File size: 4,239 Bytes
49c4d2f | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 | ---
library_name: transformers
license: apache-2.0
base_model: google/flan-t5-base
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: flan-t5-rouge-durga-3b
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# flan-t5-rouge-durga-3b
This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2135
- Rouge1: 0.4326
- Rouge2: 0.3704
- Rougel: 0.4285
- Rougelsum: 0.4288
## 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: 0.0003
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 30
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:---------:|
| 2.2165 | 1.0 | 2000 | 1.9218 | 0.2604 | 0.1215 | 0.2356 | 0.2356 |
| 1.8357 | 2.0 | 4000 | 1.6767 | 0.2496 | 0.1181 | 0.2262 | 0.2264 |
| 1.11 | 3.0 | 6000 | 1.4820 | 0.2709 | 0.1363 | 0.2460 | 0.2461 |
| 1.9651 | 4.0 | 8000 | 1.3179 | 0.2750 | 0.1410 | 0.2505 | 0.2506 |
| 1.0764 | 5.0 | 10000 | 1.1559 | 0.2768 | 0.1493 | 0.2586 | 0.2586 |
| 1.0195 | 6.0 | 12000 | 1.0138 | 0.2885 | 0.1612 | 0.2685 | 0.2689 |
| 0.7455 | 7.0 | 14000 | 0.8792 | 0.2965 | 0.1739 | 0.2788 | 0.2788 |
| 0.9384 | 8.0 | 16000 | 0.7466 | 0.2998 | 0.1841 | 0.2849 | 0.2849 |
| 1.3216 | 9.0 | 18000 | 0.6505 | 0.3168 | 0.2041 | 0.3007 | 0.3010 |
| 0.9445 | 10.0 | 20000 | 0.5516 | 0.3268 | 0.2238 | 0.3133 | 0.3140 |
| 1.2079 | 11.0 | 22000 | 0.4781 | 0.3370 | 0.2381 | 0.3236 | 0.3237 |
| 0.7909 | 12.0 | 24000 | 0.4186 | 0.3394 | 0.2439 | 0.3249 | 0.3250 |
| 0.4352 | 13.0 | 26000 | 0.3459 | 0.3630 | 0.2749 | 0.3503 | 0.3507 |
| 0.4925 | 14.0 | 28000 | 0.3134 | 0.3665 | 0.2798 | 0.3548 | 0.3553 |
| 0.0707 | 15.0 | 30000 | 0.2752 | 0.3828 | 0.3025 | 0.3724 | 0.3728 |
| 0.2351 | 16.0 | 32000 | 0.2579 | 0.3815 | 0.3009 | 0.3708 | 0.3711 |
| 0.9694 | 17.0 | 34000 | 0.2439 | 0.3874 | 0.3103 | 0.3788 | 0.3795 |
| 0.6217 | 18.0 | 36000 | 0.2273 | 0.4037 | 0.3282 | 0.3956 | 0.3959 |
| 0.0914 | 19.0 | 38000 | 0.2121 | 0.4055 | 0.3343 | 0.3981 | 0.3987 |
| 0.0534 | 20.0 | 40000 | 0.2006 | 0.4134 | 0.3444 | 0.4066 | 0.4067 |
| 0.1081 | 21.0 | 42000 | 0.2009 | 0.4181 | 0.3514 | 0.4131 | 0.4132 |
| 0.0786 | 22.0 | 44000 | 0.2086 | 0.4182 | 0.3488 | 0.4112 | 0.4118 |
| 0.0364 | 23.0 | 46000 | 0.2047 | 0.4182 | 0.3508 | 0.4110 | 0.4114 |
| 0.0568 | 24.0 | 48000 | 0.2058 | 0.4227 | 0.3570 | 0.4175 | 0.4177 |
| 0.1106 | 25.0 | 50000 | 0.2073 | 0.4211 | 0.3566 | 0.4159 | 0.4164 |
| 0.0733 | 26.0 | 52000 | 0.2075 | 0.4248 | 0.3603 | 0.4195 | 0.4199 |
| 0.1169 | 27.0 | 54000 | 0.2123 | 0.4286 | 0.3650 | 0.4244 | 0.4247 |
| 0.0892 | 28.0 | 56000 | 0.2109 | 0.4327 | 0.3704 | 0.4281 | 0.4285 |
| 0.045 | 29.0 | 58000 | 0.2107 | 0.4339 | 0.3715 | 0.4300 | 0.4302 |
| 0.0908 | 30.0 | 60000 | 0.2135 | 0.4326 | 0.3704 | 0.4285 | 0.4288 |
### Framework versions
- Transformers 4.46.0
- Pytorch 2.5.0+cu121
- Datasets 3.0.2
- Tokenizers 0.20.1
|