metadata
license: apache-2.0
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- generator
base_model: mistralai/Mistral-7B-Instruct-v0.2
model-index:
- name: mistral7binstruct_summarize
results: []
mistral7binstruct_summarize
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 1.5218
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: 0.0002
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_steps: 3
- training_steps: 200
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.6377 | 0.03 | 25 | 1.6268 |
| 1.5147 | 0.05 | 50 | 1.5521 |
| 1.5779 | 0.08 | 75 | 1.5416 |
| 1.548 | 0.11 | 100 | 1.5338 |
| 1.421 | 0.13 | 125 | 1.5284 |
| 1.6 | 0.16 | 150 | 1.5293 |
| 1.5307 | 0.19 | 175 | 1.5227 |
| 1.5756 | 0.22 | 200 | 1.5218 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2