metadata
library_name: peft
tags:
- generated_from_trainer
base_model: bigcode/starcoder2-15b
model-index:
- name: peft-lora-starcoder2-15b-unity-copilot
results: []
peft-lora-starcoder2-15b-unity-copilot
This model is a fine-tuned version of bigcode/starcoder2-15b on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2024
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 2000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.1273 | 0.05 | 100 | 0.3306 |
| 0.3286 | 0.1 | 200 | 0.2493 |
| 0.2192 | 0.15 | 300 | 0.2495 |
| 0.2075 | 0.2 | 400 | 0.2442 |
| 0.2335 | 0.25 | 500 | 0.2409 |
| 0.2676 | 0.3 | 600 | 0.2447 |
| 0.2175 | 0.35 | 700 | 0.2370 |
| 0.2681 | 0.4 | 800 | 0.2334 |
| 0.2817 | 0.45 | 900 | 0.2160 |
| 0.2387 | 0.5 | 1000 | 0.2111 |
| 0.2211 | 0.55 | 1100 | 0.2145 |
| 0.1398 | 0.6 | 1200 | 0.2142 |
| 0.1735 | 0.65 | 1300 | 0.2141 |
| 0.3307 | 0.7 | 1400 | 0.2096 |
| 0.3404 | 0.75 | 1500 | 0.2068 |
| 0.2454 | 0.8 | 1600 | 0.2048 |
| 0.1532 | 0.85 | 1700 | 0.2032 |
| 0.3629 | 0.9 | 1800 | 0.2026 |
| 0.3118 | 0.95 | 1900 | 0.2024 |
| 0.2489 | 1.0 | 2000 | 0.2024 |
Framework versions
- PEFT 0.9.1.dev0
- Transformers 4.39.0.dev0
- Pytorch 2.2.1
- Datasets 2.18.0
- Tokenizers 0.15.2