File size: 1,906 Bytes
d419245
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: allenai/longformer-base-4096
tags:
- generated_from_trainer
model-index:
- name: clinical_longformer_new_tokens_v2
  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. -->

# clinical_longformer_new_tokens_v2

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: 4.8474

## 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: 2
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 64
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1500
- num_epochs: 20

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.5192        | 1.42  | 2    | 4.9585          |
| 2.6002        | 2.84  | 4    | 4.5364          |
| 2.0541        | 4.0   | 6    | 5.1700          |
| 1.8191        | 5.0   | 8    | 4.7585          |
| 1.8123        | 6.0   | 10   | 4.3768          |
| 2.0364        | 7.11  | 12   | 4.8248          |
| 2.4587        | 8.53  | 14   | 4.9780          |
| 2.4665        | 9.96  | 16   | 4.2456          |
| 1.7881        | 11.0  | 18   | 4.1161          |
| 1.7651        | 12.0  | 20   | 4.8474          |


### Framework versions

- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1