war-eng)', 'Tatoeba (wuu-eng)', 'Tatoeba (xho-eng)', 'Tatoeba (yid-eng)', 'Tatoeba (yue-eng)', 'Tatoeba (zsm-eng)']
TASK_LIST_BITEXT_MINING_OTHER = ["BornholmBitextMining"]
TASK_LIST_CLASSIFICATION = [
"AmazonCounterfactualClassification (en)",
"AmazonPolarityClassification",
"AmazonReviewsClassification (en)",
"Banking77Classification",
"EmotionClassification",
"ImdbClassification",
"MassiveIntentClassification (en)",
"MassiveScenarioClassification (en)",
"MTOPDomainClassification (en)",
"MTOPIntentClassification (en)",
"ToxicConversationsClassification",
"TweetSentimentExtractionClassification",
]
TASK_LIST_CLASSIFICATION_NORM = [x.replace(" (en)", "") for x in TASK_LIST_CLASSIFICATION]
TASK_LIST_CLASSIFICATION_SV = [
"DalajClassification",
"MassiveIntentClassification (sv)",
"MassiveScenarioClassification (sv)",
"NordicLangClassification",
"ScalaSvClassification",
"SweRecClassification",
]
TASK_LIST_CLASSIFICATION_OTHER = ['AmazonCounterfactualClassification (de)', 'AmazonCounterfactualClassification (ja)', 'AmazonReviewsClassification (de)', 'AmazonReviewsClassification (es)', 'AmazonReviewsClassification (fr)', 'AmazonReviewsClassification (ja)', 'AmazonReviewsClassification (zh)', 'MTOPDomainClassification (de)', 'MTOPDomainClassification (es)', 'MTOPDomainClassification (fr)', 'MTOPDomainClassification (hi)', 'MTOPDomainClassification (th)', 'MTOPIntentClassification (de)', 'MTOPIntentClassification (es)', 'MTOPIntentClassification (fr)', 'MTOPIntentClassification (hi)', 'MTOPIntentClassification (th)', 'MassiveIntentClassification (af)', 'MassiveIntentClassification (am)', 'MassiveIntentClassification (ar)', 'MassiveIntentClassification (az)', 'MassiveIntentClassification (bn)', 'MassiveIntentClassification (cy)', 'MassiveIntentClassification (de)', 'MassiveIntentClassification (el)', 'MassiveIntentClassification (es)', 'MassiveIntentClassification (fa)', 'MassiveIntentClassification (fi)', 'MassiveIntentClassification (fr)', 'MassiveIntentClassification (he)', 'MassiveIntentClassification (hi)', 'MassiveIntentClassification (hu)', 'MassiveIntentClassification (hy)', 'MassiveIntentClassification (id)', 'MassiveIntentClassification (is)', 'MassiveIntentClassification (it)', 'MassiveIntentClassification (ja)', 'MassiveIntentClassification (jv)', 'MassiveIntentClassification (ka)', 'MassiveIntentClassification (km)', 'MassiveIntentClassification (kn)', 'MassiveIntentClassification (ko)', 'MassiveIntentClassification (lv)', 'MassiveIntentClassification (ml)', 'MassiveIntentClassification (mn)', 'MassiveIntentClassification (ms)', 'MassiveIntentClassification (my)', 'MassiveIntentClassification (nl)', 'MassiveIntentClassification (pt)', 'MassiveIntentClassification (ro)', 'MassiveIntentClassification (ru)', 'MassiveIntentClassification (sl)', 'MassiveIntentClassification (sq)', 'MassiveIntentClassification (sw)', 'MassiveIntentClassification (ta)', 'MassiveIntentClassification (te)', 'MassiveIntentClassification (th)', 'MassiveIntentClassification (tl)', 'MassiveIntentClassification (tr)', 'MassiveIntentClassification (ur)', 'MassiveIntentClassification (vi)', 'MassiveIntentClassification (zh-TW)', 'MassiveScenarioClassification (af)', 'MassiveScenarioClassification (am)', 'MassiveScenarioClassification (ar)', 'MassiveScenarioClassification (az)', 'MassiveScenarioClassification (bn)', 'MassiveScenarioClassification (cy)', 'MassiveScenarioClassification (de)', 'MassiveScenarioClassification (el)', 'MassiveScenarioClassification (es)', 'MassiveScenarioClassification (fa)', 'MassiveScenarioClassification (fi)', 'MassiveScenarioClassification (fr)', 'MassiveScenarioClassification (he)', 'MassiveScenarioClassification (hi)', 'MassiveScenarioClassification (hu)', 'MassiveScenarioClassification (hy)', 'MassiveScenarioClassification (id)', 'MassiveScenarioClassification (is)', 'MassiveScenarioClassification (it)', 'MassiveScenarioClassification (ja)', 'MassiveScenarioClassification (jv)', 'MassiveScenarioClassification (ka)', 'MassiveScenarioClassification (km)', 'MassiveScenarioClassification (kn)', 'MassiveScenarioClassification (ko)', 'MassiveScenarioClassification (lv)', 'MassiveScenarioClassification (ml)', 'MassiveScenarioClassification (mn)', 'MassiveScenarioClassification (ms)', 'MassiveScenarioClassification (my)', 'MassiveScenarioClassification (nl)', 'MassiveScenarioClassification (pt)', 'MassiveScenarioClassification (ro)', 'MassiveScenarioClassification (ru)', 'MassiveScenarioClassification (sl)', 'MassiveScenarioClassification (sq)', 'MassiveScenarioClassification (sw)', 'MassiveScenarioClassification (ta)', 'MassiveScenarioClassification (te)', 'MassiveScenarioClassification (th)', 'MassiveScenarioClassification (tl)', 'MassiveScenarioClassification (tr)', 'MassiveScenarioClassification (ur)', 'MassiveScenarioClassification (vi)', 'MassiveScenarioClassification (zh-TW)']
TASK_LIST_CLUSTERING = [
"ArxivClusteringP2P",
"ArxivClusteringS2S",
"BiorxivClusteringP2P",
"BiorxivClusteringS2S",
"MedrxivClusteringP2P",
"MedrxivClusteringS2S",
"RedditClustering",
"RedditClusteringP2P",
"StackExchangeClustering",
"StackExchangeClusteringP2P",
"TwentyNewsgroupsClustering",
]
TASK_LIST_PAIR_CLASSIFICATION = [
"SprintDuplicateQuestions",
"TwitterSemEval2015",
"TwitterURLCorpus",
]
TASK_LIST_RERANKING = [
"AskUbuntuDupQuestions",
"MindSmallReranking",
"SciDocsRR",
"StackOverflowDupQuestions",
]
TASK_LIST_RETRIEVAL = [
"ArguAna",
"ClimateFEVER",
"CQADupstackRetrieval",
"DBPedia",
"FEVER",
"FiQA2018",
"HotpotQA",
"MSMARCO",
"NFCorpus",
"NQ",
"QuoraRetrieval",
"SCIDOCS",
"SciFact",
"Touche2020",
"TRECCOVID",
]
TASK_LIST_RETRIEVAL_NORM = TASK_LIST_RETRIEVAL + [
"CQADupstackAndroidRetrieval",
"CQADupstackEnglishRetrieval",
"CQADupstackGamingRetrieval",
"CQADupstackGisRetrieval",
"CQADupstackMathematicaRetrieval",
"CQADupstackPhysicsRetrieval",
"CQADupstackProgrammersRetrieval",
"CQADupstackStatsRetrieval",
"CQADupstackTexRetrieval",
"CQADupstackUnixRetrieval",
"CQADupstackWebmastersRetrieval",
"CQADupstackWordpressRetrieval"
]
TASK_LIST_STS = [
"BIOSSES",
"SICK-R",
"STS12",
"STS13",
"STS14",
"STS15",
"STS16",
"STS17 (en-en)",
"STS22 (en)",
"STSBenchmark",
]
TASK_LIST_STS_OTHER = ["STS17 (ar-ar)", "STS17 (en-ar)", "STS17 (en-de)", "STS17 (en-tr)", "STS17 (es-en)", "STS17 (es-es)", "STS17 (fr-en)", "STS17 (it-en)", "STS17 (ko-ko)", "STS17 (nl-en)", "STS22 (ar)", "STS22 (de)", "STS22 (de-en)", "STS22 (de-fr)", "STS22 (de-pl)", "STS22 (es)", "STS22 (es-en)", "STS22 (es-it)", "STS22 (fr)", "STS22 (fr-pl)", "STS22 (it)", "STS22 (pl)", "STS22 (pl-en)", "STS22 (ru)", "STS22 (tr)", "STS22 (zh-en)", "STSBenchmark",]
TASK_LIST_STS_NORM = [x.replace(" (en)", "").replace(" (en-en)", "") for x in TASK_LIST_STS]
TASK_LIST_SUMMARIZATION = ["SummEval",]
TASK_LIST_EN = TASK_LIST_CLASSIFICATION + TASK_LIST_CLUSTERING + TASK_LIST_PAIR_CLASSIFICATION + TASK_LIST_RERANKING + TASK_LIST_RETRIEVAL + TASK_LIST_STS + TASK_LIST_SUMMARIZATION
TASK_TO_METRIC = {
"BitextMining": "f1",
"Clustering": "v_measure",
"Classification": "accuracy",
"PairClassification": "cos_sim_ap",
"Reranking": "map",
"Retrieval": "ndcg_at_10",
"STS": "cos_sim_spearman",
"Summarization": "cos_sim_spearman",
}
def make_clickable_model(model_name, link=None):
if link is None:
link = "https://huggingface.co/" + model_name
# Remove user from model name
return (
f'{model_name.split("/")[-1]}'
)
# Models without metadata, thus we cannot fetch their results naturally
EXTERNAL_MODELS = [
"all-MiniLM-L12-v2",
"all-MiniLM-L6-v2",
"all-mpnet-base-v2",
"allenai-specter",
"bert-base-uncased",
"contriever-base-msmarco",
"dfm-encoder-large-v1",
"dfm-sentence-encoder-large-1",
"distiluse-base-multilingual-cased-v2",
"e5-base",
"e5-large",
"e5-small",
"gbert-base",
"gbert-large",
"gelectra-base",
"gelectra-large",
"gottbert-base",
"glove.6B.300d",
"gtr-t5-base",
"gtr-t5-large",
"gtr-t5-xl",
"gtr-t5-xxl",
"herbert-base-retrieval-v2",
"komninos",
"luotuo-bert-medium",
"LASER2",
"LaBSE",
"m3e-base",
"m3e-large",
"msmarco-bert-co-condensor",
"multilingual-e5-base",
"multilingual-e5-large",
"multilingual-e5-small",
"nb-bert-base",
"nb-bert-large",
"norbert3-base",
"norbert3-large",
"paraphrase-multilingual-MiniLM-L12-v2",
"paraphrase-multilingual-mpnet-base-v2",
"sentence-t5-base",
"sentence-t5-large",
"sentence-t5-xl",
"sentence-t5-xxl",
"sup-simcse-bert-base-uncased",
"text-embedding-3-small",
"text-embedding-3-large",
"text-embedding-3-large-256",
"titan-embed-text-v1",
"unsup-simcse-bert-base-uncased",
"use-cmlm-multilingual",
"voyage-lite-01-instruct",
"voyage-lite-02-instruct",
"xlm-roberta-base",
"xlm-roberta-large",
]
EXTERNAL_MODEL_TO_LINK = {
"allenai-specter": "https://huggingface.co/sentence-transformers/allenai-specter",
"allenai-specter": "https://huggingface.co/sentence-transformers/allenai-specter",
"all-MiniLM-L12-v2": "https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2",
"all-MiniLM-L6-v2": "https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2",
"all-mpnet-base-v2": "https://huggingface.co/sentence-transformers/all-mpnet-base-v2",
"bert-base-uncased": "https://huggingface.co/bert-base-uncased",
"contriever-base-msmarco": "https://huggingface.co/nthakur/contriever-base-msmarco",
"distiluse-base-multilingual-cased-v2": "https://huggingface.co/sentence-transformers/distiluse-base-multilingual-cased-v2",
"dfm-encoder-large-v1": "https://huggingface.co/chcaa/dfm-encoder-large-v1",
"dfm-sentence-encoder-large-1": "https://huggingface.co/chcaa/dfm-encoder-large-v1",
"e5-base": "https://huggingface.co/intfloat/e5-base",
"e5-large": "https://huggingface.co/intfloat/e5-large",
"e5-small": "https://huggingface.co/intfloat/e5-small",
"gbert-base": "https://huggingface.co/deepset/gbert-base",
"gbert-large": "https://huggingface.co/deepset/gbert-large",
"gelectra-base": "https://huggingface.co/deepset/gelectra-base",
"gelectra-large": "https://huggingface.co/deepset/gelectra-large",
"glove.6B.300d": "https://huggingface.co/sentence-transformers/average_word_embeddings_glove.6B.300d",
"gottbert-base": "https://huggingface.co/uklfr/gottbert-base",
"gtr-t5-base": "https://huggingface.co/sentence-transformers/gtr-t5-base",
"gtr-t5-large": "https://huggingface.co/sentence-transformers/gtr-t5-large",
"gtr-t5-xl": "https://huggingface.co/sentence-transformers/gtr-t5-xl",
"gtr-t5-xxl": "https://huggingface.co/sentence-transformers/gtr-t5-xxl",
"herbert-base-retrieval-v2": "https://huggingface.co/ipipan/herbert-base-retrieval-v2",
"komninos": "https://huggingface.co/sentence-transformers/average_word_embeddings_komninos",
"luotuo-bert-medium": "https://huggingface.co/silk-road/luotuo-bert-medium",
"LASER2": "https://github.com/facebookresearch/LASER",
"LaBSE": "https://huggingface.co/sentence-transformers/LaBSE",
"m3e-base": "https://huggingface.co/moka-ai/m3e-base",
"m3e-large": "https://huggingface.co/moka-ai/m3e-large",
"msmarco-bert-co-condensor": "https://huggingface.co/sentence-transformers/msmarco-bert-co-condensor",
"multilingual-e5-base": "https://huggingface.co/intfloat/multilingual-e5-base",
"multilingual-e5-large": "https://huggingface.co/intfloat/multilingual-e5-large",
"multilingual-e5-small": "https://huggingface.co/intfloat/multilingual-e5-small",
"nb-bert-base": "https://huggingface.co/NbAiLab/nb-bert-base",
"nb-bert-large": "https://huggingface.co/NbAiLab/nb-bert-large",
"norbert3-base": "https://huggingface.co/ltg/norbert3-base",
"norbert3-large": "https://huggingface.co/ltg/norbert3-large",
"paraphrase-multilingual-mpnet-base-v2": "https://huggingface.co/sentence-transformers/paraphrase-multilingual-mpnet-base-v2",
"paraphrase-multilingual-MiniLM-L12-v2": "https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
"sentence-t5-base": "https://huggingface.co/sentence-transformers/sentence-t5-base",
"sentence-t5-large": "https://huggingface.co/sentence-transformers/sentence-t5-large",
"sentence-t5-xl": "https://huggingface.co/sentence-transformers/sentence-t5-xl",
"sentence-t5-xxl": "https://huggingface.co/sentence-transformers/sentence-t5-xxl",
"sup-simcse-bert-base-uncased": "https://huggingface.co/princeton-nlp/sup-simcse-bert-base-uncased",
"text-embedding-3-small": "https://openai.com/blog/new-embedding-models-and-api-updates",
"text-embedding-3-large": "https://openai.com/blog/new-embedding-models-and-api-updates",
"text-embedding-3-large-256": "https://openai.com/blog/new-embedding-models-and-api-updates",
"titan-embed-text-v1": "https://docs.aws.amazon.com/bedrock/latest/userguide/embeddings.html",
"unsup-simcse-bert-base-uncased": "https://huggingface.co/princeton-nlp/unsup-simcse-bert-base-uncased",
"use-cmlm-multilingual": "https://huggingface.co/sentence-transformers/use-cmlm-multilingual",
"voyage-lite-01-instruct": "https://docs.voyageai.com/embeddings/",
"voyage-lite-02-instruct": "https://docs.voyageai.com/embeddings/",
"xlm-roberta-base": "https://huggingface.co/xlm-roberta-base",
"xlm-roberta-large": "https://huggingface.co/xlm-roberta-large",
}
EXTERNAL_MODEL_TO_DIM = {
"all-MiniLM-L12-v2": 384,
"all-MiniLM-L6-v2": 384,
"all-mpnet-base-v2": 768,
"allenai-specter": 768,
"bert-base-uncased": 768,
"contriever-base-msmarco": 768,
"distiluse-base-multilingual-cased-v2": 512,
"dfm-encoder-large-v1": 1024,
"dfm-sentence-encoder-large-1": 1024,
"e5-base": 768,
"e5-small": 384,
"e5-large": 1024,
"luotuo-bert-medium": 768,
"LASER2": 1024,
"LaBSE": 768,
"gbert-base": 768,
"gbert-large": 1024,
"gelectra-base": 768,
"gelectra-large": 1024,
"glove.6B.300d": 300,
"gottbert-base": 768,
"gtr-t5-base": 768,
"gtr-t5-large": 768,
"gtr-t5-xl": 768,
"gtr-t5-xxl": 768,
"herbert-base-retrieval-v2": 768,
"komninos": 300,
"m3e-base": 768,
"m3e-large": 768,
"msmarco-bert-co-condensor": 768,
"multilingual-e5-base": 768,
"multilingual-e5-small": 384,
"multilingual-e5-large": 1024,
"nb-bert-base": 768,
"nb-bert-large": 1024,
"norbert3-base": 768,
"norbert3-large": 1024,
"paraphrase-multilingual-MiniLM-L12-v2": 384,
"paraphrase-multilingual-mpnet-base-v2": 768,
"sentence-t5-base": 768,
"sentence-t5-large": 768,
"sentence-t5-xl": 768,
"sentence-t5-xxl": 768,
"sup-simcse-bert-base-uncased": 768,
"text-embedding-3-large": 3072,
"text-embedding-3-large-256": 256,
"text-embedding-3-small": 1536,
"titan-embed-text-v1": 1536,
"unsup-simcse-bert-base-uncased": 768,
"use-cmlm-multilingual": 768,
"voyage-lite-01-instruct": 1024,
"voyage-lite-02-instruct": 1024,
"xlm-roberta-base": 768,
"xlm-roberta-large": 1024,
}
EXTERNAL_MODEL_TO_SEQLEN = {
"all-MiniLM-L12-v2": 512,
"all-MiniLM-L6-v2": 512,
"all-mpnet-base-v2": 514,
"allenai-specter": 512,
"bert-base-uncased": 512,
"contriever-base-msmarco": 512,
"dfm-encoder-large-v1": 512,
"dfm-sentence-encoder-large-1": 512,
"distiluse-base-multilingual-cased-v2": 512,
"e5-base": 512,
"e5-large": 512,
"e5-small": 512,
"gbert-base": 512,
"gbert-large": 512,
"gelectra-base": 512,
"gelectra-large": 512,
"gottbert-base": 512,
"glove.6B.300d": "N/A",
"gtr-t5-base": 512,
"gtr-t5-large": 512,
"gtr-t5-xl": 512,
"gtr-t5-xxl": 512,
"herbert-base-retrieval-v2": 514,
"komninos": "N/A",
"luotuo-bert-medium": 512,
"LASER2": "N/A",
"LaBSE": 512,
"m3e-base": 512,
"m3e-large": 512,
"msmarco-bert-co-condensor": 512,
"multilingual-e5-base": 514,
"multilingual-e5-large": 514,
"multilingual-e5-small": 512,
"nb-bert-base": 512,
"nb-bert-large": 512,
"norbert3-base": 512,
"norbert3-large": 512,
"paraphrase-multilingual-MiniLM-L12-v2": 512,
"paraphrase-multilingual-mpnet-base-v2": 514,
"sentence-t5-base": 512,
"sentence-t5-large": 512,
"sentence-t5-xl": 512,
"sentence-t5-xxl": 512,
"sup-simcse-bert-base-uncased": 512,
"text-embedding-3-large": 8191,
"text-embedding-3-large-256": 8191,
"text-embedding-3-small": 8191,
"titan-embed-text-v1": 8000,
"use-cmlm-multilingual": 512,
"unsup-simcse-bert-base-uncased": 512,
"voyage-lite-01-instruct": 4000,
"voyage-lite-02-instruct": 4000,
"xlm-roberta-base": 514,
"xlm-roberta-large": 514,
}
EXTERNAL_MODEL_TO_SIZE = {
"allenai-specter": 0.44,
"all-MiniLM-L12-v2": 0.13,
"all-MiniLM-L6-v2": 0.09,
"all-mpnet-base-v2": 0.44,
"bert-base-uncased": 0.44,
"contriever-base-msmarco": 0.44,
"distiluse-base-multilingual-cased-v2": 0.54,
"dfm-encoder-large-v1": 1.42,
"dfm-sentence-encoder-large-1": 1.63,
"e5-base": 0.44,
"e5-small": 0.13,
"e5-large": 1.34,
"gbert-base": 0.44,
"gbert-large": 1.35,
"gelectra-base": 0.44,
"gelectra-large": 1.34,
"glove.6B.300d": 0.48,
"gottbert-base": 0.51,
"gtr-t5-base": 0.22,
"gtr-t5-large": 0.67,
"gtr-t5-xl": 2.48,
"gtr-t5-xxl": 9.73,
"herbert-base-retrieval-v2": 0.50,
"komninos": 0.27,
"luotuo-bert-medium": 1.31,
"LASER2": 0.17,
"LaBSE": 1.88,
"m3e-base": 0.41,
"m3e-large": 0.41,
"msmarco-bert-co-condensor": 0.44,
"multilingual-e5-base": 1.11,
"multilingual-e5-small": 0.47,
"multilingual-e5-large": 2.24,
"nb-bert-base": 0.71,
"nb-bert-large": 1.42,
"norbert3-base": 0.52,
"norbert3-large": 1.47,
"paraphrase-multilingual-mpnet-base-v2": 1.11,
"paraphrase-multilingual-MiniLM-L12-v2": 0.47,
"sentence-t5-base": 0.22,
"sentence-t5-large": 0.67,
"sentence-t5-xl": 2.48,
"sentence-t5-xxl": 9.73,
"sup-simcse-bert-base-uncased": 0.44,
"unsup-simcse-bert-base-uncased": 0.44,
"use-cmlm-multilingual": 1.89,
"xlm-roberta-base": 1.12,
"xlm-roberta-large": 2.24,
}
MODELS_TO_SKIP = {
"baseplate/instructor-large-1", # Duplicate
"radames/e5-large", # Duplicate
"gentlebowl/instructor-large-safetensors", # Duplicate
"Consensus/instructor-base", # Duplicate
"GovCompete/instructor-xl", # Duplicate
"GovCompete/e5-large-v2", # Duplicate
"t12e/instructor-base", # Duplicate
"michaelfeil/ct2fast-e5-large-v2",
"michaelfeil/ct2fast-e5-large",
"michaelfeil/ct2fast-e5-small-v2",
"newsrx/instructor-xl-newsrx",
"newsrx/instructor-large-newsrx",
"fresha/e5-large-v2-endpoint",
"ggrn/e5-small-v2",
"michaelfeil/ct2fast-e5-small",
"jncraton/e5-small-v2-ct2-int8",
"anttip/ct2fast-e5-small-v2-hfie",
"newsrx/instructor-large",
"newsrx/instructor-xl",
"dmlls/all-mpnet-base-v2",
"cgldo/semanticClone",
"Malmuk1/e5-large-v2_Sharded",
"jncraton/gte-small-ct2-int8",
"Einas/einas_ashkar",
"gruber/e5-small-v2-ggml",
"jncraton/bge-small-en-ct2-int8",
"vectoriseai/bge-small-en",
"recipe/embeddings",
"dhairya0907/thenlper-get-large",
"Narsil/bge-base-en",
"kozistr/fused-large-en",
"sionic-ai/sionic-ai-v2", # Wait for https://huggingface.co/sionic-ai/sionic-ai-v2/discussions/1
"sionic-ai/sionic-ai-v1", # Wait for https://huggingface.co/sionic-ai/sionic-ai-v2/discussions/1
"BAAI/bge-large-en", # Deprecated in favor of v1.5
"BAAI/bge-base-en", # Deprecated in favor of v1.5
"BAAI/bge-small-en", # Deprecated in favor of v1.5
"d0rj/e5-large-en-ru",
"d0rj/e5-base-en-ru",
"d0rj/e5-small-en-ru",
"aident-ai/bge-base-en-onnx",
"barisaydin/bge-base-en",
"barisaydin/gte-large",
"barisaydin/gte-base",
"barisaydin/gte-small",
"barisaydin/bge-small-en",
"odunola/e5-base-v2",
"goldenrooster/multilingual-e5-large",
"davidpeer/gte-small",
"barisaydin/bge-large-en",
"jamesgpt1/english-large-v1",
"vectoriseai/bge-large-en-v1.5",
"vectoriseai/bge-base-en-v1.5",
"vectoriseai/instructor-large",
"vectoriseai/instructor-base",
"vectoriseai/gte-large",
"vectoriseai/gte-base",
"vectoriseai/e5-large-v2",
"vectoriseai/bge-small-en-v1.5",
"vectoriseai/e5-base-v2",
"vectoriseai/e5-large",
"vectoriseai/multilingual-e5-large",
"vectoriseai/gte-small",
"vectoriseai/ember-v1",
"vectoriseai/e5-base",
"vectoriseai/e5-small-v2",
"michaelfeil/ct2fast-bge-large-en-v1.5",
"michaelfeil/ct2fast-bge-large-en-v1.5",
"michaelfeil/ct2fast-bge-base-en-v1.5",
"michaelfeil/ct2fast-gte-large",
"michaelfeil/ct2fast-gte-base",
"michaelfeil/ct2fast-bge-small-en-v1.5",
"rizki/bgr-tf",
"ef-zulla/e5-multi-sml-torch",
"cherubhao/yogamodel",
"morgendigital/multilingual-e5-large-quantized",
"jncraton/gte-tiny-ct2-int8",
"Research2NLP/electrical_stella",
"Intel/bge-base-en-v1.5-sts-int8-static",
"Intel/bge-base-en-v1.5-sts-int8-dynamic",
"Intel/bge-base-en-v1.5-sst2",
"Intel/bge-base-en-v1.5-sst2-int8-static",
"Intel/bge-base-en-v1.5-sst2-int8-dynamic",
"Intel/bge-small-en-v1.5-sst2",
"Intel/bge-small-en-v1.5-sst2-int8-dynamic",
"Intel/bge-small-en-v1.5-sst2-int8-static",
"binqiangliu/EmbeddingModlebgelargeENv1.5",
"DecisionOptimizationSystem/DeepFeatEmbeddingLargeContext",
"woody72/multilingual-e5-base",
"Severian/embed",
"Frazic/udever-bloom-3b-sentence",
"jamesgpt1/zzz",
"karrar-alwaili/UAE-Large-V1",
"odunola/UAE-Large-VI",
"shubham-bgi/UAE-Large",
"retrainai/instructor-xl",
"weakit-v/bge-base-en-v1.5-onnx",
"ieasybooks/multilingual-e5-large-onnx",
"gizmo-ai/Cohere-embed-multilingual-v3.0",
"jingyeom/korean_embedding_model",
"barisaydin/text2vec-base-multilingual",
"mlx-community/multilingual-e5-large-mlx",
"mlx-community/multilingual-e5-base-mlx",
"mlx-community/multilingual-e5-small-mlx",
"maiyad/multilingual-e5-small",
"khoa-klaytn/bge-base-en-v1.5-angle",
"khoa-klaytn/bge-small-en-v1.5-angle",
"mixamrepijey/instructor-small",
"mixamrepijey/instructor-models",
"lsf1000/bge-evaluation", # Empty
}
EXTERNAL_MODEL_RESULTS = {model: {k: {v: []} for k, v in TASK_TO_METRIC.items()} for model in EXTERNAL_MODELS}
def add_lang(examples):
if not(examples["eval_language"]):
examples["mteb_dataset_name_with_lang"] = examples["mteb_dataset_name"]
else:
examples["mteb_dataset_name_with_lang"] = examples["mteb_dataset_name"] + f' ({examples["eval_language"]})'
return examples
def add_task(examples):
# Could be added to the dataset loading script instead
if examples["mteb_dataset_name"] in TASK_LIST_CLASSIFICATION_NORM + TASK_LIST_CLASSIFICATION_SV:
examples["mteb_task"] = "Classification"
elif examples["mteb_dataset_name"] in TASK_LIST_CLUSTERING:
examples["mteb_task"] = "Clustering"
elif examples["mteb_dataset_name"] in TASK_LIST_PAIR_CLASSIFICATION:
examples["mteb_task"] = "PairClassification"
elif examples["mteb_dataset_name"] in TASK_LIST_RERANKING:
examples["mteb_task"] = "Reranking"
elif examples["mteb_dataset_name"] in TASK_LIST_RETRIEVAL_NORM:
examples["mteb_task"] = "Retrieval"
elif examples["mteb_dataset_name"] in TASK_LIST_STS_NORM:
examples["mteb_task"] = "STS"
elif examples["mteb_dataset_name"] in TASK_LIST_SUMMARIZATION:
examples["mteb_task"] = "Summarization"
elif examples["mteb_dataset_name"] in [x.split(" ")[0] for x in TASK_LIST_BITEXT_MINING + TASK_LIST_BITEXT_MINING_OTHER]:
examples["mteb_task"] = "BitextMining"
else:
print("WARNING: Task not found for dataset", examples["mteb_dataset_name"])
examples["mteb_task"] = "Unknown"
return examples
pbar = tqdm(EXTERNAL_MODELS, desc="Fetching external model results")
for model in pbar:
pbar.set_description(f"Fetching external model results for {model!r}")
ds = load_dataset("mteb/results", model, trust_remote_code=True)
# For local debugging:
#, download_mode='force_redownload', verification_mode="no_checks")
ds = ds.map(add_lang)
ds = ds.map(add_task)
base_dict = {"Model": make_clickable_model(model, link=EXTERNAL_MODEL_TO_LINK.get(model, "https://huggingface.co/spaces/mteb/leaderboard"))}
# For now only one metric per task - Could add more metrics lateron
for task, metric in TASK_TO_METRIC.items():
ds_dict = ds.filter(lambda x: (x["mteb_task"] == task) and (x["metric"] == metric))["test"].to_dict()
ds_dict = {k: round(v, 2) for k, v in zip(ds_dict["mteb_dataset_name_with_lang"], ds_dict["score"])}
EXTERNAL_MODEL_RESULTS[model][task][metric].append({**base_dict, **ds_dict})
def get_dim_seq_size(model):
filenames = [sib.rfilename for sib in model.siblings]
dim, seq, size = "", "", ""
if "1_Pooling/config.json" in filenames:
st_config_path = hf_hub_download(model.modelId, filename="1_Pooling/config.json")
dim = json.load(open(st_config_path)).get("word_embedding_dimension", "")
elif "2_Pooling/config.json" in filenames:
st_config_path = hf_hub_download(model.modelId, filename="2_Pooling/config.json")
dim = json.load(open(st_config_path)).get("word_embedding_dimension", "")
if "config.json" in filenames:
config_path = hf_hub_download(model.modelId, filename="config.json")
config = json.load(open(config_path))
if not dim:
dim = config.get("hidden_dim", config.get("hidden_size", config.get("d_model", "")))
seq = config.get("n_positions", config.get("max_position_embeddings", config.get("n_ctx", config.get("seq_length", ""))))
# Get model file size without downloading
if "pytorch_model.bin" in filenames:
url = hf_hub_url(model.modelId, filename="pytorch_model.bin")
meta = get_hf_file_metadata(url)
size = round(meta.size / 1e9, 2)
elif "pytorch_model.bin.index.json" in filenames:
index_path = hf_hub_download(model.modelId, filename="pytorch_model.bin.index.json")
"""
{
"metadata": {
"total_size": 28272820224
},....
"""
size = json.load(open(index_path))
if ("metadata" in size) and ("total_size" in size["metadata"]):
size = round(size["metadata"]["total_size"] / 1e9, 2)
elif "model.safetensors" in filenames:
url = hf_hub_url(model.modelId, filename="model.safetensors")
meta = get_hf_file_metadata(url)
size = round(meta.size / 1e9, 2)
elif "model.safetensors.index.json" in filenames:
index_path = hf_hub_download(model.modelId, filename="model.safetensors.index.json")
"""
{
"metadata": {
"total_size": 14483464192
},....
"""
size = json.load(open(index_path))
if ("metadata" in size) and ("total_size" in size["metadata"]):
size = round(size["metadata"]["total_size"] / 1e9, 2)
return dim, seq, size
def make_datasets_clickable(df):
"""Does not work"""
if "BornholmBitextMining" in df.columns:
link = "https://huggingface.co/datasets/strombergnlp/bornholmsk_parallel"
df = df.rename(
columns={f'BornholmBitextMining': 'BornholmBitextMining',})
return df
def add_rank(df):
cols_to_rank = [col for col in df.columns if col not in ["Model", "Model Size (GB)", "Embedding Dimensions", "Max Tokens"]]
if len(cols_to_rank) == 1:
df.sort_values(cols_to_rank[0], ascending=False, inplace=True)
else:
df.insert(1, "Average", df[cols_to_rank].mean(axis=1, skipna=False))
df.sort_values("Average", ascending=False, inplace=True)
df.insert(0, "Rank", list(range(1, len(df) + 1)))
df = df.round(2)
# Fill NaN after averaging
df.fillna("", inplace=True)
return df
def get_mteb_data(tasks=["Clustering"], langs=[], datasets=[], fillna=True, add_emb_dim=False, task_to_metric=TASK_TO_METRIC, rank=True):
api = HfApi()
models = api.list_models(filter="mteb")
# Initialize list to models that we cannot fetch metadata from
df_list = []
for model in EXTERNAL_MODEL_RESULTS:
results_list = [res for task in tasks for res in EXTERNAL_MODEL_RESULTS[model][task][task_to_metric[task]]]
if len(datasets) > 0:
res = {k: v for d in results_list for k, v in d.items() if (k == "Model") or any([x in k for x in datasets])}
elif langs:
# Would be cleaner to rely on an extra language column instead
langs_format = [f"({lang})" for lang in langs]
res = {k: v for d in results_list for k, v in d.items() if any([k.split(" ")[-1] in (k, x) for x in langs_format])}
else:
res = {k: v for d in results_list for k, v in d.items()}
# Model & at least one result
if len(res) > 1:
if add_emb_dim:
res["Model Size (GB)"] = EXTERNAL_MODEL_TO_SIZE.get(model, "")
res["Embedding Dimensions"] = EXTERNAL_MODEL_TO_DIM.get(model, "")
res["Max Tokens"] = EXTERNAL_MODEL_TO_SEQLEN.get(model, "")
df_list.append(res)
for model in models:
if model.modelId in MODELS_TO_SKIP:
continue
print("MODEL", model)
readme_path = hf_hub_download(model.modelId, filename="README.md")
meta = metadata_load(readme_path)
if "model-index" not in meta:
continue
if len(datasets) > 0:
task_results = [sub_res for sub_res in meta["model-index"][0]["results"] if (sub_res.get("task", {}).get("type", "") in tasks) and any([x in sub_res.get("dataset", {}).get("name", "") for x in datasets])]
elif langs:
task_results = [sub_res for sub_res in meta["model-index"][0]["results"] if (sub_res.get("task", {}).get("type", "") in tasks) and (sub_res.get("dataset", {}).get("config", "default") in ("default", *langs))]
else:
task_results = [sub_res for sub_res in meta["model-index"][0]["results"] if (sub_res.get("task", {}).get("type", "") in tasks)]
out = [{res["dataset"]["name"].replace("MTEB ", ""): [round(score["value"], 2) for score in res["metrics"] if score["type"] == task_to_metric.get(res["task"]["type"])][0]} for res in task_results]
out = {k: v for d in out for k, v in d.items()}
out["Model"] = make_clickable_model(model.modelId)
# Model & at least one result
if len(out) > 1:
if add_emb_dim:
try:
# Fails on gated repos, so we only include scores for them
out["Embedding Dimensions"], out["Max Tokens"], out["Model Size (GB)"] = get_dim_seq_size(model)
except:
pass
df_list.append(out)
df = pd.DataFrame(df_list)
# If there are any models that are the same, merge them
# E.g. if out["Model"] has the same value in two places, merge & take whichever one is not NaN else just take the first one
df = df.groupby("Model", as_index=False).first()
# Put 'Model' column first
cols = sorted(list(df.columns))
cols.insert(0, cols.pop(cols.index("Model")))
df = df[cols]
if rank:
df = add_rank(df)
if fillna:
df.fillna("", inplace=True)
return df
def get_mteb_average():
global DATA_OVERALL
DATA_OVERALL = get_mteb_data(
tasks=[
"Classification",
"Clustering",
"PairClassification",
"Reranking",
"Retrieval",
"STS",
"Summarization",
],
datasets=TASK_LIST_CLASSIFICATION + TASK_LIST_CLUSTERING + TASK_LIST_PAIR_CLASSIFICATION + TASK_LIST_RERANKING + TASK_LIST_RETRIEVAL + TASK_LIST_STS + TASK_LIST_SUMMARIZATION,
fillna=False,
add_emb_dim=True,
rank=False,
)
# Debugging:
# DATA_OVERALL.to_csv("overall.csv")
DATA_OVERALL.insert(1, "Average", DATA_OVERALL[TASK_LIST_EN].mean(axis=1, skipna=False))
DATA_OVERALL.insert(2, "Classification Average", DATA_OVERALL[TASK_LIST_CLASSIFICATION].mean(axis=1, skipna=False))
DATA_OVERALL.insert(3, "Clustering Average", DATA_OVERALL[TASK_LIST_CLUSTERING].mean(axis=1, skipna=False))
DATA_OVERALL.insert(4, "Pair Classification Average", DATA_OVERALL[TASK_LIST_PAIR_CLASSIFICATION].mean(axis=1, skipna=False))
DATA_OVERALL.insert(5, "Reranking Average", DATA_OVERALL[TASK_LIST_RERANKING].mean(axis=1, skipna=False))
DATA_OVERALL.insert(6, "Retrieval Average", DATA_OVERALL[TASK_LIST_RETRIEVAL].mean(axis=1, skipna=False))
DATA_OVERALL.insert(7, "STS Average", DATA_OVERALL[TASK_LIST_STS].mean(axis=1, skipna=False))
DATA_OVERALL.insert(8, "Summarization Average", DATA_OVERALL[TASK_LIST_SUMMARIZATION].mean(axis=1, skipna=False))
DATA_OVERALL.sort_values("Average", ascending=False, inplace=True)
# Start ranking from 1
DATA_OVERALL.insert(0, "Rank", list(range(1, len(DATA_OVERALL) + 1)))
# For columns you know are numeric:
numeric_columns = DATA_OVERALL.select_dtypes(include=['float64', 'int64']).columns
DATA_OVERALL[numeric_columns] = DATA_OVERALL[numeric_columns].fillna(np.nan) # Or a placeholder like 0 or -1
# For columns expected to be strings, filling with an empty string is fine:
string_columns = DATA_OVERALL.select_dtypes(include=['object']).columns
DATA_OVERALL[string_columns] = DATA_OVERALL[string_columns].fillna('')
DATA_OVERALL = DATA_OVERALL.round(2)
# Fill NaN after averaging
DATA_OVERALL.fillna("", inplace=True)
DATA_OVERALL = DATA_OVERALL[["Rank", "Model", "Model Size (GB)", "Embedding Dimensions", "Max Tokens", "Average",
"Classification Average", "Clustering Average", "Pair Classification Average", "Reranking Average", "Retrieval Average",
"STS Average", "Summarization Average"]]
DATA_OVERALL = DATA_OVERALL[DATA_OVERALL.iloc[:, 5:].ne("").any(axis=1)]
return DATA_OVERALL
DATA_OVERALL=get_mteb_average()
import unicodedata
def is_valid_unicode(char):
try:
unicodedata.name(char)
return True # Valid Unicode character
except ValueError:
return False # Invalid Unicode character
def remove_invalid_unicode(input_string):
if isinstance(input_string, str):
valid_chars = [char for char in input_string if is_valid_unicode(char)]
return ''.join(valid_chars)
else:
return input_string # Return non-string values as is
from dataclasses import dataclass
@dataclass
class LeaderboardColumn:
name: str
type: str
DATA_OVERALL_COLUMN_TO_DATATYPE = [
LeaderboardColumn("Rank", "number"),
LeaderboardColumn("Model Size (GB)", "number"),
LeaderboardColumn("Embedding Dimensions", "number"),
LeaderboardColumn("Max Tokens", "number"),
LeaderboardColumn("Average", "number"),
LeaderboardColumn("Classification Average", "number"),
LeaderboardColumn("Clustering Average", "number"),
LeaderboardColumn("Pair Classification Average", "number"),
LeaderboardColumn("Reranking Average", "number"),
LeaderboardColumn("Retrieval Average", "number"),
LeaderboardColumn("STS Average", "number"),
LeaderboardColumn("Summarization Average", "number")
]
COLS = [col.name for col in DATA_OVERALL_COLUMN_TO_DATATYPE]
TYPES = [col.type for col in DATA_OVERALL_COLUMN_TO_DATATYPE]
data_overall = gr.components.Dataframe(
DATA_OVERALL,
headers=COLS,
datatype=TYPES,
visible=False,
line_breaks=False,
interactive=False
)
print(data_overall)
def display(x, y):
global data_overall # Ensure we're accessing the global variable
# Check if leaderboard_table is indeed a DataFrame
if isinstance(data_overall, pd.DataFrame):
for column in data_overall.columns:
if data_overall[column].dtype == 'object':
data_overall[column] = data_overall[column].apply(remove_invalid_unicode)
# Ensure COLS is defined and valid; it's used to filter columns in the DataFrame
subset_df = data_overall[COLS] # Assuming COLS is a list of column names you want to keep
return subset_df
else:
print("leaderboard_table is not a DataFrame.")
dummy1 = gr.Textbox(visible=False)
INTRODUCTION_TEXT = """
This is a copied space from LLM Trustworthy Leaderboard. Instead of displaying
the results as table this space was modified to simply provides a gradio API interface.
Using the following python script below, users can access the full leaderboard data easily.
Python on how to access the data:
```python
# Import dependencies
from gradio_client import Client
# Initialize the Gradio client with the API URL
client = Client("https://rodrigomasini-data-only-llm-perf-leaderboard.hf.space/")
try:
# Perform the API call
response = client.predict("","", api_name='/predict')
# Check if response it's directly accessible
if len(response) > 0:
print("Response received!")
headers = response.get('headers', [])
data = response.get('data', [])
print(headers)
# Remove commenst if you want to download the dataset and save in csv format
# Specify the path to your CSV file
#csv_file_path = 'llm-perf-benchmark.csv'
# Open the CSV file for writing
#with open(csv_file_path, mode='w', newline='', encoding='utf-8') as file:
# writer = csv.writer(file)
# Write the headers
# writer.writerow(headers)
# Write the data
# for row in data:
# writer.writerow(row)
#print(f"Results saved to {csv_file_path}")
# If the above line prints a string that looks like JSON, you can parse it with json.loads(response)
# Otherwise, you might need to adjust based on the actual structure of `response`
except Exception as e:
print(f"An error occurred: {e}")
```
"""
interface = gr.Interface(
fn=display,
inputs=[gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text"), dummy1],
outputs=[data_overall]
)
interface.launch()