tmp-org/tmp_cv_model_2025_09_01_0

SetFit with Alibaba-NLP/gte-multilingual-base

This is a SetFit model trained on the tmp-org/lidl-dataset-ctx-1 dataset that can be used for Text Classification. This SetFit model uses Alibaba-NLP/gte-multilingual-base 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 Type: SetFit
• Sentence Transformer body: Alibaba-NLP/gte-multilingual-base
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 8192 tokens
• Number of Classes: 33 classes
• Training Dataset: tmp-org/lidl-dataset-ctx-1

Model Sources

• Repository: SetFit on GitHub
• Paper: Efficient Few-Shot Learning Without Prompts
• Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts

Model Labels

Label	Examples
Home_Loading	'Home () [TextView
Home_Home	'Gewinnspiel () [TextView
Other_Gewinnspiel	' (0 ungesendete Lose) [View
Other_Rubbellos	'Sofortgewinne () [TextView
Other_Rabattsammler	'Dein Fortschritt () [TextView
Prospekte_Top Angebote	'Lidl Reisen () [TextView
Lidl Plus_Coupons	'Online () [TextView
Lidl Plus_Angebote	'Coupons () [TextView
Lidl Plus_Partnervorteile	'LIMITIERT: Zeitschriften für Kinder und Eltern () [TextView
Other_Partnervorteile details	'LIMITIERT: Zeitschriften für Kinder und Eltern () [TextView
Other_Other	'Dein Onlineticket für den Saurierpark () [TextView
Onlineshop_Onlineshop	' (Ein grünes Trampolin mit Sicherheitsnetz.) [View
Other_Receipt	'Onlineshop Angebote () [TextView
Mehr_Meine Kassenbons	'Raderberger Str. 211 () [TextView
Other_Lidl Plus Karte	'08. September () [TextView
Other_Lidl Pay PIN	' (Schließen) [ImageButton
Other_Lidl Logo	' (Passcode) [ImageView
Other_Mein Pfand	' (Produkt zur Einkaufsliste hinzufügen) [CheckBox
Mehr_Mehr	'Sportkleidung für die ganze Familie () [TextView
Other_Empfohlene Produkte	'Unsere Monats-Highlights () [TextView
Prospekte_Loading	'Schön, dass du da bist, Kristin :-) () [TextView
Other_Prospekt	'Schön, dass du da bist, Kristin :-) () [TextView
Other_Coupon details	'01.10.2025 – 31.10.2025 () [TextView
Other_Loading	' (Zwei Kinder in Skianzügen und Mützen, die Skier und Stöcke halten, lächeln auf einem schneebedeckten Berg.) [ImageView
Other_Empty	'Add to cart () [Button
Lidl Plus_Loading	' (Schließen) [Button
Konto_Meine Kassenbons	'Raderberger Str. 211 () [TextView
Konto_Konto	'11. Oktober () [TextView
Other_Roulette	'Gewinnspiel () [TextView
Konto_Loading	'20. Oktober () [TextView
Konto_Receipt	'Raderberger Str. 211 () [TextView
Lidl Plus_Coupon details	'Online (Coupon zur Verwendung bei Online) [TextView
Prospekte_Empty	'Home () [TextView

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("tmp-org/tmp_cv_model_2025_09_01_0")
# Run inference
preds = model("Aktuelle Magazine () [TextView|View] | Spielwarenmagazin () [TextView|View] | 13.10.2025 – 27.12.2025 () [TextView|View] | Wein & Spirituosen im Onlineshop () [TextView|View] | Unsere Monats-Highlights () [TextView|View] | 01.10.2025 – 31.10.2025 () [TextView|View] | Aktuelle Reiseprospekte () [TextView|View] | Oktober Reise-Highlights () [TextView|View] | 26.09.2025 - 31.10.2025 () [TextView|View]

[SELECTED START]

Aktuelle Magazine () [TextView|View] | Spielwarenmagazin () [TextView|View] | 13.10.2025 – 27.12.2025 () [TextView|View] | Wein & Spirituosen im Onlineshop () [TextView|View]
[CONTEXT SEPARATOR]
Aktuelle Magazine () [TextView|View] | Spielwarenmagazin () [TextView|View] | 13.10.2025 – 27.12.2025 () [TextView|View] | Wein & Spirituosen im Onlineshop () [TextView|View]")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	9	481.5685	6984

Label	Training Sample Count
Home_Home	48
Home_Loading	10
Konto_Konto	23
Konto_Loading	1
Konto_Meine Kassenbons	28
Konto_Receipt	1
Lidl Plus_Angebote	48
Lidl Plus_Coupon details	1
Lidl Plus_Coupons	48
Lidl Plus_Loading	3
Lidl Plus_Partnervorteile	48
Mehr_Mehr	2
Mehr_Meine Kassenbons	2
Onlineshop_Onlineshop	48
Other_Coupon details	12
Other_Empfohlene Produkte	12
Other_Empty	7
Other_Gewinnspiel	31
Other_Lidl Logo	2
Other_Lidl Pay PIN	12
Other_Lidl Plus Karte	20
Other_Loading	11
Other_Mein Pfand	2
Other_Other	23
Other_Partnervorteile details	9
Other_Prospekt	29
Other_Rabattsammler	48
Other_Receipt	43
Other_Roulette	2
Other_Rubbellos	10
Prospekte_Empty	1
Prospekte_Loading	2
Prospekte_Top Angebote	41

Training Hyperparameters

• batch_size: (4, 4)
• num_epochs: (1, 1)
• max_steps: -1
• sampling_strategy: undersampling
• body_learning_rate: (2e-05, 1e-05)
• 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
• l2_weight: 0.01
• seed: 4242
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0002	1	0.1168	-
0.0087	50	0.2765	-
0.0174	100	0.2022	-
0.0261	150	0.185	-
0.0349	200	0.1247	-
0.0436	250	0.1438	-
0.0523	300	0.1139	-
0.0610	350	0.1175	-
0.0697	400	0.0959	-
0.0784	450	0.0658	-
0.0872	500	0.1046	-
0.0959	550	0.104	-
0.1046	600	0.0741	-
0.1133	650	0.0647	-
0.1220	700	0.0863	-
0.1307	750	0.1213	-
0.1394	800	0.0706	-
0.1482	850	0.0699	-
0.1569	900	0.0616	-
0.1656	950	0.045	-
0.1743	1000	0.0466	-
0.1830	1050	0.0738	-
0.1917	1100	0.0505	-
0.2005	1150	0.0379	-
0.2092	1200	0.036	-
0.2179	1250	0.0509	-
0.2266	1300	0.0266	-
0.2353	1350	0.0307	-
0.2440	1400	0.0438	-
0.2527	1450	0.0395	-
0.2615	1500	0.0515	-
0.2702	1550	0.0361	-
0.2789	1600	0.0269	-
0.2876	1650	0.0265	-
0.2963	1700	0.016	-
0.3050	1750	0.0351	-
0.3138	1800	0.0482	-
0.3225	1850	0.0285	-
0.3312	1900	0.0284	-
0.3399	1950	0.0364	-
0.3486	2000	0.0225	-
0.3573	2050	0.0168	-
0.3660	2100	0.0306	-
0.3748	2150	0.0247	-
0.3835	2200	0.0189	-
0.3922	2250	0.0321	-
0.4009	2300	0.0324	-
0.4096	2350	0.0208	-
0.4183	2400	0.0205	-
0.4271	2450	0.0123	-
0.4358	2500	0.0225	-
0.4445	2550	0.0189	-
0.4532	2600	0.0234	-
0.4619	2650	0.0206	-
0.4706	2700	0.0135	-
0.4793	2750	0.026	-
0.4881	2800	0.0118	-
0.4968	2850	0.0195	-
0.5055	2900	0.0235	-
0.5142	2950	0.0256	-
0.5229	3000	0.0137	-
0.5316	3050	0.0111	-
0.5404	3100	0.02	-
0.5491	3150	0.0157	-
0.5578	3200	0.0105	-
0.5665	3250	0.0135	-
0.5752	3300	0.013	-
0.5839	3350	0.0154	-
0.5926	3400	0.018	-
0.6014	3450	0.0112	-
0.6101	3500	0.0157	-
0.6188	3550	0.0085	-
0.6275	3600	0.0113	-
0.6362	3650	0.0218	-
0.6449	3700	0.0181	-
0.6537	3750	0.0153	-
0.6624	3800	0.0135	-
0.6711	3850	0.0209	-
0.6798	3900	0.0112	-
0.6885	3950	0.0155	-
0.6972	4000	0.0098	-
0.7059	4050	0.0121	-
0.7147	4100	0.0125	-
0.7234	4150	0.0165	-
0.7321	4200	0.015	-
0.7408	4250	0.0066	-
0.7495	4300	0.0113	-
0.7582	4350	0.0045	-
0.7670	4400	0.0181	-
0.7757	4450	0.015	-
0.7844	4500	0.0091	-
0.7931	4550	0.006	-
0.8018	4600	0.0193	-
0.8105	4650	0.0164	-
0.8192	4700	0.0133	-
0.8280	4750	0.014	-
0.8367	4800	0.0113	-
0.8454	4850	0.0127	-
0.8541	4900	0.0141	-
0.8628	4950	0.0097	-
0.8715	5000	0.0069	-
0.8803	5050	0.0128	-
0.8890	5100	0.0066	-
0.8977	5150	0.02	-
0.9064	5200	0.006	-
0.9151	5250	0.0085	-
0.9238	5300	0.0067	-
0.9325	5350	0.0079	-
0.9413	5400	0.0083	-
0.9500	5450	0.0153	-
0.9587	5500	0.0086	-
0.9674	5550	0.0113	-
0.9761	5600	0.0103	-
0.9848	5650	0.0166	-
0.9936	5700	0.006	-

Framework Versions

• Python: 3.12.6
• SetFit: 1.1.2
• Sentence Transformers: 5.2.2
• Transformers: 4.57.1
• PyTorch: 2.10.0+cu128
• Datasets: 3.6.0
• Tokenizers: 0.22.2

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}
}