# Relation Extraction on TACRED v1.0 (10% train)

Source: [Relation Extraction on TACRED v1.0 (10% train)](<https://www.sota2.com/research/sota/relation-extraction-on-tacred-v1-0-10-percent-train>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:52:42.068Z

Evaluation of model performance on the 10% training subset of the TACRED v1.0 dataset for Relation Extraction, typically measured using Micro F1 score.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Relation Extraction |
| Dataset (including any stated split) | TACRED v1.0 (10% train) |
| Metric | Micro F1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 13 of 13 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Micro F1](<https://www.sota2.com/research/sota/relation-extraction-on-tacred-v1-0-10-percent-train.md?metric=Micro%20F1>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Micro F1 | Result date | Report paper |
| --- | --- | --- | --- |
| SURE-PEGASUS-large | 70.7 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| SURE-BART-large-cnn | 68.7 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| SURE-BART-large-xsum | 68 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| NLI-DeBERTa | 67.9 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| SURE-BART-large | 67.9 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| NLI-RoBERTa | 67.8 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| IRE-RoBERTa-large | 67 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| KnowPrompt | 65.2 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| LUKE | 60.6 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| K-Adapter | 56 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| RoBERTa | 55.1 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| RECENT | 54.2 | Not recorded | 68fa6dbe1e209131cea3bcb4 |
| SpanBERT | 1.6 | Not recorded | 68fa6dbe1e209131cea3bcb4 |

### SURE-PEGASUS-large

Micro F1: 70.7. Result ID: 698fe3793a6c314849fc963a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Implementation](<https://github.com/luka-group/SuRE>)

Recorded evaluation configuration: {"Method paradigm":"Summarization-based","Backbone architecture":"PEGASUS-large"}. This record does not assert independent reproduction.

### SURE-BART-large-cnn

Micro F1: 68.7. Result ID: 698fe3793a6c314849fc9638.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Implementation](<https://github.com/luka-group/SuRE>)

Recorded evaluation configuration: {"Method paradigm":"Summarization-based","Backbone architecture":"BART-large-cnn"}. This record does not assert independent reproduction.

### SURE-BART-large-xsum

Micro F1: 68. Result ID: 698fe3793a6c314849fc9639.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Implementation](<https://github.com/luka-group/SuRE>)

Recorded evaluation configuration: {"Method paradigm":"Summarization-based","Backbone architecture":"BART-large-xsum"}. This record does not assert independent reproduction.

### NLI-DeBERTa

Micro F1: 67.9. Result ID: 698fe3793a6c314849fc9635.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ab81e209131cea2a02e.md>)

[Implementation](<https://github.com/osainz59/Ask2Transformers>)

Recorded evaluation configuration: {"Method paradigm":"Indirect Supervision","Backbone architecture":"DeBERTa"}. This record does not assert independent reproduction.

### SURE-BART-large

Micro F1: 67.9. Result ID: 698fe3793a6c314849fc9637.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Implementation](<https://github.com/luka-group/SuRE>)

Recorded evaluation configuration: {"Method paradigm":"Summarization-based","Backbone architecture":"BART-large"}. This record does not assert independent reproduction.

### NLI-RoBERTa

Micro F1: 67.8. Result ID: 698fe3793a6c314849fc9634.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ab81e209131cea2a02e.md>)

[Implementation](<https://github.com/osainz59/Ask2Transformers>)

Recorded evaluation configuration: {"Method paradigm":"Indirect Supervision","Backbone architecture":"RoBERTa"}. This record does not assert independent reproduction.

### IRE-RoBERTa-large

Micro F1: 67. Result ID: 698fe3793a6c314849fc9632.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d881e209131cea3a6b1.md>)

[Implementation](<https://github.com/wzhouad/RE_improved_baseline>)

Recorded evaluation configuration: {"Method paradigm":"Classification-based","Backbone architecture":"RoBERTa-large","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

### KnowPrompt

Micro F1: 65.2. Result ID: 698fe3793a6c314849fc9636.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70981e209131cea4c11a.md>)

[Implementation](<https://github.com/zjunlp/KnowPrompt>)

Recorded evaluation configuration: {"Method paradigm":"Indirect Supervision","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

### LUKE

Micro F1: 60.6. Result ID: 698fe3793a6c314849fc9631.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa686f1e209131cea1c2db.md>)

[Implementation](<https://github.com/studio-ousia/luke>)

Recorded evaluation configuration: {"Method paradigm":"Classification-based","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

### K-Adapter

Micro F1: 56. Result ID: 698fe3793a6c314849fc9630.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68fd1e209131cea1f977.md>)

[Implementation](<https://github.com/microsoft/k-adapter>)

Recorded evaluation configuration: {"Method paradigm":"Classification-based","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

### RoBERTa

Micro F1: 55.1. Result ID: 698fe3793a6c314849fc962f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68fd1e209131cea1f977.md>)

[Implementation](<https://github.com/microsoft/k-adapter>)

Recorded evaluation configuration: {"Method paradigm":"Classification-based","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

### RECENT

Micro F1: 54.2. Result ID: 698fe3793a6c314849fc9633.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69ef1e209131cea2508b.md>)

[Implementation](<https://github.com//Saintfe/RECENT>)

Recorded evaluation configuration: {"Method paradigm":"Classification-based","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

### SpanBERT

Micro F1: 1.6. Result ID: 698fe3793a6c314849fc962e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dbe1e209131cea3bcb4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66ee1e209131cea13573.md>)

[Implementation](<https://github.com/facebookresearch/SpanBERT>)

Recorded evaluation configuration: {"Method paradigm":"Classification-based","Implementation":"Re-implemented"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.