meta_data/README_s42_e6.md

metadata

base_model: allenai/longformer-base-4096
tags:
  - generated_from_trainer
datasets:
  - essays_su_g
metrics:
  - accuracy
model-index:
  - name: longformer-spans
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: essays_su_g
          type: essays_su_g
          config: spans
          split: train[80%:100%]
          args: spans
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9382671349433361

longformer-spans

This model is a fine-tuned version of allenai/longformer-base-4096 on the essays_su_g dataset. It achieves the following results on the evaluation set:

• Loss: 0.1775
• B: {'precision': 0.8277385159010601, 'recall': 0.8983700862895494, 'f1-score': 0.8616091954022989, 'support': 1043.0}
• I: {'precision': 0.9442383361439011, 'recall': 0.9681844380403458, 'f1-score': 0.9560614684120662, 'support': 17350.0}
• O: {'precision': 0.9404392319190525, 'recall': 0.886516366789508, 'f1-score': 0.9126820286782348, 'support': 9226.0}
• Accuracy: 0.9383
• Macro avg: {'precision': 0.9041386946546712, 'recall': 0.9176902970398011, 'f1-score': 0.9101175641641999, 'support': 27619.0}
• Weighted avg: {'precision': 0.9385697801465175, 'recall': 0.9382671349433361, 'f1-score': 0.9380038837155341, 'support': 27619.0}

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	B	I	O	Accuracy	Macro avg	Weighted avg
No log	1.0	41	0.2767	{'precision': 0.8252032520325203, 'recall': 0.38926174496644295, 'f1-score': 0.5289902280130293, 'support': 1043.0}	{'precision': 0.8942471288813271, 'recall': 0.9693948126801153, 'f1-score': 0.9303058797499861, 'support': 17350.0}	{'precision': 0.9191008534679649, 'recall': 0.8287448515066117, 'f1-score': 0.8715873468224564, 'support': 9226.0}	0.9005	{'precision': 0.8795170781272708, 'recall': 0.7291338030510567, 'f1-score': 0.7769611515284907, 'support': 27619.0}	{'precision': 0.8999420381641764, 'recall': 0.9005032767297875, 'f1-score': 0.8955359963526497, 'support': 27619.0}
No log	2.0	82	0.2284	{'precision': 0.7671568627450981, 'recall': 0.900287631831256, 'f1-score': 0.8284075871195412, 'support': 1043.0}	{'precision': 0.9168915272531031, 'recall': 0.9792507204610951, 'f1-score': 0.947045707915273, 'support': 17350.0}	{'precision': 0.9646535282898919, 'recall': 0.8223498807717321, 'f1-score': 0.8878357030015798, 'support': 9226.0}	0.9239	{'precision': 0.8829006394293644, 'recall': 0.900629411021361, 'f1-score': 0.8877629993454645, 'support': 27619.0}	{'precision': 0.9271916455225395, 'recall': 0.9238567652702849, 'f1-score': 0.9227866447586169, 'support': 27619.0}
No log	3.0	123	0.1770	{'precision': 0.8351648351648352, 'recall': 0.8744007670182167, 'f1-score': 0.8543325526932084, 'support': 1043.0}	{'precision': 0.9442345644206371, 'recall': 0.9651873198847263, 'f1-score': 0.954595981188542, 'support': 17350.0}	{'precision': 0.9328935395814377, 'recall': 0.8890093214827661, 'f1-score': 0.9104229104229105, 'support': 9226.0}	0.9363	{'precision': 0.90409764638897, 'recall': 0.909532469461903, 'f1-score': 0.906450481434887, 'support': 27619.0}	{'precision': 0.9363272534108158, 'recall': 0.9363119591585503, 'f1-score': 0.9360538360419275, 'support': 27619.0}
No log	4.0	164	0.1804	{'precision': 0.8234265734265734, 'recall': 0.9031639501438159, 'f1-score': 0.8614540466392319, 'support': 1043.0}	{'precision': 0.9435452033162258, 'recall': 0.9642651296829972, 'f1-score': 0.9537926512927226, 'support': 17350.0}	{'precision': 0.9335544373284538, 'recall': 0.8847821374376761, 'f1-score': 0.9085141903171953, 'support': 9226.0}	0.9354	{'precision': 0.9001754046904177, 'recall': 0.917403739088163, 'f1-score': 0.9079202960830499, 'support': 27619.0}	{'precision': 0.9356716909523425, 'recall': 0.9354067851841124, 'f1-score': 0.9351805275513196, 'support': 27619.0}
No log	5.0	205	0.1774	{'precision': 0.8283450704225352, 'recall': 0.9022051773729626, 'f1-score': 0.8636989444699403, 'support': 1043.0}	{'precision': 0.9497974784642592, 'recall': 0.9595965417867435, 'f1-score': 0.9546718655924767, 'support': 17350.0}	{'precision': 0.9269600178691088, 'recall': 0.8996314762627358, 'f1-score': 0.9130913091309132, 'support': 9226.0}	0.9374	{'precision': 0.9017008555853011, 'recall': 0.9204777318074807, 'f1-score': 0.9104873730644435, 'support': 27619.0}	{'precision': 0.9375822182072485, 'recall': 0.9373981679278758, 'f1-score': 0.9373465833358712, 'support': 27619.0}
No log	6.0	246	0.1775	{'precision': 0.8277385159010601, 'recall': 0.8983700862895494, 'f1-score': 0.8616091954022989, 'support': 1043.0}	{'precision': 0.9442383361439011, 'recall': 0.9681844380403458, 'f1-score': 0.9560614684120662, 'support': 17350.0}	{'precision': 0.9404392319190525, 'recall': 0.886516366789508, 'f1-score': 0.9126820286782348, 'support': 9226.0}	0.9383	{'precision': 0.9041386946546712, 'recall': 0.9176902970398011, 'f1-score': 0.9101175641641999, 'support': 27619.0}	{'precision': 0.9385697801465175, 'recall': 0.9382671349433361, 'f1-score': 0.9380038837155341, 'support': 27619.0}

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2