File size: 2,873 Bytes
c375137
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: gemma
base_model: google/gemma-2-2b
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: collapse_gemma-2-2b_hs2_accumulatesubsample_iter7_sftsd0
  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. -->

# collapse_gemma-2-2b_hs2_accumulatesubsample_iter7_sftsd0

This model is a fine-tuned version of [google/gemma-2-2b](https://huggingface.co/google/gemma-2-2b) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1806
- Num Input Tokens Seen: 5046464

## 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: 8e-06
- train_batch_size: 8
- eval_batch_size: 16
- seed: 0
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_ratio: 0.05
- num_epochs: 1

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Input Tokens Seen |
|:-------------:|:------:|:----:|:---------------:|:-----------------:|
| No log        | 0      | 0    | 1.3909          | 0                 |
| 1.3617        | 0.0544 | 5    | 1.2733          | 278136            |
| 1.199         | 0.1088 | 10   | 1.2001          | 559424            |
| 1.0696        | 0.1632 | 15   | 1.1783          | 845472            |
| 1.01          | 0.2175 | 20   | 1.1976          | 1120848           |
| 0.9704        | 0.2719 | 25   | 1.1931          | 1401104           |
| 0.7928        | 0.3263 | 30   | 1.1923          | 1680784           |
| 0.8869        | 0.3807 | 35   | 1.1986          | 1956960           |
| 0.7529        | 0.4351 | 40   | 1.2010          | 2233408           |
| 0.6472        | 0.4895 | 45   | 1.2048          | 2509944           |
| 0.6316        | 0.5438 | 50   | 1.2074          | 2783928           |
| 0.6944        | 0.5982 | 55   | 1.1871          | 3057160           |
| 0.5639        | 0.6526 | 60   | 1.1970          | 3334664           |
| 0.5933        | 0.7070 | 65   | 1.1907          | 3613112           |
| 0.5994        | 0.7614 | 70   | 1.1841          | 3885304           |
| 0.5655        | 0.8158 | 75   | 1.1783          | 4165304           |
| 0.5658        | 0.8702 | 80   | 1.1869          | 4444120           |
| 0.5515        | 0.9245 | 85   | 1.1819          | 4716040           |
| 0.4924        | 0.9789 | 90   | 1.1824          | 4993184           |


### Framework versions

- Transformers 4.44.0
- Pytorch 2.4.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1