Commit
·
88988ba
1
Parent(s):
98fa273
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,91 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model: gpt2
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: MammoLLM
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# MammoLLM
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the None dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.9172
|
| 19 |
+
|
| 20 |
+
## Model description
|
| 21 |
+
|
| 22 |
+
More information needed
|
| 23 |
+
|
| 24 |
+
## Intended uses & limitations
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Training and evaluation data
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training procedure
|
| 33 |
+
|
| 34 |
+
### Training hyperparameters
|
| 35 |
+
|
| 36 |
+
The following hyperparameters were used during training:
|
| 37 |
+
- learning_rate: 0.0005
|
| 38 |
+
- train_batch_size: 64
|
| 39 |
+
- eval_batch_size: 64
|
| 40 |
+
- seed: 42
|
| 41 |
+
- gradient_accumulation_steps: 2
|
| 42 |
+
- total_train_batch_size: 128
|
| 43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
+
- lr_scheduler_type: cosine
|
| 45 |
+
- lr_scheduler_warmup_steps: 100
|
| 46 |
+
- num_epochs: 20
|
| 47 |
+
|
| 48 |
+
### Training results
|
| 49 |
+
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 51 |
+
|:-------------:|:-----:|:-----:|:---------------:|
|
| 52 |
+
| 1.5202 | 0.62 | 500 | 1.1041 |
|
| 53 |
+
| 1.1505 | 1.24 | 1000 | 1.0581 |
|
| 54 |
+
| 1.1018 | 1.87 | 1500 | 1.0286 |
|
| 55 |
+
| 1.0694 | 2.49 | 2000 | 1.0166 |
|
| 56 |
+
| 1.0574 | 3.11 | 2500 | 1.0041 |
|
| 57 |
+
| 1.0351 | 3.73 | 3000 | 0.9909 |
|
| 58 |
+
| 1.0193 | 4.36 | 3500 | 0.9865 |
|
| 59 |
+
| 1.0137 | 4.98 | 4000 | 0.9799 |
|
| 60 |
+
| 0.993 | 5.6 | 4500 | 0.9745 |
|
| 61 |
+
| 0.9813 | 6.22 | 5000 | 0.9632 |
|
| 62 |
+
| 0.9728 | 6.85 | 5500 | 0.9573 |
|
| 63 |
+
| 0.9534 | 7.47 | 6000 | 0.9521 |
|
| 64 |
+
| 0.9474 | 8.09 | 6500 | 0.9481 |
|
| 65 |
+
| 0.9264 | 8.71 | 7000 | 0.9405 |
|
| 66 |
+
| 0.9099 | 9.33 | 7500 | 0.9365 |
|
| 67 |
+
| 0.9017 | 9.96 | 8000 | 0.9292 |
|
| 68 |
+
| 0.8735 | 10.58 | 8500 | 0.9267 |
|
| 69 |
+
| 0.8623 | 11.2 | 9000 | 0.9268 |
|
| 70 |
+
| 0.8444 | 11.82 | 9500 | 0.9168 |
|
| 71 |
+
| 0.8205 | 12.45 | 10000 | 0.9148 |
|
| 72 |
+
| 0.8111 | 13.07 | 10500 | 0.9129 |
|
| 73 |
+
| 0.7842 | 13.69 | 11000 | 0.9129 |
|
| 74 |
+
| 0.767 | 14.31 | 11500 | 0.9138 |
|
| 75 |
+
| 0.759 | 14.93 | 12000 | 0.9094 |
|
| 76 |
+
| 0.7329 | 15.56 | 12500 | 0.9109 |
|
| 77 |
+
| 0.7261 | 16.18 | 13000 | 0.9145 |
|
| 78 |
+
| 0.7121 | 16.8 | 13500 | 0.9145 |
|
| 79 |
+
| 0.7038 | 17.42 | 14000 | 0.9161 |
|
| 80 |
+
| 0.699 | 18.05 | 14500 | 0.9167 |
|
| 81 |
+
| 0.6902 | 18.67 | 15000 | 0.9169 |
|
| 82 |
+
| 0.6883 | 19.29 | 15500 | 0.9172 |
|
| 83 |
+
| 0.6873 | 19.91 | 16000 | 0.9172 |
|
| 84 |
+
|
| 85 |
+
|
| 86 |
+
### Framework versions
|
| 87 |
+
|
| 88 |
+
- Transformers 4.31.0
|
| 89 |
+
- Pytorch 2.0.1+cu117
|
| 90 |
+
- Datasets 2.14.3
|
| 91 |
+
- Tokenizers 0.13.3
|