results_1026_hard

This model is a fine-tuned version of google/gemma-2b-it on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.9233

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: 0.0002
  • train_batch_size: 3
  • eval_batch_size: 2
  • seed: 42
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 24
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.03
  • num_epochs: 20
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
3.5736 0.0390 10 3.7270
3.5199 0.0780 20 3.6964
3.5506 0.1170 30 3.6179
3.5942 0.1560 40 3.4426
3.6553 0.1950 50 3.1769
3.0275 0.2340 60 2.9904
2.8814 0.2730 70 2.8504
2.7885 0.3120 80 2.7437
2.7586 0.3510 90 2.6479
2.7403 0.3901 100 2.5633
2.6222 0.4291 110 2.5077
2.5677 0.4681 120 2.4772
2.5527 0.5071 130 2.4482
2.4995 0.5461 140 2.4147
2.4424 0.5851 150 2.3864
2.516 0.6241 160 2.3641
2.5714 0.6631 170 2.3598
2.4379 0.7021 180 2.3446
2.4347 0.7411 190 2.3232
2.3534 0.7801 200 2.3023
2.5086 0.8191 210 2.2902
2.4283 0.8581 220 2.2820
2.4121 0.8971 230 2.2730
2.3684 0.9361 240 2.2586
2.201 0.9751 250 2.2538
2.4081 1.0141 260 2.2342
2.4069 1.0531 270 2.2349
2.4516 1.0922 280 2.2325
2.3317 1.1312 290 2.2203
2.2839 1.1702 300 2.2135
2.1889 1.2092 310 2.2072
2.3444 1.2482 320 2.2099
2.3374 1.2872 330 2.1991
2.2693 1.3262 340 2.1864
2.2853 1.3652 350 2.1808
2.1615 1.4042 360 2.1646
2.3807 1.4432 370 2.1669
2.3797 1.4822 380 2.1583
2.2207 1.5212 390 2.1504
2.2082 1.5602 400 2.1472
2.1436 1.5992 410 2.1371
2.3218 1.6382 420 2.1415
2.2971 1.6772 430 2.1366
2.2211 1.7162 440 2.1206
2.1516 1.7552 450 2.1175
2.1373 1.7942 460 2.1064
2.3533 1.8333 470 2.1160
2.3165 1.8723 480 2.1046
2.1929 1.9113 490 2.0939
2.1694 1.9503 500 2.0889
2.0101 1.9893 510 2.0816
2.2328 2.0283 520 2.0820
2.3023 2.0673 530 2.0838
2.2905 2.1063 540 2.0826
2.1098 2.1453 550 2.0691
2.0876 2.1843 560 2.0685
2.1142 2.2233 570 2.0720
2.2877 2.2623 580 2.0681
2.2996 2.3013 590 2.0669
2.0875 2.3403 600 2.0579
2.0833 2.3793 610 2.0567
2.1244 2.4183 620 2.0530
2.2519 2.4573 630 2.0538
2.2328 2.4963 640 2.0530
2.0509 2.5353 650 2.0461
2.0441 2.5744 660 2.0434
2.1149 2.6134 670 2.0402
2.2131 2.6524 680 2.0430
2.2375 2.6914 690 2.0395
2.0678 2.7304 700 2.0300
2.0432 2.7694 710 2.0288
2.0401 2.8084 720 2.0232
2.2512 2.8474 730 2.0257
2.1837 2.8864 740 2.0220
2.0976 2.9254 750 2.0143
2.0048 2.9644 760 2.0171
2.0055 3.0034 770 2.0105
2.1763 3.0424 780 2.0121
2.1476 3.0814 790 2.0148
2.0939 3.1204 800 2.0080
2.06 3.1594 810 2.0050
1.8336 3.1984 820 2.0098
2.1914 3.2374 830 2.0038
2.1717 3.2765 840 2.0042
2.0993 3.3155 850 2.0002
2.0274 3.3545 860 1.9945
1.882 3.3935 870 2.0022
2.1923 3.4325 880 1.9964
2.2148 3.4715 890 1.9979
2.0947 3.5105 900 1.9944
2.0078 3.5495 910 1.9884
1.8902 3.5885 920 1.9910
2.1769 3.6275 930 1.9871
2.2394 3.6665 940 1.9887
2.1078 3.7055 950 1.9832
2.0057 3.7445 960 1.9808
1.8797 3.7835 970 1.9785
2.1812 3.8225 980 1.9810
2.1957 3.8615 990 1.9831
2.0468 3.9005 1000 1.9771
2.0083 3.9395 1010 1.9747
1.864 3.9785 1020 1.9784
2.1341 4.0176 1030 1.9728
2.1277 4.0566 1040 1.9806
2.1695 4.0956 1050 1.9781
1.9919 4.1346 1060 1.9711
1.9703 4.1736 1070 1.9711
1.8744 4.2126 1080 1.9706
2.1932 4.2516 1090 1.9754
2.1219 4.2906 1100 1.9708
1.9301 4.3296 1110 1.9680
1.9821 4.3686 1120 1.9669
1.8614 4.4076 1130 1.9657
2.1329 4.4466 1140 1.9670
2.1088 4.4856 1150 1.9618
1.9898 4.5246 1160 1.9570
1.9979 4.5636 1170 1.9601
1.9033 4.6026 1180 1.9591
2.1564 4.6416 1190 1.9636
2.0839 4.6806 1200 1.9606
2.0341 4.7196 1210 1.9550
1.9504 4.7587 1220 1.9556
1.9073 4.7977 1230 1.9541
2.1465 4.8367 1240 1.9572
2.1004 4.8757 1250 1.9546
1.9396 4.9147 1260 1.9477
1.9283 4.9537 1270 1.9499
1.8075 4.9927 1280 1.9501
2.0063 5.0317 1290 1.9530
2.1161 5.0707 1300 1.9565
2.0103 5.1097 1310 1.9503
1.8916 5.1487 1320 1.9493
1.8334 5.1877 1330 1.9523
1.977 5.2267 1340 1.9471
2.054 5.2657 1350 1.9465
2.0638 5.3047 1360 1.9491
1.9736 5.3437 1370 1.9415
1.8327 5.3827 1380 1.9484
1.9745 5.4217 1390 1.9401
2.1027 5.4608 1400 1.9483
2.0582 5.4998 1410 1.9440
1.8867 5.5388 1420 1.9367
1.8296 5.5778 1430 1.9456
1.9654 5.6168 1440 1.9392
2.0957 5.6558 1450 1.9441
2.0372 5.6948 1460 1.9427
1.9088 5.7338 1470 1.9345
1.875 5.7728 1480 1.9393
2.0197 5.8118 1490 1.9352
2.0809 5.8508 1500 1.9411
2.0418 5.8898 1510 1.9376
1.8642 5.9288 1520 1.9351
1.8732 5.9678 1530 1.9369
1.8508 6.0068 1540 1.9335
2.0401 6.0458 1550 1.9413
2.0478 6.0848 1560 1.9382
1.9128 6.1238 1570 1.9375
1.8604 6.1628 1580 1.9375
1.7512 6.2019 1590 1.9432
2.0882 6.2409 1600 1.9374
2.0302 6.2799 1610 1.9393
1.9593 6.3189 1620 1.9333
1.9192 6.3579 1630 1.9312
1.7429 6.3969 1640 1.9428
2.0898 6.4359 1650 1.9323
2.0244 6.4749 1660 1.9358
1.9363 6.5139 1670 1.9320
1.8761 6.5529 1680 1.9300
1.7863 6.5919 1690 1.9343
2.0783 6.6309 1700 1.9277
2.0438 6.6699 1710 1.9286
1.9239 6.7089 1720 1.9271
1.9065 6.7479 1730 1.9270
1.7128 6.7869 1740 1.9394
2.054 6.8259 1750 1.9251
2.0033 6.8649 1760 1.9303
1.9365 6.9039 1770 1.9272
1.8432 6.9430 1780 1.9246
1.7173 6.9820 1790 1.9321
1.9681 7.0210 1800 1.9275
2.0218 7.0600 1810 1.9284
2.0277 7.0990 1820 1.9318
1.8129 7.1380 1830 1.9231
1.8421 7.1770 1840 1.9297
1.756 7.2160 1850 1.9271
1.9677 7.2550 1860 1.9298
2.0093 7.2940 1870 1.9282
1.8747 7.3330 1880 1.9236
1.8182 7.3720 1890 1.9277
1.8005 7.4110 1900 1.9244
2.0192 7.4500 1910 1.9244
2.0789 7.4890 1920 1.9260
1.8149 7.5280 1930 1.9231
1.8864 7.5670 1940 1.9239
1.7469 7.6060 1950 1.9270
2.06 7.6451 1960 1.9231
2.0537 7.6841 1970 1.9232
1.8399 7.7231 1980 1.9211
1.8329 7.7621 1990 1.9236
1.7469 7.8011 2000 1.9226
1.9746 7.8401 2010 1.9228
1.9878 7.8791 2020 1.9228
1.8066 7.9181 2030 1.9176
1.8584 7.9571 2040 1.9234
1.7306 7.9961 2050 1.9184
1.9665 8.0351 2060 1.9241
1.9003 8.0741 2070 1.9247
1.9381 8.1131 2080 1.9223
1.8019 8.1521 2090 1.9218
1.7392 8.1911 2100 1.9345
1.9122 8.2301 2110 1.9183
2.0289 8.2691 2120 1.9245
1.9618 8.3081 2130 1.9195
1.8147 8.3471 2140 1.9208
1.6113 8.3862 2150 1.9290
1.9722 8.4252 2160 1.9180
1.9899 8.4642 2170 1.9219
1.9632 8.5032 2180 1.9210
1.8129 8.5422 2190 1.9175
1.7413 8.5812 2200 1.9213
1.9669 8.6202 2210 1.9176
1.9385 8.6592 2220 1.9225
1.8858 8.6982 2230 1.9170
1.7968 8.7372 2240 1.9173
1.6389 8.7762 2250 1.9226
1.8684 8.8152 2260 1.9127
1.9646 8.8542 2270 1.9214
1.9303 8.8932 2280 1.9169
1.8307 8.9322 2290 1.9165
1.7232 8.9712 2300 1.9188
1.865 9.0102 2310 1.9131
1.9066 9.0492 2320 1.9200
2.0 9.0882 2330 1.9192
1.8236 9.1273 2340 1.9192
1.7505 9.1663 2350 1.9224
1.6129 9.2053 2360 1.9270
1.9888 9.2443 2370 1.9168
1.9675 9.2833 2380 1.9193
1.8029 9.3223 2390 1.9172
1.8321 9.3613 2400 1.9166
1.6774 9.4003 2410 1.9255
1.9432 9.4393 2420 1.9170
1.955 9.4783 2430 1.9195
1.83 9.5173 2440 1.9173
1.787 9.5563 2450 1.9175
1.6397 9.5953 2460 1.9254
1.9513 9.6343 2470 1.9109
1.9759 9.6733 2480 1.9218
1.8647 9.7123 2490 1.9130
1.8224 9.7513 2500 1.9199
1.6592 9.7903 2510 1.9250
1.9316 9.8294 2520 1.9128
1.9789 9.8684 2530 1.9187
1.8478 9.9074 2540 1.9131
1.7795 9.9464 2550 1.9176
1.6661 9.9854 2560 1.9209
1.8865 10.0244 2570 1.9161
1.909 10.0634 2580 1.9197
1.8718 10.1024 2590 1.9178
1.7341 10.1414 2600 1.9172
1.7184 10.1804 2610 1.9230
1.7985 10.2194 2620 1.9196
1.9851 10.2584 2630 1.9150
1.9196 10.2974 2640 1.9202
1.7619 10.3364 2650 1.9138
1.8059 10.3754 2660 1.9206
1.7272 10.4144 2670 1.9210
1.9646 10.4534 2680 1.9155
1.9286 10.4924 2690 1.9222
1.813 10.5314 2700 1.9144
1.7469 10.5705 2710 1.9255
1.7245 10.6095 2720 1.9180
1.9804 10.6485 2730 1.9138
1.9858 10.6875 2740 1.9181
1.7625 10.7265 2750 1.9143
1.7629 10.7655 2760 1.9200
1.7465 10.8045 2770 1.9179
1.8676 10.8435 2780 1.9119
1.8867 10.8825 2790 1.9181
1.6718 10.9215 2800 1.9134
1.702 10.9605 2810 1.9191
1.6299 10.9995 2820 1.9160
1.9335 11.0385 2830 1.9129
1.886 11.0775 2840 1.9169
1.8241 11.1165 2850 1.9167
1.7694 11.1555 2860 1.9178
1.5334 11.1945 2870 1.9289
1.9443 11.2335 2880 1.9132
1.8983 11.2725 2890 1.9186
1.8696 11.3116 2900 1.9184
1.7524 11.3506 2910 1.9164
1.5569 11.3896 2920 1.9265
1.8525 11.4286 2930 1.9143
1.9238 11.4676 2940 1.9148
1.8338 11.5066 2950 1.9172
1.7537 11.5456 2960 1.9155
1.5637 11.5846 2970 1.9269
1.9556 11.6236 2980 1.9172
1.9149 11.6626 2990 1.9140
1.8231 11.7016 3000 1.9158
1.7019 11.7406 3010 1.9177
1.6225 11.7796 3020 1.9238
1.881 11.8186 3030 1.9145
1.9226 11.8576 3040 1.9195
1.8623 11.8966 3050 1.9168
1.7823 11.9356 3060 1.9134
1.6013 11.9746 3070 1.9238
1.8802 12.0137 3080 1.9156
1.9009 12.0527 3090 1.9137
1.901 12.0917 3100 1.9186
1.8332 12.1307 3110 1.9138
1.6932 12.1697 3120 1.9233
1.6191 12.2087 3130 1.9314
1.8903 12.2477 3140 1.9130
1.9178 12.2867 3150 1.9212
1.7168 12.3257 3160 1.9218
1.7454 12.3647 3170 1.9182
1.6673 12.4037 3180 1.9234
1.8559 12.4427 3190 1.9157
1.9028 12.4817 3200 1.9182
1.7831 12.5207 3210 1.9175
1.7204 12.5597 3220 1.9188
1.6052 12.5987 3230 1.9255
1.903 12.6377 3240 1.9152
1.9326 12.6767 3250 1.9200
1.8399 12.7157 3260 1.9144
1.7603 12.7548 3270 1.9183
1.5705 12.7938 3280 1.9264
1.9442 12.8328 3290 1.9142
1.8617 12.8718 3300 1.9191
1.7211 12.9108 3310 1.9161
1.7292 12.9498 3320 1.9176
1.6164 12.9888 3330 1.9236
1.7743 13.0278 3340 1.9167
1.8782 13.0668 3350 1.9186
1.8963 13.1058 3360 1.9194
1.7273 13.1448 3370 1.9176
1.6749 13.1838 3380 1.9228
1.7136 13.2228 3390 1.9224
1.8888 13.2618 3400 1.9148
1.9772 13.3008 3410 1.9192
1.6517 13.3398 3420 1.9187
1.6801 13.3788 3430 1.9231
1.7207 13.4178 3440 1.9198
1.9426 13.4569 3450 1.9155
1.8399 13.4959 3460 1.9193
1.7387 13.5349 3470 1.9175
1.6471 13.5739 3480 1.9209
1.6685 13.6129 3490 1.9217
1.7974 13.6519 3500 1.9155
1.9056 13.6909 3510 1.9162
1.7132 13.7299 3520 1.9158
1.7372 13.7689 3530 1.9189
1.7267 13.8079 3540 1.9205
1.907 13.8469 3550 1.9172
1.8128 13.8859 3560 1.9204
1.695 13.9249 3570 1.9174
1.6292 13.9639 3580 1.9240
1.6518 14.0029 3590 1.9240
1.8934 14.0419 3600 1.9160
1.8894 14.0809 3610 1.9192
1.7708 14.1199 3620 1.9224
1.7299 14.1589 3630 1.9198
1.5334 14.1980 3640 1.9260
1.9007 14.2370 3650 1.9207
1.9238 14.2760 3660 1.9158
1.7738 14.3150 3670 1.9207
1.698 14.3540 3680 1.9230
1.5367 14.3930 3690 1.9276
1.8589 14.4320 3700 1.9210
1.8649 14.4710 3710 1.9190
1.7754 14.5100 3720 1.9189
1.7097 14.5490 3730 1.9194
1.5139 14.5880 3740 1.9258
1.9137 14.6270 3750 1.9215
1.8741 14.6660 3760 1.9177
1.846 14.7050 3770 1.9178
1.7331 14.7440 3780 1.9167
1.5727 14.7830 3790 1.9249
1.8976 14.8220 3800 1.9253
1.8668 14.8610 3810 1.9189
1.757 14.9000 3820 1.9170
1.6765 14.9391 3830 1.9184
1.4913 14.9781 3840 1.9260
1.805 15.0171 3850 1.9232
1.7945 15.0561 3860 1.9191
1.9473 15.0951 3870 1.9186
1.7805 15.1341 3880 1.9189
1.6793 15.1731 3890 1.9205
1.6189 15.2121 3900 1.9265
1.866 15.2511 3910 1.9222
1.8517 15.2901 3920 1.9206
1.7299 15.3291 3930 1.9210
1.7081 15.3681 3940 1.9210
1.5948 15.4071 3950 1.9258
1.8608 15.4461 3960 1.9231
1.8593 15.4851 3970 1.9203
1.7276 15.5241 3980 1.9202
1.6947 15.5631 3990 1.9221
1.6253 15.6021 4000 1.9253
1.8292 15.6412 4010 1.9215
1.8675 15.6802 4020 1.9198
1.6399 15.7192 4030 1.9199
1.6357 15.7582 4040 1.9231
1.6157 15.7972 4050 1.9271
1.8636 15.8362 4060 1.9218
1.9543 15.8752 4070 1.9186
1.7454 15.9142 4080 1.9182
1.7537 15.9532 4090 1.9191
1.5771 15.9922 4100 1.9249
1.7834 16.0312 4110 1.9229
1.8561 16.0702 4120 1.9214
1.8247 16.1092 4130 1.9202
1.6602 16.1482 4140 1.9188
1.5985 16.1872 4150 1.9223
1.7295 16.2262 4160 1.9242
1.8385 16.2652 4170 1.9226
1.8439 16.3042 4180 1.9206
1.6876 16.3432 4190 1.9194
1.6098 16.3823 4200 1.9209
1.7531 16.4213 4210 1.9237
1.8777 16.4603 4220 1.9228
1.8005 16.4993 4230 1.9199
1.7005 16.5383 4240 1.9192
1.6033 16.5773 4250 1.9217
1.7276 16.6163 4260 1.9250
1.8966 16.6553 4270 1.9232
1.8298 16.6943 4280 1.9210
1.7147 16.7333 4290 1.9189
1.6529 16.7723 4300 1.9204
1.7152 16.8113 4310 1.9226
1.9019 16.8503 4320 1.9216
1.8542 16.8893 4330 1.9207
1.678 16.9283 4340 1.9202
1.6445 16.9673 4350 1.9218
1.6286 17.0063 4360 1.9249
1.8551 17.0453 4370 1.9246
1.9001 17.0843 4380 1.9228
1.7704 17.1234 4390 1.9214
1.6516 17.1624 4400 1.9208
1.5215 17.2014 4410 1.9230
1.8635 17.2404 4420 1.9238
1.8312 17.2794 4430 1.9234
1.7603 17.3184 4440 1.9223
1.6772 17.3574 4450 1.9220
1.4826 17.3964 4460 1.9241
1.8174 17.4354 4470 1.9250
1.845 17.4744 4480 1.9243
1.7784 17.5134 4490 1.9231
1.706 17.5524 4500 1.9220
1.5611 17.5914 4510 1.9228
1.876 17.6304 4520 1.9234
1.8972 17.6694 4530 1.9224
1.8434 17.7084 4540 1.9210
1.7421 17.7474 4550 1.9201
1.52 17.7864 4560 1.9212
1.8748 17.8255 4570 1.9223
1.8471 17.8645 4580 1.9232
1.7265 17.9035 4590 1.9233
1.6646 17.9425 4600 1.9228
1.5362 17.9815 4610 1.9234
1.8238 18.0205 4620 1.9238
1.879 18.0595 4630 1.9235
1.8606 18.0985 4640 1.9228
1.6117 18.1375 4650 1.9223
1.6915 18.1765 4660 1.9219
1.6033 18.2155 4670 1.9224
1.8238 18.2545 4680 1.9228
1.8833 18.2935 4690 1.9226
1.7583 18.3325 4700 1.9218
1.7145 18.3715 4710 1.9214
1.554 18.4105 4720 1.9221
1.9294 18.4495 4730 1.9225
1.8302 18.4885 4740 1.9225
1.7174 18.5275 4750 1.9221
1.6752 18.5666 4760 1.9222
1.6211 18.6056 4770 1.9228
1.8542 18.6446 4780 1.9232
1.8592 18.6836 4790 1.9230
1.6617 18.7226 4800 1.9227
1.6886 18.7616 4810 1.9228
1.6375 18.8006 4820 1.9230
1.8799 18.8396 4830 1.9230
1.8634 18.8786 4840 1.9231
1.6909 18.9176 4850 1.9230
1.6734 18.9566 4860 1.9231
1.5097 18.9956 4870 1.9233
1.874 19.0346 4880 1.9234
1.8426 19.0736 4890 1.9234
1.8063 19.1126 4900 1.9234
1.6723 19.1516 4910 1.9233
1.5852 19.1906 4920 1.9234
1.7843 19.2296 4930 1.9235
1.9114 19.2686 4940 1.9235
1.8529 19.3077 4950 1.9234
1.6797 19.3467 4960 1.9233
1.5724 19.3857 4970 1.9234
1.7458 19.4247 4980 1.9234
1.8628 19.4637 4990 1.9234
1.7605 19.5027 5000 1.9233
1.7124 19.5417 5010 1.9233
1.5613 19.5807 5020 1.9233
1.7912 19.6197 5030 1.9233
1.8772 19.6587 5040 1.9233
1.8407 19.6977 5050 1.9233
1.6174 19.7367 5060 1.9233
1.5523 19.7757 5070 1.9233
1.7137 19.8147 5080 1.9233
1.8212 19.8537 5090 1.9233
1.8225 19.8927 5100 1.9233
1.7476 19.9317 5110 1.9233
1.5684 19.9707 5120 1.9233

Framework versions

  • PEFT 0.12.0
  • Transformers 4.45.0
  • Pytorch 2.4.0+cu121
  • Datasets 2.21.0
  • Tokenizers 0.20.1
Downloads last month
1
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no pipeline_tag.

Model tree for SangMoone/results_1026_hard

Base model

google/gemma-2b-it
Adapter
(553)
this model