---
base_model: roneneldan/TinyStories-33M
library_name: Distily
tags:
- generated_from_trainer
model-index:
- name: distily_bench_obj_cross
  results: []
---

# distily_bench_obj_cross

This student model is distilled from the teacher model [roneneldan/TinyStories-33M](https://huggingface.co/roneneldan/TinyStories-33M) 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: 24580.0566
- eval_frwikippl: 58429.5703
- eval_zhwikippl: 90638.1875
- eval_tinystoriesppl: 13633.8428
- eval_loss: 18.8988
- eval_runtime: 32.6253
- eval_samples_per_second: 76.628
- eval_steps_per_second: 9.594

<!-- 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=10.0, loss_fn=raw_mse, layer_mapper=None, projector=None), attn_loss_component=LossComponent(label=attn, weight=10.0, loss_fn=raw_mse, layer_mapper=None, projector=None))
- train_embeddings: True
- learning_rate: 4e-05
- 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
- num_epochs: 1.0

### Resource Usage
Peak GPU Memory: 16.2498 GB

### Eval-Phase Metrics
| step | epoch | enwikippl | frwikippl | loss | runtime | samples_per_second | steps_per_second | tinystoriesppl | zhwikippl |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| **teacher eval** |  | 169.9865 | 47377.9414 |  |  |  |  | 3.9789 | 4998.1294 |
| 0 | 0 | 30500.8262 | 64429.8789 | 19.1222 | 32.5358 | 76.838 | 9.62 | 17883.2402 | 92396.1641 |
| 2000 | 0.1293 | 24580.0566 | 58429.5703 | 18.8980 | 32.4735 | 76.986 | 9.639 | 13633.8428 | 90638.1875 |
| 4000 | 0.2586 | 24580.0566 | 58429.5703 | 18.8980 | 32.5203 | 76.875 | 9.625 | 13633.8428 | 90638.1875 |
| 6000 | 0.3879 | 24580.0566 | 58429.5703 | 18.8988 | 32.628 | 76.621 | 9.593 | 13633.8428 | 90638.1875 |
| 8000 | 0.5172 | 24580.0566 | 58429.5703 | 18.8988 | 32.6253 | 76.628 | 9.594 | 13633.8428 | 90638.1875 |
| 10000 | 0.6465 | 24580.0566 | 58429.5703 | 18.8988 | 32.4883 | 76.951 | 9.634 | 13633.8428 | 90638.1875 |
| 12000 | 0.7757 | 24580.0566 | 58429.5703 | 18.8980 | 32.4949 | 76.935 | 9.632 | 13633.8428 | 90638.1875 |
| 14000 | 0.9050 | 24580.0566 | 58429.5703 | 18.8988 | 32.507 | 76.906 | 9.629 | 13633.8428 | 90638.1875 |
| 15469 | 1.0 | 24580.0566 | 58429.5703 | 18.8988 | 32.6353 | 76.604 | 9.591 | 13633.8428 | 90638.1875 |

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