File size: 3,002 Bytes
9465ccf
 
 
 
 
 
 
6eab4d4
9465ccf
 
 
 
 
 
 
 
d5ff321
9465ccf
 
 
6eab4d4
9465ccf
6394933
e336b22
9465ccf
 
f504d80
 
 
 
 
9465ccf
 
 
 
 
 
 
f504d80
 
dac3bf6
 
 
f504d80
 
 
9579804
dac3bf6
 
 
f504d80
 
 
 
9579804
dac3bf6
 
 
f504d80
 
 
 
9579804
dac3bf6
 
 
9465ccf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d5ff321
9465ccf
 
 
 
 
 
 
 
 
 
 
 
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
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: mistralai/Mistral-7B-Instruct-v0.1
datasets:
- generator
model-index:
- name: Mistral-7B-v0.1_Emotion
  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. -->

# Mistral-7B-v0.1_Emotion

This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset.

Dataset: [dair-ai/emotion](https://huggingface.co/datasets/dair-ai/emotion)

## Model description

Article: https://ai.plainenglish.io/fine-tuning-the-mistral-7b-instruct-v0-1-model-with-the-emotion-dataset-c84c50b553dc

Fine tunning: https://github.com/frank-morales2020/MLxDL/blob/main/FineTuning_Mistral_7b_hfdeployment_dataset_Emotion.ipynb

Evaluation: https://github.com/frank-morales2020/MLxDL/blob/main/FineTunning_Testing_For_EmotionQADataset.ipynb

## Intended uses & limitations

More information needed

## Training and evaluation data

Evaluation: https://github.com/frank-morales2020/MLxDL/blob/main/FineTunning_Testing_For_EmotionQADataset.ipynb


*************

The following hyperparameters were used during training:
learning_rate: 0.0002 train_batch_size: 3 eval_batch_size: 8 seed: 42 gradient_accumulation_steps: 2 total_train_batch_size: 6 optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 lr_scheduler_type: constant lr_scheduler_warmup_ratio: 0.03
num_epochs: 1

Accuracy (Eval dataset and predict) for a sample of 2000: 59.45%

*************

The following hyperparameters were used during training:
learning_rate: 0.0002 train_batch_size: 3 eval_batch_size: 8 seed: 42 gradient_accumulation_steps: 2 total_train_batch_size: 6 optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 lr_scheduler_type: constant lr_scheduler_warmup_ratio: 0.03
num_epochs: 25

Accuracy (Eval dataset and predict) for a sample of 2000: 79.95%

*************

The following hyperparameters were used during training:
learning_rate: 0.0002 train_batch_size: 3 eval_batch_size: 8 seed: 42 gradient_accumulation_steps: 2 total_train_batch_size: 6 optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 lr_scheduler_type: constant lr_scheduler_warmup_ratio: 0.03
num_epochs: 40

 Accuracy (Eval dataset and predict) for a sample of 2000: 80.70%

*************

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 3
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 6
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 40

### Training results



### Framework versions

- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1