GTE Large EN v1.5
Alibaba-NLP/gte-large-en-v1.5
published Apr 2024 · updated Aug 2024
GTE Large EN v1.5 is a text embedding model that supports long context up to 8192 tokens and achieves state-of-the-art performance on the MTEB benchmark.
specs
| Task | Text Embeddings |
| Architecture | Transformer++ (BERT + RoPE + GLU) |
| Parameters | 434M |
| License | Apache-2.0 |
about this model
Key Strengths and Benchmark Results
The model achieves a state-of-the-art average score of 65.39 on the MTEB benchmark (56 tasks) within its size category, and a competitive 86.71 on the LoCo long-context retrieval benchmark. Its embedding dimension is 1024.
MTEB Benchmark Comparison
Results from the MTEB leaderboard (evaluation using mteb==1.2.0, fp16 auto mix precision, max_length=8192, ntk scaling factor of 2):
| Model Name | Param Size (M) | Dimension | Sequence Length | Average (56) | Class. (12) | Clust. (11) | Pair Class. (3) | Reran. (4) | Retr. (15) | STS (10) | Summ. (1) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| gte-large-en-v1.5 | 409 | 1024 | 8192 | 65.39 | 77.75 | 47.95 | 84.63 | 58.50 | 57.91 | 81.43 | 30.91 |
| mxbai-embed-large-v1 | 335 | 1024 | 512 | 64.68 | 75.64 | 46.71 | 87.2 | 60.11 | 54.39 | 85 | 32.71 |
| multilingual-e5-large-instruct | 560 | 1024 | 514 | 64.41 | 77.56 | 47.1 | 86.19 | 58.58 | 52.47 | 84.78 | 30.39 |
| bge-large-en-v1.5 | 335 | 1024 | 512 | 64.23 | 75.97 | 46.08 | 87.12 | 60.03 | 54.29 | 83.11 | 31.61 |
| gte-base-en-v1.5 | 137 | 768 | 8192 | 64.11 | 77.17 | 46.82 | 85.33 | 57.66 | 54.09 | 81.97 | 31.17 |
| bge-base-en-v1.5 | 109 | 768 | 512 | 63.55 | 75.53 | 45.77 | 86.55 | 58.86 | 53.25 | 82.4 | 31.07 |
LoCo Long-Context Retrieval
| Model Name | Dimension | Sequence Length | Average (5) | QsmsumRetrieval | SummScreenRetrieval | QasperAbstractRetrieval | QasperTitleRetrieval | GovReportRetrieval |
|---|---|---|---|---|---|---|---|---|
| gte-large-v1.5 | 1024 | 8192 | 86.71 | 44.55 | 92.61 | 99.82 | 97.81 | 98.74 |
| gte-base-v1.5 | 768 | 8192 | 87.44 | 49.91 | 91.78 | 99.82 | 97.13 | 98.58 |
| gte-qwen1.5-7b | 4096 | 32768 | 87.57 | 49.37 | 93.10 | 99.67 | 97.54 | 98.21 |
For further details, see the mGTE paper and GTE paper.
best for
- ·Long-context document retrieval
- ·Semantic search with large passages
- ·Clustering and classification of long texts
FAQ
It supports up to 8192 tokens.
It achieves 65.39 on MTEB (56 tasks), outperforming BGE Large v1.5 (64.23) and has a longer context length (8192 vs 512).
It is released under the Apache-2.0 license.
Use the OpenAI-compatible endpoint with your API key, sending text inputs and receiving normalized embeddings.
Yes, it supports unpadding and xformers for faster inference; set trust_remote_code=True and use float16.
We're benchmarking and onboarding GTE Large EN v1.5 as a hosted, OpenAI-compatible API. Sign in for free credit and be ready when it lands, or tell us you want it and we'll prioritize it.