File size: 2,610 Bytes
53f7f4a
 
 
6b76655
53f7f4a
 
 
 
6b76655
 
53f7f4a
 
 
 
 
 
 
 
 
 
6b76655
53f7f4a
6b76655
 
 
 
 
 
 
 
 
53f7f4a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f8069b4
 
 
53f7f4a
 
 
 
 
 
 
 
 
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
---
library_name: peft
tags:
- alignment-handbook
- trl
- dpo
- generated_from_trainer
base_model: norallm/normistral-7b-warm
datasets:
- hugodk-sch/aftonposten_title_prefs
model-index:
- name: ap-normistral-7b-align-scan
  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. -->

# ap-normistral-7b-align-scan

This model is a fine-tuned version of [data/ap-normistral-7b-sft-qlora](https://huggingface.co/data/ap-normistral-7b-sft-qlora) on the hugodk-sch/aftonposten_title_prefs dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9790
- Rewards/chosen: -0.0904
- Rewards/rejected: -0.1295
- Rewards/accuracies: 0.5216
- Rewards/margins: 0.0391
- Logps/rejected: -36.2902
- Logps/chosen: -32.6691
- Logits/rejected: 98.5856
- Logits/chosen: 98.6051

## 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: 5e-06
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
| 0.8779        | 0.26  | 100  | 0.9411          | 0.0053         | -0.0639          | 0.5461             | 0.0692          | -36.1263       | -32.4300     | 98.7518         | 98.7613       |
| 0.683         | 0.52  | 200  | 0.9410          | -0.1143        | -0.1935          | 0.5569             | 0.0793          | -36.4504       | -32.7289     | 98.5890         | 98.6121       |
| 0.7349        | 0.78  | 300  | 0.9809          | -0.0923        | -0.1298          | 0.5158             | 0.0375          | -36.2911       | -32.6740     | 98.5972         | 98.6205       |


### Framework versions

- PEFT 0.10.0
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1