PartAI/Tooka-SBERT

`PartAI/Tooka-SBERT` powered by Text Embeddings Inference (TEI) - [Original Model Card](https://huggingface.co/PartAI/Tooka-SBERT) - [Text Embeddings Inference Documentation](https://huggingface.co/d…

azureFoundry/HuggingFace/partai-tooka-sbert@1

PartAI/Tooka-SBERT powered by Text Embeddings Inference (TEI)

Send Request

You can use cURL or any REST Client to send a request to the AzureML endpoint with your AzureML token.

curl <AZUREML_ENDPOINT_URL> \
    -X POST \
    -d '{"inputs":"What is Deep Learning?"}' \
    -H "Authorization: Bearer <AZUREML_TOKEN>" \
    -H "Content-Type: application/json"

Supported Parameters

  • inputs (string): The input sentence to create the embeddings for.
  • parameters (object):
    • normalize: Whether to normalize returned vectors to have length 1. Defaults to False.
    • prompt_name: The name of the prompt to use for encoding. Must be a key in the prompts dictionary, which is either set in the constructor or loaded from the model configuration. For example if prompt_name is "query" and the prompts is {"query": "query: ", ...}, then the sentence "What is the capital of France?" will be encoded as "query: What is the capital of France?" because the sentence is appended to the prompt. If prompt is also set, this argument is ignored. Defaults to null.
    • truncate: Truncate the inputs that are longer than the maximum supported size. Defaults to False.
    • truncation_direction: Can either be "left" or "right", defaults to "right". Truncating to the "right" means that tokens are removed from the end of the sequence until the maximum supported size is matched, whilst truncating to the "left" means from the beginning of the sequence.

Example payload

{
  "inputs": "What is Deep Learning?",
  "normalize": false,
  "prompt_name": null,
  "truncate": false,
  "truncation_direction": "right"
}

OpenAI Embeddings API compatibility

Additionally, Text Embeddings Inference (TEI) offers an OpenAI Chat Completion API compatible layer under the endpoint /v1/embeddings, check the full specification in the OpenAI Embeddings Create Documentation.

Specifications & capabilities

Input
text
Output
embeddings
Published model ID
Tooka-SBERT
Capabilities
embeddings

Evaluations

1–7 / 7

MTEB

89.085%

source id
mteb
leaderboard name
MTEB Semantic Similarity
dimension
embedding
metric
Average task score
rank
9
sample size
1
aggregation scope
splits
  • test
subsets
  • default
task_count
1
task_names
  • SynPerSTS
sample_size
1
subset_count
1
subset_sample
  • default
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

MTEB

59.63%

source id
mteb
leaderboard name
MTEB Clustering
dimension
embedding
metric
Average task score
rank
92
sample size
3
aggregation scope
splits
  • test
subsets
  • default
task_count
3
task_names
  • BeytooteClustering
  • DigikalamagClustering
  • NLPTwitterAnalysisClustering
sample_size
3
subset_count
1
subset_sample
  • default
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

MTEB

58.293%

source id
mteb
leaderboard name
MTEB Reranking
dimension
embedding
metric
Average task score
rank
232
sample size
2
aggregation scope
splits
  • dev
  • test
subsets
  • fa
task_count
2
task_names
  • MIRACLReranking
  • WikipediaRerankingMultilingual
sample_size
2
subset_count
1
subset_sample
  • fa
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

MTEB

61.904%

source id
mteb
leaderboard name
MTEB Classification
dimension
embedding
metric
Average task score
rank
282
sample size
28
aggregation scope
splits
  • test
  • validation
subsets
  • default
  • fa
  • fas
task_count
25
task_names
  • DigikalamagClassification
  • MassiveIntentClassification
  • MassiveScenarioClassification
  • MultilingualSentimentClassification
  • NLPTwitterAnalysisClassification
  • NLPTwitterAnalysisClassification.v2
  • PerShopDomainClassification
  • PerShopIntentClassification
  • PersianFoodSentimentClassification
  • SIDClassification
  • SIDClassification.v2
  • StyleClassification
  • SynPerChatbotConvSAClassification
  • SynPerChatbotConvSAToneChatbotClassification
  • SynPerChatbotConvSAToneClassification
  • SynPerChatbotConvSAToneUserClassification
  • SynPerChatbotRAGToneChatbotClassification
  • SynPerChatbotRAGToneClassification
  • SynPerChatbotRAGToneUserClassification
  • SynPerChatbotSatisfactionLevelClassification
  • SynPerChatbotToneChatbotClassification
  • SynPerChatbotToneClassification
  • SynPerChatbotToneUserClassification
  • SynPerTextToneClassification
  • SynPerTextToneClassification.v3
sample_size
28
subset_count
3
subset_sample
  • default
  • fa
  • fas
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

MTEB

51.845%

source id
mteb
leaderboard name
MTEB Other
dimension
embedding
metric
Average task score
rank
442
sample size
49
aggregation scope
splits
  • dev
  • test
subsets
  • default
task_count
49
task_names
  • ArguAna-Fa
  • ArguAna-Fa.v2
  • CExaPPC
  • ClimateFEVER-Fa
  • DBPedia-Fa
  • DeepSentiPers
  • DeepSentiPers.v2
  • FEVER-FaHardNegatives
  • FarsTail
  • FarsiParaphraseDetection
  • Farsick
  • FiQA2018-Fa
  • FiQA2018-Fa.v2
  • HamshahriClustring
  • HotpotQA-Fa
  • HotpotQA-FaHardNegatives
  • MSMARCO-Fa
  • MSMARCO-FaHardNegatives
  • NFCorpus-Fa
  • NQ-Fa
  • NQ-FaHardNegatives
  • ParsinluEntail
  • ParsinluQueryParaphPC
  • PersianTextEmotion
  • PersianTextEmotion.v2
  • Query2Query
  • SAMSumFa
  • SCIDOCS-Fa
  • SCIDOCS-Fa.v2
  • SIDClustring
  • SciFact-Fa
  • SciFact-Fa.v2
  • SentimentDKSF
  • SynPerChatbotConvSAAnger
  • SynPerChatbotConvSAFear
  • SynPerChatbotConvSAFriendship
  • SynPerChatbotConvSAHappiness
  • SynPerChatbotConvSAJealousy
  • SynPerChatbotConvSALove
  • SynPerChatbotConvSASadness
  • SynPerChatbotConvSASatisfaction
  • SynPerChatbotConvSASurprise
  • SynPerChatbotRAGFAQPC
  • SynPerQAPC
  • SynPerTextKeywordsPC
  • TRECCOVID-Fa
  • TRECCOVID-Fa.v2
  • Touche2020-Fa
  • Touche2020-Fa.v2
sample_size
49
subset_count
1
subset_sample
  • default
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

MTEB

49.808%

source id
mteb
leaderboard name
MTEB Overall
dimension
embedding
metric
Average task score
rank
579
sample size
111
aggregation scope
splits
  • dev
  • test
  • validation
subsets
  • default
  • fa
  • fas
task_count
108
task_names
  • ArguAna-Fa
  • ArguAna-Fa.v2
  • BeytooteClustering
  • CExaPPC
  • CQADupstackAndroidRetrieval-Fa
  • CQADupstackEnglishRetrieval-Fa
  • CQADupstackGamingRetrieval-Fa
  • CQADupstackGisRetrieval-Fa
  • CQADupstackMathematicaRetrieval-Fa
  • CQADupstackPhysicsRetrieval-Fa
  • CQADupstackProgrammersRetrieval-Fa
  • CQADupstackRetrieval-Fa
  • CQADupstackStatsRetrieval-Fa
  • CQADupstackTexRetrieval-Fa
  • CQADupstackUnixRetrieval-Fa
  • CQADupstackWebmastersRetrieval-Fa
  • CQADupstackWordpressRetrieval-Fa
  • ClimateFEVER-Fa
  • DBPedia-Fa
  • DeepSentiPers
  • DeepSentiPers.v2
  • DigikalamagClassification
  • DigikalamagClustering
  • FEVER-FaHardNegatives
  • FarsTail
  • FarsiParaphraseDetection
  • Farsick
  • FiQA2018-Fa
  • FiQA2018-Fa.v2
  • HamshahriClustring
  • HotpotQA-Fa
  • HotpotQA-FaHardNegatives
  • MIRACLReranking
  • MIRACLRetrieval
  • MIRACLRetrievalHardNegatives
  • MSMARCO-Fa
  • MSMARCO-FaHardNegatives
  • MassiveIntentClassification
  • MassiveScenarioClassification
  • MultilingualSentimentClassification
  • NFCorpus-Fa
  • NLPTwitterAnalysisClassification
  • NLPTwitterAnalysisClassification.v2
  • NLPTwitterAnalysisClustering
  • NQ-Fa
  • NQ-FaHardNegatives
  • NeuCLIR2023Retrieval
  • NeuCLIR2023RetrievalHardNegatives
  • ParsinluEntail
  • ParsinluQueryParaphPC
  • PerShopDomainClassification
  • PerShopIntentClassification
  • PersianFoodSentimentClassification
  • PersianTextEmotion
  • PersianTextEmotion.v2
  • PersianWebDocumentRetrieval
  • Query2Query
  • QuoraRetrieval-Fa
  • QuoraRetrieval-Fa.v2
  • SAMSumFa
  • SCIDOCS-Fa
  • SCIDOCS-Fa.v2
  • SIDClassification
  • SIDClassification.v2
  • SIDClustring
  • SciFact-Fa
  • SciFact-Fa.v2
  • SentimentDKSF
  • StyleClassification
  • SynPerChatbotConvSAAnger
  • SynPerChatbotConvSAClassification
  • SynPerChatbotConvSAFear
  • SynPerChatbotConvSAFriendship
  • SynPerChatbotConvSAHappiness
  • SynPerChatbotConvSAJealousy
  • SynPerChatbotConvSALove
  • SynPerChatbotConvSASadness
  • SynPerChatbotConvSASatisfaction
  • SynPerChatbotConvSASurprise
  • SynPerChatbotConvSAToneChatbotClassification
  • SynPerChatbotConvSAToneClassification
  • SynPerChatbotConvSAToneUserClassification
  • SynPerChatbotRAGFAQPC
  • SynPerChatbotRAGFAQRetrieval
  • SynPerChatbotRAGSumSRetrieval
  • SynPerChatbotRAGToneChatbotClassification
  • SynPerChatbotRAGToneClassification
  • SynPerChatbotRAGToneUserClassification
  • SynPerChatbotRAGTopicsRetrieval
  • SynPerChatbotSatisfactionLevelClassification
  • SynPerChatbotSumSRetrieval
  • SynPerChatbotToneChatbotClassification
  • SynPerChatbotToneClassification
  • SynPerChatbotToneUserClassification
  • SynPerChatbotTopicsRetrieval
  • SynPerQAPC
  • SynPerQARetrieval
  • SynPerSTS
  • SynPerTextKeywordsPC
  • SynPerTextToneClassification
  • SynPerTextToneClassification.v3
  • TRECCOVID-Fa
  • TRECCOVID-Fa.v2
  • Touche2020-Fa
  • Touche2020-Fa.v2
  • WebFAQRetrieval
  • WikipediaRerankingMultilingual
  • WikipediaRetrievalMultilingual
sample_size
111
subset_count
3
subset_sample
  • default
  • fa
  • fas
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

MTEB

32.381%

source id
mteb
leaderboard name
MTEB Retrieval
dimension
embedding
metric
Average task score
rank
691
sample size
28
aggregation scope
splits
  • dev
  • test
subsets
  • default
  • fa
  • fas
task_count
28
task_names
  • CQADupstackAndroidRetrieval-Fa
  • CQADupstackEnglishRetrieval-Fa
  • CQADupstackGamingRetrieval-Fa
  • CQADupstackGisRetrieval-Fa
  • CQADupstackMathematicaRetrieval-Fa
  • CQADupstackPhysicsRetrieval-Fa
  • CQADupstackProgrammersRetrieval-Fa
  • CQADupstackRetrieval-Fa
  • CQADupstackStatsRetrieval-Fa
  • CQADupstackTexRetrieval-Fa
  • CQADupstackUnixRetrieval-Fa
  • CQADupstackWebmastersRetrieval-Fa
  • CQADupstackWordpressRetrieval-Fa
  • MIRACLRetrieval
  • MIRACLRetrievalHardNegatives
  • NeuCLIR2023Retrieval
  • NeuCLIR2023RetrievalHardNegatives
  • PersianWebDocumentRetrieval
  • QuoraRetrieval-Fa
  • QuoraRetrieval-Fa.v2
  • SynPerChatbotRAGFAQRetrieval
  • SynPerChatbotRAGSumSRetrieval
  • SynPerChatbotRAGTopicsRetrieval
  • SynPerChatbotSumSRetrieval
  • SynPerChatbotTopicsRetrieval
  • SynPerQARetrieval
  • WebFAQRetrieval
  • WikipediaRetrievalMultilingual
sample_size
28
subset_count
3
subset_sample
  • default
  • fa
  • fas
aggregation_method
mean-of-task-means
model_revision_count
1
model_revision_sample
  • 5d07f0c543aca654373b931ae07cd197769110fd
source model id
PartAI/Tooka-SBERT
Original result ↗

Provider specifications & access

1–1 / 1

PartAI/Tooka-SBERT ↗

source model id
HuggingFace/partai-tooka-sbert@1
hosting
foundry-managed-compute
platforms
  • Microsoft Foundry
license
apache-2.0
architecture
input_modalities
  • text
output_modalities
  • embeddings
capabilities

    Original sources

    1–5 / 5

    Hugging Face · PartAI/Tooka-SBERT ↗

    Microsoft Foundry · Model page ↗

    Microsoft Foundry · Model documentation ↗

    Hugging Face · Catalog ↗

    Microsoft Foundry · Catalog ↗

    Source links

    1–2 / 2

    Explore more models

    Models

    Loading
    EVIDENCE / COMPARISON

    Benchmarks

    Explore every benchmark

    Loading

    These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

    Install TensorTrace