metadata
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: models/fluent-capybara-27-replication-merged
model-index:
- name: Sheared-Llama-1.3b-koalpaca-instruction-ft-lora-240328-1
results: []
Sheared-Llama-1.3b-koalpaca-instruction-ft-lora-240328-1
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.3978
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: 2.5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 100
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 200
- num_epochs: 5.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.3713 | 0.99 | 148 | 2.4699 |
| 2.3834 | 2.0 | 297 | 2.4172 |
| 2.2654 | 3.0 | 446 | 2.3985 |
| 2.186 | 4.0 | 595 | 2.3957 |
| 2.1947 | 4.97 | 740 | 2.3978 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.2.0+cu118
- Datasets 2.18.0
- Tokenizers 0.15.2