Llama-3.3-70B-Instruct-3d-1M-100K-0.1-reverse-padzero-plus-mul-sub-99-128D-1L-4H-512I
This model is a fine-tuned version of meta-llama/Llama-3.3-70B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.3622
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: 0.001
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.05
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 0 | 0 | 3.0591 |
| 1.8122 | 0.0640 | 500 | 1.7977 |
| 1.5897 | 0.1280 | 1000 | 1.5791 |
| 1.5392 | 0.1920 | 1500 | 1.5382 |
| 1.4605 | 0.2560 | 2000 | 1.4562 |
| 1.4384 | 0.3200 | 2500 | 1.4356 |
| 1.4288 | 0.3840 | 3000 | 1.4266 |
| 1.4222 | 0.4480 | 3500 | 1.4205 |
| 1.4048 | 0.5120 | 4000 | 1.4108 |
| 1.3962 | 0.5760 | 4500 | 1.3939 |
| 1.3928 | 0.6400 | 5000 | 1.3884 |
| 1.3855 | 0.7040 | 5500 | 1.3866 |
| 1.3855 | 0.7680 | 6000 | 1.3831 |
| 1.3797 | 0.8319 | 6500 | 1.3809 |
| 1.3799 | 0.8959 | 7000 | 1.3785 |
| 1.3788 | 0.9599 | 7500 | 1.3774 |
| 1.378 | 1.0239 | 8000 | 1.3782 |
| 1.376 | 1.0879 | 8500 | 1.3755 |
| 1.3751 | 1.1519 | 9000 | 1.3756 |
| 1.3739 | 1.2159 | 9500 | 1.3747 |
| 1.3748 | 1.2799 | 10000 | 1.3729 |
| 1.373 | 1.3439 | 10500 | 1.3723 |
| 1.3718 | 1.4079 | 11000 | 1.3719 |
| 1.3718 | 1.4719 | 11500 | 1.3718 |
| 1.3707 | 1.5359 | 12000 | 1.3699 |
| 1.3706 | 1.5999 | 12500 | 1.3696 |
| 1.3693 | 1.6639 | 13000 | 1.3691 |
| 1.369 | 1.7279 | 13500 | 1.3686 |
| 1.3674 | 1.7919 | 14000 | 1.3675 |
| 1.3666 | 1.8559 | 14500 | 1.3678 |
| 1.3691 | 1.9199 | 15000 | 1.3676 |
| 1.3682 | 1.9839 | 15500 | 1.3668 |
| 1.3667 | 2.0479 | 16000 | 1.3667 |
| 1.3657 | 2.1119 | 16500 | 1.3660 |
| 1.3655 | 2.1759 | 17000 | 1.3659 |
| 1.3646 | 2.2399 | 17500 | 1.3657 |
| 1.365 | 2.3039 | 18000 | 1.3654 |
| 1.366 | 2.3678 | 18500 | 1.3650 |
| 1.3647 | 2.4318 | 19000 | 1.3647 |
| 1.3648 | 2.4958 | 19500 | 1.3645 |
| 1.365 | 2.5598 | 20000 | 1.3645 |
| 1.3657 | 2.6238 | 20500 | 1.3643 |
| 1.3634 | 2.6878 | 21000 | 1.3639 |
| 1.3645 | 2.7518 | 21500 | 1.3641 |
| 1.3638 | 2.8158 | 22000 | 1.3636 |
| 1.3641 | 2.8798 | 22500 | 1.3638 |
| 1.3635 | 2.9438 | 23000 | 1.3634 |
| 1.3631 | 3.0078 | 23500 | 1.3634 |
| 1.3635 | 3.0718 | 24000 | 1.3631 |
| 1.3634 | 3.1358 | 24500 | 1.3632 |
| 1.3629 | 3.1998 | 25000 | 1.3630 |
| 1.3623 | 3.2638 | 25500 | 1.3629 |
| 1.364 | 3.3278 | 26000 | 1.3628 |
| 1.363 | 3.3918 | 26500 | 1.3629 |
| 1.3618 | 3.4558 | 27000 | 1.3626 |
| 1.3632 | 3.5198 | 27500 | 1.3626 |
| 1.3626 | 3.5838 | 28000 | 1.3625 |
| 1.3623 | 3.6478 | 28500 | 1.3625 |
| 1.3627 | 3.7118 | 29000 | 1.3624 |
| 1.3613 | 3.7758 | 29500 | 1.3624 |
| 1.3614 | 3.8398 | 30000 | 1.3624 |
| 1.362 | 3.9038 | 30500 | 1.3623 |
| 1.3627 | 3.9677 | 31000 | 1.3623 |
| 1.3625 | 4.0317 | 31500 | 1.3623 |
| 1.3631 | 4.0957 | 32000 | 1.3623 |
| 1.3615 | 4.1597 | 32500 | 1.3623 |
| 1.363 | 4.2237 | 33000 | 1.3623 |
| 1.3635 | 4.2877 | 33500 | 1.3623 |
| 1.3634 | 4.3517 | 34000 | 1.3622 |
| 1.3612 | 4.4157 | 34500 | 1.3622 |
| 1.3622 | 4.4797 | 35000 | 1.3622 |
| 1.3635 | 4.5437 | 35500 | 1.3622 |
| 1.3612 | 4.6077 | 36000 | 1.3622 |
| 1.3613 | 4.6717 | 36500 | 1.3622 |
| 1.3627 | 4.7357 | 37000 | 1.3622 |
| 1.3622 | 4.7997 | 37500 | 1.3622 |
| 1.3634 | 4.8637 | 38000 | 1.3622 |
| 1.3618 | 4.9277 | 38500 | 1.3622 |
| 1.3608 | 4.9917 | 39000 | 1.3622 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.9.0+cu128
- Datasets 4.5.0
- Tokenizers 0.22.1