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: 1.5138
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: 0.0005
- 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: cosine
- lr_scheduler_warmup_steps: 3
- training_steps: 300
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.5958 | 0.03 | 25 | 1.5827 |
| 1.5245 | 0.05 | 50 | 1.5574 |
| 1.5865 | 0.08 | 75 | 1.5528 |
| 1.5586 | 0.11 | 100 | 1.5438 |
| 1.4311 | 0.13 | 125 | 1.5382 |
| 1.6084 | 0.16 | 150 | 1.5360 |
| 1.531 | 0.19 | 175 | 1.5254 |
| 1.5787 | 0.22 | 200 | 1.5217 |
| 1.4842 | 0.24 | 225 | 1.5182 |
| 1.4974 | 0.27 | 250 | 1.5154 |
| 1.5111 | 0.3 | 275 | 1.5140 |
| 1.4698 | 0.32 | 300 | 1.5138 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2