File size: 2,692 Bytes
570dae1
dac6bdf
 
 
 
 
 
 
 
570dae1
 
dac6bdf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
base_model: TheBloke/Mistral-7B-Instruct-v0.2-GPTQ
model-index:
- name: brevity_ss-promt_e30
  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. -->

# brevity_ss-promt_e30

This model is a fine-tuned version of [TheBloke/Mistral-7B-Instruct-v0.2-GPTQ](https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.2-GPTQ) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4920

## 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: 0.0002
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2
- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 3.9164        | 0.8   | 2    | 3.1402          |
| 2.3902        | 2.0   | 5    | 2.6152          |
| 3.053         | 2.8   | 7    | 2.3220          |
| 1.7893        | 4.0   | 10   | 1.9873          |
| 2.3626        | 4.8   | 12   | 1.8054          |
| 1.3909        | 6.0   | 15   | 1.5493          |
| 1.8073        | 6.8   | 17   | 1.3868          |
| 1.0338        | 8.0   | 20   | 1.1445          |
| 1.2723        | 8.8   | 22   | 0.9763          |
| 0.6819        | 10.0  | 25   | 0.7926          |
| 0.8371        | 10.8  | 27   | 0.7085          |
| 0.4868        | 12.0  | 30   | 0.6329          |
| 0.6522        | 12.8  | 32   | 0.5975          |
| 0.4052        | 14.0  | 35   | 0.5653          |
| 0.5732        | 14.8  | 37   | 0.5482          |
| 0.3641        | 16.0  | 40   | 0.5303          |
| 0.5239        | 16.8  | 42   | 0.5213          |
| 0.3372        | 18.0  | 45   | 0.5101          |
| 0.4896        | 18.8  | 47   | 0.5041          |
| 0.3174        | 20.0  | 50   | 0.4974          |
| 0.4645        | 20.8  | 52   | 0.4949          |
| 0.3046        | 22.0  | 55   | 0.4929          |
| 0.4505        | 22.8  | 57   | 0.4923          |
| 0.2972        | 24.0  | 60   | 0.4920          |


### Framework versions

- PEFT 0.10.0
- Transformers 4.40.2
- Pytorch 2.1.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1