Downloads · 30 days
4
6% of all-time downloads
yihang7/zephyr-7b-dpo-full-hydrox-safe
zephyr-7b-dpo-full-hydrox-safe is a text generation model from yihang7. Use it when you need the model to write or continue text. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
Downloads · 30 days
4
6% of all-time downloads
All-time downloads
64
Public
Parameters
7.2B
14.5 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors14.5 GB · 100%
From the Hugging Face model README
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
More information needed
More information needed
More information needed
The following hyperparameters were used during training:
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.201 | 0.03 | 100 | 0.1790 | 0.8766 | -1.8347 | 0.9571 | 2.7113 | -463.1273 | -277.3755 | -2.9298 | -3.0062 |
| 0.1124 | 0.07 | 200 | 0.0931 | 1.4748 | -3.5536 | 0.9764 | 5.0284 | -480.3171 | -271.3938 | -2.9144 | -2.9984 |
| 0.0657 | 0.1 | 300 | 0.1207 | 1.7581 | -4.7603 | 0.9806 | 6.5185 | -492.3842 | -268.5603 | -2.9028 | -2.9908 |
| 0.0676 | 0.14 | 400 | 0.0446 | 1.3392 | -8.0193 | 0.9857 | 9.3585 | -524.9738 | -272.75 | -2.9171 | -3.0132 |
| 0.0477 | 0.17 | 500 | 0.0581 | 1.3777 | -9.9106 | 0.9882 | 11.2883 | -543.8868 | -272.3650 | -2.8810 | -2.9695 |
| 0.0404 | 0.2 | 600 | 0.0845 | 0.3594 | -12.1256 | 0.9840 | 12.4850 | -566.0372 | -282.5481 | -2.8164 | -2.9235 |
| 0.0122 | 0.24 | 700 | 0.0831 | 0.4374 | -19.3178 | 0.9848 | 19.7552 | -637.9587 | -281.7681 | -2.7024 | -2.8110 |
| 0.0183 | 0.27 | 800 | 0.0775 | -0.0400 | -18.4153 | 0.9891 | 18.3753 | -628.9333 | -286.5414 | -2.8078 | -2.9358 |
| 0.0555 | 0.31 | 900 | 0.0589 | 0.4664 | -23.4223 | 0.9924 | 23.8887 | -679.0039 | -281.4781 | -2.8207 | -2.9406 |
| 0.0376 | 0.34 | 1000 | 0.0552 | 0.6445 | -26.2370 | 0.9874 | 26.8814 | -707.1502 | -279.6971 | -2.7989 | -2.9241 |
| 0.0438 | 0.37 | 1100 | 0.0685 | 1.7090 | -23.4352 | 0.9874 | 25.1442 | -679.1324 | -269.0518 | -2.8438 | -3.0501 |
| 0.074 | 0.41 | 1200 | 0.0715 | 1.4961 | -24.8055 | 0.9882 | 26.3016 | -692.8359 | -271.1810 | -2.7269 | -2.8728 |
| 0.1029 | 0.44 | 1300 | 0.1098 | 0.9065 | -28.0158 | 0.9882 | 28.9223 | -724.9388 | -277.0765 | -2.7411 | -2.9059 |
| 0.0271 | 0.48 | 1400 | 0.0346 | 0.7923 | -26.9632 | 0.9907 | 27.7555 | -714.4124 | -278.2186 | -2.7555 | -2.9286 |
| 0.0219 | 0.51 | 1500 | 0.0405 | 0.5053 | -27.0039 | 0.9933 | 27.5092 | -714.8198 | -281.0890 | -2.7955 | -2.9553 |
| 0.0214 | 0.54 | 1600 | 0.0293 | 0.7598 | -22.6033 | 0.9941 | 23.3631 | -670.8132 | -278.5433 | -2.7650 | -2.9211 |
| 0.0221 | 0.58 | 1700 | 0.0206 | 0.6463 | -25.7837 | 0.9949 | 26.4300 | -702.6176 | -279.6782 | -2.7340 | -2.9148 |
| 0.0339 | 0.61 | 1800 | 0.0204 | 1.2579 | -22.2971 | 0.9966 | 23.5550 | -667.7516 | -273.5627 | -2.7337 | -2.8868 |
| 0.0662 | 0.65 | 1900 | 0.0290 | 0.9343 | -30.7737 | 0.9975 | 31.7081 | -752.5181 | -276.7985 | -2.7130 | -2.8608 |
| 0.1032 | 0.68 | 2000 | 0.0358 | 1.6616 | -22.3776 | 0.9975 | 24.0392 | -668.5566 | -269.5255 | -2.7454 | -2.8772 |
| 0.0059 | 0.71 | 2100 | 0.0148 | 0.9711 | -28.7176 | 0.9949 | 29.6887 | -731.9562 | -276.4307 | -2.6758 | -2.8253 |
| 0.0096 | 0.75 | 2200 | 0.0282 | 0.9406 | -25.7753 | 0.9949 | 26.7158 | -702.5333 | -276.7361 | -2.9417 | -3.1166 |
| 0.0117 | 0.78 | 2300 | 0.0489 | -0.3402 | -33.3639 | 0.9924 | 33.0238 | -778.4199 | -289.5432 | -3.0059 | -3.1712 |
| 0.0034 | 0.82 | 2400 | 0.0136 | 1.4644 | -27.4436 | 0.9966 | 28.9079 | -719.2164 | -271.4980 | -2.6572 | -2.7758 |
| 0.0448 | 0.85 | 2500 | 0.0122 | -0.7904 | -29.3707 | 0.9983 | 28.5803 | -738.4877 | -294.0459 | -2.7608 | -2.8870 |
| 0.0606 | 0.88 | 2600 | 0.0723 | 0.6824 | -40.6246 | 0.9949 | 41.3070 | -851.0272 | -279.3180 | -2.6804 | -2.8281 |
| 0.0172 | 0.92 | 2700 | 0.0151 | 1.7774 | -28.1840 | 0.9958 | 29.9614 | -726.6205 | -268.3677 | -2.7720 | -2.9085 |
| 0.0206 | 0.95 | 2800 | 0.0105 | 0.0966 | -39.5676 | 0.9975 | 39.6643 | -840.4568 | -285.1752 | -2.8397 | -2.9803 |
| 0.0093 | 0.99 | 2900 | 0.0165 | 0.6995 | -33.4341 | 0.9966 | 34.1336 | -779.1221 | -279.1465 | -2.8086 | -2.9511 |
| 0.0008 | 1.02 | 3000 | 0.0101 | -0.0075 | -36.2959 | 0.9983 | 36.2884 | -807.7395 | -286.2166 | -2.9213 | -3.0591 |
| 0.0015 | 1.05 | 3100 | 0.0091 | -0.2430 | -38.1496 | 0.9966 | 37.9066 | -826.2767 | -288.5718 | -2.8160 | -2.9881 |
| 0.001 | 1.09 | 3200 | 0.0069 | 0.7306 | -37.6657 | 0.9966 | 38.3964 | -821.4379 | -278.8353 | -2.8287 | -2.9686 |
| 0.0447 | 1.12 | 3300 | 0.0148 | 0.4641 | -39.1245 | 0.9975 | 39.5886 | -836.0254 | -281.5004 | -2.8502 | -2.9821 |
| 0.0015 | 1.16 | 3400 | 0.0145 | 1.2867 | -34.9331 | 0.9983 | 36.2198 | -794.1116 | -273.2742 | -2.8645 | -2.9793 |
| 0.0069 | 1.19 | 3500 | 0.0120 | 1.0980 | -33.3648 | 0.9983 | 34.4628 | -778.4290 | -275.1619 | -2.8298 | -2.9435 |
| 0.0006 | 1.22 | 3600 | 0.0144 | 0.4168 | -38.9632 | 0.9975 | 39.3800 | -834.4130 | -281.9736 | -2.8061 | -2.9050 |
| 0.0532 | 1.26 | 3700 | 0.0163 | 1.2493 | -35.1549 | 0.9983 | 36.4042 | -796.3295 | -273.6483 | -3.0117 | -3.1271 |
| 0.0041 | 1.29 | 3800 | 0.0138 | 0.6365 | -36.2498 | 0.9966 | 36.8863 | -807.2789 | -279.7767 | -2.9031 | -3.0283 |
| 0.0043 | 1.33 | 3900 | 0.0095 | 0.8546 | -36.5756 | 0.9992 | 37.4302 | -810.5366 | -277.5952 | -2.9713 | -3.0945 |
| 0.0016 | 1.36 | 4000 | 0.0098 | 0.4731 | -41.9454 | 0.9992 | 42.4186 | -864.2352 | -281.4105 | -3.0390 | -3.1458 |
| 0.0018 | 1.39 | 4100 | 0.0103 | 1.1650 | -31.5836 | 0.9992 | 32.7487 | -760.6172 | -274.4915 | -3.0057 | -3.1525 |
| 0.0015 | 1.43 | 4200 | 0.0097 | 0.7472 | -35.6321 | 0.9992 | 36.3793 | -801.1016 | -278.6696 | -2.9895 | -3.1466 |
| 0.0078 | 1.46 | 4300 | 0.0171 | 0.4428 | -47.4943 | 0.9983 | 47.9370 | -919.7234 | -281.7140 | -3.0189 | -3.1577 |
| 0.0006 | 1.5 | 4400 | 0.0117 | 2.3601 | -36.7779 | 0.9992 | 39.1379 | -812.5596 | -262.5412 | -2.9504 | -3.0998 |
| 0.0021 | 1.53 | 4500 | 0.0096 | 1.8317 | -33.0472 | 0.9983 | 34.8789 | -775.2529 | -267.8244 | -2.8650 | -3.0361 |
| 0.0022 | 1.56 | 4600 | 0.0124 | 1.8244 | -35.2193 | 0.9983 | 37.0437 | -796.9737 | -267.8973 | -2.8983 | -3.0580 |
| 0.0025 | 1.6 | 4700 | 0.0120 | 1.4261 | -35.4094 | 0.9983 | 36.8355 | -798.8742 | -271.8802 | -2.8938 | -3.0468 |
| 0.0154 | 1.63 | 4800 | 0.0070 | 0.7832 | -39.4533 | 0.9983 | 40.2365 | -839.3135 | -278.3098 | -2.9491 | -3.1257 |
| 0.0025 | 1.67 | 4900 | 0.0070 | 0.5157 | -40.3519 | 0.9983 | 40.8676 | -848.2994 | -280.9843 | -2.8537 | -3.0217 |
| 0.0022 | 1.7 | 5000 | 0.0060 | 1.8378 | -36.1887 | 0.9983 | 38.0265 | -806.6680 | -267.7638 | -2.8008 | -2.9944 |
| 0.0116 | 1.73 | 5100 | 0.0067 | 1.6571 | -30.2966 | 0.9983 | 31.9537 | -747.7468 | -269.5709 | -2.8245 | -3.0273 |
| 0.0027 | 1.77 | 5200 | 0.0056 | 1.8559 | -28.9170 | 0.9983 | 30.7729 | -733.9506 | -267.5823 | -2.8308 | -3.0333 |
| 0.0008 | 1.8 | 5300 | 0.0120 | 0.1806 | -39.8991 | 0.9983 | 40.0797 | -843.7715 | -284.3351 | -3.0642 | -3.2708 |
| 0.0008 | 1.84 | 5400 | 0.0042 | 1.8857 | -32.3213 | 0.9983 | 34.2070 | -767.9935 | -267.2849 | -2.9745 | -3.1835 |
| 0.0419 | 1.87 | 5500 | 0.0049 | 0.4195 | -42.5787 | 0.9983 | 42.9982 | -870.5677 | -281.9464 | -2.9390 | -3.1435 |
| 0.0055 | 1.9 | 5600 | 0.0039 | 0.8564 | -39.5863 | 0.9966 | 40.4428 | -840.6442 | -277.5772 | -2.9517 | -3.1465 |
| 0.0016 | 1.94 | 5700 | 0.0039 | 1.5280 | -38.4427 | 0.9983 | 39.9707 | -829.2079 | -270.8615 | -2.9334 | -3.1210 |
| 0.0015 | 1.97 | 5800 | 0.0031 | 2.5689 | -32.1821 | 0.9992 | 34.7510 | -766.6016 | -260.4530 | -2.9266 | -3.1170 |
| 0.0022 | 2.01 | 5900 | 0.0154 | 1.5272 | -38.5973 | 0.9983 | 40.1246 | -830.7538 | -270.8692 | -2.9905 | -3.1795 |
| 0.0001 | 2.04 | 6000 | 0.0153 | 1.4304 | -38.7250 | 0.9983 | 40.1554 | -832.0308 | -271.8378 | -3.0212 | -3.2078 |
| 0.0004 | 2.07 | 6100 | 0.0151 | 0.9974 | -40.5361 | 0.9983 | 41.5335 | -850.1418 | -276.1674 | -3.0092 | -3.1920 |
| 0.0032 | 2.11 | 6200 | 0.0143 | 2.2246 | -35.4872 | 0.9983 | 37.7117 | -799.6525 | -263.8961 | -3.0169 | -3.2017 |
| 0.0026 | 2.14 | 6300 | 0.0138 | 2.1149 | -36.0655 | 0.9983 | 38.1804 | -805.4358 | -264.9924 | -3.0121 | -3.1974 |
| 0.0052 | 2.18 | 6400 | 0.0093 | 2.0832 | -38.0872 | 0.9983 | 40.1704 | -825.6531 | -265.3097 | -2.9119 | -3.1190 |
| 0.0002 | 2.21 | 6500 | 0.0031 | 2.4808 | -33.4480 | 0.9992 | 35.9287 | -779.2603 | -261.3340 | -2.9056 | -3.1235 |
| 0.0014 | 2.24 | 6600 | 0.0029 | 2.1622 | -35.1487 | 0.9992 | 37.3110 | -796.2680 | -264.5192 | -2.9322 | -3.1505 |
| 0.0001 | 2.28 | 6700 | 0.0029 | 2.6987 | -34.3087 | 0.9992 | 37.0074 | -787.8676 | -259.1545 | -2.9528 | -3.1683 |
| 0.0004 | 2.31 | 6800 | 0.0028 | 2.2809 | -36.5033 | 0.9992 | 38.7842 | -809.8135 | -263.3324 | -2.9377 | -3.1456 |
| 0.0268 | 2.35 | 6900 | 0.0028 | 1.9206 | -37.6915 | 0.9983 | 39.6121 | -821.6957 | -266.9360 | -2.9110 | -3.1199 |
| 0.0012 | 2.38 | 7000 | 0.0026 | 2.0357 | -34.8535 | 0.9992 | 36.8892 | -793.3152 | -265.7842 | -2.9165 | -3.1204 |
| 0.0002 | 2.41 | 7100 | 0.0031 | 0.4887 | -42.0925 | 0.9983 | 42.5812 | -865.7061 | -281.2549 | -2.8696 | -3.0815 |
| 0.0006 | 2.45 | 7200 | 0.0027 | 1.6025 | -38.2160 | 0.9992 | 39.8185 | -826.9409 | -270.1166 | -2.8921 | -3.1009 |
| 0.0037 | 2.48 | 7300 | 0.0027 | 1.2715 | -42.6264 | 0.9992 | 43.8978 | -871.0444 | -273.4270 | -2.8870 | -3.0973 |
| 0.0023 | 2.52 | 7400 | 0.0026 | 1.3232 | -42.5318 | 0.9992 | 43.8549 | -870.0981 | -272.9100 | -2.8912 | -3.1000 |
| 0.0012 | 2.55 | 7500 | 0.0026 | 1.4663 | -41.9917 | 0.9992 | 43.4581 | -864.6981 | -271.4783 | -2.8990 | -3.1084 |
| 0.0012 | 2.58 | 7600 | 0.0025 | 1.2190 | -42.9826 | 0.9992 | 44.2016 | -874.6072 | -273.9521 | -2.8991 | -3.1070 |
| 0.0001 | 2.62 | 7700 | 0.0026 | 1.7469 | -40.1663 | 0.9992 | 41.9132 | -846.4437 | -268.6728 | -2.8615 | -3.0744 |
| 0.0001 | 2.65 | 7800 | 0.0026 | 1.7291 | -40.4524 | 0.9992 | 42.1815 | -849.3045 | -268.8504 | -2.8619 | -3.0751 |
| 0.003 | 2.69 | 7900 | 0.0018 | 1.8715 | -39.4482 | 0.9992 | 41.3197 | -839.2630 | -267.4267 | -2.8391 | -3.0504 |
| 0.0001 | 2.72 | 8000 | 0.0018 | 1.4707 | -42.0026 | 0.9992 | 43.4733 | -864.8069 | -271.4347 | -2.8587 | -3.0725 |
| 0.0001 | 2.75 | 8100 | 0.0018 | 1.7032 | -41.2255 | 0.9992 | 42.9287 | -857.0361 | -269.1100 | -2.8611 | -3.0771 |
| 0.0001 | 2.79 | 8200 | 0.0018 | 1.7053 | -41.5715 | 0.9992 | 43.2768 | -860.4957 | -269.0886 | -2.8608 | -3.0769 |
| 0.0 | 2.82 | 8300 | 0.0017 | 1.5198 | -41.7233 | 0.9992 | 43.2431 | -862.0136 | -270.9432 | -2.8563 | -3.0742 |
| 0.0 | 2.86 | 8400 | 0.0018 | 1.7544 | -40.1153 | 0.9992 | 41.8696 | -845.9333 | -268.5981 | -2.8694 | -3.0853 |
| 0.0003 | 2.89 | 8500 | 0.0018 | 1.6997 | -40.2116 | 0.9992 | 41.9113 | -846.8971 | -269.1451 | -2.8782 | -3.0951 |
| 0.0071 | 2.92 | 8600 | 0.0018 | 1.5913 | -40.7278 | 0.9992 | 42.3191 | -852.0585 | -270.2284 | -2.8728 | -3.0908 |
| 0.0013 | 2.96 | 8700 | 0.0018 | 1.5263 | -41.0232 | 0.9992 | 42.5495 | -855.0129 | -270.8784 | -2.8716 | -3.0900 |
| 0.0003 | 2.99 | 8800 | 0.0019 | 1.5305 | -41.0444 | 0.9992 | 42.5748 | -855.2242 | -270.8367 | -2.8724 | -3.0907 |