File size: 2,371 Bytes
f70aff9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: gpt2
library_name: Distily
license: mit
tags:
- generated_from_trainer
model-index:
- name: distily_modelcard_try
  results: []
---

# distily_modelcard_try

This student model is distilled from the teacher model [gpt2](https://huggingface.co/gpt2) using the dataset (unspecified).

The [Distily](https://github.com/lapp0/distily) library was used for this distillation.

It achieves the following results on the evaluation set:
- eval_enwikippl: 38656.0
- eval_frwikippl: 218112.0
- eval_zhwikippl: 54001664.0
- eval_tinystoriesppl: 12160.0
- eval_loss: 6.4375
- eval_runtime: 0.0668
- eval_samples_per_second: 29.948
- eval_steps_per_second: 14.974

<!-- 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.

## 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:
- distillation_objective: DistillationObjective(logits_loss_component=LossComponent(label=logits, weight=1, loss_fn=kl, layer_mapper=None, projector=None), hs_loss_component=LossComponent(label=hs, weight=0, loss_fn=mse, layer_mapper=last, projector=None), attn_loss_component=LossComponent(label=attn, weight=0, loss_fn=mse, layer_mapper=layer-2, projector=None))
- train_embeddings: True
- learning_rate: 0.0001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 1.0

### Resource Usage
Peak GPU Memory: 15.4263 GB

### Eval-Phase Metrics
| step | epoch | enwikippl | frwikippl | loss | runtime | samples_per_second | steps_per_second | tinystoriesppl | zhwikippl |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| **teacher eval** |  | 43.75 | 61.75 |  |  |  |  | 11.8125 | 19.125 |
| 0 | 0 | 738734374912.0 | 47828755808256.0 | 20.375 | 0.128 | 15.619 | 7.81 | 2617245696.0 | 12232066859008.0 |
| 10 | 1.0 | 38656.0 | 218112.0 | 6.4375 | 0.0668 | 29.948 | 14.974 | 12160.0 | 54001664.0 |

### Framework versions
- Distily 0.2.0
- Transformers 4.44.0
- Pytorch 2.3.0
- Datasets 2.21.0