Hugging Face's logo

Kurrant
/
RevenueStreamJP

Text Classification
setfit
Safetensors
sentence-transformers
bert
generated_from_setfit_trainer
Model card Files Community
RevenueStreamJP / README.md
MaiDD's picture
MaiDD
Add SetFit model
73913dc
preview code
|
raw
history blame
18.2 kB
metadata
library_name: setfit
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
metrics:
  - accuracy
widget:
  - text: スマホやタブレットPC、Oculus GOやVIVE、Apple Watchなど新しいデバイス向けアプリの企画・開発を行うスタートアップ。
  - text: ベンチャー企業へのハンズオン投資などを行うベンチャーキャピタル。
  - text: GoogleカレンダーやZoomと連携してスケジュール調整を自動化する日程調整ツール「Jicoo」を開発、提供するスタートアップ
  - text: 住まい探しに特化したウェブサイト「TOKYO APARTMENTS」を提供する企業。
  - text: 医療機器、産業機器の研究開発・製造販売を行う企業。
pipeline_tag: text-classification
inference: false
model-index:
  - name: SetFit
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: accuracy
            value: 0.7272727272727273
            name: Accuracy

SetFit

This is a SetFit model that can be used for Text Classification. A OneVsRestClassifier 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 Type: SetFit
  • Classification head: a OneVsRestClassifier instance
  • Maximum Sequence Length: 512 tokens

Model Sources

Evaluation

Metrics

Label Accuracy
all 0.7273

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("Ekohe/RevenueStreamJP")
# Run inference
preds = model("医療機器、産業機器の研究開発・製造販売を行う企業。")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 1 1.9824 57

Training Hyperparameters

  • batch_size: (10, 10)
  • num_epochs: (35, 35)
  • max_steps: -1
  • sampling_strategy: oversampling
  • num_iterations: 3
  • body_learning_rate: (2e-05, 2e-05)
  • head_learning_rate: 2e-05
  • loss: CosineSimilarityLoss
  • distance_metric: cosine_distance
  • margin: 0.25
  • end_to_end: False
  • use_amp: False
  • warmup_proportion: 0.1
  • seed: 42
  • eval_max_steps: -1
  • load_best_model_at_end: False

Training Results

Epoch Step Training Loss Validation Loss
0.0029 1 0.2602 -
0.1462 50 0.25 -
0.2924 100 0.1712 -
0.4386 150 0.2671 -
0.5848 200 0.2288 -
0.7310 250 0.2253 -
0.8772 300 0.2675 -
1.0234 350 0.1204 -
1.1696 400 0.1185 -
1.3158 450 0.1884 -
1.4620 500 0.2311 -
1.6082 550 0.0659 -
1.7544 600 0.1719 -
1.9006 650 0.0094 -
2.0468 700 0.0237 -
2.1930 750 0.0007 -
2.3392 800 0.0021 -
2.4854 850 0.0013 -
2.6316 900 0.1887 -
2.7778 950 0.0004 -
2.9240 1000 0.0001 -
3.0702 1050 0.0003 -
3.2164 1100 0.0764 -
3.3626 1150 0.0025 -
3.5088 1200 0.0001 -
3.6550 1250 0.0001 -
3.8012 1300 0.0001 -
3.9474 1350 0.0001 -
4.0936 1400 0.0 -
4.2398 1450 0.0001 -
4.3860 1500 0.0001 -
4.5322 1550 0.0 -
4.6784 1600 0.0 -
4.8246 1650 0.0 -
4.9708 1700 0.0 -
5.1170 1750 0.0001 -
5.2632 1800 0.0 -
5.4094 1850 0.0 -
5.5556 1900 0.0 -
5.7018 1950 0.0883 -
5.8480 2000 0.0 -
5.9942 2050 0.0 -
6.1404 2100 0.0 -
6.2865 2150 0.0 -
6.4327 2200 0.0 -
6.5789 2250 0.0 -
6.7251 2300 0.0 -
6.8713 2350 0.0 -
7.0175 2400 0.0 -
7.1637 2450 0.0 -
7.3099 2500 0.0 -
7.4561 2550 0.0 -
7.6023 2600 0.0 -
7.7485 2650 0.0 -
7.8947 2700 0.0 -
8.0409 2750 0.0 -
8.1871 2800 0.0 -
8.3333 2850 0.0 -
8.4795 2900 0.0 -
8.6257 2950 0.0 -
8.7719 3000 0.0 -
8.9181 3050 0.0 -
9.0643 3100 0.0 -
9.2105 3150 0.0 -
9.3567 3200 0.0 -
9.5029 3250 0.0618 -
9.6491 3300 0.3522 -
9.7953 3350 0.0051 -
9.9415 3400 0.0002 -
10.0877 3450 0.0018 -
10.2339 3500 0.0027 -
10.3801 3550 0.0001 -
10.5263 3600 0.0 -
10.6725 3650 0.0 -
10.8187 3700 0.0001 -
10.9649 3750 0.0 -
11.1111 3800 0.0 -
11.2573 3850 0.0001 -
11.4035 3900 0.0001 -
11.5497 3950 0.0 -
11.6959 4000 0.0 -
11.8421 4050 0.0 -
11.9883 4100 0.0 -
12.1345 4150 0.0 -
12.2807 4200 0.0001 -
12.4269 4250 0.0 -
12.5731 4300 0.0 -
12.7193 4350 0.0003 -
12.8655 4400 0.0 -
13.0117 4450 0.0 -
13.1579 4500 0.0 -
13.3041 4550 0.0 -
13.4503 4600 0.0 -
13.5965 4650 0.0 -
13.7427 4700 0.0 -
13.8889 4750 0.0 -
14.0351 4800 0.0 -
14.1813 4850 0.0 -
14.3275 4900 0.0 -
14.4737 4950 0.0 -
14.6199 5000 0.0 -
14.7661 5050 0.0 -
14.9123 5100 0.0 -
15.0585 5150 0.0 -
15.2047 5200 0.0 -
15.3509 5250 0.0 -
15.4971 5300 0.0 -
15.6433 5350 0.0 -
15.7895 5400 0.0 -
15.9357 5450 0.0 -
16.0819 5500 0.0 -
16.2281 5550 0.0 -
16.3743 5600 0.0 -
16.5205 5650 0.0 -
16.6667 5700 0.0 -
16.8129 5750 0.0 -
16.9591 5800 0.0 -
17.1053 5850 0.0 -
17.2515 5900 0.0 -
17.3977 5950 0.0 -
17.5439 6000 0.0 -
17.6901 6050 0.0 -
17.8363 6100 0.0 -
17.9825 6150 0.0 -
18.1287 6200 0.0 -
18.2749 6250 0.0 -
18.4211 6300 0.0 -
18.5673 6350 0.0 -
18.7135 6400 0.0 -
18.8596 6450 0.0 -
19.0058 6500 0.0 -
19.1520 6550 0.0 -
19.2982 6600 0.0 -
19.4444 6650 0.0 -
19.5906 6700 0.0 -
19.7368 6750 0.0 -
19.8830 6800 0.0 -
20.0292 6850 0.0 -
20.1754 6900 0.0 -
20.3216 6950 0.0 -
20.4678 7000 0.0 -
20.6140 7050 0.0 -
20.7602 7100 0.0 -
20.9064 7150 0.0 -
21.0526 7200 0.0 -
21.1988 7250 0.0 -
21.3450 7300 0.0 -
21.4912 7350 0.0 -
21.6374 7400 0.0 -
21.7836 7450 0.0 -
21.9298 7500 0.0 -
22.0760 7550 0.0 -
22.2222 7600 0.0 -
22.3684 7650 0.0 -
22.5146 7700 0.0 -
22.6608 7750 0.0 -
22.8070 7800 0.0 -
22.9532 7850 0.0 -
23.0994 7900 0.0 -
23.2456 7950 0.0 -
23.3918 8000 0.0 -
23.5380 8050 0.0 -
23.6842 8100 0.0 -
23.8304 8150 0.0 -
23.9766 8200 0.0 -
24.1228 8250 0.0858 -
24.2690 8300 0.0 -
24.4152 8350 0.0001 -
24.5614 8400 0.0 -
24.7076 8450 0.0005 -
24.8538 8500 0.0992 -
25.0 8550 0.0 -
25.1462 8600 0.0 -
25.2924 8650 0.0 -
25.4386 8700 0.0 -
25.5848 8750 0.0 -
25.7310 8800 0.0 -
25.8772 8850 0.0 -
26.0234 8900 0.0 -
26.1696 8950 0.0 -
26.3158 9000 0.0 -
26.4620 9050 0.0 -
26.6082 9100 0.0 -
26.7544 9150 0.0 -
26.9006 9200 0.0 -
27.0468 9250 0.0 -
27.1930 9300 0.0 -
27.3392 9350 0.0 -
27.4854 9400 0.0 -
27.6316 9450 0.0 -
27.7778 9500 0.0 -
27.9240 9550 0.0 -
28.0702 9600 0.0 -
28.2164 9650 0.0 -
28.3626 9700 0.0 -
28.5088 9750 0.0 -
28.6550 9800 0.0 -
28.8012 9850 0.0 -
28.9474 9900 0.0 -
29.0936 9950 0.0 -
29.2398 10000 0.0 -
29.3860 10050 0.0 -
29.5322 10100 0.0 -
29.6784 10150 0.0 -
29.8246 10200 0.0 -
29.9708 10250 0.0 -
30.1170 10300 0.0 -
30.2632 10350 0.0 -
30.4094 10400 0.0 -
30.5556 10450 0.0 -
30.7018 10500 0.0 -
30.8480 10550 0.0 -
30.9942 10600 0.0 -
31.1404 10650 0.0 -
31.2865 10700 0.0 -
31.4327 10750 0.0 -
31.5789 10800 0.0 -
31.7251 10850 0.0 -
31.8713 10900 0.0 -
32.0175 10950 0.0 -
32.1637 11000 0.0 -
32.3099 11050 0.0 -
32.4561 11100 0.0 -
32.6023 11150 0.0 -
32.7485 11200 0.0 -
32.8947 11250 0.0 -
33.0409 11300 0.0 -
33.1871 11350 0.0 -
33.3333 11400 0.0 -
33.4795 11450 0.0 -
33.6257 11500 0.0 -
33.7719 11550 0.0 -
33.9181 11600 0.0 -
34.0643 11650 0.0 -
34.2105 11700 0.0 -
34.3567 11750 0.0 -
34.5029 11800 0.0 -
34.6491 11850 0.0 -
34.7953 11900 0.0 -
34.9415 11950 0.0 -

Framework Versions

  • Python: 3.10.12
  • SetFit: 1.0.1
  • Sentence Transformers: 2.2.2
  • Transformers: 4.35.2
  • PyTorch: 2.1.0+cu118
  • Datasets: 2.15.0
  • Tokenizers: 0.15.0

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}
}