Mistral_Sparse_refined_web_50p_2024-03-11
This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.0911
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: 1
- eval_batch_size: 1
- seed: 0
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2600
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
2.4148 | 0.0 | 25 | 2.5704 |
2.2319 | 0.01 | 50 | 2.5232 |
2.2571 | 0.01 | 75 | 2.4830 |
2.2167 | 0.02 | 100 | 2.4616 |
2.2224 | 0.02 | 125 | 2.4487 |
2.2966 | 0.02 | 150 | 2.4342 |
2.2266 | 0.03 | 175 | 2.4288 |
2.2225 | 0.03 | 200 | 2.4205 |
2.2524 | 0.04 | 225 | 2.4123 |
2.2948 | 0.04 | 250 | 2.4103 |
2.2072 | 0.04 | 275 | 2.4080 |
2.2138 | 0.05 | 300 | 2.4053 |
2.2304 | 0.05 | 325 | 2.4038 |
2.0709 | 0.06 | 350 | 2.4033 |
2.233 | 0.06 | 375 | 2.4019 |
2.2723 | 0.06 | 400 | 2.4004 |
2.2638 | 0.07 | 425 | 2.3994 |
2.2223 | 0.07 | 450 | 2.3998 |
2.1447 | 0.08 | 475 | 2.3966 |
2.2069 | 0.08 | 500 | 2.3932 |
2.2706 | 0.08 | 525 | 2.3929 |
2.1619 | 0.09 | 550 | 2.3912 |
2.3014 | 0.09 | 575 | 2.3938 |
2.0855 | 0.1 | 600 | 2.3922 |
2.1869 | 0.1 | 625 | 2.3884 |
2.2745 | 0.1 | 650 | 2.3872 |
2.2586 | 0.11 | 675 | 2.3835 |
2.2286 | 0.11 | 700 | 2.3862 |
2.2134 | 0.12 | 725 | 2.3860 |
2.1182 | 0.12 | 750 | 2.3871 |
2.1639 | 0.12 | 775 | 2.3828 |
2.253 | 0.13 | 800 | 2.3852 |
2.2617 | 0.13 | 825 | 2.3870 |
2.2428 | 0.14 | 850 | 2.3852 |
2.1117 | 0.14 | 875 | 2.3864 |
2.1737 | 0.14 | 900 | 2.3868 |
2.1223 | 0.15 | 925 | 2.3841 |
2.2612 | 0.15 | 950 | 2.3843 |
2.1923 | 0.16 | 975 | 2.3853 |
2.2358 | 0.16 | 1000 | 2.3824 |
2.2425 | 0.16 | 1025 | 2.3774 |
2.2375 | 0.17 | 1050 | 2.3802 |
2.2797 | 0.17 | 1075 | 2.3826 |
2.2075 | 0.18 | 1100 | 2.3808 |
2.2427 | 0.18 | 1125 | 2.3784 |
2.2007 | 0.18 | 1150 | 2.3792 |
2.2646 | 0.19 | 1175 | 2.3785 |
2.1105 | 0.19 | 1200 | 2.3771 |
2.2026 | 0.2 | 1225 | 2.3799 |
2.1858 | 0.2 | 1250 | 2.3762 |
2.2267 | 0.2 | 1275 | 2.3794 |
2.2582 | 0.21 | 1300 | 2.3765 |
2.2358 | 0.21 | 1325 | 2.3788 |
2.1314 | 0.22 | 1350 | 2.3758 |
2.2052 | 0.22 | 1375 | 2.3767 |
2.2176 | 0.22 | 1400 | 2.3773 |
2.1586 | 0.23 | 1425 | 2.3772 |
2.1425 | 0.23 | 1450 | 2.3760 |
2.1108 | 0.24 | 1475 | 2.3766 |
2.1591 | 0.24 | 1500 | 2.3762 |
2.1012 | 0.24 | 1525 | 2.3777 |
2.1696 | 0.25 | 1550 | 2.3745 |
2.2862 | 0.25 | 1575 | 2.3749 |
2.2379 | 0.26 | 1600 | 2.3771 |
2.2022 | 0.26 | 1625 | 2.3763 |
2.288 | 0.26 | 1650 | 2.3764 |
2.1911 | 0.27 | 1675 | 2.3783 |
2.2301 | 0.27 | 1700 | 2.3794 |
2.2186 | 0.28 | 1725 | 2.3803 |
2.1986 | 0.28 | 1750 | 2.3785 |
2.1809 | 0.28 | 1775 | 2.3770 |
2.1431 | 0.29 | 1800 | 2.3755 |
2.2225 | 0.29 | 1825 | 2.3780 |
2.2156 | 0.3 | 1850 | 2.3759 |
2.2012 | 0.3 | 1875 | 2.3767 |
2.2225 | 0.3 | 1900 | 2.3765 |
2.2747 | 0.31 | 1925 | 2.3762 |
2.2206 | 0.31 | 1950 | 2.3754 |
2.2461 | 0.32 | 1975 | 2.3724 |
2.1601 | 0.32 | 2000 | 2.3752 |
2.1831 | 0.32 | 2025 | 2.3759 |
2.1657 | 0.33 | 2050 | 2.3750 |
2.1661 | 0.33 | 2075 | 2.3762 |
2.262 | 0.34 | 2100 | 2.3747 |
2.3022 | 0.34 | 2125 | 2.3724 |
2.2074 | 0.34 | 2150 | 2.3739 |
2.1682 | 0.35 | 2175 | 2.3731 |
2.1887 | 0.35 | 2200 | 2.3737 |
2.1387 | 0.36 | 2225 | 2.3725 |
2.1311 | 0.36 | 2250 | 2.3723 |
2.2672 | 0.36 | 2275 | 2.3691 |
2.2492 | 0.37 | 2300 | 2.3713 |
2.1457 | 0.37 | 2325 | 2.3697 |
2.1854 | 0.38 | 2350 | 2.3693 |
2.2247 | 0.38 | 2375 | 2.3675 |
2.2369 | 0.38 | 2400 | 2.3695 |
2.0793 | 0.39 | 2425 | 2.3704 |
2.0802 | 0.39 | 2450 | 2.3666 |
2.1658 | 0.4 | 2475 | 2.3673 |
2.2502 | 0.4 | 2500 | 2.3692 |
2.259 | 0.4 | 2525 | 2.3716 |
2.1085 | 0.41 | 2550 | 2.3692 |
2.1335 | 0.41 | 2575 | 2.3689 |
2.214 | 0.42 | 2600 | 2.3671 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 9
Inference API (serverless) does not yet support model repos that contain custom code.
Model tree for thrunlab/Mistral_Sparse_refined_web_50p_2024-03-11
Base model
mistralai/Mistral-7B-v0.1