Llama-2-7b-hf_oasst1_l0.0002_64
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: 1.3565
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: 0
- gradient_accumulation_steps: 16
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.03
- training_steps: 1875
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
1.401 | 0.0018 | 1 | 1.6289 |
1.4429 | 0.3392 | 187 | 1.2571 |
1.1074 | 0.6783 | 374 | 1.2502 |
1.3471 | 1.0175 | 561 | 1.2494 |
1.2256 | 1.3566 | 748 | 1.2493 |
1.2047 | 1.6958 | 935 | 1.2460 |
0.8848 | 2.0349 | 1122 | 1.2662 |
1.0369 | 2.3741 | 1309 | 1.3172 |
0.8789 | 2.7132 | 1496 | 1.3079 |
0.736 | 3.0524 | 1683 | 1.3552 |
0.7258 | 3.3915 | 1870 | 1.3486 |
Framework versions
- PEFT 0.12.1.dev0
- Transformers 4.45.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1
- Downloads last month
- 6
Model tree for alexander-hm/Llama-2-7b-hf_oasst1_l0.0002_64
Base model
meta-llama/Llama-2-7b-hf