Add vultr/VultronRetrieverCore-Qwen3.5-4.5B ViDoRe results#579
Merged
Samoed merged 1 commit intoJun 22, 2026
Merged
Conversation
Model Results ComparisonReference models: Results for
|
| task_name | vultr/VultronRetrieverCore-Qwen3.5-4.5B | Max result | Model with max result | In Training Data |
|---|---|---|---|---|
| Vidore2BioMedicalLecturesRetrieval | .665 | .670 | DataScience-UIBK/Argus-Colqwen3.5-9b-v0-bf16 | False |
| Vidore2ESGReportsHLRetrieval | .729 | .791 | DataScience-UIBK/Argus-Colqwen3.5-9b-v0-bf16 | False |
| Vidore2ESGReportsRetrieval | .639 | .660 | OpenSearch-AI/Ops-Colqwen3-4B | False |
| Vidore2EconomicsReportsRetrieval | .612 | .658 | DataScience-UIBK/Argus-Colqwen3.5-9b-v0 | False |
| Vidore3ComputerScienceRetrieval | .798 | .809 | webAI-Official/webAI-ColVec1-9b | False |
| Vidore3EnergyRetrieval | .692 | .703 | vultr/VultronRetrieverPrime-Qwen3.5-8B | False |
| Vidore3FinanceEnRetrieval | .689 | .690 | vultr/VultronRetrieverPrime-Qwen3.5-8B | False |
| Vidore3FinanceFrRetrieval | .520 | .545 | vultr/VultronRetrieverPrime-Qwen3.5-8B | False |
| Vidore3HrRetrieval | .661 | .700 | webAI-Official/webAI-ColVec1-9b | False |
| Vidore3IndustrialRetrieval | .561 | .574 | vultr/VultronRetrieverPrime-Qwen3.5-8B | False |
| Vidore3PharmaceuticalsRetrieval | .674 | .682 | vultr/VultronRetrieverPrime-Qwen3.5-8B | False |
| Vidore3PhysicsRetrieval | .502 | .517 | vultr/VultronRetrieverPrime-Qwen3.5-8B | False |
| VidoreArxivQARetrieval | .920 | .938 | VAGOsolutions/SauerkrautLM-ColQwen3-8b-v0.1 | True |
| VidoreDocVQARetrieval | .673 | .687 | webAI-Official/webAI-ColVec1-9b | True |
| VidoreInfoVQARetrieval | .940 | .952 | webAI-Official/webAI-ColVec1-9b | True |
| VidoreShiftProjectRetrieval | .938 | .947 | DataScience-UIBK/Argus-Colqwen3.5-4b-v0 | False |
| VidoreSyntheticDocQAAIRetrieval | .996 | 1.000 | ApsaraStackMaaS/EvoQwen2.5-VL-Retriever-3B-v1 | True |
| VidoreSyntheticDocQAEnergyRetrieval | .983 | .980 | nvidia/llama-nemotron-colembed-vl-3b-v2 | True |
| VidoreSyntheticDocQAGovernmentReportsRetrieval | .985 | .989 | DataScience-UIBK/Argus-Colqwen3.5-9b-v0 | True |
| VidoreSyntheticDocQAHealthcareIndustryRetrieval | 1.000 | 1.000 | vultr/VultronRetrieverPrime-Qwen3.5-8B | True |
| VidoreTabfquadRetrieval | .962 | .981 | nvidia/nemotron-colembed-vl-4b-v2 | True |
| VidoreTatdqaRetrieval | .824 | .857 | DataScience-UIBK/Argus-Colqwen3.5-9b-v0-bf16 | True |
| Average | .771 | .788 | nan | - |
Model have high performance on these tasks: VidoreSyntheticDocQAEnergyRetrieval
Contributor
Author
|
@Samoed ready to roll |
Samoed
approved these changes
Jun 22, 2026
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
2 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 the 22 per-task ViDoRe V1/V2/V3 result JSONs (+
model_meta.json) forvultr/VultronRetrieverCore-Qwen3.5-4.5Bat revision5b63301ce5a49993f9ec1cf36645840b8cbd8120, the mid-tier of the VultronRetriever family.Means: ViDoRe V1 ndcg@5 0.9221 / V2 ndcg@5 0.6612 / V3 ndcg@10 0.6372.
Depends on the ModelMeta PR embeddings-benchmark/mteb#4850 merging first, so the model id resolves via
get_model_meta.model_meta.jsonis generated from that entry'sModelMeta.to_dict()at this SHA. JSONs pulled verbatim from the model repo'seval_results/.