metadata
license: apache-2.0
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: HuggingFaceTB/SmolLM-360M-Instruct
datasets:
- generator
model-index:
- name: smolLM
results: []
smolLM
This model is a fine-tuned version of HuggingFaceTB/SmolLM-360M-Instruct on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 1.8076
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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.2932 | 0.9524 | 10 | 2.1445 |
| 2.105 | 2.0 | 21 | 2.0315 |
| 2.017 | 2.9524 | 31 | 1.9665 |
| 1.9535 | 4.0 | 42 | 1.9197 |
| 1.9104 | 4.9524 | 52 | 1.8906 |
| 1.888 | 6.0 | 63 | 1.8669 |
| 1.8552 | 6.9524 | 73 | 1.8511 |
| 1.8491 | 8.0 | 84 | 1.8384 |
| 1.8228 | 8.9524 | 94 | 1.8296 |
| 1.8198 | 10.0 | 105 | 1.8224 |
| 1.8073 | 10.9524 | 115 | 1.8173 |
| 1.7958 | 12.0 | 126 | 1.8131 |
| 1.7958 | 12.9524 | 136 | 1.8106 |
| 1.792 | 14.0 | 147 | 1.8088 |
| 1.7843 | 14.9524 | 157 | 1.8080 |
| 1.7873 | 16.0 | 168 | 1.8077 |
| 1.7848 | 16.9524 | 178 | 1.8077 |
| 1.7837 | 18.0 | 189 | 1.8076 |
| 1.7828 | 18.9524 | 199 | 1.8076 |
| 1.7827 | 19.0476 | 200 | 1.8076 |
Framework versions
- PEFT 0.10.0
- Transformers 4.40.0
- Pytorch 2.1.0
- Datasets 2.18.0
- Tokenizers 0.19.1