metadata
license: apache-2.0
library_name: peft
tags:
- trl
- sft
- SFT
- WeniGPT
- generated_from_trainer
base_model: mistralai/Mistral-7B-Instruct-v0.2
datasets:
- generator
model-index:
- name: WeniGPT-Agents-Mistral-1.0.11-SFT
results: []
WeniGPT-Agents-Mistral-1.0.11-SFT
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: 1.1287
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: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.03
- training_steps: 330
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0466 | 0.5357 | 30 | 1.1547 |
| 0.8783 | 1.0714 | 60 | 1.1288 |
| 0.6778 | 1.6071 | 90 | 1.1287 |
| 0.4609 | 2.1429 | 120 | 1.1453 |
| 0.3946 | 2.6786 | 150 | 1.1766 |
| 0.2669 | 3.2143 | 180 | 1.2127 |
| 0.2871 | 3.75 | 210 | 1.2313 |
| 0.1831 | 4.2857 | 240 | 1.2560 |
| 0.2145 | 4.8214 | 270 | 1.2665 |
| 0.167 | 5.3571 | 300 | 1.2968 |
| 0.1569 | 5.8929 | 330 | 1.2957 |
Framework versions
- PEFT 0.10.0
- Transformers 4.40.0
- Pytorch 2.1.0+cu118
- Datasets 2.18.0
- Tokenizers 0.19.1