metadata
license: apache-2.0
base_model: EleutherAI/pythia-70m
tags:
- generated_from_trainer
model-index:
- name: pythia70M-finetuned-on-lamini-docs
results: []
inference: false
datasets:
- lamini/lamini_docs
pythia70M-finetuned-on-lamini-docs
This model is a fine-tuned version of EleutherAI/pythia-70m on lamini/lamini_docs dataset. It achieves the following results on the evaluation set:
- Loss: 2.0582
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: 1e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.3194 | 0.16 | 50 | 2.4776 |
| 2.3198 | 0.32 | 100 | 2.3780 |
| 2.16 | 0.48 | 150 | 2.2905 |
| 2.2156 | 0.63 | 200 | 2.2450 |
| 2.3342 | 0.79 | 250 | 2.1934 |
| 2.7634 | 0.95 | 300 | 2.1748 |
| 2.4963 | 1.11 | 350 | 2.1500 |
| 2.1493 | 1.27 | 400 | 2.1413 |
| 1.8731 | 1.43 | 450 | 2.1200 |
| 2.0132 | 1.59 | 500 | 2.1030 |
| 1.9606 | 1.75 | 550 | 2.0848 |
| 1.37 | 1.9 | 600 | 2.0659 |
| 1.7681 | 2.06 | 650 | 2.0744 |
| 1.8926 | 2.22 | 700 | 2.0779 |
| 1.2409 | 2.38 | 750 | 2.0683 |
| 1.489 | 2.54 | 800 | 2.0616 |
| 1.5143 | 2.7 | 850 | 2.0604 |
| 1.3736 | 2.86 | 900 | 2.0582 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1