--- license: other base_model: Gryphe/MythoMax-L2-13b tags: - generated_from_trainer model-index: - name: lora-out-10 results: [] --- [Built with Axolotl](https://github.com/OpenAccess-AI-Collective/axolotl) # lora-out-10 This model is a fine-tuned version of [Gryphe/MythoMax-L2-13b](https://huggingface.co/Gryphe/MythoMax-L2-13b) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.3963 ## 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: 2 - eval_batch_size: 2 - seed: 42 - gradient_accumulation_steps: 3 - total_train_batch_size: 6 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 20 - num_epochs: 1 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:-----:|:---------------:| | 1.9065 | 0.01 | 200 | 1.9889 | | 1.764 | 0.01 | 400 | 1.8015 | | 1.6177 | 0.02 | 600 | 1.6777 | | 1.5756 | 0.03 | 800 | 1.5862 | | 1.4847 | 0.03 | 1000 | 1.5134 | | 1.5603 | 0.04 | 1200 | 1.4480 | | 1.347 | 0.05 | 1400 | 1.3939 | | 1.2312 | 0.06 | 1600 | 1.3450 | | 1.3749 | 0.06 | 1800 | 1.3037 | | 1.2194 | 0.07 | 2000 | 1.2657 | | 1.2434 | 0.08 | 2200 | 1.2297 | | 1.2037 | 0.08 | 2400 | 1.1983 | | 1.2709 | 0.09 | 2600 | 1.1679 | | 1.1537 | 0.1 | 2800 | 1.1405 | | 1.1408 | 0.1 | 3000 | 1.1154 | | 1.0502 | 0.11 | 3200 | 1.0921 | | 1.0453 | 0.12 | 3400 | 1.0717 | | 0.9332 | 0.13 | 3600 | 1.0462 | | 1.1569 | 0.13 | 3800 | 1.0284 | | 1.146 | 0.14 | 4000 | 1.0092 | | 1.0325 | 0.15 | 4200 | 0.9916 | | 0.9936 | 0.15 | 4400 | 0.9748 | | 0.9131 | 0.16 | 4600 | 0.9592 | | 0.9426 | 0.17 | 4800 | 0.9416 | | 0.8971 | 0.17 | 5000 | 0.9278 | | 0.8258 | 0.18 | 5200 | 0.9124 | | 0.8038 | 0.19 | 5400 | 0.8991 | | 0.9663 | 0.2 | 5600 | 0.8841 | | 0.7214 | 0.2 | 5800 | 0.8726 | | 0.8789 | 0.21 | 6000 | 0.8621 | | 0.8095 | 0.22 | 6200 | 0.8492 | | 0.8446 | 0.22 | 6400 | 0.8379 | | 0.8704 | 0.23 | 6600 | 0.8261 | | 0.8175 | 0.24 | 6800 | 0.8165 | | 0.7169 | 0.24 | 7000 | 0.8066 | | 0.6899 | 0.25 | 7200 | 0.7952 | | 0.8256 | 0.26 | 7400 | 0.7860 | | 0.7186 | 0.27 | 7600 | 0.7759 | | 0.8891 | 0.27 | 7800 | 0.7664 | | 0.7919 | 0.28 | 8000 | 0.7580 | | 0.8378 | 0.29 | 8200 | 0.7487 | | 0.6804 | 0.29 | 8400 | 0.7390 | | 0.6921 | 0.3 | 8600 | 0.7308 | | 0.7263 | 0.31 | 8800 | 0.7221 | | 0.6268 | 0.31 | 9000 | 0.7145 | | 0.6954 | 0.32 | 9200 | 0.7067 | | 0.8179 | 0.33 | 9400 | 0.6992 | | 0.5709 | 0.34 | 9600 | 0.6929 | | 0.7092 | 0.34 | 9800 | 0.6858 | | 0.7396 | 0.35 | 10000 | 0.6774 | | 0.5967 | 0.36 | 10200 | 0.6702 | | 0.6984 | 0.36 | 10400 | 0.6636 | | 0.5621 | 0.37 | 10600 | 0.6578 | | 0.5342 | 0.38 | 10800 | 0.6512 | | 0.6695 | 0.38 | 11000 | 0.6442 | | 0.7242 | 0.39 | 11200 | 0.6374 | | 0.6046 | 0.4 | 11400 | 0.6317 | | 0.5994 | 0.41 | 11600 | 0.6260 | | 0.6133 | 0.41 | 11800 | 0.6201 | | 0.6707 | 0.42 | 12000 | 0.6138 | | 0.6454 | 0.43 | 12200 | 0.6078 | | 0.6534 | 0.43 | 12400 | 0.6013 | | 0.6382 | 0.44 | 12600 | 0.5963 | | 0.6659 | 0.45 | 12800 | 0.5908 | | 0.6608 | 0.45 | 13000 | 0.5851 | | 0.6864 | 0.46 | 13200 | 0.5793 | | 0.6046 | 0.47 | 13400 | 0.5748 | | 0.5331 | 0.48 | 13600 | 0.5692 | | 0.5314 | 0.48 | 13800 | 0.5640 | | 0.5528 | 0.49 | 14000 | 0.5594 | | 0.5617 | 0.5 | 14200 | 0.5545 | | 0.6572 | 0.5 | 14400 | 0.5488 | | 0.6108 | 0.51 | 14600 | 0.5441 | | 0.4776 | 0.52 | 14800 | 0.5387 | | 0.493 | 0.52 | 15000 | 0.5357 | | 0.5464 | 0.53 | 15200 | 0.5305 | | 0.5519 | 0.54 | 15400 | 0.5260 | | 0.4209 | 0.55 | 15600 | 0.5225 | | 0.4759 | 0.55 | 15800 | 0.5173 | | 0.5357 | 0.56 | 16000 | 0.5129 | | 0.6064 | 0.57 | 16200 | 0.5091 | | 0.4835 | 0.57 | 16400 | 0.5048 | | 0.4951 | 0.58 | 16600 | 0.5017 | | 0.3621 | 0.59 | 16800 | 0.4971 | | 0.5166 | 0.59 | 17000 | 0.4935 | | 0.5464 | 0.6 | 17200 | 0.4896 | | 0.5093 | 0.61 | 17400 | 0.4858 | | 0.443 | 0.62 | 17600 | 0.4828 | | 0.4323 | 0.62 | 17800 | 0.4787 | | 0.5066 | 0.63 | 18000 | 0.4754 | | 0.4388 | 0.64 | 18200 | 0.4717 | | 0.5436 | 0.64 | 18400 | 0.4682 | | 0.3881 | 0.65 | 18600 | 0.4649 | | 0.6051 | 0.66 | 18800 | 0.4623 | | 0.5628 | 0.66 | 19000 | 0.4589 | | 0.4372 | 0.67 | 19200 | 0.4560 | | 0.4748 | 0.68 | 19400 | 0.4529 | | 0.5461 | 0.69 | 19600 | 0.4499 | | 0.4313 | 0.69 | 19800 | 0.4471 | | 0.4353 | 0.7 | 20000 | 0.4445 | | 0.4988 | 0.71 | 20200 | 0.4419 | | 0.4037 | 0.71 | 20400 | 0.4394 | | 0.446 | 0.72 | 20600 | 0.4368 | | 0.4381 | 0.73 | 20800 | 0.4347 | | 0.4849 | 0.73 | 21000 | 0.4324 | | 0.4726 | 0.74 | 21200 | 0.4303 | | 0.4842 | 0.75 | 21400 | 0.4279 | | 0.3508 | 0.76 | 21600 | 0.4259 | | 0.4452 | 0.76 | 21800 | 0.4236 | | 0.3565 | 0.77 | 22000 | 0.4216 | | 0.4634 | 0.78 | 22200 | 0.4196 | | 0.3925 | 0.78 | 22400 | 0.4179 | | 0.4086 | 0.79 | 22600 | 0.4164 | | 0.4149 | 0.8 | 22800 | 0.4145 | | 0.3856 | 0.8 | 23000 | 0.4128 | | 0.4053 | 0.81 | 23200 | 0.4113 | | 0.47 | 0.82 | 23400 | 0.4099 | | 0.3918 | 0.83 | 23600 | 0.4086 | | 0.4021 | 0.83 | 23800 | 0.4074 | | 0.376 | 0.84 | 24000 | 0.4063 | | 0.5067 | 0.85 | 24200 | 0.4052 | | 0.4721 | 0.85 | 24400 | 0.4041 | | 0.4015 | 0.86 | 24600 | 0.4031 | | 0.3576 | 0.87 | 24800 | 0.4021 | | 0.3975 | 0.87 | 25000 | 0.4015 | | 0.3898 | 0.88 | 25200 | 0.4006 | | 0.4235 | 0.89 | 25400 | 0.4000 | | 0.3808 | 0.9 | 25600 | 0.3992 | | 0.3811 | 0.9 | 25800 | 0.3989 | | 0.4011 | 0.91 | 26000 | 0.3983 | | 0.4219 | 0.92 | 26200 | 0.3981 | | 0.3764 | 0.92 | 26400 | 0.3977 | | 0.4046 | 0.93 | 26600 | 0.3974 | | 0.4342 | 0.94 | 26800 | 0.3972 | | 0.4209 | 0.94 | 27000 | 0.3969 | | 0.4549 | 0.95 | 27200 | 0.3967 | | 0.3316 | 0.96 | 27400 | 0.3967 | | 0.4648 | 0.97 | 27600 | 0.3967 | | 0.4657 | 0.97 | 27800 | 0.3965 | | 0.3959 | 0.98 | 28000 | 0.3964 | | 0.3666 | 0.99 | 28200 | 0.3964 | | 0.3973 | 0.99 | 28400 | 0.3963 | ### Framework versions - Transformers 4.35.0.dev0 - Pytorch 2.1.0+cu118 - Datasets 2.14.5 - Tokenizers 0.14.1