metadata
base_model: sentence-transformers/all-mpnet-base-v2
library_name: setfit
metrics:
- accuracy
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
widget:
- text: >
Having previously lived in D.C., Rochester and Detroit and having made
regular trips on the thruways and turnpikes in-between, I can truly say
that the rest stops along the New York Thruway are the least desirable for
food offerings. Even the NJ Turnpike offers a much better selection, with
Ohio striking the best balance overall. Delaware has the largest rest
stop, which offers a great selection but at the cost of having to
negotiate a mall-size parking lot. Although I don't begrudge those who
like McDonald's, I can honestly say I've never eaten at a rest stop or
airport McDonalds, even when there were no other options. There's nothing
wrong with wanting better food, so long as there are options available at
reasonable prices.If there's one thing for which I can give credit to the
New York Thruway rest stops, it's in forcing us to seek out roadside
alternatives in the many communities along the way. As a result, my wife
has an extensive collection of books on diners that has morphed into
somewhat of an obsession over the years. Of course with smartphones and
apps such as Yelp, finding exceptional food along the way has never been
easier. Put another way, I see the thruway rest stop as a place for an
early morning snack or cup of coffee when we're desperate. Unfortunately,
the options are at their worst at 2 am, no matter where one stops.
- text: >
Now that Iran is actively funneling missiles, warheads and drones to
Russia for use in Ukraine, and Russia is funneling technical expertise and
supplies to Iran to make more weapons, things are quickly heating up and
the clock is approaching midnight as Iran get closer and closer to
weaponizing a nuclear MIRV ICBM.The no so cold war between Iran and
Israel, Egypt, Saudi Arabia and the UAE is about to get very hot and
Israel's efforts to avoid aligning against Russia in Syrian airspace
(thank you President Obama) is about to fail as the Russo-Nato proxy war
in Ukraine spills into the Middle East and a heavily armed and nuclear
Israel gets drawn into a very open conflict with Iran and Russia. The
bombing of an Iranian plant inside Iran is major escalation and I doubt
that the CIA and DIA were blindsided by the IDF operation as such a strike
was likely meant to cripple Iranian efforts to resupply Russia as much as
Iranian efforts to resupply Hizbollah in Lebanon. With the Turks waging
war in Syria, the air space over Syria is clearly going to become very
crowded and very dangerous very quickly as Russia is stumbling into a
second war with Israel through its Iranian proxy and Israel unlike Ukraine
can take out both Russian and Iranian offensive capabilities. We just
witnessed the opening salvo of a hot war which is why the DIA, CIA have
been in Tel Aviv and Cairo recently - it is not really about the
Palestinian territories.
- text: >
It's the year of our Lord, 2023; it's hard to believe that we are having
this conversation about the urgent necessity of ammo and lethal weapons.
WWI, WWII, the Korean War, Gulf Wars I & II, Afghanistan, ISIS, etc., have
come and gone. This does not include the multitude of conflicts in Africa,
Georgia, and other hot spots. Mankind has not changed a bit. We are still
driven by fear, greed, and the curse of the ego and its lust for power.
Another article in today's edition discusses the Doomsday Clock and its
relentless ticking toward oblivion. It's just a matter of time -and Boom!
- text: >
i'd go further than the correct interpretation that putin's "cease fire"
was nothing more than "propaganda."i suggest that the russian attack on
kramatorsk on january 7, which russia falsely claimed killed 600 ukrainian
soldiers, reveals the expectation that a cease fire would gather
ukrainians in a rest area where they could be killed en masse. the
headline was preplanned before the event.i point readers to the Institute
for the Study of War (ISW) as an excellent daily summary of open source
information by highly skilled military analysts. they point out that putin
is using a "grievance-revenge" framing of russian military activities
(e.g., kramatorsk was revenge for the grievance of russians killed in
makiivka). the ISW points out that this has only worsened the antagonism
toward the kremlin and military from pro-invasion russian commentators,
who ask why any "grievance event" was allowed to occur in the first place.
- text: >
I cannot entirely agree with this. If there's a disconnect between what's
being taught, and what the student really wants to learn, that can be a
problem. I, for example, learned a _LOT_ about computers, back in '84 --
and a fair bit of other stuff, too. (I speak what I'll term
"conversational" Spanish; I can't claim to be fluent, but I can absolutely
carry on modest conversations and express myself.)But the teachers in my
core subjects were uninspired or flatly failed me (e.g., the CompSci prof
who lost my test, and gave me a zero; that really took the wind out of my
sails, considering I thought I nailed it). So I was having far more fun
at 11:00 p.m. in the computer lab than I was doing school work. Bombed
out of college, but I've now worked at four Fortune 500 companies, and am
currently a senior cloud admin. Students _do_ need to have a desire to
learn, yes, but teachers need to be equipped properly to teach them, too.
inference: true
model-index:
- name: SetFit with sentence-transformers/all-mpnet-base-v2
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: accuracy
value: 0.9
name: Accuracy
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:
- 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/all-mpnet-base-v2
- Classification head: a LogisticRegression instance
- Maximum Sequence Length: 384 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 |
---|---|
yes |
|
no |
|
Evaluation
Metrics
Label | Accuracy |
---|---|
all | 0.9 |
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("davidadamczyk/setfit-model-4")
# Run inference
preds = model("It's the year of our Lord, 2023; it's hard to believe that we are having this conversation about the urgent necessity of ammo and lethal weapons. WWI, WWII, the Korean War, Gulf Wars I & II, Afghanistan, ISIS, etc., have come and gone. This does not include the multitude of conflicts in Africa, Georgia, and other hot spots. Mankind has not changed a bit. We are still driven by fear, greed, and the curse of the ego and its lust for power. Another article in today's edition discusses the Doomsday Clock and its relentless ticking toward oblivion. It's just a matter of time -and Boom!
")
Training Details
Training Set Metrics
Training set | Min | Median | Max |
---|---|---|---|
Word count | 18 | 133.075 | 255 |
Label | Training Sample Count |
---|---|
no | 18 |
yes | 22 |
Training Hyperparameters
- batch_size: (16, 16)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 120
- 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
- l2_weight: 0.01
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
Training Results
Epoch | Step | Training Loss | Validation Loss |
---|---|---|---|
0.0017 | 1 | 0.4133 | - |
0.0833 | 50 | 0.188 | - |
0.1667 | 100 | 0.0071 | - |
0.25 | 150 | 0.0002 | - |
0.3333 | 200 | 0.0001 | - |
0.4167 | 250 | 0.0001 | - |
0.5 | 300 | 0.0001 | - |
0.5833 | 350 | 0.0001 | - |
0.6667 | 400 | 0.0001 | - |
0.75 | 450 | 0.0001 | - |
0.8333 | 500 | 0.0001 | - |
0.9167 | 550 | 0.0001 | - |
1.0 | 600 | 0.0001 | - |
Framework Versions
- Python: 3.10.13
- SetFit: 1.1.0
- Sentence Transformers: 3.0.1
- Transformers: 4.45.2
- PyTorch: 2.4.0+cu124
- Datasets: 2.21.0
- Tokenizers: 0.20.0
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}
}