jetri20/ABSA_review_game_geometry-aspect

SetFit Aspect Model

This is a SetFit model that can be used for Aspect Based Sentiment Analysis (ABSA). A LogisticRegression instance is used for classification. In particular, this model is in charge of filtering aspect span candidates.

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.

This model was trained within the context of a larger system for ABSA, which looks like so:

1. Use a spaCy model to select possible aspect span candidates.
2. Use this SetFit model to filter these possible aspect span candidates.
3. Use a SetFit model to classify the filtered aspect span candidates.

Model Details

Model Description

• Model Type: SetFit
• Classification head: a LogisticRegression instance
• spaCy Model: id_core_news_trf
• SetFitABSA Aspect Model: jetri20/ABSA_review_game_geometry-aspect
• SetFitABSA Polarity Model: jetri20/ABSA_review_game_geometry-polarity
• Maximum Sequence Length: 512 tokens
• Number of Classes: 2 classes

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
aspect	'level:sih game level nya sulit banget level clutter funk susah nya ampun level 11 nya level sulit nya level 10 sulit menyerah sabar banting hp saking sulit nya ya geometri dash kebanyakan level nya sulit mah geometri dash game stress saking susah nya' 'iklan:iklan emang sih level iklan muncul masuk level mending kayak mah' 'game:game apasih sampe strees gitu final boss level the tower susahnya ampun kenapasih kalo hijau pas udah nya jatuh mati sih cube nya cuman 1 hp ya ngeselin sih'
no aspect	'sih game level:sih game level nya sulit banget level clutter funk susah nya ampun level 11 nya level sulit nya level 10 sulit menyerah sabar banting hp saking sulit nya ya geometri dash kebanyakan level nya sulit mah geometri dash game stress saking susah nya' 'level clutter funk:sih game level nya sulit banget level clutter funk susah nya ampun level 11 nya level sulit nya level 10 sulit menyerah sabar banting hp saking sulit nya ya geometri dash kebanyakan level nya sulit mah geometri dash game stress saking susah nya' 'level:sih game level nya sulit banget level clutter funk susah nya ampun level 11 nya level sulit nya level 10 sulit menyerah sabar banting hp saking sulit nya ya geometri dash kebanyakan level nya sulit mah geometri dash game stress saking susah nya'

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 AbsaModel

# Download from the 🤗 Hub
model = AbsaModel.from_pretrained(
    "jetri20/ABSA_review_game_geometry-aspect",
    "jetri20/ABSA_review_game_geometry-polarity",
)
# Run inference
preds = model("The food was great, but the venue is just way too busy.")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	2	23.5963	67

Label	Training Sample Count
no aspect	754
aspect	321

Training Hyperparameters

• batch_size: (4, 4)
• num_epochs: (1, 1)
• max_steps: -1
• sampling_strategy: oversampling
• num_iterations: 5
• 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
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0004	1	0.3713	-
0.0186	50	0.2045	-
0.0372	100	0.1548	-
0.0558	150	0.3116	-
0.0744	200	0.2066	-
0.0930	250	0.2932	-
0.1116	300	0.3138	-
0.1302	350	0.1258	-
0.1488	400	0.3442	-
0.1674	450	0.0558	-
0.1860	500	0.2819	-
0.2046	550	0.2211	-
0.2232	600	0.1269	-
0.2418	650	0.0098	-
0.2604	700	0.2395	-
0.2790	750	0.4382	-
0.2976	800	0.488	-
0.3162	850	0.6662	-
0.3348	900	0.1811	-
0.3534	950	0.2431	-
0.3720	1000	0.2032	-
0.3906	1050	0.0475	-
0.4092	1100	0.177	-
0.4278	1150	0.0556	-
0.4464	1200	0.3048	-
0.4650	1250	0.0015	-
0.4836	1300	0.0841	-
0.5022	1350	0.0105	-
0.5208	1400	0.0036	-
0.5394	1450	0.2296	-
0.5580	1500	0.0045	-
0.5766	1550	0.0134	-
0.5952	1600	0.0367	-
0.6138	1650	0.0044	-
0.6324	1700	0.0068	-
0.6510	1750	0.1408	-
0.6696	1800	0.0092	-
0.6882	1850	0.1926	-
0.7068	1900	0.0014	-
0.7254	1950	0.0003	-
0.7440	2000	0.2094	-
0.7626	2050	0.0329	-
0.7812	2100	0.0028	-
0.7999	2150	0.0144	-
0.8185	2200	0.1555	-
0.8371	2250	0.0005	-
0.8557	2300	0.0067	-
0.8743	2350	0.1485	-
0.8929	2400	0.0034	-
0.9115	2450	0.0044	-
0.9301	2500	0.2752	-
0.9487	2550	0.1342	-
0.9673	2600	0.0108	-
0.9859	2650	0.0106	-
1.0	2688	-	0.2236

Framework Versions

• Python: 3.10.13
• SetFit: 1.0.3
• Sentence Transformers: 3.0.1
• spaCy: 3.7.5
• Transformers: 4.36.2
• PyTorch: 2.1.2
• Datasets: 2.19.2
• Tokenizers: 0.15.2

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}
}