Hugging Face's logo

pacoreyes
/
StanceFit

Text Classification
setfit
PyTorch
sentence-transformers
mpnet
generated_from_setfit_trainer
Model card Files Community
Edit model card

SetFit with sentence-transformers/all-mpnet-base-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/all-mpnet-base-v2 as the Sentence Transformer embedding model. A LogisticRegression instance is used for classification.

The model has been trained using an efficient few-shot learning technique that involves:

  1. Fine-tuning a Sentence Transformer with contrastive learning.
  2. Training a classification head with features from the fine-tuned Sentence Transformer.

Model Details

Model Description

Model Sources

Model Labels

Label Examples
0
  • 'We in the United States believe if we can promote democracy around the world, there will be more peace.'
  • 'We recognise the transformative power of technology, including digital public infrastructure, to support sustainable development in the Indo-Pacific and deliver economic and social benefits.'
  • 'This program strengthens democracy, transparency, and the rule of law in developing nations, and I ask you to fully fund this important initiative.'
1
  • 'I do not ever want to ever fight a war that is unconstitutional and I am the dangerous person.'
  • "And so, we are at a moment where I really think threats to our democracy, threats to our core freedoms are very much on people's minds."
  • 'My views in opposition to the cancellation of the war debt are a matter of detailed record in many public statements and in a recent message to the Congress.'

Uses

Direct Use for Inference

First install the SetFit library:

pip install setfit

Then you can load this model and run inference.

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("setfit_model_id")
# Run inference
preds = model("We cannot allow the world's leading sponsor of terrorism to possess the planet's most dangerous weapons.")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 3 23.4393 46
Label Training Sample Count
0 486
1 486

Training Hyperparameters

  • batch_size: (16, 16)
  • num_epochs: (1, 1)
  • max_steps: -1
  • sampling_strategy: oversampling
  • body_learning_rate: (1.003444469523018e-06, 1.003444469523018e-06)
  • head_learning_rate: 0.01
  • loss: CosineSimilarityLoss
  • distance_metric: cosine_distance
  • margin: 0.25
  • end_to_end: False
  • use_amp: False
  • warmup_proportion: 0.1
  • seed: 37
  • eval_max_steps: -1
  • load_best_model_at_end: True

Training Results

Epoch Step Training Loss Validation Loss
0.0000 1 0.3295 -
0.0017 50 0.3132 -
0.0034 100 0.274 -
0.0051 150 0.2774 -
0.0068 200 0.2578 -
0.0084 250 0.2536 -
0.0101 300 0.3353 -
0.0118 350 0.253 -
0.0135 400 0.2865 -
0.0152 450 0.2894 -
0.0169 500 0.2554 0.2632
0.0186 550 0.2487 -
0.0203 600 0.2713 -
0.0220 650 0.2841 -
0.0237 700 0.2251 -
0.0253 750 0.2534 -
0.0270 800 0.2489 -
0.0287 850 0.2297 -
0.0304 900 0.2288 -
0.0321 950 0.211 -
0.0338 1000 0.188 0.2073
0.0355 1050 0.1488 -
0.0372 1100 0.2103 -
0.0389 1150 0.1607 -
0.0406 1200 0.0793 -
0.0422 1250 0.0968 -
0.0439 1300 0.0987 -
0.0456 1350 0.0786 -
0.0473 1400 0.0267 -
0.0490 1450 0.0432 -
0.0507 1500 0.0262 0.064
0.0524 1550 0.1269 -
0.0541 1600 0.039 -
0.0558 1650 0.0266 -
0.0575 1700 0.0455 -
0.0591 1750 0.0175 -
0.0608 1800 0.0157 -
0.0625 1850 0.0063 -
0.0642 1900 0.0146 -
0.0659 1950 0.0046 -
0.0676 2000 0.0046 0.0464
0.0693 2050 0.0035 -
0.0710 2100 0.0073 -
0.0727 2150 0.0012 -
0.0744 2200 0.0025 -
0.0760 2250 0.0023 -
0.0777 2300 0.0017 -
0.0794 2350 0.0012 -
0.0811 2400 0.0017 -
0.0828 2450 0.0016 -
0.0845 2500 0.0014 0.0535
0.0862 2550 0.0011 -
0.0879 2600 0.0021 -
0.0896 2650 0.0009 -
0.0913 2700 0.0008 -
0.0929 2750 0.0006 -
0.0946 2800 0.0007 -
0.0963 2850 0.0012 -
0.0980 2900 0.001 -
0.0997 2950 0.0005 -
0.1014 3000 0.0006 0.0575
0.1031 3050 0.0006 -
0.1048 3100 0.0004 -
0.1065 3150 0.0006 -
0.1082 3200 0.0005 -
0.1098 3250 0.0006 -
0.1115 3300 0.0005 -
0.1132 3350 0.0008 -
0.1149 3400 0.0003 -
0.1166 3450 0.0005 -
0.1183 3500 0.0004 0.0642
0.1200 3550 0.0006 -
0.1217 3600 0.0003 -
0.1234 3650 0.0009 -
0.1251 3700 0.0002 -
0.1267 3750 0.0003 -
0.1284 3800 0.0005 -
0.1301 3850 0.0002 -
0.1318 3900 0.0002 -
0.1335 3950 0.0005 -
0.1352 4000 0.0003 0.0697
0.1369 4050 0.0002 -
0.1386 4100 0.0002 -
0.1403 4150 0.0004 -
0.1420 4200 0.0012 -
0.1436 4250 0.0002 -
0.1453 4300 0.0002 -
0.1470 4350 0.0001 -
0.1487 4400 0.0002 -
0.1504 4450 0.0002 -
0.1521 4500 0.0003 0.0718
0.1538 4550 0.0003 -
0.1555 4600 0.0002 -
0.1572 4650 0.0002 -
0.1589 4700 0.0003 -
0.1605 4750 0.0002 -
0.1622 4800 0.0002 -
0.1639 4850 0.0002 -
0.1656 4900 0.0002 -
0.1673 4950 0.0002 -
0.1690 5000 0.0002 0.0684
0.1707 5050 0.0002 -
0.1724 5100 0.0002 -
0.1741 5150 0.0002 -
0.1758 5200 0.0003 -
0.1774 5250 0.0002 -
0.1791 5300 0.0001 -
0.1808 5350 0.0002 -
0.1825 5400 0.0001 -
0.1842 5450 0.0001 -
0.1859 5500 0.0001 0.0731
0.1876 5550 0.0002 -
0.1893 5600 0.0002 -
0.1910 5650 0.0001 -
0.1927 5700 0.0001 -
0.1943 5750 0.0001 -
0.1960 5800 0.0002 -
0.1977 5850 0.0001 -
0.1994 5900 0.0003 -
0.2011 5950 0.0002 -
0.2028 6000 0.0002 0.0724
0.2045 6050 0.0001 -
0.2062 6100 0.0001 -
0.2079 6150 0.0001 -
0.2096 6200 0.0001 -
0.2112 6250 0.0001 -
0.2129 6300 0.0002 -
0.2146 6350 0.0001 -
0.2163 6400 0.0001 -
0.2180 6450 0.0001 -
0.2197 6500 0.0001 0.0784
0.2214 6550 0.0001 -
0.2231 6600 0.0001 -
0.2248 6650 0.0001 -
0.2265 6700 0.0001 -
0.2281 6750 0.0001 -
0.2298 6800 0.0001 -
0.2315 6850 0.0001 -
0.2332 6900 0.0001 -
0.2349 6950 0.0002 -
0.2366 7000 0.0001 0.0672
0.2383 7050 0.0001 -
0.2400 7100 0.0001 -
0.2417 7150 0.0001 -
0.2434 7200 0.0001 -
0.2450 7250 0.0001 -
0.2467 7300 0.0001 -
0.2484 7350 0.0001 -
0.2501 7400 0.0001 -
0.2518 7450 0.0001 -
0.2535 7500 0.0001 0.0627
0.2552 7550 0.0001 -
0.2569 7600 0.0001 -
0.2586 7650 0.0 -
0.2603 7700 0.0001 -
0.2619 7750 0.0 -
0.2636 7800 0.0001 -
0.2653 7850 0.0001 -
0.2670 7900 0.0001 -
0.2687 7950 0.0001 -
0.2704 8000 0.0 0.0754
0.2721 8050 0.0001 -
0.2738 8100 0.0001 -
0.2755 8150 0.0 -
0.2772 8200 0.0 -
0.2788 8250 0.0 -
0.2805 8300 0.0001 -
0.2822 8350 0.0001 -
0.2839 8400 0.0001 -
0.2856 8450 0.0 -
0.2873 8500 0.0 0.0748
0.2890 8550 0.0 -
0.2907 8600 0.0 -
0.2924 8650 0.0 -
0.2941 8700 0.0 -
0.2957 8750 0.0001 -
0.2974 8800 0.0001 -
0.2991 8850 0.0001 -
0.3008 8900 0.0 -
0.3025 8950 0.0001 -
0.3042 9000 0.0001 0.057
0.3059 9050 0.0 -
0.3076 9100 0.0 -
0.3093 9150 0.0002 -
0.3110 9200 0.0 -
0.3126 9250 0.0 -
0.3143 9300 0.0 -
0.3160 9350 0.0001 -
0.3177 9400 0.0002 -
0.3194 9450 0.0 -
0.3211 9500 0.0 0.0781
0.3228 9550 0.0 -
0.3245 9600 0.0 -
0.3262 9650 0.0 -
0.3279 9700 0.0 -
0.3295 9750 0.0 -
0.3312 9800 0.0 -
0.3329 9850 0.0 -
0.3346 9900 0.0001 -
0.3363 9950 0.0 -
0.3380 10000 0.0 0.0698
0.3397 10050 0.0 -
0.3414 10100 0.0 -
0.3431 10150 0.0 -
0.3448 10200 0.0 -
0.3464 10250 0.0022 -
0.3481 10300 0.0 -
0.3498 10350 0.0001 -
0.3515 10400 0.0 -
0.3532 10450 0.0 -
0.3549 10500 0.0 0.0698
0.3566 10550 0.0 -
0.3583 10600 0.0 -
0.3600 10650 0.0 -
0.3617 10700 0.0 -
0.3633 10750 0.0 -
0.3650 10800 0.0 -
0.3667 10850 0.0 -
0.3684 10900 0.0001 -
0.3701 10950 0.0 -
0.3718 11000 0.0 0.0746
0.3735 11050 0.0 -
0.3752 11100 0.0 -
0.3769 11150 0.0001 -
0.3786 11200 0.0 -
0.3802 11250 0.0 -
0.3819 11300 0.0 -
0.3836 11350 0.0 -
0.3853 11400 0.0 -
0.3870 11450 0.0 -
0.3887 11500 0.0 0.0753
0.3904 11550 0.0 -
0.3921 11600 0.0001 -
0.3938 11650 0.0 -
0.3955 11700 0.0 -
0.3971 11750 0.0 -
0.3988 11800 0.0 -
0.4005 11850 0.0 -
0.4022 11900 0.0 -
0.4039 11950 0.0 -
0.4056 12000 0.0 0.0743
0.4073 12050 0.0 -
0.4090 12100 0.0 -
0.4107 12150 0.0 -
0.4124 12200 0.0 -
0.4140 12250 0.0 -
0.4157 12300 0.0 -
0.4174 12350 0.0 -
0.4191 12400 0.0 -
0.4208 12450 0.0 -
0.4225 12500 0.0 0.0733
0.4242 12550 0.0 -
0.4259 12600 0.0 -
0.4276 12650 0.0 -
0.4293 12700 0.0 -
0.4309 12750 0.0 -
0.4326 12800 0.0 -
0.4343 12850 0.0 -
0.4360 12900 0.0 -
0.4377 12950 0.0 -
0.4394 13000 0.0 0.072
0.4411 13050 0.0 -
0.4428 13100 0.0 -
0.4445 13150 0.0 -
0.4462 13200 0.0 -
0.4478 13250 0.0 -
0.4495 13300 0.0 -
0.4512 13350 0.0 -
0.4529 13400 0.0 -
0.4546 13450 0.0 -
0.4563 13500 0.0 0.0753
0.4580 13550 0.0 -
0.4597 13600 0.0 -
0.4614 13650 0.0 -
0.4631 13700 0.0 -
0.4647 13750 0.0 -
0.4664 13800 0.0 -
0.4681 13850 0.0 -
0.4698 13900 0.0 -
0.4715 13950 0.0 -
0.4732 14000 0.0 0.0756
0.4749 14050 0.0 -
0.4766 14100 0.0 -
0.4783 14150 0.0 -
0.4800 14200 0.0 -
0.4816 14250 0.0 -
0.4833 14300 0.0 -
0.4850 14350 0.0 -
0.4867 14400 0.0 -
0.4884 14450 0.0 -
0.4901 14500 0.0 0.0622
0.4918 14550 0.0 -
0.4935 14600 0.0 -
0.4952 14650 0.0 -
0.4969 14700 0.0 -
0.4985 14750 0.0 -
0.5002 14800 0.0 -
0.5019 14850 0.0 -
0.5036 14900 0.0 -
0.5053 14950 0.0 -
0.5070 15000 0.0 0.0676
0.5087 15050 0.0 -
0.5104 15100 0.0 -
0.5121 15150 0.0 -
0.5138 15200 0.0 -
0.5154 15250 0.0 -
0.5171 15300 0.0 -
0.5188 15350 0.0 -
0.5205 15400 0.0 -
0.5222 15450 0.0 -
0.5239 15500 0.0 0.0668
0.5256 15550 0.0 -
0.5273 15600 0.0 -
0.5290 15650 0.0 -
0.5307 15700 0.0 -
0.5323 15750 0.0 -
0.5340 15800 0.0 -
0.5357 15850 0.0 -
0.5374 15900 0.0 -
0.5391 15950 0.0 -
0.5408 16000 0.0 0.0707
0.5425 16050 0.0 -
0.5442 16100 0.0 -
0.5459 16150 0.0 -
0.5476 16200 0.0 -
0.5492 16250 0.0 -
0.5509 16300 0.0 -
0.5526 16350 0.0 -
0.5543 16400 0.0 -
0.5560 16450 0.0 -
0.5577 16500 0.0 0.0644
0.5594 16550 0.0 -
0.5611 16600 0.0 -
0.5628 16650 0.0 -
0.5645 16700 0.0 -
0.5661 16750 0.0 -
0.5678 16800 0.0 -
0.5695 16850 0.0 -
0.5712 16900 0.0 -
0.5729 16950 0.0 -
0.5746 17000 0.0 0.0742
0.5763 17050 0.0 -
0.5780 17100 0.0 -
0.5797 17150 0.0 -
0.5814 17200 0.0 -
0.5830 17250 0.0 -
0.5847 17300 0.0 -
0.5864 17350 0.0 -
0.5881 17400 0.0 -
0.5898 17450 0.0 -
0.5915 17500 0.0 0.0738
0.5932 17550 0.0 -
0.5949 17600 0.0 -
0.5966 17650 0.0 -
0.5983 17700 0.0 -
0.5999 17750 0.0 -
0.6016 17800 0.0 -
0.6033 17850 0.0 -
0.6050 17900 0.0 -
0.6067 17950 0.0 -
0.6084 18000 0.0 0.0725
0.6101 18050 0.0 -
0.6118 18100 0.0 -
0.6135 18150 0.0 -
0.6152 18200 0.0 -
0.6168 18250 0.0 -
0.6185 18300 0.0 -
0.6202 18350 0.0 -
0.6219 18400 0.0 -
0.6236 18450 0.0 -
0.6253 18500 0.0 0.0724
0.6270 18550 0.0 -
0.6287 18600 0.0 -
0.6304 18650 0.0 -
0.6321 18700 0.0 -
0.6337 18750 0.0 -
0.6354 18800 0.0 -
0.6371 18850 0.0 -
0.6388 18900 0.0 -
0.6405 18950 0.0 -
0.6422 19000 0.0 0.0622
0.6439 19050 0.0 -
0.6456 19100 0.0 -
0.6473 19150 0.0 -
0.6490 19200 0.0 -
0.6506 19250 0.0 -
0.6523 19300 0.0 -
0.6540 19350 0.0 -
0.6557 19400 0.0 -
0.6574 19450 0.0 -
0.6591 19500 0.0 0.0754
0.6608 19550 0.0 -
0.6625 19600 0.0 -
0.6642 19650 0.0 -
0.6659 19700 0.0 -
0.6675 19750 0.0 -
0.6692 19800 0.0 -
0.6709 19850 0.0 -
0.6726 19900 0.0 -
0.6743 19950 0.0 -
0.6760 20000 0.0 0.0723
0.6777 20050 0.0 -
0.6794 20100 0.0 -
0.6811 20150 0.0 -
0.6828 20200 0.0 -
0.6844 20250 0.0 -
0.6861 20300 0.0 -
0.6878 20350 0.0 -
0.6895 20400 0.0 -
0.6912 20450 0.0 -
0.6929 20500 0.0 0.0741
0.6946 20550 0.0 -
0.6963 20600 0.0 -
0.6980 20650 0.0 -
0.6997 20700 0.0 -
0.7013 20750 0.0 -
0.7030 20800 0.0 -
0.7047 20850 0.0 -
0.7064 20900 0.0 -
0.7081 20950 0.0 -
0.7098 21000 0.0 0.0733
0.7115 21050 0.0 -
0.7132 21100 0.0 -
0.7149 21150 0.0 -
0.7166 21200 0.0 -
0.7182 21250 0.0 -
0.7199 21300 0.0 -
0.7216 21350 0.0 -
0.7233 21400 0.0 -
0.7250 21450 0.0 -
0.7267 21500 0.0 0.0757
0.7284 21550 0.0 -
0.7301 21600 0.0 -
0.7318 21650 0.0 -
0.7335 21700 0.0 -
0.7351 21750 0.0 -
0.7368 21800 0.0 -
0.7385 21850 0.0 -
0.7402 21900 0.0 -
0.7419 21950 0.0 -
0.7436 22000 0.0 0.0766
0.7453 22050 0.0 -
0.7470 22100 0.0 -
0.7487 22150 0.0 -
0.7504 22200 0.0 -
0.7520 22250 0.0 -
0.7537 22300 0.0 -
0.7554 22350 0.0 -
0.7571 22400 0.0 -
0.7588 22450 0.0 -
0.7605 22500 0.0 0.0757
0.7622 22550 0.0 -
0.7639 22600 0.0 -
0.7656 22650 0.0 -
0.7673 22700 0.0 -
0.7689 22750 0.0 -
0.7706 22800 0.0 -
0.7723 22850 0.0 -
0.7740 22900 0.0 -
0.7757 22950 0.0 -
0.7774 23000 0.0 0.0755
0.7791 23050 0.0 -
0.7808 23100 0.0 -
0.7825 23150 0.0 -
0.7842 23200 0.0 -
0.7858 23250 0.0 -
0.7875 23300 0.0 -
0.7892 23350 0.0 -
0.7909 23400 0.0 -
0.7926 23450 0.0 -
0.7943 23500 0.0 0.076
0.7960 23550 0.0 -
0.7977 23600 0.0 -
0.7994 23650 0.0 -
0.8011 23700 0.0 -
0.8027 23750 0.0 -
0.8044 23800 0.0 -
0.8061 23850 0.0 -
0.8078 23900 0.0 -
0.8095 23950 0.0 -
0.8112 24000 0.0 0.0756
0.8129 24050 0.0 -
0.8146 24100 0.0 -
0.8163 24150 0.0 -
0.8180 24200 0.0 -
0.8196 24250 0.0 -
0.8213 24300 0.0 -
0.8230 24350 0.0 -
0.8247 24400 0.0 -
0.8264 24450 0.0 -
0.8281 24500 0.0 0.0759
0.8298 24550 0.0 -
0.8315 24600 0.0 -
0.8332 24650 0.0 -
0.8349 24700 0.0 -
0.8365 24750 0.0 -
0.8382 24800 0.0 -
0.8399 24850 0.0 -
0.8416 24900 0.0 -
0.8433 24950 0.0 -
0.8450 25000 0.0 0.0762
0.8467 25050 0.0 -
0.8484 25100 0.0 -
0.8501 25150 0.0 -
0.8518 25200 0.0 -
0.8534 25250 0.0 -
0.8551 25300 0.0 -
0.8568 25350 0.0 -
0.8585 25400 0.0 -
0.8602 25450 0.0 -
0.8619 25500 0.0 0.0733
0.8636 25550 0.0 -
0.8653 25600 0.0 -
0.8670 25650 0.0 -
0.8687 25700 0.0 -
0.8703 25750 0.0 -
0.8720 25800 0.0 -
0.8737 25850 0.0 -
0.8754 25900 0.0 -
0.8771 25950 0.0 -
0.8788 26000 0.0 0.0742
0.8805 26050 0.0 -
0.8822 26100 0.0 -
0.8839 26150 0.0 -
0.8856 26200 0.0 -
0.8872 26250 0.0 -
0.8889 26300 0.0 -
0.8906 26350 0.0 -
0.8923 26400 0.0 -
0.8940 26450 0.0 -
0.8957 26500 0.0 0.0756
0.8974 26550 0.0 -
0.8991 26600 0.0 -
0.9008 26650 0.0 -
0.9025 26700 0.0 -
0.9041 26750 0.0 -
0.9058 26800 0.0 -
0.9075 26850 0.0 -
0.9092 26900 0.0 -
0.9109 26950 0.0 -
0.9126 27000 0.0 0.0751
0.9143 27050 0.0 -
0.9160 27100 0.0 -
0.9177 27150 0.0 -
0.9194 27200 0.0 -
0.9210 27250 0.0 -
0.9227 27300 0.0 -
0.9244 27350 0.0 -
0.9261 27400 0.0 -
0.9278 27450 0.0 -
0.9295 27500 0.0 0.075
0.9312 27550 0.0 -
0.9329 27600 0.0 -
0.9346 27650 0.0 -
0.9363 27700 0.0 -
0.9379 27750 0.0 -
0.9396 27800 0.0 -
0.9413 27850 0.0 -
0.9430 27900 0.0 -
0.9447 27950 0.0 -
0.9464 28000 0.0 0.0725
0.9481 28050 0.0 -
0.9498 28100 0.0 -
0.9515 28150 0.0 -
0.9532 28200 0.0 -
0.9548 28250 0.0 -
0.9565 28300 0.0 -
0.9582 28350 0.0 -
0.9599 28400 0.0 -
0.9616 28450 0.0 -
0.9633 28500 0.0 0.0761
0.9650 28550 0.0 -
0.9667 28600 0.0 -
0.9684 28650 0.0 -
0.9701 28700 0.0 -
0.9717 28750 0.0 -
0.9734 28800 0.0 -
0.9751 28850 0.0 -
0.9768 28900 0.0 -
0.9785 28950 0.0 -
0.9802 29000 0.0 0.0759
0.9819 29050 0.0 -
0.9836 29100 0.0 -
0.9853 29150 0.0 -
0.9870 29200 0.0 -
0.9886 29250 0.0 -
0.9903 29300 0.0 -
0.9920 29350 0.0 -
0.9937 29400 0.0 -
0.9954 29450 0.0 -
0.9971 29500 0.0 0.0761
0.9988 29550 0.0 -
  • The bold row denotes the saved checkpoint.

Framework Versions

  • Python: 3.10.11
  • SetFit: 1.0.1
  • Sentence Transformers: 2.2.2
  • Transformers: 4.25.1
  • PyTorch: 2.1.2
  • Datasets: 2.15.0
  • Tokenizers: 0.13.3

Citation

BibTeX 

@article{https://doi.org/10.48550/arxiv.2209.11055,
    doi = {10.48550/ARXIV.2209.11055},
    url = {https://arxiv.org/abs/2209.11055},
    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
    title = {Efficient Few-Shot Learning Without Prompts},
    publisher = {arXiv},
    year = {2022},
    copyright = {Creative Commons Attribution 4.0 International}
}
Downloads last month
5
Inference API
Text Classification
Examples
This model can be loaded on Inference API (serverless).

Finetuned from