README.md

---
license: apache-2.0
base_model: allenai/longformer-base-4096
tags:
- generated_from_trainer
datasets:
- essays_su_g
metrics:
- accuracy
model-index:
- name: longformer-sep_tok
  results:
  - task:
      name: Token Classification
      type: token-classification
    dataset:
      name: essays_su_g
      type: essays_su_g
      config: sep_tok
      split: test
      args: sep_tok
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.8970593132847104
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# longformer-sep_tok

This model is a fine-tuned version of [allenai/longformer-base-4096](https://huggingface.co/allenai/longformer-base-4096) on the essays_su_g dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2727
- Claim: {'precision': 0.6323296354992076, 'recall': 0.6568673565380997, 'f1-score': 0.6443649786595916, 'support': 4252.0}
- Majorclaim: {'precision': 0.8782371649250341, 'recall': 0.885884509624198, 'f1-score': 0.8820442619210586, 'support': 2182.0}
- O: {'precision': 1.0, 'recall': 0.9996489072237339, 'f1-score': 0.9998244227899218, 'support': 11393.0}
- Premise: {'precision': 0.9002495840266223, 'recall': 0.8869672131147541, 'f1-score': 0.8935590421139553, 'support': 12200.0}
- Accuracy: 0.8971
- Macro avg: {'precision': 0.852704096112716, 'recall': 0.8573419966251964, 'f1-score': 0.8549481763711319, 'support': 30027.0}
- Weighted avg: {'precision': 0.8985587647495203, 'recall': 0.8970593132847104, 'f1-score': 0.897754701815305, 'support': 30027.0}

## 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: 7

### Training results

| Training Loss | Epoch | Step | Validation Loss | Claim                                                                                                               | Majorclaim                                                                                                         | O                                                                                                                   | Premise                                                                                                             | Accuracy | Macro avg                                                                                                           | Weighted avg                                                                                                        |
|:-------------:|:-----:|:----:|:---------------:|:-------------------------------------------------------------------------------------------------------------------:|:------------------------------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------------------------------:|:--------:|:-------------------------------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------------------------------:|
| No log        | 1.0   | 41   | 0.3545          | {'precision': 0.5265911072362686, 'recall': 0.28410159924741296, 'f1-score': 0.3690803544149099, 'support': 4252.0} | {'precision': 0.5903767014878126, 'recall': 0.8547204399633364, 'f1-score': 0.6983710915558883, 'support': 2182.0} | {'precision': 0.9956778689247596, 'recall': 0.9907838146230141, 'f1-score': 0.9932248130224374, 'support': 11393.0} | {'precision': 0.8420336934350684, 'recall': 0.9136065573770492, 'f1-score': 0.8763612061170736, 'support': 12200.0} | 0.8495   | {'precision': 0.7386698427709772, 'recall': 0.7608031028027031, 'f1-score': 0.7342593662775774, 'support': 30027.0} | {'precision': 0.837374242221422, 'recall': 0.8494688114030706, 'f1-score': 0.8359340726059904, 'support': 30027.0}  |
| No log        | 2.0   | 82   | 0.2887          | {'precision': 0.5331588132635253, 'recall': 0.5747883349012229, 'f1-score': 0.5531914893617021, 'support': 4252.0}  | {'precision': 0.9024745269286754, 'recall': 0.5682859761686526, 'f1-score': 0.6974128233970753, 'support': 2182.0} | {'precision': 0.9994723419224343, 'recall': 0.997542350566137, 'f1-score': 0.9985064136355649, 'support': 11393.0}  | {'precision': 0.8662781540400063, 'recall': 0.9016393442622951, 'f1-score': 0.8836051088440838, 'support': 12200.0} | 0.8675   | {'precision': 0.8253459590386604, 'recall': 0.7605640014745769, 'f1-score': 0.7831789588096065, 'support': 30027.0} | {'precision': 0.8722740387839361, 'recall': 0.8675192326905785, 'f1-score': 0.8668828351772135, 'support': 30027.0} |
| No log        | 3.0   | 123  | 0.2610          | {'precision': 0.6448462929475588, 'recall': 0.4193320790216369, 'f1-score': 0.5081943850648425, 'support': 4252.0}  | {'precision': 0.8409090909090909, 'recall': 0.847846012832264, 'f1-score': 0.8443633044272022, 'support': 2182.0}  | {'precision': 0.9999121959785758, 'recall': 0.9995611340296673, 'f1-score': 0.9997366341848828, 'support': 11393.0} | {'precision': 0.8441453960359834, 'recall': 0.9460655737704918, 'f1-score': 0.8922042283461523, 'support': 12200.0} | 0.8846   | {'precision': 0.8324532439678022, 'recall': 0.803201199913515, 'f1-score': 0.81112463800577, 'support': 30027.0}    | {'precision': 0.8747901406867009, 'recall': 0.8846371598894328, 'f1-score': 0.8751501753304457, 'support': 30027.0} |
| No log        | 4.0   | 164  | 0.2530          | {'precision': 0.6281010374379793, 'recall': 0.6549858889934148, 'f1-score': 0.6412618005986644, 'support': 4252.0}  | {'precision': 0.8315485996705108, 'recall': 0.9252978918423465, 'f1-score': 0.8759219088937094, 'support': 2182.0} | {'precision': 1.0, 'recall': 0.9996489072237339, 'f1-score': 0.9998244227899218, 'support': 11393.0}                | {'precision': 0.9083729619565217, 'recall': 0.8768032786885246, 'f1-score': 0.8923089756423088, 'support': 12200.0} | 0.8955   | {'precision': 0.8420056497662529, 'recall': 0.8641839916870049, 'f1-score': 0.8523292769811512, 'support': 30027.0} | {'precision': 0.8978677454136914, 'recall': 0.8955273587104939, 'f1-score': 0.896362471543389, 'support': 30027.0}  |
| No log        | 5.0   | 205  | 0.2707          | {'precision': 0.6235240690281563, 'recall': 0.6458137347130762, 'f1-score': 0.6344731977818855, 'support': 4252.0}  | {'precision': 0.873358348968105, 'recall': 0.8533455545371219, 'f1-score': 0.8632359758924432, 'support': 2182.0}  | {'precision': 1.0, 'recall': 1.0, 'f1-score': 1.0, 'support': 11393.0}                                              | {'precision': 0.8975037196230782, 'recall': 0.89, 'f1-score': 0.8937361099678987, 'support': 12200.0}               | 0.8945   | {'precision': 0.848596534404835, 'recall': 0.8472898223125496, 'f1-score': 0.8478613209105568, 'support': 30027.0}  | {'precision': 0.8958416637811863, 'recall': 0.8944949545409132, 'f1-score': 0.8951257694066757, 'support': 30027.0} |
| No log        | 6.0   | 246  | 0.2700          | {'precision': 0.631960692559663, 'recall': 0.6352304797742239, 'f1-score': 0.6335913675815154, 'support': 4252.0}   | {'precision': 0.885956644674835, 'recall': 0.8615948670944088, 'f1-score': 0.8736059479553903, 'support': 2182.0}  | {'precision': 1.0, 'recall': 0.9995611340296673, 'f1-score': 0.9997805188534304, 'support': 11393.0}                | {'precision': 0.8923466470636282, 'recall': 0.8954918032786885, 'f1-score': 0.8939164586998323, 'support': 12200.0} | 0.8957   | {'precision': 0.8525659960745315, 'recall': 0.8479695710442472, 'f1-score': 0.850223573272542, 'support': 30027.0}  | {'precision': 0.8958565077303907, 'recall': 0.8956605721517301, 'f1-score': 0.8957444606797332, 'support': 30027.0} |
| No log        | 7.0   | 287  | 0.2727          | {'precision': 0.6323296354992076, 'recall': 0.6568673565380997, 'f1-score': 0.6443649786595916, 'support': 4252.0}  | {'precision': 0.8782371649250341, 'recall': 0.885884509624198, 'f1-score': 0.8820442619210586, 'support': 2182.0}  | {'precision': 1.0, 'recall': 0.9996489072237339, 'f1-score': 0.9998244227899218, 'support': 11393.0}                | {'precision': 0.9002495840266223, 'recall': 0.8869672131147541, 'f1-score': 0.8935590421139553, 'support': 12200.0} | 0.8971   | {'precision': 0.852704096112716, 'recall': 0.8573419966251964, 'f1-score': 0.8549481763711319, 'support': 30027.0}  | {'precision': 0.8985587647495203, 'recall': 0.8970593132847104, 'f1-score': 0.897754701815305, 'support': 30027.0}  |


### Framework versions

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