tinyllama-1.1b-sum-dpo-full_LR2e-8_3epochs_old

This model is a fine-tuned version of martimfasantos/tinyllama-1.1b-sum-sft-full_old on the openai/summarize_from_feedback dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6872
  • Rewards/chosen: -0.0423
  • Rewards/rejected: -0.0554
  • Rewards/accuracies: 0.5943
  • Rewards/margins: 0.0131
  • Logps/rejected: -68.7246
  • Logps/chosen: -62.9434
  • Logits/rejected: -3.0648
  • Logits/chosen: -3.0704

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: 2e-08
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 3

Training results

Training Loss Epoch Step Logits/chosen Logits/rejected Logps/chosen Logps/rejected Validation Loss Rewards/accuracies Rewards/chosen Rewards/margins Rewards/rejected
0.6931 0.0172 100 -3.1635 -3.1579 -58.7134 -63.1764 0.6932 0.4868 -0.0000 -0.0001 0.0000
0.6931 0.0345 200 -3.1635 -3.1578 -58.7115 -63.1768 0.6932 0.4823 0.0000 -0.0000 0.0000
0.6932 0.0517 300 -3.1633 -3.1576 -58.7154 -63.1759 0.6932 0.4777 -0.0000 -0.0001 0.0000
0.6931 0.0689 400 -3.1632 -3.1575 -58.7141 -63.1718 0.6932 0.4954 -0.0000 -0.0001 0.0001
0.6932 0.0861 500 -3.1630 -3.1574 -58.7011 -63.1630 0.6932 0.4735 0.0001 -0.0001 0.0002
0.6932 0.1034 600 -3.1630 -3.1572 -58.7157 -63.1771 0.6932 0.4893 -0.0000 -0.0001 0.0000
0.6931 0.1206 700 -3.1632 -3.1575 -58.7038 -63.1662 0.6932 0.4937 0.0001 -0.0001 0.0001
0.693 0.1378 800 -3.1633 -3.1576 -58.7029 -63.1637 0.6932 0.4826 0.0001 -0.0001 0.0002
0.6933 0.1551 900 -3.1632 -3.1575 -58.7115 -63.1671 0.6932 0.4921 0.0000 -0.0001 0.0001
0.6932 0.1723 1000 -3.1633 -3.1577 -58.7041 -63.1666 0.6932 0.4823 0.0001 -0.0001 0.0001
0.6933 0.1895 1100 -3.1628 -3.1572 -58.7041 -63.1678 0.6932 0.4879 0.0001 -0.0000 0.0001
0.6931 0.2068 1200 -3.1629 -3.1573 -58.6888 -63.1590 0.6931 0.4984 0.0002 0.0000 0.0002
0.6931 0.2240 1300 -3.1629 -3.1572 -58.6903 -63.1651 0.6931 0.5123 0.0002 0.0001 0.0002
0.693 0.2412 1400 -3.1628 -3.1571 -58.6913 -63.1595 0.6931 0.5112 0.0002 -0.0000 0.0002
0.693 0.2584 1500 -3.1625 -3.1569 -58.6790 -63.1508 0.6931 0.5116 0.0003 0.0000 0.0003
0.6928 0.2757 1600 -3.1625 -3.1569 -58.6717 -63.1467 0.6931 0.4888 0.0004 0.0001 0.0003
0.6928 0.2929 1700 -3.1622 -3.1565 -58.6647 -63.1465 0.6931 0.5074 0.0005 0.0001 0.0003
0.6927 0.3101 1800 -3.1617 -3.1561 -58.6570 -63.1432 0.6931 0.5197 0.0005 0.0002 0.0004
0.6927 0.3274 1900 -3.1615 -3.1559 -58.6514 -63.1370 0.6931 0.5260 0.0006 0.0002 0.0004
0.6927 0.3446 2000 -3.1611 -3.1555 -58.6395 -63.1397 0.6930 0.5448 0.0007 0.0003 0.0004
0.6926 0.3618 2100 -3.1606 -3.1550 -58.6333 -63.1319 0.6930 0.5344 0.0008 0.0003 0.0005
0.692 0.3790 2200 -3.1602 -3.1545 -58.6257 -63.1267 0.6930 0.5295 0.0009 0.0003 0.0005
0.692 0.3963 2300 -3.1600 -3.1543 -58.6131 -63.1255 0.6929 0.5437 0.0010 0.0004 0.0005
0.6924 0.4135 2400 -3.1597 -3.1540 -58.6085 -63.1220 0.6929 0.5407 0.0010 0.0005 0.0006
0.6923 0.4307 2500 -3.1590 -3.1533 -58.5965 -63.1095 0.6929 0.5344 0.0012 0.0004 0.0007
0.6914 0.4480 2600 -3.1586 -3.1529 -58.5858 -63.1145 0.6928 0.5520 0.0013 0.0006 0.0007
0.6916 0.4652 2700 -3.1581 -3.1524 -58.5732 -63.1112 0.6928 0.5500 0.0014 0.0007 0.0007
0.692 0.4824 2800 -3.1575 -3.1518 -58.5713 -63.1121 0.6928 0.5504 0.0014 0.0007 0.0007
0.6921 0.4997 2900 -3.1570 -3.1514 -58.5498 -63.1157 0.6927 0.5548 0.0016 0.0010 0.0006
0.692 0.5169 3000 -3.1564 -3.1508 -58.5515 -63.1176 0.6927 0.5609 0.0016 0.0010 0.0006
0.6912 0.5341 3100 -3.1558 -3.1502 -58.5477 -63.1245 0.6926 0.5500 0.0016 0.0011 0.0006
0.6922 0.5513 3200 -3.1553 -3.1497 -58.5271 -63.1111 0.6926 0.5616 0.0018 0.0012 0.0007
0.6907 0.5686 3300 -3.1541 -3.1485 -58.5238 -63.1269 0.6925 0.5627 0.0019 0.0013 0.0005
0.6902 0.5858 3400 -3.1535 -3.1479 -58.5243 -63.1298 0.6925 0.5527 0.0019 0.0014 0.0005
0.6907 0.6030 3500 -3.1527 -3.1471 -58.5080 -63.1429 0.6923 0.5620 0.0020 0.0017 0.0004
0.6908 0.6203 3600 -3.1519 -3.1462 -58.5280 -63.1556 0.6924 0.5676 0.0018 0.0016 0.0002
0.6882 0.6375 3700 -3.1510 -3.1454 -58.5192 -63.1647 0.6923 0.5737 0.0019 0.0018 0.0002
0.6912 0.6547 3800 -3.1502 -3.1446 -58.5257 -63.1704 0.6923 0.5583 0.0019 0.0018 0.0001
0.6884 0.6720 3900 -3.1495 -3.1439 -58.5325 -63.1916 0.6922 0.5653 0.0018 0.0019 -0.0001
0.6893 0.6892 4000 -3.1483 -3.1427 -58.5317 -63.2120 0.6921 0.5748 0.0018 0.0021 -0.0003
0.6882 0.7064 4100 -3.1470 -3.1414 -58.5366 -63.2363 0.6920 0.5767 0.0018 0.0023 -0.0006
0.6897 0.7236 4200 -3.1459 -3.1403 -58.5300 -63.2312 0.6920 0.5683 0.0018 0.0023 -0.0005
0.6898 0.7409 4300 -3.1451 -3.1394 -58.5430 -63.2659 0.6919 0.5743 0.0017 0.0025 -0.0009
0.6907 0.7581 4400 -3.1442 -3.1386 -58.5652 -63.2958 0.6919 0.5718 0.0015 0.0026 -0.0012
0.6907 0.7753 4500 -3.1433 -3.1376 -58.5910 -63.3274 0.6919 0.5685 0.0012 0.0027 -0.0015
0.6884 0.7926 4600 -3.1425 -3.1369 -58.5812 -63.3476 0.6917 0.5829 0.0013 0.0030 -0.0017
0.6899 0.8098 4700 -3.1417 -3.1360 -58.6116 -63.3823 0.6917 0.5723 0.0010 0.0030 -0.0020
0.6905 0.8270 4800 -3.1404 -3.1348 -58.6264 -63.4076 0.6916 0.5755 0.0009 0.0031 -0.0023
0.6874 0.8442 4900 -3.1397 -3.1340 -58.6553 -63.4466 0.6916 0.5750 0.0006 0.0032 -0.0027
0.69 0.8615 5000 -3.1387 -3.1330 -58.6609 -63.4651 0.6915 0.5727 0.0005 0.0034 -0.0028
0.6862 0.8787 5100 -3.1378 -3.1321 -58.6905 -63.5114 0.6915 0.5720 0.0002 0.0035 -0.0033
0.6863 0.8959 5200 -3.1369 -3.1312 -58.7184 -63.5646 0.6913 0.5771 -0.0001 0.0038 -0.0038
0.6876 0.9132 5300 -3.1359 -3.1303 -58.7430 -63.6003 0.6913 0.5704 -0.0003 0.0039 -0.0042
0.6888 0.9304 5400 -3.1349 -3.1292 -58.7667 -63.6265 0.6913 0.5804 -0.0005 0.0039 -0.0045
0.6877 0.9476 5500 -3.1341 -3.1284 -58.7894 -63.6709 0.6912 0.5732 -0.0008 0.0041 -0.0049
0.6899 0.9649 5600 -3.1337 -3.1280 -58.8098 -63.7023 0.6911 0.5755 -0.0010 0.0042 -0.0052
0.6868 0.9821 5700 -3.1323 -3.1267 -58.8416 -63.7534 0.6910 0.5762 -0.0013 0.0044 -0.0057
0.6854 0.9993 5800 -3.1315 -3.1258 -58.8782 -63.7916 0.6910 0.5748 -0.0017 0.0045 -0.0061
0.6859 1.0165 5900 -3.1300 -3.1244 -58.9206 -63.8560 0.6909 0.5781 -0.0021 0.0047 -0.0068
0.6885 1.0338 6000 -3.1292 -3.1235 -58.9602 -63.9056 0.6909 0.5799 -0.0025 0.0048 -0.0073
0.6853 1.0510 6100 -3.1272 -3.1215 -58.9989 -63.9666 0.6908 0.5795 -0.0029 0.0050 -0.0079
0.6886 1.0682 6200 -3.1262 -3.1205 -59.0433 -64.0141 0.6908 0.5820 -0.0033 0.0050 -0.0083
0.6859 1.0855 6300 -3.1255 -3.1199 -59.0883 -64.0800 0.6907 0.5846 -0.0038 0.0052 -0.0090
0.6879 1.1027 6400 -3.1255 -3.1198 -59.1255 -64.1257 0.6906 0.5809 -0.0041 0.0053 -0.0095
0.6889 1.1199 6500 -3.1235 -3.1178 -59.1877 -64.2032 0.6906 0.5804 -0.0048 0.0055 -0.0102
0.6881 1.1371 6600 -3.1233 -3.1176 -59.2408 -64.2806 0.6904 0.5871 -0.0053 0.0057 -0.0110
0.6856 1.1544 6700 -3.1209 -3.1152 -59.2896 -64.3343 0.6904 0.5792 -0.0058 0.0058 -0.0115
0.6856 1.1716 6800 -3.1199 -3.1142 -59.3530 -64.4170 0.6903 0.5857 -0.0064 0.0060 -0.0124
0.6876 1.1888 6900 -3.1196 -3.1140 -59.4390 -64.4999 0.6904 0.5846 -0.0073 0.0059 -0.0132
0.6846 1.2061 7000 -3.1178 -3.1121 -59.5252 -64.6186 0.6902 0.5897 -0.0081 0.0063 -0.0144
0.6837 1.2233 7100 -3.1168 -3.1111 -59.5873 -64.6919 0.6902 0.5864 -0.0088 0.0064 -0.0151
0.6846 1.2405 7200 -3.1156 -3.1099 -59.6617 -64.7766 0.6901 0.5825 -0.0095 0.0065 -0.0160
0.6901 1.2578 7300 -3.1146 -3.1089 -59.7055 -64.8355 0.6901 0.5811 -0.0099 0.0066 -0.0166
0.6842 1.2750 7400 -3.1143 -3.1086 -59.7743 -64.9195 0.6900 0.5785 -0.0106 0.0068 -0.0174
0.6862 1.2922 7500 -3.1130 -3.1073 -59.8530 -65.0033 0.6900 0.5778 -0.0114 0.0068 -0.0182
0.6862 1.3094 7600 -3.1114 -3.1057 -59.9138 -65.0788 0.6899 0.5809 -0.0120 0.0070 -0.0190
0.6848 1.3267 7700 -3.1101 -3.1044 -59.9958 -65.1799 0.6898 0.5829 -0.0128 0.0072 -0.0200
0.6874 1.3439 7800 -3.1095 -3.1038 -60.0529 -65.2523 0.6897 0.5818 -0.0134 0.0073 -0.0207
0.6863 1.3611 7900 -3.1087 -3.1030 -60.1189 -65.3403 0.6896 0.5825 -0.0141 0.0075 -0.0216
0.6861 1.3784 8000 -3.1078 -3.1021 -60.2116 -65.4348 0.6896 0.5836 -0.0150 0.0075 -0.0225
0.6868 1.3956 8100 -3.1069 -3.1012 -60.2774 -65.5086 0.6896 0.5843 -0.0157 0.0076 -0.0233
0.6843 1.4128 8200 -3.1059 -3.1002 -60.3531 -65.6035 0.6895 0.5834 -0.0164 0.0078 -0.0242
0.6821 1.4300 8300 -3.1054 -3.0997 -60.3991 -65.6808 0.6894 0.5846 -0.0169 0.0081 -0.0250
0.6844 1.4473 8400 -3.1046 -3.0989 -60.4732 -65.7510 0.6894 0.5890 -0.0176 0.0081 -0.0257
0.6817 1.4645 8500 -3.1035 -3.0979 -60.5483 -65.8390 0.6894 0.5820 -0.0184 0.0082 -0.0266
0.6845 1.4817 8600 -3.1020 -3.0962 -60.6134 -65.9166 0.6893 0.5911 -0.0190 0.0084 -0.0274
0.6817 1.4990 8700 -3.1018 -3.0962 -60.6714 -65.9962 0.6892 0.5876 -0.0196 0.0086 -0.0282
0.6847 1.5162 8800 -3.1011 -3.0955 -60.7533 -66.0791 0.6892 0.5895 -0.0204 0.0086 -0.0290
0.6823 1.5334 8900 -3.1000 -3.0943 -60.8401 -66.1949 0.6891 0.5899 -0.0213 0.0089 -0.0301
0.6758 1.5507 9000 -3.0991 -3.0934 -60.8754 -66.2333 0.6891 0.5911 -0.0216 0.0089 -0.0305
0.6809 1.5679 9100 -3.0981 -3.0923 -60.9570 -66.3274 0.6890 0.5881 -0.0225 0.0090 -0.0315
0.6818 1.5851 9200 -3.0977 -3.0919 -61.0578 -66.4371 0.6890 0.5874 -0.0235 0.0091 -0.0326
0.6832 1.6023 9300 -3.0969 -3.0911 -61.0783 -66.4729 0.6889 0.5860 -0.0237 0.0093 -0.0329
0.6855 1.6196 9400 -3.0955 -3.0898 -61.1367 -66.5481 0.6888 0.5948 -0.0242 0.0094 -0.0337
0.6839 1.6368 9500 -3.0946 -3.0889 -61.1814 -66.5979 0.6888 0.5925 -0.0247 0.0095 -0.0342
0.682 1.6540 9600 -3.0937 -3.0880 -61.2442 -66.6790 0.6887 0.5918 -0.0253 0.0097 -0.0350
0.6826 1.6713 9700 -3.0929 -3.0872 -61.2947 -66.7452 0.6886 0.5964 -0.0258 0.0098 -0.0357
0.678 1.6885 9800 -3.0924 -3.0867 -61.3594 -66.8017 0.6887 0.5911 -0.0265 0.0097 -0.0362
0.6778 1.7057 9900 -3.0915 -3.0858 -61.4006 -66.8464 0.6887 0.5980 -0.0269 0.0098 -0.0367
0.6802 1.7229 10000 -3.0913 -3.0856 -61.4128 -66.8861 0.6885 0.5950 -0.0270 0.0101 -0.0371
0.6794 1.7402 10100 -3.0902 -3.0845 -61.4624 -66.9483 0.6885 0.5904 -0.0275 0.0102 -0.0377
0.6826 1.7574 10200 -3.0891 -3.0834 -61.5260 -67.0267 0.6884 0.5943 -0.0281 0.0103 -0.0385
0.6855 1.7746 10300 -3.0885 -3.0828 -61.5950 -67.1042 0.6884 0.5943 -0.0288 0.0104 -0.0392
0.6811 1.7919 10400 -3.0877 -3.0820 -61.6404 -67.1605 0.6883 0.5960 -0.0293 0.0105 -0.0398
0.6821 1.8091 10500 -3.0870 -3.0813 -61.7017 -67.2424 0.6882 0.5950 -0.0299 0.0107 -0.0406
0.6848 1.8263 10600 -3.0865 -3.0808 -61.7690 -67.3148 0.6882 0.5953 -0.0306 0.0108 -0.0413
0.6784 1.8436 10700 -3.0858 -3.0801 -61.8288 -67.3826 0.6882 0.5948 -0.0312 0.0109 -0.0420
0.6802 1.8608 10800 -3.0852 -3.0795 -61.8876 -67.4650 0.6881 0.5964 -0.0318 0.0111 -0.0428
0.6788 1.8780 10900 -3.0848 -3.0790 -61.9505 -67.5242 0.6881 0.5929 -0.0324 0.0111 -0.0434
0.6883 1.8952 11000 -3.0842 -3.0785 -61.9829 -67.5779 0.6880 0.5892 -0.0327 0.0113 -0.0440
0.6814 1.9125 11100 -3.0834 -3.0777 -62.0617 -67.6717 0.6879 0.5941 -0.0335 0.0114 -0.0449
0.6796 1.9297 11200 -3.0830 -3.0773 -62.1302 -67.7374 0.6880 0.5950 -0.0342 0.0114 -0.0456
0.6839 1.9469 11300 -3.0824 -3.0767 -62.1736 -67.8008 0.6879 0.5899 -0.0346 0.0116 -0.0462
0.6825 1.9642 11400 -3.0815 -3.0758 -62.2222 -67.8551 0.6878 0.5892 -0.0351 0.0116 -0.0467
0.6855 1.9814 11500 -3.0812 -3.0755 -62.2490 -67.8931 0.6878 0.5920 -0.0354 0.0118 -0.0471
0.6827 1.9986 11600 -3.0808 -3.0750 -62.3037 -67.9551 0.6878 0.5932 -0.0359 0.0118 -0.0477
0.6771 2.0159 11700 -3.0803 -3.0746 -62.3479 -68.0038 0.6877 0.6004 -0.0364 0.0119 -0.0482
0.6814 2.0331 11800 -3.0795 -3.0737 -62.3687 -68.0385 0.6877 0.5929 -0.0366 0.0120 -0.0486
0.679 2.0503 11900 -3.0788 -3.0731 -62.4235 -68.0889 0.6877 0.5964 -0.0371 0.0120 -0.0491
0.6823 2.0675 12000 -3.0788 -3.0731 -62.4228 -68.1068 0.6876 0.5980 -0.0371 0.0122 -0.0493
0.6811 2.0848 12100 -3.0785 -3.0728 -62.4527 -68.1324 0.6876 0.5922 -0.0374 0.0121 -0.0495
0.6774 2.1020 12200 -3.0782 -3.0725 -62.4816 -68.1725 0.6876 0.5915 -0.0377 0.0122 -0.0499
0.6787 2.1192 12300 -3.0778 -3.0721 -62.5033 -68.1982 0.6876 0.5848 -0.0379 0.0123 -0.0502
0.6816 2.1365 12400 -3.0775 -3.0718 -62.5353 -68.2405 0.6875 0.5953 -0.0382 0.0124 -0.0506
0.6767 2.1537 12500 -3.0770 -3.0712 -62.5716 -68.2769 0.6875 0.5922 -0.0386 0.0124 -0.0510
0.6797 2.1709 12600 -3.0766 -3.0709 -62.6046 -68.3137 0.6875 0.5974 -0.0389 0.0124 -0.0513
0.6777 2.1881 12700 -3.0762 -3.0705 -62.6237 -68.3343 0.6875 0.5936 -0.0391 0.0124 -0.0515
0.6832 2.2054 12800 -3.0758 -3.0701 -62.6200 -68.3353 0.6875 0.5939 -0.0391 0.0125 -0.0516
0.6782 2.2226 12900 -3.0755 -3.0698 -62.6173 -68.3392 0.6875 0.5922 -0.0391 0.0125 -0.0516
0.6834 2.2398 13000 -3.0752 -3.0695 -62.6487 -68.3802 0.6874 0.5946 -0.0394 0.0126 -0.0520
0.6793 2.2571 13100 -3.0749 -3.0692 -62.6767 -68.4021 0.6874 0.5925 -0.0396 0.0126 -0.0522
0.679 2.2743 13200 -3.0748 -3.0691 -62.6739 -68.3968 0.6875 0.5913 -0.0396 0.0125 -0.0522
0.685 2.2915 13300 -3.0744 -3.0687 -62.6910 -68.4248 0.6874 0.5915 -0.0398 0.0127 -0.0524
0.6766 2.3088 13400 -3.0740 -3.0684 -62.7184 -68.4492 0.6874 0.5922 -0.0401 0.0126 -0.0527
0.6747 2.3260 13500 -3.0737 -3.0680 -62.7332 -68.4680 0.6874 0.5913 -0.0402 0.0127 -0.0529
0.6812 2.3432 13600 -3.0732 -3.0675 -62.7365 -68.4875 0.6873 0.5927 -0.0402 0.0128 -0.0531
0.6801 2.3604 13700 -3.0732 -3.0675 -62.7692 -68.5067 0.6874 0.5911 -0.0406 0.0127 -0.0533
0.6809 2.3777 13800 -3.0728 -3.0671 -62.7696 -68.5208 0.6873 0.5946 -0.0406 0.0128 -0.0534
0.6816 2.3949 13900 -3.0730 -3.0674 -62.7933 -68.5354 0.6874 0.5953 -0.0408 0.0127 -0.0536
0.6821 2.4121 14000 -3.0728 -3.0670 -62.8191 -68.5660 0.6873 0.5897 -0.0411 0.0128 -0.0539
0.6775 2.4294 14100 -3.0726 -3.0669 -62.8322 -68.5844 0.6873 0.5946 -0.0412 0.0128 -0.0540
0.6787 2.4466 14200 -3.0723 -3.0666 -62.8444 -68.6121 0.6873 0.5976 -0.0413 0.0130 -0.0543
0.6777 2.4638 14300 -3.0725 -3.0667 -62.8538 -68.6067 0.6873 0.5927 -0.0414 0.0128 -0.0543
0.6785 2.4810 14400 -3.0720 -3.0663 -62.8758 -68.6288 0.6873 0.5922 -0.0416 0.0128 -0.0545
0.6723 2.4983 14500 -3.0719 -3.0662 -62.8855 -68.6587 0.6872 0.5927 -0.0417 0.0130 -0.0548
0.6776 2.5155 14600 -3.0718 -3.0661 -62.8962 -68.6640 0.6873 0.5936 -0.0418 0.0130 -0.0548
0.677 2.5327 14700 -3.0717 -3.0660 -62.8845 -68.6537 0.6872 0.5915 -0.0417 0.0130 -0.0547
0.6804 2.5500 14800 -3.0713 -3.0656 -62.9039 -68.6914 0.6872 0.5925 -0.0419 0.0132 -0.0551
0.6793 2.5672 14900 -3.0715 -3.0658 -62.9045 -68.6895 0.6872 0.5978 -0.0419 0.0132 -0.0551
0.6836 2.5844 15000 -3.0714 -3.0657 -62.9167 -68.6898 0.6872 0.5883 -0.0420 0.0130 -0.0551
0.6829 2.6017 15100 -3.0713 -3.0656 -62.9200 -68.7007 0.6872 0.5929 -0.0421 0.0131 -0.0552
0.6745 2.6189 15200 -3.0708 -3.0651 -62.9332 -68.7246 0.6871 0.5939 -0.0422 0.0132 -0.0554
0.6792 2.6361 15300 -3.0714 -3.0657 -62.9299 -68.7169 0.6872 0.5925 -0.0422 0.0132 -0.0554
0.6756 2.6533 15400 -3.0711 -3.0654 -62.9311 -68.7170 0.6872 0.5899 -0.0422 0.0132 -0.0554
0.6876 2.6706 15500 -3.0710 -3.0653 -62.9352 -68.7054 0.6872 0.5939 -0.0422 0.0130 -0.0553
0.6793 2.6878 15600 -3.0708 -3.0651 -62.9398 -68.7274 0.6872 0.5939 -0.0423 0.0132 -0.0555
0.6814 2.7050 15700 -3.0709 -3.0652 -62.9370 -68.7273 0.6871 0.5901 -0.0423 0.0132 -0.0555
0.6771 2.7223 15800 -3.0711 -3.0654 -62.9376 -68.7270 0.6872 0.5950 -0.0423 0.0132 -0.0555
0.6732 2.7395 15900 -3.0708 -3.0651 -62.9390 -68.7115 0.6872 0.5929 -0.0423 0.0130 -0.0553
0.6784 2.7567 16000 -3.0706 -3.0649 -62.9350 -68.7192 0.6872 0.5925 -0.0422 0.0132 -0.0554
0.6783 2.7739 16100 -3.0708 -3.0651 -62.9475 -68.7257 0.6872 0.5876 -0.0424 0.0131 -0.0555
0.6733 2.7912 16200 -3.0709 -3.0652 -62.9490 -68.7286 0.6872 0.5899 -0.0424 0.0131 -0.0555
0.6816 2.8084 16300 -3.0708 -3.0651 -62.9483 -68.7357 0.6872 0.5997 -0.0424 0.0132 -0.0556
0.6797 2.8256 16400 -3.0709 -3.0652 -62.9484 -68.7255 0.6872 0.5911 -0.0424 0.0131 -0.0555
0.6758 2.8429 16500 -3.0711 -3.0654 -62.9445 -68.7217 0.6872 0.5915 -0.0423 0.0131 -0.0554
0.6868 2.8601 16600 -3.0709 -3.0652 -62.9350 -68.7219 0.6872 0.5920 -0.0422 0.0132 -0.0554
0.6821 2.8773 16700 -3.0711 -3.0654 -62.9312 -68.7231 0.6871 0.5915 -0.0422 0.0132 -0.0554
0.6862 2.8946 16800 -3.0705 -3.0648 -62.9449 -68.7250 0.6872 0.5929 -0.0423 0.0131 -0.0554
0.6855 2.9118 16900 -3.0710 -3.0653 -62.9327 -68.7176 0.6872 0.5908 -0.0422 0.0132 -0.0554
0.6815 2.9290 17000 -3.0706 -3.0650 -62.9358 -68.7221 0.6872 0.5906 -0.0422 0.0132 -0.0554
0.6795 2.9462 17100 -3.0706 -3.0649 -62.9485 -68.7265 0.6872 0.5929 -0.0424 0.0131 -0.0555
0.6849 2.9635 17200 -3.0705 -3.0649 -62.9499 -68.7367 0.6872 0.5925 -0.0424 0.0132 -0.0556
0.6857 2.9807 17300 -3.0704 -3.0648 -62.9481 -68.7256 0.6872 0.5934 -0.0424 0.0131 -0.0555
0.6745 2.9979 17400 -3.0704 -3.0648 -62.9434 -68.7246 0.6872 0.5943 -0.0423 0.0131 -0.0554

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.1.2
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
34
Safetensors
Model size
1.1B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for martimfasantos/tinyllama-1.1b-sum-dpo-full_LR2e-8_3epochs_old

Dataset used to train martimfasantos/tinyllama-1.1b-sum-dpo-full_LR2e-8_3epochs_old