Mistral_Sparse_refined_web_90p_2024-03-12
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: 2.2527
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 |
---|---|---|---|
6.2537 | 0.0 | 25 | 5.6955 |
3.6435 | 0.01 | 50 | 3.6962 |
2.9245 | 0.01 | 75 | 3.0857 |
2.7206 | 0.02 | 100 | 2.9030 |
2.5957 | 0.02 | 125 | 2.8198 |
2.6569 | 0.02 | 150 | 2.7748 |
2.5613 | 0.03 | 175 | 2.7355 |
2.506 | 0.03 | 200 | 2.7097 |
2.528 | 0.04 | 225 | 2.6859 |
2.5391 | 0.04 | 250 | 2.6627 |
2.4585 | 0.04 | 275 | 2.6523 |
2.4632 | 0.05 | 300 | 2.6416 |
2.4795 | 0.05 | 325 | 2.6272 |
2.3057 | 0.06 | 350 | 2.6283 |
2.458 | 0.06 | 375 | 2.6262 |
2.499 | 0.06 | 400 | 2.6085 |
2.4815 | 0.07 | 425 | 2.6019 |
2.4348 | 0.07 | 450 | 2.6016 |
2.3424 | 0.08 | 475 | 2.5921 |
2.4052 | 0.08 | 500 | 2.5833 |
2.4802 | 0.08 | 525 | 2.5826 |
2.3818 | 0.09 | 550 | 2.5817 |
2.5175 | 0.09 | 575 | 2.5777 |
2.3259 | 0.1 | 600 | 2.5741 |
2.4037 | 0.1 | 625 | 2.5720 |
2.4633 | 0.1 | 650 | 2.5701 |
2.4678 | 0.11 | 675 | 2.5670 |
2.4137 | 0.11 | 700 | 2.5590 |
2.4163 | 0.12 | 725 | 2.5577 |
2.3292 | 0.12 | 750 | 2.5579 |
2.3493 | 0.12 | 775 | 2.5574 |
2.446 | 0.13 | 800 | 2.5544 |
2.4328 | 0.13 | 825 | 2.5492 |
2.4239 | 0.14 | 850 | 2.5520 |
2.2832 | 0.14 | 875 | 2.5554 |
2.3791 | 0.14 | 900 | 2.5515 |
2.3112 | 0.15 | 925 | 2.5531 |
2.4415 | 0.15 | 950 | 2.5509 |
2.3655 | 0.16 | 975 | 2.5469 |
2.4154 | 0.16 | 1000 | 2.5425 |
2.4198 | 0.16 | 1025 | 2.5475 |
2.4526 | 0.17 | 1050 | 2.5451 |
2.4692 | 0.17 | 1075 | 2.5417 |
2.3865 | 0.18 | 1100 | 2.5433 |
2.4104 | 0.18 | 1125 | 2.5405 |
2.3916 | 0.18 | 1150 | 2.5419 |
2.4311 | 0.19 | 1175 | 2.5394 |
2.2988 | 0.19 | 1200 | 2.5368 |
2.3728 | 0.2 | 1225 | 2.5353 |
2.3802 | 0.2 | 1250 | 2.5343 |
2.4077 | 0.2 | 1275 | 2.5456 |
2.4275 | 0.21 | 1300 | 2.5390 |
2.3964 | 0.21 | 1325 | 2.5362 |
2.2976 | 0.22 | 1350 | 2.5360 |
2.3733 | 0.22 | 1375 | 2.5376 |
2.3842 | 0.22 | 1400 | 2.5402 |
2.3268 | 0.23 | 1425 | 2.5312 |
2.3077 | 0.23 | 1450 | 2.5265 |
2.2747 | 0.24 | 1475 | 2.5344 |
2.347 | 0.24 | 1500 | 2.5348 |
2.2611 | 0.24 | 1525 | 2.5358 |
2.322 | 0.25 | 1550 | 2.5296 |
2.4424 | 0.25 | 1575 | 2.5330 |
2.3987 | 0.26 | 1600 | 2.5281 |
2.3923 | 0.26 | 1625 | 2.5314 |
2.4507 | 0.26 | 1650 | 2.5315 |
2.3536 | 0.27 | 1675 | 2.5344 |
2.4019 | 0.27 | 1700 | 2.5365 |
2.392 | 0.28 | 1725 | 2.5346 |
2.3721 | 0.28 | 1750 | 2.5284 |
2.3612 | 0.28 | 1775 | 2.5327 |
2.2935 | 0.29 | 1800 | 2.5279 |
2.3774 | 0.29 | 1825 | 2.5268 |
2.3739 | 0.3 | 1850 | 2.5257 |
2.3647 | 0.3 | 1875 | 2.5300 |
2.3867 | 0.3 | 1900 | 2.5290 |
2.4372 | 0.31 | 1925 | 2.5296 |
2.3871 | 0.31 | 1950 | 2.5257 |
2.3842 | 0.32 | 1975 | 2.5292 |
2.3272 | 0.32 | 2000 | 2.5255 |
2.3526 | 0.32 | 2025 | 2.5289 |
2.3392 | 0.33 | 2050 | 2.5287 |
2.3201 | 0.33 | 2075 | 2.5265 |
2.4196 | 0.34 | 2100 | 2.5305 |
2.4652 | 0.34 | 2125 | 2.5260 |
2.377 | 0.34 | 2150 | 2.5230 |
2.3212 | 0.35 | 2175 | 2.5254 |
2.3602 | 0.35 | 2200 | 2.5231 |
2.3185 | 0.36 | 2225 | 2.5257 |
2.3209 | 0.36 | 2250 | 2.5238 |
2.4438 | 0.36 | 2275 | 2.5233 |
2.4092 | 0.37 | 2300 | 2.5190 |
2.3106 | 0.37 | 2325 | 2.5217 |
2.357 | 0.38 | 2350 | 2.5211 |
2.3904 | 0.38 | 2375 | 2.5212 |
2.3879 | 0.38 | 2400 | 2.5205 |
2.2494 | 0.39 | 2425 | 2.5195 |
2.2334 | 0.39 | 2450 | 2.5198 |
2.3426 | 0.4 | 2475 | 2.5161 |
2.4106 | 0.4 | 2500 | 2.5104 |
2.41 | 0.4 | 2525 | 2.5169 |
2.2716 | 0.41 | 2550 | 2.5128 |
2.3065 | 0.41 | 2575 | 2.5133 |
2.3745 | 0.42 | 2600 | 2.5268 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 6
Inference API (serverless) does not yet support model repos that contain custom code.
Model tree for thrunlab/Mistral_Sparse_refined_web_90p_2024-03-12
Base model
mistralai/Mistral-7B-v0.1
Finetuned
this model