chesspythia-70m-random_1M
This model is a fine-tuned version of EleutherAI/pythia-70m-deduped on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.9894
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 1
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
1.4951 | 0.01 | 25 | 1.5347 |
1.356 | 0.02 | 50 | 1.3895 |
1.316 | 0.03 | 75 | 1.3312 |
1.2512 | 0.04 | 100 | 1.2981 |
1.2981 | 0.05 | 125 | 1.2738 |
1.2777 | 0.06 | 150 | 1.2508 |
1.2913 | 0.07 | 175 | 1.2362 |
1.2591 | 0.08 | 200 | 1.2201 |
1.1763 | 0.09 | 225 | 1.2136 |
1.1835 | 0.1 | 250 | 1.2001 |
1.1913 | 0.11 | 275 | 1.1864 |
1.1828 | 0.12 | 300 | 1.1787 |
1.1523 | 0.13 | 325 | 1.1708 |
1.2038 | 0.14 | 350 | 1.1653 |
1.1555 | 0.15 | 375 | 1.1591 |
1.1235 | 0.16 | 400 | 1.1611 |
1.1431 | 0.17 | 425 | 1.1484 |
1.1355 | 0.18 | 450 | 1.1470 |
1.1464 | 0.19 | 475 | 1.1354 |
1.146 | 0.2 | 500 | 1.1335 |
1.1471 | 0.21 | 525 | 1.1241 |
1.1568 | 0.22 | 550 | 1.1243 |
1.1565 | 0.23 | 575 | 1.1247 |
1.1253 | 0.24 | 600 | 1.1160 |
1.1002 | 0.25 | 625 | 1.1081 |
1.1356 | 0.26 | 650 | 1.1067 |
1.0997 | 0.27 | 675 | 1.1074 |
1.1092 | 0.28 | 700 | 1.1007 |
1.1016 | 0.29 | 725 | 1.1019 |
1.0578 | 0.3 | 750 | 1.0897 |
1.0588 | 0.31 | 775 | 1.0867 |
1.063 | 0.32 | 800 | 1.0901 |
1.0559 | 0.33 | 825 | 1.0885 |
1.1272 | 0.34 | 850 | 1.0841 |
1.0762 | 0.35 | 875 | 1.0797 |
1.0842 | 0.36 | 900 | 1.0793 |
1.0566 | 0.37 | 925 | 1.0744 |
1.0712 | 0.38 | 950 | 1.0682 |
1.076 | 0.39 | 975 | 1.0692 |
1.0624 | 0.4 | 1000 | 1.0670 |
1.0541 | 0.41 | 1025 | 1.0599 |
1.0292 | 0.42 | 1050 | 1.0601 |
1.0788 | 0.43 | 1075 | 1.0536 |
1.0846 | 0.44 | 1100 | 1.0530 |
1.0547 | 0.45 | 1125 | 1.0562 |
0.9978 | 0.46 | 1150 | 1.0483 |
1.0261 | 0.47 | 1175 | 1.0470 |
1.0031 | 0.48 | 1200 | 1.0475 |
1.0141 | 0.49 | 1225 | 1.0442 |
1.064 | 0.5 | 1250 | 1.0416 |
1.0154 | 0.51 | 1275 | 1.0408 |
1.0474 | 0.52 | 1300 | 1.0383 |
1.0392 | 0.53 | 1325 | 1.0380 |
1.036 | 0.54 | 1350 | 1.0360 |
1.0325 | 0.55 | 1375 | 1.0312 |
1.062 | 0.56 | 1400 | 1.0308 |
1.0601 | 0.57 | 1425 | 1.0282 |
1.0287 | 0.58 | 1450 | 1.0274 |
0.9944 | 0.59 | 1475 | 1.0249 |
1.0302 | 0.6 | 1500 | 1.0241 |
1.0188 | 0.61 | 1525 | 1.0210 |
0.8968 | 0.62 | 1550 | 1.0211 |
0.9661 | 0.63 | 1575 | 1.0176 |
1.0741 | 0.64 | 1600 | 1.0182 |
1.0666 | 0.65 | 1625 | 1.0156 |
1.0623 | 0.66 | 1650 | 1.0138 |
1.0435 | 0.67 | 1675 | 1.0117 |
0.9892 | 0.68 | 1700 | 1.0117 |
0.9941 | 0.69 | 1725 | 1.0097 |
0.967 | 0.7 | 1750 | 1.0092 |
1.0387 | 0.71 | 1775 | 1.0074 |
0.9718 | 0.72 | 1800 | 1.0056 |
0.9841 | 0.73 | 1825 | 1.0049 |
1.0086 | 0.74 | 1850 | 1.0038 |
0.9855 | 0.75 | 1875 | 1.0025 |
1.0174 | 0.76 | 1900 | 1.0022 |
0.9725 | 0.77 | 1925 | 1.0003 |
0.987 | 0.78 | 1950 | 0.9989 |
1.0133 | 0.79 | 1975 | 0.9979 |
0.9978 | 0.8 | 2000 | 0.9974 |
1.0193 | 0.81 | 2025 | 0.9960 |
0.9842 | 0.82 | 2050 | 0.9955 |
0.9316 | 0.83 | 2075 | 0.9952 |
1.0233 | 0.84 | 2100 | 0.9940 |
1.0054 | 0.85 | 2125 | 0.9930 |
0.922 | 0.86 | 2150 | 0.9933 |
1.0309 | 0.87 | 2175 | 0.9925 |
0.975 | 0.88 | 2200 | 0.9917 |
1.0108 | 0.89 | 2225 | 0.9915 |
0.9418 | 0.9 | 2250 | 0.9911 |
0.9445 | 0.91 | 2275 | 0.9907 |
0.9973 | 0.92 | 2300 | 0.9903 |
1.0052 | 0.93 | 2325 | 0.9900 |
0.9347 | 0.94 | 2350 | 0.9899 |
0.954 | 0.95 | 2375 | 0.9897 |
0.9857 | 0.96 | 2400 | 0.9896 |
0.958 | 0.97 | 2425 | 0.9895 |
1.0051 | 0.98 | 2450 | 0.9894 |
0.9999 | 0.99 | 2475 | 0.9894 |
0.959 | 1.0 | 2500 | 0.9894 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
- Downloads last month
- 6
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for superlazycoder/chesspythia-70m-random_1M
Base model
EleutherAI/pythia-70m-deduped