metadata
base_model: microsoft/Phi-3.5-mini-instruct
library_name: peft
license: mit
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: Phi-3.5-MultiCap-tool-embedding-step1
results: []
Phi-3.5-MultiCap-tool-embedding-step1
This model is a fine-tuned version of microsoft/Phi-3.5-mini-instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5820
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.7683 | 0.2256 | 50 | 0.7572 |
| 0.5568 | 0.4512 | 100 | 0.5648 |
| 0.5259 | 0.6768 | 150 | 0.5344 |
| 0.5689 | 0.9024 | 200 | 0.5199 |
| 0.4983 | 1.1280 | 250 | 0.5107 |
| 0.4835 | 1.3536 | 300 | 0.5050 |
| 0.4492 | 1.5792 | 350 | 0.5019 |
| 0.4918 | 1.8049 | 400 | 0.4996 |
| 0.4735 | 2.0305 | 450 | 0.4997 |
| 0.4139 | 2.2561 | 500 | 0.5017 |
| 0.451 | 2.4817 | 550 | 0.5025 |
| 0.4516 | 2.7073 | 600 | 0.5047 |
| 0.4586 | 2.9329 | 650 | 0.5086 |
| 0.4393 | 3.1585 | 700 | 0.5176 |
| 0.4207 | 3.3841 | 750 | 0.5206 |
| 0.3999 | 3.6097 | 800 | 0.5249 |
| 0.414 | 3.8353 | 850 | 0.5327 |
| 0.4002 | 4.0609 | 900 | 0.5408 |
| 0.3651 | 4.2865 | 950 | 0.5498 |
| 0.3775 | 4.5121 | 1000 | 0.5528 |
| 0.4012 | 4.7377 | 1050 | 0.5595 |
| 0.3676 | 4.9633 | 1100 | 0.5668 |
| 0.3634 | 5.1889 | 1150 | 0.5741 |
| 0.3821 | 5.4146 | 1200 | 0.5793 |
| 0.3903 | 5.6402 | 1250 | 0.5815 |
| 0.3655 | 5.8658 | 1300 | 0.5820 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.2
- Pytorch 2.4.1+cu124
- Datasets 3.0.0
- Tokenizers 0.19.1