You can deploy this model and start earning money today!

bhaskars113/
whiskey-recipe-type-model

Name: whiskey-recipe-type-model
Availability: InStock

setfit

safetensors

mpnet

sentence-transformers

text-classification

generated_from_setfit_trainer

2209.11055

sentence-transformers/paraphrase-mpnet-base-v2

finetune

Views

No views yet

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

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-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:

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

Model Details

Model Description

Model Type: SetFit
Sentence Transformer body: sentence-transformers/paraphrase-mpnet-base-v2
Classification head: a LogisticRegression instance
Maximum Sequence Length: 512 tokens
Number of Classes: 6 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
4	"@vurnt22 Ginger beer and bourbon is one of two times I actually drink anything ginger-y. The other is ginger ale on an airplane because it just seems like you're supposed to." "It's not for everyone, but almost everywhere sells Moscow mules. Ginger ale is also good. If there were Ginger wines, Ginger stouts, and Ginger Whiskeys, I'd probably drink those too." 'I just like the smell of cinnamon. I like the taste too. My favorite candy is cinnamon-flavored, my favorite tea is cinnamon-flavored, my favorite whiskey is cinnamon-flavored.'
0	"Bourbon Chocolate Ice Cream. It fluffs up beautifully, doesn't melt rapidly during serving and it is one of the best chocolate flavours I have ever had. You know where the recipe is, go and get it." 'Beautiful Lady.... Now the question is, what did you put in it? I prefer Chocolate Whiskey myself....lol' "Bourbon S'mores~Maybe it’s the promise of summer, the nostalgia of a campfire and roasting marshmallows, or the memories of childhood… but S’mores are one of my favorite treats. The way the toasty marshmallow melts the chocolate and the texture of them sandwiched between graham crackers just makes me happy. The Bourbon S’mores Bundt is a grown up version of a childhood favorite. Chocolate graham cracker cake, soaked with bourbon and topped with marshmallow sauce, a fudgy bourbon glaze and toasted marshmallows. One bite and you’ll want “some more."
1	'That could be it. Helps the smoke stick to the meat and it almost doesn’t matter what you use. I use apple cider vinegar with a little bourbon mixed in. I have zero evidence the bourbon has any effect, it just sounds cool, lol. Try that next time. Just a quick spritz to keep the edges from drying out every hour or so until you wrap it or wherever it’s about 165F.' 'My brother in law makes the absolute best smoky old fashioned. #whiskey #oldfashioned #smoky #drinks' 'Smoked to perfection ?? Bridge Street BBQ Platter
3	'A pastry that not only looks like the fruit it’s meant to showcase but also bursts with the fresh flavor of it. In my mind it is a fusion of two classics - a cocktail Whiskey Sour and a Lemon Meringue pie. ▫️Candied lemon and orange peel is suspended in a lemon gel made with freshly squeezed lemon juice and bourbon. ▫️the fruity core is surrounded by white chocolate ganache made with Italian meringue.' 'They also do some interesting stuff like they have a summer whiskey where it is infused tea and lemon.' 'Cheers to Peach Whiskey! This peach whiskey from olesmoky goes perfect with BBQ as a refreshing cocktail or on the rocks. I mixed mine with pineapple juice and ginger beer. The perfect refreshing smooth texture, and all the citrus notes of the peach come through. I love drinking Ole Smoky Whiskey, as it’s the best on the market. '
2	'I soaked walnuts in like 4 shots of bourbon with brown sugar and cinnamon' 'Figured pecans and bourbon both like a little smoke so decided to smoke my Bourbon Pecan pie recipe for tomorrow. Lick test on the thermometer probe says its delicious. Will find out for sure tomorrow.' 'I looooove pecan pie. I found a delicious recipe for bourbon pecan pie with homemade bourbon whip cream. I may need to make one soon'
5	'Just in Nola we have Roulaison, Lula, seven three, wetlands sake, Atelier Vie and happy raptor. A lot of bars have one or two of these available but I rarely see them featured in cocktails. I’d especially love to try flights of local rums or whiskeys alongside common brands so you can see what makes the local stuff unique' 'I have some Milk Chocolate Truffle right now and that shit is good.' 'We celebrated our one-year anniversary here and the staff made us feel so loved and celebrated. The butternut bisque for the fall menu was incredible. My whiskey sour was also phenomenal. The room was loud and cold but not uncommon for indoor restaurant.'

Uses

Direct Use for Inference

First install the SetFit library:

pip install setfit

Then you can load this model and run inference.

python

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("bhaskars113/whiskey-recipe-type-model")
# Run inference
preds = model("Four roses small batch select. Milk chocolate on the nose and palate for me")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	7	50.0446	362

Label	Training Sample Count
0	20
1	20
2	20
3	20
4	16
5	16

Training Hyperparameters

batch_size: (16, 16)
num_epochs: (1, 1)
max_steps: -1
sampling_strategy: oversampling
num_iterations: 20
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
0.0036	1	0.239	-
0.1786	50	0.1855	-
0.3571	100	0.0275	-
0.5357	150	0.0397	-
0.7143	200	0.0063	-
0.8929	250	0.0034	-

Framework Versions

Python: 3.10.12
SetFit: 1.0.3
Sentence Transformers: 2.6.1
Transformers: 4.38.2
PyTorch: 2.2.1+cu121
Datasets: 2.18.0
Tokenizers: 0.15.2

Citation

BibTeX

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}
}