t5-small-finetuned-thai-informal-to-formal

This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.3091
  • Bleu: 20.5964
  • Gen Len: 19.9981

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-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 300
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Bleu Gen Len
2.2862 1.0 1011 2.2028 31.6678 20.0
2.1228 2.0 2022 2.0339 32.3643 20.0
2.0581 3.0 3033 1.9386 32.3784 20.0
1.9714 4.0 4044 1.8899 31.9728 20.0
1.9169 5.0 5055 1.8318 32.1064 20.0
1.8969 6.0 6066 1.8005 31.4324 20.0
1.8486 7.0 7077 1.7813 31.7758 20.0
1.802 8.0 8088 1.7464 31.9055 20.0
1.7654 9.0 9099 1.7352 31.9598 20.0
1.7439 10.0 10110 1.7009 32.1696 20.0
1.7603 11.0 11121 1.6873 31.8118 20.0
1.7288 12.0 12132 1.6678 31.5711 20.0
1.7004 13.0 13143 1.6482 31.4575 20.0
1.6851 14.0 14154 1.6374 31.9579 20.0
1.6497 15.0 15165 1.6290 31.4299 20.0
1.656 16.0 16176 1.6130 31.2145 20.0
1.6423 17.0 17187 1.5931 31.365 20.0
1.6024 18.0 18198 1.5797 31.2247 20.0
1.6064 19.0 19209 1.5736 31.1535 20.0
1.5974 20.0 20220 1.5609 31.431 20.0
1.5961 21.0 21231 1.5578 30.9905 20.0
1.5621 22.0 22242 1.5466 30.8979 20.0
1.5307 23.0 23253 1.5285 31.277 20.0
1.5359 24.0 24264 1.5370 31.4321 20.0
1.5558 25.0 25275 1.5215 31.2769 20.0
1.513 26.0 26286 1.5173 30.9782 19.9997
1.5241 27.0 27297 1.5105 30.6717 20.0
1.5133 28.0 28308 1.4973 30.3152 20.0
1.4713 29.0 29319 1.4927 30.276 19.9997
1.478 30.0 30330 1.4887 30.1004 19.9989
1.4572 31.0 31341 1.4845 29.8939 19.9983
1.4485 32.0 32352 1.4653 30.0169 19.9986
1.4404 33.0 33363 1.4648 28.9061 19.9989
1.4408 34.0 34374 1.4586 29.598 19.9994
1.4296 35.0 35385 1.4585 28.9821 19.9981
1.408 36.0 36396 1.4517 29.6025 19.9986
1.4004 37.0 37407 1.4456 27.8564 19.9992
1.3991 38.0 38418 1.4411 28.8947 19.9994
1.401 39.0 39429 1.4309 27.6809 19.9994
1.391 40.0 40440 1.4278 29.1687 19.9994
1.3709 41.0 41451 1.4217 28.2947 19.9989
1.3726 42.0 42462 1.4247 27.2108 19.9983
1.3702 43.0 43473 1.4144 25.9973 19.9981
1.3636 44.0 44484 1.4163 26.0146 19.9953
1.3673 45.0 45495 1.4118 25.8126 19.9978
1.3539 46.0 46506 1.4076 25.5185 19.9981
1.3434 47.0 47517 1.4023 26.2123 19.9947
1.3428 48.0 48528 1.4008 25.8932 19.9955
1.3325 49.0 49539 1.4003 25.7762 19.9969
1.3258 50.0 50550 1.3896 24.8206 19.9961
1.3151 51.0 51561 1.3852 24.4683 19.9978
1.3035 52.0 52572 1.3843 24.9821 19.9992
1.2931 53.0 53583 1.3847 24.715 19.9989
1.2707 54.0 54594 1.3776 24.4374 19.9986
1.2792 55.0 55605 1.3801 23.7683 19.9967
1.284 56.0 56616 1.3781 23.6961 19.9975
1.2664 57.0 57627 1.3680 23.6677 19.9975
1.2783 58.0 58638 1.3695 23.3193 19.9986
1.2762 59.0 59649 1.3741 22.613 19.9972
1.2759 60.0 60660 1.3629 23.9067 19.9964
1.2618 61.0 61671 1.3687 23.7587 19.9967
1.2614 62.0 62682 1.3613 23.2615 19.9975
1.2455 63.0 63693 1.3623 23.8722 19.9986
1.1977 64.0 64704 1.3528 23.1421 19.9981
1.2199 65.0 65715 1.3520 22.6977 19.9975
1.2368 66.0 66726 1.3552 23.2495 19.9989
1.2087 67.0 67737 1.3404 22.6422 19.9989
1.214 68.0 68748 1.3499 21.979 19.9972
1.2322 69.0 69759 1.3453 22.1766 19.9978
1.2028 70.0 70770 1.3402 21.8311 19.9975
1.2163 71.0 71781 1.3399 22.1417 19.9989
1.1769 72.0 72792 1.3446 22.253 19.9972
1.221 73.0 73803 1.3413 22.1546 19.9986
1.1768 74.0 74814 1.3335 21.8914 19.9972
1.1829 75.0 75825 1.3323 21.7763 19.9947
1.1687 76.0 76836 1.3344 21.4495 19.9964
1.1873 77.0 77847 1.3337 21.7655 19.9964
1.1807 78.0 78858 1.3308 21.4564 19.9967
1.1735 79.0 79869 1.3282 21.233 19.9967
1.1693 80.0 80880 1.3240 21.0794 19.9955
1.1714 81.0 81891 1.3262 21.1856 19.9969
1.154 82.0 82902 1.3282 20.5583 19.9964
1.1572 83.0 83913 1.3229 20.9262 19.995
1.1473 84.0 84924 1.3233 20.5432 19.995
1.1315 85.0 85935 1.3227 20.4939 19.9942
1.1567 86.0 86946 1.3203 21.3354 19.9964
1.1485 87.0 87957 1.3211 20.9952 19.9939
1.1313 88.0 88968 1.3202 20.1199 19.9961
1.1428 89.0 89979 1.3188 20.414 19.9925
1.1374 90.0 90990 1.3220 20.003 19.993
1.1274 91.0 92001 1.3153 20.7172 19.9953
1.1174 92.0 93012 1.3126 20.5997 19.9953
1.1155 93.0 94023 1.3131 20.0402 19.993
1.1167 94.0 95034 1.3140 20.219 19.9905
1.1301 95.0 96045 1.3142 19.8332 19.9922
1.0975 96.0 97056 1.3096 19.6051 19.9942
1.1025 97.0 98067 1.3148 20.4323 19.993
1.0932 98.0 99078 1.3134 20.0839 19.9942
1.0871 99.0 100089 1.3071 20.0202 19.9939
1.102 100.0 101100 1.3091 20.0454 19.9947
1.0969 101.0 102111 1.3090 19.4474 19.9947
1.0988 102.0 103122 1.3117 20.1905 19.9922
1.0816 103.0 104133 1.3048 20.3346 19.9928
1.0809 104.0 105144 1.3058 20.323 19.9953
1.0861 105.0 106155 1.3052 20.6984 19.9944
1.0907 106.0 107166 1.3076 20.3413 19.9947
1.0747 107.0 108177 1.3050 20.3362 19.9955
1.0839 108.0 109188 1.3060 20.5379 19.9936
1.0755 109.0 110199 1.3071 20.3886 19.9939
1.0463 110.0 111210 1.3058 19.9524 19.9953
1.0644 111.0 112221 1.3033 19.7226 19.9972
1.0771 112.0 113232 1.3089 19.9861 19.9958
1.0819 113.0 114243 1.3031 20.5527 19.9942
1.0483 114.0 115254 1.3063 20.0048 19.9978
1.04 115.0 116265 1.3020 20.2327 19.9969
1.0574 116.0 117276 1.3025 19.6818 19.995
1.0356 117.0 118287 1.3077 20.1054 19.9967
1.0525 118.0 119298 1.3022 20.14 19.9967
1.0409 119.0 120309 1.2983 19.7657 19.9972
1.0431 120.0 121320 1.2945 20.1315 19.9975
1.0419 121.0 122331 1.3035 19.8364 19.9972
1.0411 122.0 123342 1.2951 20.204 19.9981
1.0396 123.0 124353 1.3019 20.6711 19.9955
1.0424 124.0 125364 1.2950 20.6527 19.9969
1.0203 125.0 126375 1.3008 20.4314 19.9972
1.0351 126.0 127386 1.3008 20.0237 19.9978
1.0424 127.0 128397 1.2993 20.3024 19.9983
1.0165 128.0 129408 1.2960 20.1769 19.9978
1.0216 129.0 130419 1.2977 19.8483 19.9972
1.0207 130.0 131430 1.2939 20.0639 19.9969
1.0119 131.0 132441 1.2985 19.731 19.9972
0.9965 132.0 133452 1.3006 19.5983 19.9969
1.0034 133.0 134463 1.2974 19.6943 19.9989
1.0241 134.0 135474 1.3015 20.0083 19.9981
1.0181 135.0 136485 1.2982 19.6057 19.9989
1.0112 136.0 137496 1.2931 19.3408 19.9986
0.9927 137.0 138507 1.2999 19.5222 19.9983
1.0134 138.0 139518 1.2909 19.42 19.9989
0.9921 139.0 140529 1.2951 19.8604 19.9989
0.9891 140.0 141540 1.2916 20.0752 19.9989
0.9896 141.0 142551 1.2910 19.7536 19.9992
1.0034 142.0 143562 1.2934 20.0064 19.9986
0.9718 143.0 144573 1.2973 19.9304 19.9989
1.0141 144.0 145584 1.2940 20.5053 19.9986
0.99 145.0 146595 1.2980 20.0913 19.9975
0.9729 146.0 147606 1.2927 19.7229 19.9978
0.9732 147.0 148617 1.2920 20.2104 19.9975
0.9778 148.0 149628 1.2947 20.1365 19.9981
0.987 149.0 150639 1.3007 20.3436 19.9972
0.987 150.0 151650 1.3003 20.2827 19.9983
0.9788 151.0 152661 1.2953 20.2941 19.9972
0.9899 152.0 153672 1.2951 20.5454 19.9978
0.978 153.0 154683 1.2946 20.7448 19.9969
0.9614 154.0 155694 1.2975 20.5359 19.9969
0.9759 155.0 156705 1.2925 20.3661 19.9975
0.9627 156.0 157716 1.2954 20.5535 19.9969
0.9692 157.0 158727 1.2930 20.1919 19.9969
0.9737 158.0 159738 1.2922 20.484 19.9972
0.9642 159.0 160749 1.2952 20.5444 19.9975
0.9679 160.0 161760 1.2930 20.3731 19.9983
0.9571 161.0 162771 1.2933 20.4158 19.9978
0.9542 162.0 163782 1.2937 20.4823 19.9978
0.9537 163.0 164793 1.2997 20.6457 19.9964
0.951 164.0 165804 1.2982 20.0897 19.9986
0.9556 165.0 166815 1.2944 20.45 19.9986
0.9534 166.0 167826 1.2961 20.2743 19.9967
0.9381 167.0 168837 1.2922 19.8311 19.9969
0.9347 168.0 169848 1.2938 19.9427 19.9978
0.9514 169.0 170859 1.2968 20.2039 19.9983
0.9439 170.0 171870 1.3014 19.9784 19.9961
0.9379 171.0 172881 1.3000 20.1213 19.9964
0.9326 172.0 173892 1.2930 20.0882 19.9969
0.9178 173.0 174903 1.2942 20.1997 19.9972
0.9511 174.0 175914 1.2931 20.6471 19.9969
0.9438 175.0 176925 1.2945 20.7321 19.9983
0.929 176.0 177936 1.2967 20.5813 19.9964
0.9343 177.0 178947 1.2940 20.2307 19.9978
0.9344 178.0 179958 1.2949 20.2401 19.9969
0.9319 179.0 180969 1.2974 19.9881 19.9972
0.9286 180.0 181980 1.2974 20.2666 19.9961
0.9074 181.0 182991 1.2939 20.2549 19.9969
0.93 182.0 184002 1.2990 20.0121 19.9969
0.9303 183.0 185013 1.2944 20.056 19.9978
0.9259 184.0 186024 1.3003 19.9021 19.9953
0.9014 185.0 187035 1.2962 20.0381 19.9958
0.9288 186.0 188046 1.2976 20.1909 19.9947
0.9086 187.0 189057 1.2969 20.2923 19.9969
0.9183 188.0 190068 1.2941 20.1649 19.9967
0.9141 189.0 191079 1.3028 20.0891 19.9958
0.9264 190.0 192090 1.2935 20.0164 19.9958
0.9307 191.0 193101 1.2956 19.8606 19.9964
0.9179 192.0 194112 1.2933 19.9815 19.9961
0.9123 193.0 195123 1.2977 20.1232 19.9953
0.9221 194.0 196134 1.3014 20.0674 19.995
0.9195 195.0 197145 1.3031 19.9839 19.9944
0.9139 196.0 198156 1.2947 20.0344 19.9953
0.9074 197.0 199167 1.2956 20.1076 19.9961
0.9149 198.0 200178 1.2963 20.0898 19.9955
0.9219 199.0 201189 1.2990 20.171 19.9964
0.8989 200.0 202200 1.2983 20.1548 19.9961
0.9004 201.0 203211 1.2977 20.2135 19.9955
0.9043 202.0 204222 1.3023 20.3024 19.9964
0.917 203.0 205233 1.3014 20.5967 19.9967
0.9012 204.0 206244 1.3001 20.5489 19.9961
0.9136 205.0 207255 1.2963 20.5013 19.9969
0.897 206.0 208266 1.3016 20.3285 19.9969
0.9036 207.0 209277 1.2981 20.3278 19.9967
0.9225 208.0 210288 1.3055 20.4756 19.9967
0.8959 209.0 211299 1.2987 20.3112 19.9972
0.903 210.0 212310 1.2977 20.5512 19.9961
0.9012 211.0 213321 1.3026 20.4304 19.9964
0.8906 212.0 214332 1.2998 20.4206 19.9967
0.8906 213.0 215343 1.3031 20.4499 19.9964
0.9049 214.0 216354 1.3029 20.6908 19.9958
0.9034 215.0 217365 1.2980 20.3614 19.9969
0.8971 216.0 218376 1.2985 20.6196 19.9972
0.885 217.0 219387 1.3019 20.584 19.9972
0.8799 218.0 220398 1.3041 20.5843 19.9967
0.8805 219.0 221409 1.3035 20.5123 19.9972
0.8896 220.0 222420 1.3006 20.7331 19.9975
0.8851 221.0 223431 1.2973 20.6914 19.9975
0.893 222.0 224442 1.3004 20.7484 19.9978
0.8903 223.0 225453 1.3001 20.5207 19.9981
0.8924 224.0 226464 1.3026 20.6635 19.9972
0.8839 225.0 227475 1.3056 20.6999 19.9978
0.8631 226.0 228486 1.3042 20.9581 19.9967
0.8677 227.0 229497 1.3037 20.8283 19.9964
0.867 228.0 230508 1.3042 20.8781 19.9978
0.8878 229.0 231519 1.3035 20.6884 19.9981
0.8805 230.0 232530 1.3092 20.716 19.9975
0.8769 231.0 233541 1.2988 20.6323 19.9975
0.8833 232.0 234552 1.3039 20.5529 19.9978
0.8798 233.0 235563 1.3028 20.5848 19.9981
0.8694 234.0 236574 1.3037 20.4147 19.9983
0.8888 235.0 237585 1.3022 20.5179 19.9983
0.8724 236.0 238596 1.3027 20.4379 19.9978
0.8864 237.0 239607 1.3024 20.3993 19.9972
0.8684 238.0 240618 1.3043 20.5063 19.9969
0.8753 239.0 241629 1.3072 20.4079 19.9969
0.8734 240.0 242640 1.3026 20.5173 19.9967
0.867 241.0 243651 1.3044 20.6249 19.9972
0.8671 242.0 244662 1.3094 20.6827 19.9972
0.8721 243.0 245673 1.3045 20.5017 19.9978
0.8726 244.0 246684 1.3065 20.5748 19.9967
0.8741 245.0 247695 1.3063 20.5345 19.9972
0.8634 246.0 248706 1.3036 20.6084 19.9972
0.8527 247.0 249717 1.3045 20.535 19.9972
0.8662 248.0 250728 1.3089 20.5306 19.9972
0.8681 249.0 251739 1.3081 20.6414 19.9967
0.8711 250.0 252750 1.3061 20.6039 19.9975
0.8653 251.0 253761 1.3018 20.5632 19.9975
0.8697 252.0 254772 1.3090 20.5056 19.9978
0.8655 253.0 255783 1.3082 20.5235 19.9978
0.8636 254.0 256794 1.3067 20.5607 19.9972
0.8667 255.0 257805 1.3066 20.6694 19.9964
0.8596 256.0 258816 1.3073 20.617 19.9967
0.8507 257.0 259827 1.3083 20.6035 19.9964
0.8677 258.0 260838 1.3077 20.6196 19.9975
0.8614 259.0 261849 1.3094 20.6928 19.9969
0.8677 260.0 262860 1.3098 20.7181 19.9969
0.8628 261.0 263871 1.3065 20.679 19.9975
0.8636 262.0 264882 1.3055 20.7476 19.9975
0.8624 263.0 265893 1.3065 20.7045 19.9972
0.8594 264.0 266904 1.3093 20.5442 19.9964
0.8658 265.0 267915 1.3105 20.7153 19.9972
0.8476 266.0 268926 1.3076 20.677 19.9972
0.858 267.0 269937 1.3091 20.6701 19.9969
0.8707 268.0 270948 1.3111 20.5985 19.9975
0.8613 269.0 271959 1.3092 20.6108 19.9975
0.8497 270.0 272970 1.3070 20.5836 19.9964
0.8654 271.0 273981 1.3082 20.5806 19.9983
0.8621 272.0 274992 1.3088 20.6817 19.9975
0.8619 273.0 276003 1.3090 20.5567 19.9975
0.8638 274.0 277014 1.3087 20.6233 19.9975
0.8642 275.0 278025 1.3092 20.667 19.9967
0.8498 276.0 279036 1.3069 20.6295 19.9969
0.8572 277.0 280047 1.3107 20.6376 19.9969
0.8543 278.0 281058 1.3114 20.6473 19.9964
0.8453 279.0 282069 1.3105 20.6931 19.9967
0.8575 280.0 283080 1.3077 20.691 19.9972
0.8492 281.0 284091 1.3101 20.7528 19.9969
0.8519 282.0 285102 1.3094 20.6812 19.9981
0.8431 283.0 286113 1.3114 20.6608 19.9969
0.8546 284.0 287124 1.3093 20.6336 19.9981
0.86 285.0 288135 1.3108 20.6077 19.9967
0.8674 286.0 289146 1.3096 20.6742 19.9978
0.8493 287.0 290157 1.3106 20.6674 19.9981
0.8521 288.0 291168 1.3099 20.5915 19.9981
0.856 289.0 292179 1.3102 20.6448 19.9978
0.8614 290.0 293190 1.3096 20.6515 19.9981
0.8628 291.0 294201 1.3108 20.6679 19.9978
0.8498 292.0 295212 1.3104 20.6623 19.9978
0.8617 293.0 296223 1.3097 20.6591 19.9978
0.8563 294.0 297234 1.3098 20.6266 19.9978
0.856 295.0 298245 1.3095 20.6536 19.9978
0.8493 296.0 299256 1.3095 20.6273 19.9978
0.8498 297.0 300267 1.3092 20.5942 19.9978
0.8539 298.0 301278 1.3092 20.5942 19.9978
0.8608 299.0 302289 1.3091 20.5915 19.9981
0.8437 300.0 303300 1.3091 20.5964 19.9981

Framework versions

  • Transformers 4.21.0
  • Pytorch 1.12.0+cu113
  • Datasets 2.4.0
  • Tokenizers 0.12.1
Downloads last month
15
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.