File size: 7,410 Bytes
9d76f33 3123a57 9d76f33 3123a57 9d76f33 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 |
---
license: apache-2.0
base_model: allenai/longformer-base-4096
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: longformer-one-step
results: []
---
<!-- 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-one-step
This model is a fine-tuned version of [allenai/longformer-base-4096](https://huggingface.co/allenai/longformer-base-4096) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6204
- B-claim: {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 147.0}
- B-majorclaim: {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 77.0}
- B-premise: {'precision': 0.6069364161849711, 'recall': 0.5526315789473685, 'f1-score': 0.5785123966942148, 'support': 380.0}
- I-claim: {'precision': 0.5083841463414634, 'recall': 0.3100883310088331, 'f1-score': 0.3852151313889692, 'support': 2151.0}
- I-majorclaim: {'precision': 0.4941275167785235, 'recall': 0.562559694364852, 'f1-score': 0.5261277355962484, 'support': 1047.0}
- I-premise: {'precision': 0.8047369129323106, 'recall': 0.9100593516968498, 'f1-score': 0.8541636909012997, 'support': 6571.0}
- O: {'precision': 0.854797733046707, 'recall': 0.8704477611940299, 'f1-score': 0.862551764937882, 'support': 5025.0}
- Accuracy: 0.7676
- Macro avg: {'precision': 0.46699753218342505, 'recall': 0.45796953103027616, 'f1-score': 0.45808153135980195, 'support': 15398.0}
- Weighted avg: {'precision': 0.7419669119649179, 'recall': 0.7676321600207819, 'f1-score': 0.74985845104923, 'support': 15398.0}
## 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: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | B-claim | B-majorclaim | B-premise | I-claim | I-majorclaim | I-premise | O | Accuracy | Macro avg | Weighted avg |
|:-------------:|:-----:|:----:|:---------------:|:--------------------------------------------------------------------:|:-------------------------------------------------------------------:|:------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------------------------------:|:------------------------------------------------------------------------------------------------------------------:|:------------------------------------------------------------------------------------------------------------------:|:--------:|:----------------------------------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------------------------------:|
| No log | 1.0 | 36 | 0.8375 | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 147.0} | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 77.0} | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 380.0} | {'precision': 0.3307174887892377, 'recall': 0.13714551371455136, 'f1-score': 0.1938876109102859, 'support': 2151.0} | {'precision': 0.25, 'recall': 0.0009551098376313276, 'f1-score': 0.0019029495718363464, 'support': 1047.0} | {'precision': 0.6899198931909212, 'recall': 0.9436919799117334, 'f1-score': 0.7970949289800116, 'support': 6571.0} | {'precision': 0.7767500906782735, 'recall': 0.8523383084577114, 'f1-score': 0.8127905873422525, 'support': 5025.0} | 0.7001 | {'precision': 0.29248392466549034, 'recall': 0.2763044159888039, 'f1-score': 0.25795372525776944, 'support': 15398.0} | {'precision': 0.6111024900767319, 'recall': 0.7000909208988181, 'f1-score': 0.6326164514217569, 'support': 15398.0} |
| No log | 2.0 | 72 | 0.6930 | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 147.0} | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 77.0} | {'precision': 0.6489795918367347, 'recall': 0.41842105263157897, 'f1-score': 0.5087999999999999, 'support': 380.0} | {'precision': 0.4376956793988729, 'recall': 0.32496513249651326, 'f1-score': 0.37299893276414087, 'support': 2151.0} | {'precision': 0.3946384039900249, 'recall': 0.6045845272206304, 'f1-score': 0.47755563938136547, 'support': 1047.0} | {'precision': 0.83792191631669, 'recall': 0.8198143357175468, 'f1-score': 0.8287692307692307, 'support': 6571.0} | {'precision': 0.8073510773130546, 'recall': 0.887363184079602, 'f1-score': 0.8454683352294274, 'support': 5025.0} | 0.7363 | {'precision': 0.446655238407911, 'recall': 0.4364497474494102, 'f1-score': 0.4333703054491663, 'support': 15398.0} | {'precision': 0.7250426117598104, 'recall': 0.7362644499285621, 'f1-score': 0.7267168761345917, 'support': 15398.0} |
| No log | 3.0 | 108 | 0.6204 | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 147.0} | {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 77.0} | {'precision': 0.6069364161849711, 'recall': 0.5526315789473685, 'f1-score': 0.5785123966942148, 'support': 380.0} | {'precision': 0.5083841463414634, 'recall': 0.3100883310088331, 'f1-score': 0.3852151313889692, 'support': 2151.0} | {'precision': 0.4941275167785235, 'recall': 0.562559694364852, 'f1-score': 0.5261277355962484, 'support': 1047.0} | {'precision': 0.8047369129323106, 'recall': 0.9100593516968498, 'f1-score': 0.8541636909012997, 'support': 6571.0} | {'precision': 0.854797733046707, 'recall': 0.8704477611940299, 'f1-score': 0.862551764937882, 'support': 5025.0} | 0.7676 | {'precision': 0.46699753218342505, 'recall': 0.45796953103027616, 'f1-score': 0.45808153135980195, 'support': 15398.0} | {'precision': 0.7419669119649179, 'recall': 0.7676321600207819, 'f1-score': 0.74985845104923, 'support': 15398.0} |
### Framework versions
- Transformers 4.33.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
|