File size: 1,514 Bytes
ada4f10
 
 
 
 
a5deb3f
ada4f10
 
 
 
 
 
a5deb3f
ada4f10
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a5deb3f
ada4f10
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: None
tags:
- generated_from_trainer
model-index:
- name: checkpoints-mistral-150M-FA2
  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. -->

# checkpoints-mistral-150M-FA2

This model is a fine-tuned version of [None](https://huggingface.co/None) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 8.3607

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

[graelo/wikipedia 20230901 jp only](https://huggingface.co/datasets/graelo/wikipedia/tree/main/data/20230901/ja)

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0006
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 256
- total_train_batch_size: 4096
- optimizer: Adam with betas=(0.9,0.95) and epsilon=0.0001
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 1000
- training_steps: 100
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 8.361         | 2.87  | 100  | 8.3607          |


### Framework versions

- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.14.5
- Tokenizers 0.14.1