flant5-tuned-3 / README.md
mixtralyanis's picture
End of training
b104e78 verified
|
raw
history blame
8.46 kB
metadata
license: apache-2.0
base_model: google/flan-t5-small
tags:
  - generated_from_trainer
model-index:
  - name: flant5-tuned-3
    results: []

flant5-tuned-3

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

  • Loss: 1.2189

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: 0.0001
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 8
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss
1.9667 0.02 1 1.6464
2.4545 0.04 2 1.6360
2.4307 0.06 3 1.6170
2.1502 0.09 4 1.5949
1.8514 0.11 5 1.5724
1.7189 0.13 6 1.5528
1.9036 0.15 7 1.5352
2.0255 0.17 8 1.5151
2.2073 0.19 9 1.4944
2.1577 0.21 10 1.4748
1.6581 0.23 11 1.4545
1.9323 0.26 12 1.4363
1.4871 0.28 13 1.4198
1.574 0.3 14 1.4032
1.7671 0.32 15 1.3898
1.567 0.34 16 1.3782
1.5162 0.36 17 1.3686
1.9622 0.38 18 1.3599
1.8378 0.4 19 1.3525
1.7199 0.43 20 1.3460
1.3917 0.45 21 1.3402
1.4417 0.47 22 1.3345
1.4023 0.49 23 1.3293
1.5427 0.51 24 1.3239
1.2344 0.53 25 1.3192
2.281 0.55 26 1.3136
1.9236 0.57 27 1.3077
1.4392 0.6 28 1.3029
1.9168 0.62 29 1.2976
2.1688 0.64 30 1.2930
1.2504 0.66 31 1.2890
1.5946 0.68 32 1.2853
1.6979 0.7 33 1.2820
1.6712 0.72 34 1.2789
1.7862 0.74 35 1.2759
1.534 0.77 36 1.2734
1.6904 0.79 37 1.2712
1.6023 0.81 38 1.2692
1.6756 0.83 39 1.2667
2.0195 0.85 40 1.2640
1.2913 0.87 41 1.2618
1.1534 0.89 42 1.2607
1.5612 0.91 43 1.2597
1.3159 0.94 44 1.2586
1.6303 0.96 45 1.2582
1.3721 0.98 46 1.2584
2.703 1.0 47 1.2583
1.3063 1.02 48 1.2585
1.1093 1.04 49 1.2594
1.1362 1.06 50 1.2609
1.6691 1.09 51 1.2619
1.44 1.11 52 1.2620
1.8026 1.13 53 1.2612
1.8663 1.15 54 1.2601
1.3662 1.17 55 1.2584
1.7172 1.19 56 1.2562
1.554 1.21 57 1.2535
1.0628 1.23 58 1.2514
1.389 1.26 59 1.2494
1.0307 1.28 60 1.2481
1.5557 1.3 61 1.2462
1.6536 1.32 62 1.2438
1.652 1.34 63 1.2415
1.51 1.36 64 1.2396
1.5407 1.38 65 1.2374
1.6681 1.4 66 1.2349
1.4797 1.43 67 1.2323
1.326 1.45 68 1.2304
1.8683 1.47 69 1.2285
1.3007 1.49 70 1.2270
1.5261 1.51 71 1.2256
1.6908 1.53 72 1.2241
1.4631 1.55 73 1.2226
1.5474 1.57 74 1.2213
1.0559 1.6 75 1.2209
1.5217 1.62 76 1.2206
1.7606 1.64 77 1.2201
1.5246 1.66 78 1.2197
1.8001 1.68 79 1.2192
1.4414 1.7 80 1.2185
1.4168 1.72 81 1.2182
1.2429 1.74 82 1.2180
1.7092 1.77 83 1.2178
1.4605 1.79 84 1.2178
1.2242 1.81 85 1.2180
1.6583 1.83 86 1.2180
1.7079 1.85 87 1.2181
0.9831 1.87 88 1.2186
1.6504 1.89 89 1.2191
1.7244 1.91 90 1.2194
1.2895 1.94 91 1.2196
1.03 1.96 92 1.2201
1.377 1.98 93 1.2205
1.0463 2.0 94 1.2210
1.3759 2.02 95 1.2214
1.7144 2.04 96 1.2218
1.6047 2.06 97 1.2220
1.6515 2.09 98 1.2222
1.2909 2.11 99 1.2221
1.6717 2.13 100 1.2219
1.1318 2.15 101 1.2220
1.3417 2.17 102 1.2219
1.3242 2.19 103 1.2219
1.4135 2.21 104 1.2221
1.3863 2.23 105 1.2222
1.2301 2.26 106 1.2222
1.5481 2.28 107 1.2220
1.0813 2.3 108 1.2219
1.4198 2.32 109 1.2218
1.4751 2.34 110 1.2214
1.4133 2.36 111 1.2212
0.8784 2.38 112 1.2212
1.514 2.4 113 1.2211
1.2913 2.43 114 1.2210
1.1341 2.45 115 1.2209
1.262 2.47 116 1.2210
1.3282 2.49 117 1.2209
1.5217 2.51 118 1.2206
1.6127 2.53 119 1.2203
1.5625 2.55 120 1.2201
1.4603 2.57 121 1.2199
1.8532 2.6 122 1.2196
1.3278 2.62 123 1.2194
1.0632 2.64 124 1.2192
1.5837 2.66 125 1.2190
1.4593 2.68 126 1.2188
1.2919 2.7 127 1.2188
1.1228 2.72 128 1.2187
1.3098 2.74 129 1.2188
1.6073 2.77 130 1.2188
1.1484 2.79 131 1.2189
1.6054 2.81 132 1.2190
1.5228 2.83 133 1.2190
1.5577 2.85 134 1.2190
1.4234 2.87 135 1.2191
1.7341 2.89 136 1.2191
1.6164 2.91 137 1.2190
1.6621 2.94 138 1.2190
1.5781 2.96 139 1.2189
1.0756 2.98 140 1.2189
1.8596 3.0 141 1.2189

Framework versions

  • Transformers 4.38.1
  • Pytorch 2.1.0+cu121
  • Datasets 2.17.0
  • Tokenizers 0.15.2