File size: 3,051 Bytes
0175b5f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
---
license: apache-2.0
base_model: allenai/OLMo-1B
tags:
- generated_from_trainer
model-index:
- name: O0428HMA24
  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. -->

# O0428HMA24

This model is a fine-tuned version of [allenai/OLMo-1B](https://huggingface.co/allenai/OLMo-1B) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0551

## 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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine_with_restarts
- lr_scheduler_warmup_steps: 80
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.3605        | 0.09  | 10   | 0.1809          |
| 0.1688        | 0.18  | 20   | 0.1604          |
| 0.1494        | 0.27  | 30   | 0.1601          |
| 0.1569        | 0.36  | 40   | 0.1538          |
| 0.1533        | 0.45  | 50   | 0.1535          |
| 0.1529        | 0.54  | 60   | 0.1502          |
| 0.1499        | 0.63  | 70   | 0.1480          |
| 0.15          | 0.73  | 80   | 0.1548          |
| 0.1475        | 0.82  | 90   | 0.1495          |
| 0.1479        | 0.91  | 100  | 0.1459          |
| 0.1355        | 1.0   | 110  | 0.1022          |
| 0.2371        | 1.09  | 120  | 0.1226          |
| 0.1134        | 1.18  | 130  | 0.0893          |
| 0.0964        | 1.27  | 140  | 0.0853          |
| 0.0865        | 1.36  | 150  | 0.0728          |
| 0.0896        | 1.45  | 160  | 0.0597          |
| 0.0643        | 1.54  | 170  | 0.0606          |
| 0.0606        | 1.63  | 180  | 0.0574          |
| 0.0631        | 1.72  | 190  | 0.0569          |
| 0.0577        | 1.81  | 200  | 0.0625          |
| 0.0584        | 1.9   | 210  | 0.0613          |
| 0.0601        | 1.99  | 220  | 0.0564          |
| 0.0582        | 2.08  | 230  | 0.0578          |
| 0.0548        | 2.18  | 240  | 0.0587          |
| 0.0561        | 2.27  | 250  | 0.0592          |
| 0.061         | 2.36  | 260  | 0.0571          |
| 0.0534        | 2.45  | 270  | 0.0559          |
| 0.052         | 2.54  | 280  | 0.0556          |
| 0.0549        | 2.63  | 290  | 0.0571          |
| 0.0568        | 2.72  | 300  | 0.0551          |
| 0.0567        | 2.81  | 310  | 0.0549          |
| 0.0577        | 2.9   | 320  | 0.0551          |
| 0.0607        | 2.99  | 330  | 0.0551          |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1