metadata
base_model: bigcode/starcoder
library_name: peft
license: bigcode-openrail-m
tags:
- generated_from_trainer
model-index:
- name: starcoder-adb-sdk
results: []
starcoder-adb-sdk
This model is a fine-tuned version of bigcode/starcoder on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.3868
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: 0.0003
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 2000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.2629 | 0.05 | 100 | 0.7794 |
| 0.0853 | 0.1 | 200 | 1.0164 |
| 0.0548 | 0.15 | 300 | 1.1201 |
| 0.0409 | 0.2 | 400 | 1.1509 |
| 0.0372 | 0.25 | 500 | 1.1885 |
| 0.0346 | 0.3 | 600 | 1.2198 |
| 0.0299 | 0.35 | 700 | 1.2742 |
| 0.0288 | 0.4 | 800 | 1.2339 |
| 0.0264 | 0.45 | 900 | 1.2372 |
| 0.028 | 0.5 | 1000 | 1.3072 |
| 0.0278 | 0.55 | 1100 | 1.3365 |
| 0.023 | 0.6 | 1200 | 1.3397 |
| 0.0253 | 0.65 | 1300 | 1.2971 |
| 0.0242 | 0.7 | 1400 | 1.3585 |
| 0.0228 | 0.75 | 1500 | 1.3603 |
| 0.0206 | 0.8 | 1600 | 1.3733 |
| 0.0209 | 0.85 | 1700 | 1.3809 |
| 0.0226 | 0.9 | 1800 | 1.3890 |
| 0.0225 | 0.95 | 1900 | 1.3855 |
| 0.024 | 1.0 | 2000 | 1.3868 |
Framework versions
- PEFT 0.13.3.dev0
- Transformers 4.46.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.1