|
--- |
|
tags: |
|
- sparse sparsity quantized onnx embeddings int8 |
|
- mteb |
|
model-index: |
|
- name: gte-large-sparse |
|
results: |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/biosses-sts |
|
name: MTEB BIOSSES |
|
config: default |
|
split: test |
|
revision: d3fb88f8f02e40887cd149695127462bbcf29b4a |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 88.64253410928214 |
|
- type: cos_sim_spearman |
|
value: 85.83388349410652 |
|
- type: euclidean_pearson |
|
value: 86.86126159318735 |
|
- type: euclidean_spearman |
|
value: 85.61580623591163 |
|
- type: manhattan_pearson |
|
value: 86.6901132883383 |
|
- type: manhattan_spearman |
|
value: 85.60255292187769 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sickr-sts |
|
name: MTEB SICK-R |
|
config: default |
|
split: test |
|
revision: a6ea5a8cab320b040a23452cc28066d9beae2cee |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 85.23314640591607 |
|
- type: cos_sim_spearman |
|
value: 79.00078545104338 |
|
- type: euclidean_pearson |
|
value: 83.48009254500714 |
|
- type: euclidean_spearman |
|
value: 78.95413001389939 |
|
- type: manhattan_pearson |
|
value: 83.46945566025941 |
|
- type: manhattan_spearman |
|
value: 78.9241707208135 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts12-sts |
|
name: MTEB STS12 |
|
config: default |
|
split: test |
|
revision: a0d554a64d88156834ff5ae9920b964011b16384 |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 81.77526666043804 |
|
- type: cos_sim_spearman |
|
value: 73.4849063285867 |
|
- type: euclidean_pearson |
|
value: 78.04477932740524 |
|
- type: euclidean_spearman |
|
value: 73.01394205771743 |
|
- type: manhattan_pearson |
|
value: 78.08836684503294 |
|
- type: manhattan_spearman |
|
value: 73.05074711098149 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts13-sts |
|
name: MTEB STS13 |
|
config: default |
|
split: test |
|
revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 84.57839215661352 |
|
- type: cos_sim_spearman |
|
value: 86.13854767345153 |
|
- type: euclidean_pearson |
|
value: 85.12712609946449 |
|
- type: euclidean_spearman |
|
value: 85.52497994789026 |
|
- type: manhattan_pearson |
|
value: 85.06833141611173 |
|
- type: manhattan_spearman |
|
value: 85.45003068636466 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts14-sts |
|
name: MTEB STS14 |
|
config: default |
|
split: test |
|
revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375 |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 83.30485126978374 |
|
- type: cos_sim_spearman |
|
value: 80.36497172462357 |
|
- type: euclidean_pearson |
|
value: 82.91977909424605 |
|
- type: euclidean_spearman |
|
value: 80.16995106297438 |
|
- type: manhattan_pearson |
|
value: 82.88200991402184 |
|
- type: manhattan_spearman |
|
value: 80.14259757215227 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts15-sts |
|
name: MTEB STS15 |
|
config: default |
|
split: test |
|
revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3 |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 86.99883111314007 |
|
- type: cos_sim_spearman |
|
value: 88.531352572377 |
|
- type: euclidean_pearson |
|
value: 87.96834578059067 |
|
- type: euclidean_spearman |
|
value: 88.44800718542935 |
|
- type: manhattan_pearson |
|
value: 87.94889391725033 |
|
- type: manhattan_spearman |
|
value: 88.45467695837115 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts16-sts |
|
name: MTEB STS16 |
|
config: default |
|
split: test |
|
revision: 4d8694f8f0e0100860b497b999b3dbed754a0513 |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 82.4636984892402 |
|
- type: cos_sim_spearman |
|
value: 84.0808920789148 |
|
- type: euclidean_pearson |
|
value: 83.70613486028309 |
|
- type: euclidean_spearman |
|
value: 84.35941626905009 |
|
- type: manhattan_pearson |
|
value: 83.70259457073782 |
|
- type: manhattan_spearman |
|
value: 84.35496521501604 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts17-crosslingual-sts |
|
name: MTEB STS17 (en-en) |
|
config: en-en |
|
split: test |
|
revision: af5e6fb845001ecf41f4c1e033ce921939a2a68d |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 88.76172944971023 |
|
- type: cos_sim_spearman |
|
value: 89.4190945039165 |
|
- type: euclidean_pearson |
|
value: 89.47263005347381 |
|
- type: euclidean_spearman |
|
value: 89.49228360724095 |
|
- type: manhattan_pearson |
|
value: 89.49959868816694 |
|
- type: manhattan_spearman |
|
value: 89.5314536157954 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/sts22-crosslingual-sts |
|
name: MTEB STS22 (en) |
|
config: en |
|
split: test |
|
revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80 |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 64.57158223787549 |
|
- type: cos_sim_spearman |
|
value: 66.75053533168037 |
|
- type: euclidean_pearson |
|
value: 66.45526604831747 |
|
- type: euclidean_spearman |
|
value: 66.14567667353113 |
|
- type: manhattan_pearson |
|
value: 66.47352000151176 |
|
- type: manhattan_spearman |
|
value: 66.21099856852885 |
|
- task: |
|
type: STS |
|
dataset: |
|
type: mteb/stsbenchmark-sts |
|
name: MTEB STSBenchmark |
|
config: default |
|
split: test |
|
revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831 |
|
metrics: |
|
- type: cos_sim_pearson |
|
value: 85.055653571006 |
|
- type: cos_sim_spearman |
|
value: 85.45387832634702 |
|
- type: euclidean_pearson |
|
value: 86.31667154906651 |
|
- type: euclidean_spearman |
|
value: 85.66079590537946 |
|
- type: manhattan_pearson |
|
value: 86.2806853257308 |
|
- type: manhattan_spearman |
|
value: 85.63700636713952 |
|
- task: |
|
type: PairClassification |
|
dataset: |
|
type: mteb/sprintduplicatequestions-pairclassification |
|
name: MTEB SprintDuplicateQuestions |
|
config: default |
|
split: test |
|
revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46 |
|
metrics: |
|
- type: cos_sim_accuracy |
|
value: 99.78811881188119 |
|
- type: cos_sim_ap |
|
value: 94.67027715905307 |
|
- type: cos_sim_f1 |
|
value: 89.33074684772066 |
|
- type: cos_sim_precision |
|
value: 86.7231638418079 |
|
- type: cos_sim_recall |
|
value: 92.10000000000001 |
|
- type: dot_accuracy |
|
value: 99.47128712871287 |
|
- type: dot_ap |
|
value: 78.41478815918727 |
|
- type: dot_f1 |
|
value: 73.30049261083744 |
|
- type: dot_precision |
|
value: 72.23300970873787 |
|
- type: dot_recall |
|
value: 74.4 |
|
- type: euclidean_accuracy |
|
value: 99.78415841584159 |
|
- type: euclidean_ap |
|
value: 94.60075930867181 |
|
- type: euclidean_f1 |
|
value: 89.12175648702593 |
|
- type: euclidean_precision |
|
value: 88.94422310756973 |
|
- type: euclidean_recall |
|
value: 89.3 |
|
- type: manhattan_accuracy |
|
value: 99.78415841584159 |
|
- type: manhattan_ap |
|
value: 94.62867439278095 |
|
- type: manhattan_f1 |
|
value: 89.2337536372454 |
|
- type: manhattan_precision |
|
value: 86.62900188323917 |
|
- type: manhattan_recall |
|
value: 92.0 |
|
- type: max_accuracy |
|
value: 99.78811881188119 |
|
- type: max_ap |
|
value: 94.67027715905307 |
|
- type: max_f1 |
|
value: 89.33074684772066 |
|
- task: |
|
type: PairClassification |
|
dataset: |
|
type: mteb/twittersemeval2015-pairclassification |
|
name: MTEB TwitterSemEval2015 |
|
config: default |
|
split: test |
|
revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1 |
|
metrics: |
|
- type: cos_sim_accuracy |
|
value: 85.09864695714371 |
|
- type: cos_sim_ap |
|
value: 70.33704198164713 |
|
- type: cos_sim_f1 |
|
value: 66.22893954410307 |
|
- type: cos_sim_precision |
|
value: 62.42410088743577 |
|
- type: cos_sim_recall |
|
value: 70.52770448548813 |
|
- type: dot_accuracy |
|
value: 79.11426357513263 |
|
- type: dot_ap |
|
value: 49.15484584572233 |
|
- type: dot_f1 |
|
value: 51.12580243364951 |
|
- type: dot_precision |
|
value: 40.13840830449827 |
|
- type: dot_recall |
|
value: 70.3957783641161 |
|
- type: euclidean_accuracy |
|
value: 85.15825236931514 |
|
- type: euclidean_ap |
|
value: 70.51017350854076 |
|
- type: euclidean_f1 |
|
value: 66.45416294785159 |
|
- type: euclidean_precision |
|
value: 64.29805082654823 |
|
- type: euclidean_recall |
|
value: 68.7598944591029 |
|
- type: manhattan_accuracy |
|
value: 85.1403707456637 |
|
- type: manhattan_ap |
|
value: 70.47587863399994 |
|
- type: manhattan_f1 |
|
value: 66.4576802507837 |
|
- type: manhattan_precision |
|
value: 63.32138590203107 |
|
- type: manhattan_recall |
|
value: 69.92084432717678 |
|
- type: max_accuracy |
|
value: 85.15825236931514 |
|
- type: max_ap |
|
value: 70.51017350854076 |
|
- type: max_f1 |
|
value: 66.4576802507837 |
|
- task: |
|
type: PairClassification |
|
dataset: |
|
type: mteb/twitterurlcorpus-pairclassification |
|
name: MTEB TwitterURLCorpus |
|
config: default |
|
split: test |
|
revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf |
|
metrics: |
|
- type: cos_sim_accuracy |
|
value: 88.8539604921023 |
|
- type: cos_sim_ap |
|
value: 85.71869912577101 |
|
- type: cos_sim_f1 |
|
value: 78.00535626720983 |
|
- type: cos_sim_precision |
|
value: 76.46232344893885 |
|
- type: cos_sim_recall |
|
value: 79.61194949183862 |
|
- type: dot_accuracy |
|
value: 84.57717235223348 |
|
- type: dot_ap |
|
value: 74.89496650237145 |
|
- type: dot_f1 |
|
value: 69.05327823892932 |
|
- type: dot_precision |
|
value: 65.75666829166377 |
|
- type: dot_recall |
|
value: 72.69787496150293 |
|
- type: euclidean_accuracy |
|
value: 88.89471028835332 |
|
- type: euclidean_ap |
|
value: 85.75169460500409 |
|
- type: euclidean_f1 |
|
value: 78.17055393586006 |
|
- type: euclidean_precision |
|
value: 74.21118184334348 |
|
- type: euclidean_recall |
|
value: 82.57622420696026 |
|
- type: manhattan_accuracy |
|
value: 88.92187681918733 |
|
- type: manhattan_ap |
|
value: 85.7496679471825 |
|
- type: manhattan_f1 |
|
value: 78.11088295687884 |
|
- type: manhattan_precision |
|
value: 75.82083061535117 |
|
- type: manhattan_recall |
|
value: 80.5435786880197 |
|
- type: max_accuracy |
|
value: 88.92187681918733 |
|
- type: max_ap |
|
value: 85.75169460500409 |
|
- type: max_f1 |
|
value: 78.17055393586006 |
|
license: mit |
|
language: |
|
- en |
|
--- |
|
|
|
# gte-large-sparse |
|
|
|
This is the sparse ONNX variant of the [gte-large](https://huggingface.co/thenlper/gte-large) embeddings model created with [DeepSparse Optimum](https://github.com/neuralmagic/optimum-deepsparse) for ONNX export/inference and Neural Magic's [Sparsify](https://github.com/neuralmagic/sparsify) for one-shot quantization (INT8) and unstructured pruning 50%. |
|
|
|
Current list of sparse and quantized gte ONNX models: |
|
|
|
| Links | Sparsification Method | |
|
| --------------------------------------------------------------------------------------------------- | ---------------------- | |
|
| [zeroshot/gte-large-sparse](https://huggingface.co/zeroshot/gte-large-sparse) | Quantization (INT8) & 50% Pruning | |
|
| [zeroshot/gte-large-quant](https://huggingface.co/zeroshot/gte-large-quant) | Quantization (INT8) | |
|
| [zeroshot/gte-base-sparse](https://huggingface.co/zeroshot/gte-base-sparse) | Quantization (INT8) & 50% Pruning | |
|
| [zeroshot/gte-base-quant](https://huggingface.co/zeroshot/gte-base-quant) | Quantization (INT8) | |
|
| [zeroshot/gte-small-sparse](https://huggingface.co/zeroshot/gte-small-sparse) | Quantization (INT8) & 50% Pruning | |
|
| [zeroshot/gte-small-quant](https://huggingface.co/zeroshot/gte-small-quant) | Quantization (INT8) | |
|
|
|
```bash |
|
pip install -U deepsparse-nightly[sentence_transformers] |
|
``` |
|
|
|
```python |
|
from deepsparse.sentence_transformers import SentenceTransformer |
|
model = SentenceTransformer('zeroshot/gte-large-sparse', export=False) |
|
|
|
# Our sentences we like to encode |
|
sentences = ['This framework generates embeddings for each input sentence', |
|
'Sentences are passed as a list of string.', |
|
'The quick brown fox jumps over the lazy dog.'] |
|
|
|
# Sentences are encoded by calling model.encode() |
|
embeddings = model.encode(sentences) |
|
|
|
# Print the embeddings |
|
for sentence, embedding in zip(sentences, embeddings): |
|
print("Sentence:", sentence) |
|
print("Embedding:", embedding.shape) |
|
print("") |
|
``` |
|
|
|
For further details regarding DeepSparse & Sentence Transformers integration, refer to the [DeepSparse README](https://github.com/neuralmagic/deepsparse/tree/main/src/deepsparse/sentence_transformers). |
|
|
|
For general questions on these models and sparsification methods, reach out to the engineering team on our [community Slack](https://join.slack.com/t/discuss-neuralmagic/shared_invite/zt-q1a1cnvo-YBoICSIw3L1dmQpjBeDurQ). |
|
|
|
 |
