File size: 3,162 Bytes
b47538a 1fa4b92 041f7c8 1fa4b92 041f7c8 b47538a 041f7c8 b47538a 29c575f b47538a 041f7c8 b47538a |
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 |
---
base_model: gpt2
library_name: Distily
license: mit
tags:
- generated_from_trainer
model-index:
- name: distily_bench_obj_cross_v2.13_gpt2
results: []
---
# distily_bench_obj_cross_v2.13_gpt2
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: 1272.0
- eval_frwikippl: 5952.0
- eval_zhwikippl: 44800.0
- eval_tinystoriesppl: 816.0
- eval_loss: 2.4588
- eval_runtime: 12.986
- eval_samples_per_second: 46.204
- eval_steps_per_second: 11.551
<!-- 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=1.0, loss_fn=mse, layer_mapper=last, projector=None), attn_loss_component=LossComponent(label=attn, weight=0, loss_fn=None, layer_mapper=None, projector=None))
- train_embeddings: True
- learning_rate: 0.0001
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.5
- num_epochs: 1.0
### Resource Usage
Peak GPU Memory: 8.0905 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 | 1821066133504.0 | 158329674399744.0 | 19.3254 | 12.8793 | 46.586 | 11.647 | 12079595520.0 | 98956046499840.0 |
| 750 | 0.1010 | 1272.0 | 5952.0 | 2.4588 | 12.986 | 46.204 | 11.551 | 816.0 | 44800.0 |
| 1500 | 0.2020 | 506.0 | 3248.0 | 1.7942 | 13.1654 | 45.574 | 11.393 | 350.0 | 736.0 |
| 2250 | 0.3030 | 338.0 | 1632.0 | 1.5655 | 13.13 | 45.697 | 11.424 | 278.0 | 306.0 |
| 3000 | 0.4040 | 241.0 | 916.0 | 1.3548 | 13.0817 | 45.865 | 11.466 | 209.0 | 188.0 |
| 3750 | 0.5051 | 199.0 | 624.0 | 1.2057 | 13.1347 | 45.68 | 11.42 | 169.0 | 174.0 |
| 4500 | 0.6061 | 160.0 | 510.0 | 1.0606 | 13.0674 | 45.916 | 11.479 | 129.0 | 171.0 |
| 5250 | 0.7071 | 128.0 | 468.0 | 0.9370 | 13.0344 | 46.032 | 11.508 | 106.5 | 132.0 |
| 6000 | 0.8081 | 118.0 | 418.0 | 0.8787 | 13.3306 | 45.009 | 11.252 | 97.0 | 131.0 |
| 6750 | 0.9091 | 112.0 | 400.0 | 0.8501 | 13.0719 | 45.9 | 11.475 | 93.0 | 129.0 |
| 7425 | 1.0 | 112.0 | 396.0 | 0.8454 | 13.0273 | 46.057 | 11.514 | 91.5 | 128.0 |
### Framework versions
- Distily 0.2.0
- Transformers 4.44.0
- Pytorch 2.3.0
- Datasets 2.21.0
|