update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- generator
|
| 7 |
+
model-index:
|
| 8 |
+
- name: distilgpt2-concat
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# distilgpt2-concat
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the generator dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 4.3325
|
| 20 |
+
|
| 21 |
+
## Model description
|
| 22 |
+
|
| 23 |
+
More information needed
|
| 24 |
+
|
| 25 |
+
## Intended uses & limitations
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Training and evaluation data
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training procedure
|
| 34 |
+
|
| 35 |
+
### Training hyperparameters
|
| 36 |
+
|
| 37 |
+
The following hyperparameters were used during training:
|
| 38 |
+
- learning_rate: 0.0005
|
| 39 |
+
- train_batch_size: 64
|
| 40 |
+
- eval_batch_size: 64
|
| 41 |
+
- seed: 42
|
| 42 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 43 |
+
- lr_scheduler_type: cosine
|
| 44 |
+
- lr_scheduler_warmup_steps: 1000
|
| 45 |
+
- num_epochs: 10
|
| 46 |
+
- mixed_precision_training: Native AMP
|
| 47 |
+
|
| 48 |
+
### Training results
|
| 49 |
+
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 51 |
+
|:-------------:|:-----:|:-----:|:---------------:|
|
| 52 |
+
| 6.7514 | 0.29 | 500 | 5.6224 |
|
| 53 |
+
| 5.3454 | 0.58 | 1000 | 5.1814 |
|
| 54 |
+
| 4.9931 | 0.87 | 1500 | 4.9290 |
|
| 55 |
+
| 4.7222 | 1.16 | 2000 | 4.7811 |
|
| 56 |
+
| 4.5672 | 1.45 | 2500 | 4.6657 |
|
| 57 |
+
| 4.4669 | 1.74 | 3000 | 4.5721 |
|
| 58 |
+
| 4.3738 | 2.02 | 3500 | 4.4939 |
|
| 59 |
+
| 4.175 | 2.31 | 4000 | 4.4613 |
|
| 60 |
+
| 4.1659 | 2.6 | 4500 | 4.4128 |
|
| 61 |
+
| 4.1369 | 2.89 | 5000 | 4.3666 |
|
| 62 |
+
| 3.9858 | 3.18 | 5500 | 4.3656 |
|
| 63 |
+
| 3.9337 | 3.47 | 6000 | 4.3419 |
|
| 64 |
+
| 3.9348 | 3.76 | 6500 | 4.3095 |
|
| 65 |
+
| 3.8826 | 4.05 | 7000 | 4.3066 |
|
| 66 |
+
| 3.7106 | 4.34 | 7500 | 4.3104 |
|
| 67 |
+
| 3.7404 | 4.63 | 8000 | 4.2893 |
|
| 68 |
+
| 3.7459 | 4.92 | 8500 | 4.2648 |
|
| 69 |
+
| 3.5695 | 5.21 | 9000 | 4.2984 |
|
| 70 |
+
| 3.536 | 5.49 | 9500 | 4.2887 |
|
| 71 |
+
| 3.5604 | 5.78 | 10000 | 4.2711 |
|
| 72 |
+
| 3.5007 | 6.07 | 10500 | 4.2900 |
|
| 73 |
+
| 3.3477 | 6.36 | 11000 | 4.3013 |
|
| 74 |
+
| 3.3629 | 6.65 | 11500 | 4.2906 |
|
| 75 |
+
| 3.3771 | 6.94 | 12000 | 4.2814 |
|
| 76 |
+
| 3.211 | 7.23 | 12500 | 4.3131 |
|
| 77 |
+
| 3.1938 | 7.52 | 13000 | 4.3124 |
|
| 78 |
+
| 3.21 | 7.81 | 13500 | 4.3093 |
|
| 79 |
+
| 3.159 | 8.1 | 14000 | 4.3204 |
|
| 80 |
+
| 3.0726 | 8.39 | 14500 | 4.3257 |
|
| 81 |
+
| 3.0762 | 8.68 | 15000 | 4.3269 |
|
| 82 |
+
| 3.0834 | 8.96 | 15500 | 4.3257 |
|
| 83 |
+
| 3.0173 | 9.25 | 16000 | 4.3311 |
|
| 84 |
+
| 3.0116 | 9.54 | 16500 | 4.3325 |
|
| 85 |
+
| 3.0155 | 9.83 | 17000 | 4.3325 |
|
| 86 |
+
|
| 87 |
+
|
| 88 |
+
### Framework versions
|
| 89 |
+
|
| 90 |
+
- Transformers 4.26.1
|
| 91 |
+
- Pytorch 1.11.0+cu113
|
| 92 |
+
- Datasets 2.13.0
|
| 93 |
+
- Tokenizers 0.13.3
|