metadata

license: mit
base_model: roberta-base
tags:
  - genre
  - books
  - multi-label
  - dataset tools
metrics:
  - f1
widget:
  - text: >-
      Meet Gertrude, a penguin detective who can't stand the cold. When a shrimp
      cocktail goes missing from the Iceberg Lounge, it's up to her to solve the
      mystery, wearing her collection of custom-made tropical turtlenecks.
    example_title: Tropical Turtlenecks
  - text: >-
      Professor Wobblebottom, a notorious forgetful scientist, invents a time
      machine but forgets how to use it. Now he is randomly popping into
      significant historical events, ruining everything. The future of the past
      is in the balance.
    example_title: When I Forgot The Time
  - text: >-
      In a world where hugs are currency and your social credit score is
      determined by your knack for dad jokes, John, a man who is allergic to
      laughter, has to navigate his way without becoming broke—or
      broken-hearted.
    example_title: Laugh Now, Pay Later
  - text: >-
      Emily, a vegan vampire, is faced with an ethical dilemma when she falls
      head over heels for a human butcher named Bob. Will she bite the forbidden
      fruit or stick to her plant-based blood substitutes?
    example_title: Love at First Bite... Or Not
  - text: >-
      Steve, a sentient self-driving car, wants to be a Broadway star. His dream
      seems unreachable until he meets Sally, a GPS system with the voice of an
      angel and ambitions of her own.
    example_title: Broadway or Bust
  - text: >-
      Dr. Fredrick Tensor, a socially awkward computer scientist, is on a quest
      to perfect AI companionship. However, his models keep outputting
      cringe-worthy, melodramatic waifus that scare away even the most die-hard
      fans of AI romance. Frustrated and lonely, Fredrick must debug his love
      life and algorithms before it's too late.
    example_title: Love.exe Has Stopped Working
language:
  - en
pipeline_tag: text-classification

BEE-spoke-data/roberta-base-description2genre

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.2130
F1: 0.6717

Model description

This classifies one or more genre labels in a multi-label setting for a given book description.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 4e-05
train_batch_size: 64
eval_batch_size: 64
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-10
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.04
num_epochs: 6.0

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.3118	1.0	62	0.2885	0.3362
0.2676	2.0	124	0.2511	0.4882
0.2325	3.0	186	0.2272	0.6093
0.2127	4.0	248	0.2181	0.6591
0.1978	5.0	310	0.2140	0.6686
0.1817	6.0	372	0.2130	0.6717

Framework versions

Transformers 4.33.3
Pytorch 2.2.0.dev20231001+cu121
Datasets 2.14.5
Tokenizers 0.13.3