Add Byrne-Embed (Quazim0t0/Byrne-Embed) MTEB(eng, v2) results#577
Open
SeanceTable wants to merge 2 commits into
Open
Add Byrne-Embed (Quazim0t0/Byrne-Embed) MTEB(eng, v2) results#577SeanceTable wants to merge 2 commits into
SeanceTable wants to merge 2 commits into
Conversation
Member
|
To add results you need to add model implementation to mteb repo. You can follow this guide https://docs.mteb.org/contributing/adding_a_model/ |
KennethEnevoldsen
added
the
waiting for review of implementation
Model Results ComparisonReference models: Results for
|
| task_name | Quazim0t0/Byrne-Embed | google/gemini-embedding-001 | intfloat/multilingual-e5-large | Max result | Model with max result | In Training Data |
|---|---|---|---|---|---|---|
| AmazonCounterfactualClassification | .801 | .927 | nan | .989 | Bytedance/Seed1.6-embedding-1215 | False |
| ArXivHierarchicalClusteringP2P | .531 | .649 | .557 | .687 | NovaSearch/jasper_en_vision_language_v1 | False |
| ArXivHierarchicalClusteringS2S | .504 | .638 | .537 | .655 | Qwen/Qwen3-Embedding-8B | False |
| ArguAna | .377 | .864 | .544 | .898 | voyageai/voyage-3-m-exp | False |
| AskUbuntuDupQuestions | .529 | .642 | .592 | .753 | IEITYuan/Yuan-embedding-2.0-en | False |
| BIOSSES | .756 | .890 | .846 | .969 | Gameselo/STS-multilingual-mpnet-base-v2 | False |
| Banking77Classification | .746 | .943 | .749 | .943 | google/gemini-embedding-001 | False |
| BiorxivClusteringP2P.v2 | .337 | .539 | .372 | .842 | codefuse-ai/F2LLM-4B | False |
| CQADupstackGamingRetrieval | .371 | .707 | .587 | .816 | IEITYuan/Yuan-embedding-2.0-en | False |
| CQADupstackUnixRetrieval | .235 | .537 | .399 | .720 | voyageai/voyage-3-m-exp | False |
| ClimateFEVERHardNegatives | .136 | .311 | .260 | .591 | IEITYuan/Yuan-embedding-2.0-en | False |
| FEVERHardNegatives | .287 | .890 | .838 | .945 | ByteDance-Seed/Seed1.5-Embedding | False |
| FiQA2018 | .114 | .618 | .438 | .821 | ai-sage/Giga-Embeddings-instruct | False |
| HotpotQAHardNegatives | .305 | .870 | .706 | .870 | google/gemini-embedding-001 | False |
| ImdbClassification | .610 | .950 | .887 | .974 | Qwen/Qwen3-Embedding-8B | False |
| MTOPDomainClassification | .923 | .993 | .910 | 1.000 | voyageai/voyage-3-m-exp | False |
| MassiveIntentClassification | .632 | .885 | .680 | .919 | voyageai/voyage-3-m-exp | False |
| MassiveScenarioClassification | .731 | .921 | .718 | .993 | voyageai/voyage-3-m-exp | False |
| MedrxivClusteringP2P.v2 | .327 | .472 | .343 | .720 | codefuse-ai/F2LLM-4B | False |
| MedrxivClusteringS2S.v2 | .290 | .450 | .315 | .702 | codefuse-ai/F2LLM-4B | False |
| MindSmallReranking | .281 | .329 | .302 | .344 | Kingsoft-LLM/QZhou-Embedding | False |
| SCIDOCS | .101 | .252 | .174 | .599 | IEITYuan/Yuan-embedding-2.0-en | False |
| SICK-R | .691 | .827 | .802 | .947 | Gameselo/STS-multilingual-mpnet-base-v2 | False |
| STS12 | .649 | .815 | .800 | .955 | Gameselo/STS-multilingual-mpnet-base-v2 | False |
| STS13 | .721 | .899 | .816 | .978 | Gameselo/STS-multilingual-mpnet-base-v2 | False |
| STS14 | .678 | .854 | .777 | .975 | Gameselo/STS-multilingual-mpnet-base-v2 | False |
| STS15 | .771 | .904 | .893 | .981 | Gameselo/STS-multilingual-mpnet-base-v2 | False |
| STS17 | .832 | .916 | .881 | .959 | Qwen/Qwen3-Embedding-8B | False |
| STS22.v2 | .605 | .680 | .637 | .798 | Kingsoft-LLM/QZhou-Embedding | False |
| STSBenchmark | .771 | .891 | .873 | .950 | Kingsoft-LLM/QZhou-Embedding | False |
| SprintDuplicateQuestions | .865 | .969 | .931 | .984 | Kingsoft-LLM/QZhou-Embedding | False |
| StackExchangeClustering.v2 | .419 | .921 | .464 | .921 | google/gemini-embedding-001 | False |
| StackExchangeClusteringP2P.v2 | .352 | .509 | .385 | .551 | Kingsoft-LLM/QZhou-Embedding | False |
| SummEvalSummarization.v2 | .224 | .383 | .314 | .389 | annamodels/LGAI-Embedding-Preview | False |
| TRECCOVID | .293 | .863 | .712 | .983 | IEITYuan/Yuan-embedding-2.0-en | False |
| Touche2020Retrieval.v3 | .245 | .524 | .496 | .762 | jcorners/ingot-8b-r3 | False |
| ToxicConversationsClassification | .629 | .887 | .660 | .976 | voyageai/voyage-3-m-exp | False |
| TweetSentimentExtractionClassification | .573 | .699 | .628 | .882 | voyageai/voyage-3-m-exp | False |
| TwentyNewsgroupsClustering.v2 | .224 | .574 | .392 | .876 | GeoGPT-Research-Project/GeoEmbedding | False |
| TwitterSemEval2015 | .532 | .792 | .753 | .901 | jcorners/ingot-8b-r3 | False |
| TwitterURLCorpus | .826 | .870 | .858 | .957 | TencentBAC/Conan-embedding-v2 | False |
| Average | .508 | .733 | .621 | .841 | nan | - |
Samoed
reviewed
Jul 7, 2026
| @@ -0,0 +1 @@ | |||
| {"loader_kwargs": {}, "name": "Quazim0t0/Byrne-Embed", "revision": "b1dd73482beb385c295d8237a8c9c845680f8f75", "release_date": "2026-06-20", "languages": ["eng-Latn"], "n_parameters": 85700000, "n_active_parameters_override": null, "n_embedding_parameters": null, "memory_usage_mb": 327.0, "max_tokens": 128.0, "embed_dim": 768, "license": "apache-2.0", "open_weights": true, "public_training_code": null, "public_training_data": null, "framework": ["PyTorch"], "reference": null, "similarity_fn_name": "cosine", "use_instructions": false, "training_datasets": null, "adapted_from": null, "superseded_by": null, "modalities": ["text"], "model_type": ["dense"], "citation": null, "contacts": null, "experiment_kwargs": null, "output_dtypes": null, "extra_requirements_groups": null, "loader": "ByrneEmbedMTEB", "is_cross_encoder": false} No newline at end of file | |||
Member
There was a problem hiding this comment.
Revision in model_meta is not matching folder name. How did you run you model?
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds full MTEB(eng, v2) results (41/41 tasks, overall 50.79) for Byrne-Embed, a compact 85M-parameter sentence-embedding model.
Highlights: strong STS (71.9), Classification (70.6), PairClassification (74.1); retrieval is the expected weak spot at this size.