File size: 1,821 Bytes
678cfe7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: gptj-6b-gpteacher-role-play-chatml
  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. -->

# gptj-6b-gpteacher-role-play-chatml

This model is a fine-tuned version of [EleutherAI/gpt-j-6b](https://huggingface.co/EleutherAI/gpt-j-6b) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.9268
- Accuracy: 0.1592
- Entropy: 1.9050

## 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: 1e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 99
- distributed_type: multi-GPU
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 5.0

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | Entropy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:-------:|
| 2.0652        | 0.99  | 59   | 2.0918          | 0.1549   | 2.3687  |
| 2.083         | 2.0   | 119  | 2.0078          | 0.1581   | 2.1805  |
| 2.083         | 2.99  | 178  | 1.9658          | 0.1594   | 2.0026  |
| 1.9201        | 4.0   | 238  | 1.9424          | 0.1596   | 1.9390  |
| 1.9201        | 4.96  | 295  | 1.9268          | 0.1592   | 1.9050  |


### Framework versions

- Transformers 4.28.1
- Pytorch 1.13.1
- Datasets 2.7.1
- Tokenizers 0.13.3