L-Eval
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Long-context language understanding | L-Eval | Coursera58.28 | 26 | |
| Long-context language understanding | L-Eval (test) | Coursera58.28 | 26 | |
| Long-context Summarization | L-Eval Sum | QMS22.66 | 13 | |
| Long-context Question Answering | L-Eval QA | NQ80.73 | 13 | |
| Long-context evaluation | L-Eval | Close Score68.8 | 13 | |
| Long-context understanding | L-Eval 32K | P95 Latency (ms)215 | 12 | |
| Closed-ended Task Evaluation | L-Eval closed-ended tasks | Coursera Score41.86 | 12 | |
| Closed-ended Long-Text Understanding | L-Eval Closed-ended | Coursera Score48.51 | 6 | |
| Long-context understanding | L-Eval | Coursera Accuracy70.6 | 6 | |
| Prompt Compression | L-Eval (test) | Coursera QA Accuracy64.4 | 5 | |
| Long-context Question Answering | L-Eval | Coursera QA30.2 | 4 |
Showing 11 of 11 rows