Llama2-7B-lora-r-32-generic-step-400-lr-1e-5-labels_40.0
This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 3.2840
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 20
- training_steps: 400
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 5.5837 | 0.3653 | 20 | 5.4090 |
| 4.8334 | 0.7306 | 40 | 4.7528 |
| 4.3088 | 1.0959 | 60 | 4.2463 |
| 3.9893 | 1.4612 | 80 | 3.9851 |
| 3.8077 | 1.8265 | 100 | 3.8169 |
| 3.6972 | 2.1918 | 120 | 3.6900 |
| 3.5444 | 2.5571 | 140 | 3.5944 |
| 3.5291 | 2.9224 | 160 | 3.5185 |
| 3.3636 | 3.2877 | 180 | 3.4609 |
| 3.3943 | 3.6530 | 200 | 3.4153 |
| 3.2779 | 4.0183 | 220 | 3.3784 |
| 3.3483 | 4.3836 | 240 | 3.3493 |
| 3.3144 | 4.7489 | 260 | 3.3271 |
| 3.247 | 5.1142 | 280 | 3.3116 |
| 3.2224 | 5.4795 | 300 | 3.3001 |
| 3.213 | 5.8447 | 320 | 3.2926 |
| 3.1827 | 6.2100 | 340 | 3.2878 |
| 3.1775 | 6.5753 | 360 | 3.2852 |
| 3.1628 | 6.9406 | 380 | 3.2842 |
| 3.2233 | 7.3059 | 400 | 3.2840 |
Framework versions
- PEFT 0.15.2
- Transformers 4.45.2
- Pytorch 2.5.0+cu121
- Datasets 3.2.0
- Tokenizers 0.20.3
- Downloads last month
- 1
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for Siqi-Hu/Llama2-7B-lora-r-32-generic-step-400-lr-1e-5-labels_40.0
Base model
meta-llama/Llama-2-7b-hf