|
--- |
|
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: 1280.0 |
|
- eval_frwikippl: 5952.0 |
|
- eval_zhwikippl: 45056.0 |
|
- eval_tinystoriesppl: 816.0 |
|
- eval_loss: 2.4587 |
|
- eval_runtime: 12.9968 |
|
- eval_samples_per_second: 46.165 |
|
- eval_steps_per_second: 11.541 |
|
|
|
<!-- 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=uniform_cons, 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 | 13.0242 | 46.068 | 11.517 | 12079595520.0 | 98956046499840.0 | |
|
| 750 | 0.1010 | 1280.0 | 5952.0 | 2.4587 | 12.9968 | 46.165 | 11.541 | 816.0 | 45056.0 | |
|
| 1500 | 0.2020 | 506.0 | 3248.0 | 1.7942 | 13.0259 | 46.062 | 11.515 | 352.0 | 736.0 | |
|
| 2250 | 0.3030 | 338.0 | 1624.0 | 1.5654 | 13.2237 | 45.373 | 11.343 | 278.0 | 306.0 | |
|
| 3000 | 0.4040 | 241.0 | 920.0 | 1.3552 | 13.084 | 45.857 | 11.464 | 210.0 | 186.0 | |
|
| 3750 | 0.5051 | 200.0 | 624.0 | 1.2059 | 13.1502 | 45.627 | 11.407 | 169.0 | 175.0 | |
|
| 4500 | 0.6061 | 159.0 | 510.0 | 1.0607 | 12.9758 | 46.24 | 11.56 | 128.0 | 173.0 | |
|
| 5250 | 0.7071 | 128.0 | 464.0 | 0.9374 | 13.0443 | 45.997 | 11.499 | 106.0 | 130.0 | |
|
| 6000 | 0.8081 | 117.5 | 414.0 | 0.8786 | 12.9003 | 46.51 | 11.628 | 96.0 | 129.0 | |
|
| 6750 | 0.9091 | 112.5 | 398.0 | 0.8508 | 12.8837 | 46.571 | 11.643 | 93.0 | 125.0 | |
|
| 7425 | 1.0 | 112.0 | 396.0 | 0.8456 | 12.9485 | 46.337 | 11.584 | 91.5 | 125.0 | |
|
|
|
### Framework versions |
|
- Distily 0.2.0 |
|
- Transformers 4.44.0 |
|
- Pytorch 2.3.0 |
|
- Datasets 2.21.0 |
|
|
