Downloads · 30 days
15
32% of all-time downloads
lig_model_1
ldm0612/lig_model_1
lig_model_1 is a text generation model from ldm0612. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as mit.
should probably proofread and complete it, then remove this comment. --
All-time downloads
47
Public
Parameters
87.8M
771 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors351 MB · 100%
At a glance
- Task
- Text Generation
- Library
- transformers
- License
- mit
- Model type
- gpt2
- Access
- Public
- Created
- May 14, 2024
- Updated
- May 27, 2024
- SHA
- fd0635c6
Try a prompt
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
lig_model_1
This model is a fine-tuned version of gpt2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.7688
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: 0.0005
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.8679 | 0.2310 | 8 | 6.6211 |
| 6.3894 | 0.4621 | 16 | 6.3666 |
| 6.2641 | 0.6931 | 24 | 6.2481 |
| 6.1285 | 0.9242 | 32 | 6.0829 |
| 5.9436 | 1.1552 | 40 | 5.8900 |
| 5.8073 | 1.3863 | 48 | 5.7490 |
| 5.7164 | 1.6173 | 56 | 5.6617 |
| 5.6019 | 1.8484 | 64 | 5.5778 |
| 5.5427 | 2.0794 | 72 | 5.4886 |
| 5.454 | 2.3105 | 80 | 5.3954 |
| 5.3546 | 2.5415 | 88 | 5.3066 |
| 5.3014 | 2.7726 | 96 | 5.2124 |
| 5.2448 | 3.0036 | 104 | 5.1365 |
| 5.1185 | 3.2347 | 112 | 5.0765 |
| 5.0938 | 3.4657 | 120 | 5.0071 |
| 5.0347 | 3.6968 | 128 | 4.9339 |
| 4.9681 | 3.9278 | 136 | 4.8552 |
| 4.8323 | 4.1588 | 144 | 4.7821 |
| 4.7912 | 4.3899 | 152 | 4.7215 |
| 4.7225 | 4.6209 | 160 | 4.6431 |
| 4.6433 | 4.8520 | 168 | 4.5701 |
| 4.5309 | 5.0830 | 176 | 4.5002 |
| 4.4506 | 5.3141 | 184 | 4.4442 |
| 4.4097 | 5.5451 | 192 | 4.3820 |
| 4.3871 | 5.7762 | 200 | 4.3290 |
| 4.3345 | 6.0072 | 208 | 4.2869 |
| 4.2004 | 6.2383 | 216 | 4.2412 |
| 4.1716 | 6.4693 | 224 | 4.1978 |
| 4.1536 | 6.7004 | 232 | 4.1607 |
| 4.0975 | 6.9314 | 240 | 4.1294 |
| 3.9743 | 7.1625 | 248 | 4.1014 |
| 3.922 | 7.3935 | 256 | 4.0654 |
| 3.939 | 7.6245 | 264 | 4.0378 |
| 3.9208 | 7.8556 | 272 | 4.0102 |
| 3.8083 | 8.0866 | 280 | 3.9812 |
| 3.7611 | 8.3177 | 288 | 3.9630 |
| 3.7668 | 8.5487 | 296 | 3.9407 |
| 3.7285 | 8.7798 | 304 | 3.9183 |
| 3.6996 | 9.0108 | 312 | 3.8958 |
| 3.5754 | 9.2419 | 320 | 3.8825 |
| 3.5708 | 9.4729 | 328 | 3.8702 |
| 3.5607 | 9.7040 | 336 | 3.8510 |
| 3.5688 | 9.9350 | 344 | 3.8387 |
| 3.4188 | 10.1661 | 352 | 3.8350 |
| 3.432 | 10.3971 | 360 | 3.8261 |
| 3.4236 | 10.6282 | 368 | 3.8131 |
| 3.3985 | 10.8592 | 376 | 3.8026 |
| 3.306 | 11.0903 | 384 | 3.7934 |
| 3.3196 | 11.3213 | 392 | 3.7919 |
| 3.3031 | 11.5523 | 400 | 3.7908 |
| 3.2851 | 11.7834 | 408 | 3.7817 |
| 3.2703 | 12.0144 | 416 | 3.7789 |
| 3.2132 | 12.2455 | 424 | 3.7818 |
| 3.1829 | 12.4765 | 432 | 3.7778 |
| 3.1968 | 12.7076 | 440 | 3.7749 |
| 3.2206 | 12.9386 | 448 | 3.7711 |
| 3.1521 | 13.1697 | 456 | 3.7694 |
| 3.1412 | 13.4007 | 464 | 3.7700 |
| 3.1415 | 13.6318 | 472 | 3.7709 |
| 3.1402 | 13.8628 | 480 | 3.7694 |
| 3.129 | 14.0939 | 488 | 3.7689 |
| 3.1221 | 14.3249 | 496 | 3.7687 |
| 3.1576 | 14.5560 | 504 | 3.7688 |
Framework versions
- Transformers 4.41.0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
- Task
- Text Generation
- Library
- transformers
- Type
- gpt2
- License
- mit
- Created
- May 14, 2024
- Updated
- May 27, 2024