Training in progress, epoch 1
Browse files- README.md +427 -0
- adapter_config.json +33 -0
- adapter_model.safetensors +3 -0
- added_tokens.json +5 -0
- all_results.json +13 -0
- eval_results.json +8 -0
- merges.txt +0 -0
- special_tokens_map.json +28 -0
- tokenizer.json +0 -0
- tokenizer_config.json +49 -0
- train_results.json +8 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,427 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
license: mit
|
| 4 |
+
base_model: gpt2
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
model-index:
|
| 8 |
+
- name: Se124M100KInfPrompt_NT
|
| 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 |
+
# Se124M100KInfPrompt_NT
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.3899
|
| 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: 2e-05
|
| 39 |
+
- train_batch_size: 16
|
| 40 |
+
- eval_batch_size: 16
|
| 41 |
+
- seed: 42
|
| 42 |
+
- gradient_accumulation_steps: 2
|
| 43 |
+
- total_train_batch_size: 32
|
| 44 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 45 |
+
- lr_scheduler_type: cosine
|
| 46 |
+
- lr_scheduler_warmup_ratio: 0.03
|
| 47 |
+
- num_epochs: 3
|
| 48 |
+
|
| 49 |
+
### Training results
|
| 50 |
+
|
| 51 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 52 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 53 |
+
| 2.9983 | 0.0082 | 20 | 2.6302 |
|
| 54 |
+
| 2.9256 | 0.0164 | 40 | 2.6331 |
|
| 55 |
+
| 2.9534 | 0.0246 | 60 | 2.6305 |
|
| 56 |
+
| 2.9277 | 0.0327 | 80 | 2.6052 |
|
| 57 |
+
| 2.8694 | 0.0409 | 100 | 2.5836 |
|
| 58 |
+
| 2.879 | 0.0491 | 120 | 2.5278 |
|
| 59 |
+
| 2.7972 | 0.0573 | 140 | 2.4722 |
|
| 60 |
+
| 2.7112 | 0.0655 | 160 | 2.4048 |
|
| 61 |
+
| 2.5739 | 0.0737 | 180 | 2.3244 |
|
| 62 |
+
| 2.4522 | 0.0819 | 200 | 2.2167 |
|
| 63 |
+
| 2.3121 | 0.0901 | 220 | 2.0842 |
|
| 64 |
+
| 2.1652 | 0.0982 | 240 | 1.9278 |
|
| 65 |
+
| 2.0135 | 0.1064 | 260 | 1.7658 |
|
| 66 |
+
| 1.8352 | 0.1146 | 280 | 1.5877 |
|
| 67 |
+
| 1.6331 | 0.1228 | 300 | 1.3988 |
|
| 68 |
+
| 1.4721 | 0.1310 | 320 | 1.2257 |
|
| 69 |
+
| 1.3347 | 0.1392 | 340 | 1.0901 |
|
| 70 |
+
| 1.202 | 0.1474 | 360 | 0.9639 |
|
| 71 |
+
| 1.125 | 0.1555 | 380 | 0.8691 |
|
| 72 |
+
| 1.002 | 0.1637 | 400 | 0.8003 |
|
| 73 |
+
| 0.9698 | 0.1719 | 420 | 0.7525 |
|
| 74 |
+
| 0.8963 | 0.1801 | 440 | 0.7148 |
|
| 75 |
+
| 0.8571 | 0.1883 | 460 | 0.6803 |
|
| 76 |
+
| 0.7983 | 0.1965 | 480 | 0.6542 |
|
| 77 |
+
| 0.7838 | 0.2047 | 500 | 0.6332 |
|
| 78 |
+
| 0.7689 | 0.2129 | 520 | 0.6118 |
|
| 79 |
+
| 0.7256 | 0.2210 | 540 | 0.5931 |
|
| 80 |
+
| 0.7146 | 0.2292 | 560 | 0.5799 |
|
| 81 |
+
| 0.686 | 0.2374 | 580 | 0.5673 |
|
| 82 |
+
| 0.6729 | 0.2456 | 600 | 0.5565 |
|
| 83 |
+
| 0.6628 | 0.2538 | 620 | 0.5445 |
|
| 84 |
+
| 0.6525 | 0.2620 | 640 | 0.5406 |
|
| 85 |
+
| 0.6298 | 0.2702 | 660 | 0.5328 |
|
| 86 |
+
| 0.6345 | 0.2783 | 680 | 0.5237 |
|
| 87 |
+
| 0.6171 | 0.2865 | 700 | 0.5169 |
|
| 88 |
+
| 0.6052 | 0.2947 | 720 | 0.5113 |
|
| 89 |
+
| 0.5862 | 0.3029 | 740 | 0.5066 |
|
| 90 |
+
| 0.5767 | 0.3111 | 760 | 0.5021 |
|
| 91 |
+
| 0.5777 | 0.3193 | 780 | 0.4966 |
|
| 92 |
+
| 0.5689 | 0.3275 | 800 | 0.4939 |
|
| 93 |
+
| 0.5677 | 0.3357 | 820 | 0.4894 |
|
| 94 |
+
| 0.5567 | 0.3438 | 840 | 0.4878 |
|
| 95 |
+
| 0.5547 | 0.3520 | 860 | 0.4817 |
|
| 96 |
+
| 0.5516 | 0.3602 | 880 | 0.4808 |
|
| 97 |
+
| 0.5577 | 0.3684 | 900 | 0.4787 |
|
| 98 |
+
| 0.5461 | 0.3766 | 920 | 0.4740 |
|
| 99 |
+
| 0.5449 | 0.3848 | 940 | 0.4712 |
|
| 100 |
+
| 0.5301 | 0.3930 | 960 | 0.4711 |
|
| 101 |
+
| 0.5313 | 0.4011 | 980 | 0.4682 |
|
| 102 |
+
| 0.5278 | 0.4093 | 1000 | 0.4676 |
|
| 103 |
+
| 0.518 | 0.4175 | 1020 | 0.4643 |
|
| 104 |
+
| 0.531 | 0.4257 | 1040 | 0.4621 |
|
| 105 |
+
| 0.5302 | 0.4339 | 1060 | 0.4624 |
|
| 106 |
+
| 0.5238 | 0.4421 | 1080 | 0.4581 |
|
| 107 |
+
| 0.5179 | 0.4503 | 1100 | 0.4572 |
|
| 108 |
+
| 0.5167 | 0.4585 | 1120 | 0.4577 |
|
| 109 |
+
| 0.5181 | 0.4666 | 1140 | 0.4534 |
|
| 110 |
+
| 0.5207 | 0.4748 | 1160 | 0.4536 |
|
| 111 |
+
| 0.5037 | 0.4830 | 1180 | 0.4533 |
|
| 112 |
+
| 0.5117 | 0.4912 | 1200 | 0.4517 |
|
| 113 |
+
| 0.5066 | 0.4994 | 1220 | 0.4500 |
|
| 114 |
+
| 0.5023 | 0.5076 | 1240 | 0.4487 |
|
| 115 |
+
| 0.4903 | 0.5158 | 1260 | 0.4470 |
|
| 116 |
+
| 0.4916 | 0.5239 | 1280 | 0.4462 |
|
| 117 |
+
| 0.4908 | 0.5321 | 1300 | 0.4460 |
|
| 118 |
+
| 0.4956 | 0.5403 | 1320 | 0.4443 |
|
| 119 |
+
| 0.5059 | 0.5485 | 1340 | 0.4438 |
|
| 120 |
+
| 0.4908 | 0.5567 | 1360 | 0.4427 |
|
| 121 |
+
| 0.4978 | 0.5649 | 1380 | 0.4416 |
|
| 122 |
+
| 0.4861 | 0.5731 | 1400 | 0.4410 |
|
| 123 |
+
| 0.4865 | 0.5813 | 1420 | 0.4404 |
|
| 124 |
+
| 0.4916 | 0.5894 | 1440 | 0.4381 |
|
| 125 |
+
| 0.4832 | 0.5976 | 1460 | 0.4352 |
|
| 126 |
+
| 0.4811 | 0.6058 | 1480 | 0.4381 |
|
| 127 |
+
| 0.4779 | 0.6140 | 1500 | 0.4364 |
|
| 128 |
+
| 0.4792 | 0.6222 | 1520 | 0.4381 |
|
| 129 |
+
| 0.4755 | 0.6304 | 1540 | 0.4346 |
|
| 130 |
+
| 0.4797 | 0.6386 | 1560 | 0.4358 |
|
| 131 |
+
| 0.4769 | 0.6467 | 1580 | 0.4321 |
|
| 132 |
+
| 0.4682 | 0.6549 | 1600 | 0.4323 |
|
| 133 |
+
| 0.4797 | 0.6631 | 1620 | 0.4338 |
|
| 134 |
+
| 0.4754 | 0.6713 | 1640 | 0.4332 |
|
| 135 |
+
| 0.4687 | 0.6795 | 1660 | 0.4325 |
|
| 136 |
+
| 0.4629 | 0.6877 | 1680 | 0.4330 |
|
| 137 |
+
| 0.478 | 0.6959 | 1700 | 0.4312 |
|
| 138 |
+
| 0.4693 | 0.7041 | 1720 | 0.4291 |
|
| 139 |
+
| 0.4746 | 0.7122 | 1740 | 0.4305 |
|
| 140 |
+
| 0.4626 | 0.7204 | 1760 | 0.4300 |
|
| 141 |
+
| 0.4641 | 0.7286 | 1780 | 0.4317 |
|
| 142 |
+
| 0.4606 | 0.7368 | 1800 | 0.4287 |
|
| 143 |
+
| 0.4678 | 0.7450 | 1820 | 0.4278 |
|
| 144 |
+
| 0.4736 | 0.7532 | 1840 | 0.4267 |
|
| 145 |
+
| 0.4739 | 0.7614 | 1860 | 0.4270 |
|
| 146 |
+
| 0.4627 | 0.7695 | 1880 | 0.4269 |
|
| 147 |
+
| 0.4596 | 0.7777 | 1900 | 0.4247 |
|
| 148 |
+
| 0.4617 | 0.7859 | 1920 | 0.4245 |
|
| 149 |
+
| 0.4663 | 0.7941 | 1940 | 0.4238 |
|
| 150 |
+
| 0.4569 | 0.8023 | 1960 | 0.4243 |
|
| 151 |
+
| 0.4683 | 0.8105 | 1980 | 0.4229 |
|
| 152 |
+
| 0.4664 | 0.8187 | 2000 | 0.4231 |
|
| 153 |
+
| 0.4711 | 0.8269 | 2020 | 0.4203 |
|
| 154 |
+
| 0.4712 | 0.8350 | 2040 | 0.4201 |
|
| 155 |
+
| 0.4579 | 0.8432 | 2060 | 0.4186 |
|
| 156 |
+
| 0.4688 | 0.8514 | 2080 | 0.4221 |
|
| 157 |
+
| 0.4566 | 0.8596 | 2100 | 0.4222 |
|
| 158 |
+
| 0.4573 | 0.8678 | 2120 | 0.4179 |
|
| 159 |
+
| 0.4606 | 0.8760 | 2140 | 0.4183 |
|
| 160 |
+
| 0.456 | 0.8842 | 2160 | 0.4189 |
|
| 161 |
+
| 0.4684 | 0.8923 | 2180 | 0.4180 |
|
| 162 |
+
| 0.4522 | 0.9005 | 2200 | 0.4183 |
|
| 163 |
+
| 0.4591 | 0.9087 | 2220 | 0.4171 |
|
| 164 |
+
| 0.457 | 0.9169 | 2240 | 0.4194 |
|
| 165 |
+
| 0.4714 | 0.9251 | 2260 | 0.4160 |
|
| 166 |
+
| 0.4637 | 0.9333 | 2280 | 0.4173 |
|
| 167 |
+
| 0.4454 | 0.9415 | 2300 | 0.4190 |
|
| 168 |
+
| 0.4579 | 0.9497 | 2320 | 0.4133 |
|
| 169 |
+
| 0.4567 | 0.9578 | 2340 | 0.4153 |
|
| 170 |
+
| 0.4479 | 0.9660 | 2360 | 0.4152 |
|
| 171 |
+
| 0.4523 | 0.9742 | 2380 | 0.4138 |
|
| 172 |
+
| 0.4559 | 0.9824 | 2400 | 0.4147 |
|
| 173 |
+
| 0.4493 | 0.9906 | 2420 | 0.4131 |
|
| 174 |
+
| 0.4568 | 0.9988 | 2440 | 0.4145 |
|
| 175 |
+
| 0.4494 | 1.0070 | 2460 | 0.4120 |
|
| 176 |
+
| 0.4549 | 1.0151 | 2480 | 0.4120 |
|
| 177 |
+
| 0.4491 | 1.0233 | 2500 | 0.4130 |
|
| 178 |
+
| 0.454 | 1.0315 | 2520 | 0.4143 |
|
| 179 |
+
| 0.4474 | 1.0397 | 2540 | 0.4134 |
|
| 180 |
+
| 0.4541 | 1.0479 | 2560 | 0.4134 |
|
| 181 |
+
| 0.4458 | 1.0561 | 2580 | 0.4117 |
|
| 182 |
+
| 0.4469 | 1.0643 | 2600 | 0.4108 |
|
| 183 |
+
| 0.4502 | 1.0725 | 2620 | 0.4120 |
|
| 184 |
+
| 0.4447 | 1.0806 | 2640 | 0.4102 |
|
| 185 |
+
| 0.445 | 1.0888 | 2660 | 0.4107 |
|
| 186 |
+
| 0.4496 | 1.0970 | 2680 | 0.4080 |
|
| 187 |
+
| 0.445 | 1.1052 | 2700 | 0.4097 |
|
| 188 |
+
| 0.4549 | 1.1134 | 2720 | 0.4071 |
|
| 189 |
+
| 0.4476 | 1.1216 | 2740 | 0.4095 |
|
| 190 |
+
| 0.4427 | 1.1298 | 2760 | 0.4111 |
|
| 191 |
+
| 0.4412 | 1.1379 | 2780 | 0.4091 |
|
| 192 |
+
| 0.441 | 1.1461 | 2800 | 0.4111 |
|
| 193 |
+
| 0.4465 | 1.1543 | 2820 | 0.4080 |
|
| 194 |
+
| 0.4427 | 1.1625 | 2840 | 0.4076 |
|
| 195 |
+
| 0.4417 | 1.1707 | 2860 | 0.4080 |
|
| 196 |
+
| 0.4409 | 1.1789 | 2880 | 0.4080 |
|
| 197 |
+
| 0.4573 | 1.1871 | 2900 | 0.4078 |
|
| 198 |
+
| 0.443 | 1.1953 | 2920 | 0.4067 |
|
| 199 |
+
| 0.4412 | 1.2034 | 2940 | 0.4079 |
|
| 200 |
+
| 0.4384 | 1.2116 | 2960 | 0.4079 |
|
| 201 |
+
| 0.4426 | 1.2198 | 2980 | 0.4083 |
|
| 202 |
+
| 0.4407 | 1.2280 | 3000 | 0.4056 |
|
| 203 |
+
| 0.4487 | 1.2362 | 3020 | 0.4059 |
|
| 204 |
+
| 0.4421 | 1.2444 | 3040 | 0.4064 |
|
| 205 |
+
| 0.4412 | 1.2526 | 3060 | 0.4057 |
|
| 206 |
+
| 0.4354 | 1.2607 | 3080 | 0.4073 |
|
| 207 |
+
| 0.4454 | 1.2689 | 3100 | 0.4056 |
|
| 208 |
+
| 0.4376 | 1.2771 | 3120 | 0.4064 |
|
| 209 |
+
| 0.4469 | 1.2853 | 3140 | 0.4043 |
|
| 210 |
+
| 0.4437 | 1.2935 | 3160 | 0.4038 |
|
| 211 |
+
| 0.4412 | 1.3017 | 3180 | 0.4031 |
|
| 212 |
+
| 0.4354 | 1.3099 | 3200 | 0.4053 |
|
| 213 |
+
| 0.4413 | 1.3181 | 3220 | 0.4050 |
|
| 214 |
+
| 0.4344 | 1.3262 | 3240 | 0.4048 |
|
| 215 |
+
| 0.4471 | 1.3344 | 3260 | 0.4022 |
|
| 216 |
+
| 0.4347 | 1.3426 | 3280 | 0.4049 |
|
| 217 |
+
| 0.4367 | 1.3508 | 3300 | 0.4019 |
|
| 218 |
+
| 0.4391 | 1.3590 | 3320 | 0.4033 |
|
| 219 |
+
| 0.4424 | 1.3672 | 3340 | 0.4019 |
|
| 220 |
+
| 0.4391 | 1.3754 | 3360 | 0.4009 |
|
| 221 |
+
| 0.4377 | 1.3835 | 3380 | 0.4014 |
|
| 222 |
+
| 0.4413 | 1.3917 | 3400 | 0.4015 |
|
| 223 |
+
| 0.4382 | 1.3999 | 3420 | 0.4006 |
|
| 224 |
+
| 0.4298 | 1.4081 | 3440 | 0.4015 |
|
| 225 |
+
| 0.4503 | 1.4163 | 3460 | 0.4019 |
|
| 226 |
+
| 0.4413 | 1.4245 | 3480 | 0.4015 |
|
| 227 |
+
| 0.4343 | 1.4327 | 3500 | 0.3996 |
|
| 228 |
+
| 0.4373 | 1.4409 | 3520 | 0.4002 |
|
| 229 |
+
| 0.4338 | 1.4490 | 3540 | 0.4016 |
|
| 230 |
+
| 0.4292 | 1.4572 | 3560 | 0.4000 |
|
| 231 |
+
| 0.4444 | 1.4654 | 3580 | 0.4004 |
|
| 232 |
+
| 0.4342 | 1.4736 | 3600 | 0.3996 |
|
| 233 |
+
| 0.4339 | 1.4818 | 3620 | 0.4004 |
|
| 234 |
+
| 0.4291 | 1.4900 | 3640 | 0.4006 |
|
| 235 |
+
| 0.435 | 1.4982 | 3660 | 0.3993 |
|
| 236 |
+
| 0.445 | 1.5063 | 3680 | 0.3999 |
|
| 237 |
+
| 0.4389 | 1.5145 | 3700 | 0.4009 |
|
| 238 |
+
| 0.4316 | 1.5227 | 3720 | 0.3988 |
|
| 239 |
+
| 0.4363 | 1.5309 | 3740 | 0.3994 |
|
| 240 |
+
| 0.4384 | 1.5391 | 3760 | 0.3995 |
|
| 241 |
+
| 0.4355 | 1.5473 | 3780 | 0.4006 |
|
| 242 |
+
| 0.436 | 1.5555 | 3800 | 0.3983 |
|
| 243 |
+
| 0.4384 | 1.5637 | 3820 | 0.3981 |
|
| 244 |
+
| 0.4394 | 1.5718 | 3840 | 0.3985 |
|
| 245 |
+
| 0.4392 | 1.5800 | 3860 | 0.3978 |
|
| 246 |
+
| 0.4456 | 1.5882 | 3880 | 0.3991 |
|
| 247 |
+
| 0.4359 | 1.5964 | 3900 | 0.3984 |
|
| 248 |
+
| 0.4328 | 1.6046 | 3920 | 0.4004 |
|
| 249 |
+
| 0.4272 | 1.6128 | 3940 | 0.3992 |
|
| 250 |
+
| 0.4352 | 1.6210 | 3960 | 0.3993 |
|
| 251 |
+
| 0.4262 | 1.6291 | 3980 | 0.3994 |
|
| 252 |
+
| 0.4406 | 1.6373 | 4000 | 0.3979 |
|
| 253 |
+
| 0.4291 | 1.6455 | 4020 | 0.3991 |
|
| 254 |
+
| 0.4262 | 1.6537 | 4040 | 0.3975 |
|
| 255 |
+
| 0.4337 | 1.6619 | 4060 | 0.3978 |
|
| 256 |
+
| 0.4404 | 1.6701 | 4080 | 0.3964 |
|
| 257 |
+
| 0.4408 | 1.6783 | 4100 | 0.3983 |
|
| 258 |
+
| 0.4378 | 1.6865 | 4120 | 0.3977 |
|
| 259 |
+
| 0.4322 | 1.6946 | 4140 | 0.3973 |
|
| 260 |
+
| 0.4343 | 1.7028 | 4160 | 0.3970 |
|
| 261 |
+
| 0.43 | 1.7110 | 4180 | 0.3961 |
|
| 262 |
+
| 0.4343 | 1.7192 | 4200 | 0.3958 |
|
| 263 |
+
| 0.4308 | 1.7274 | 4220 | 0.3965 |
|
| 264 |
+
| 0.4355 | 1.7356 | 4240 | 0.3952 |
|
| 265 |
+
| 0.4371 | 1.7438 | 4260 | 0.3966 |
|
| 266 |
+
| 0.4342 | 1.7519 | 4280 | 0.3956 |
|
| 267 |
+
| 0.4364 | 1.7601 | 4300 | 0.3962 |
|
| 268 |
+
| 0.434 | 1.7683 | 4320 | 0.3953 |
|
| 269 |
+
| 0.4335 | 1.7765 | 4340 | 0.3965 |
|
| 270 |
+
| 0.4317 | 1.7847 | 4360 | 0.3953 |
|
| 271 |
+
| 0.4298 | 1.7929 | 4380 | 0.3954 |
|
| 272 |
+
| 0.4307 | 1.8011 | 4400 | 0.3942 |
|
| 273 |
+
| 0.4345 | 1.8093 | 4420 | 0.3952 |
|
| 274 |
+
| 0.433 | 1.8174 | 4440 | 0.3943 |
|
| 275 |
+
| 0.4261 | 1.8256 | 4460 | 0.3955 |
|
| 276 |
+
| 0.4338 | 1.8338 | 4480 | 0.3950 |
|
| 277 |
+
| 0.4263 | 1.8420 | 4500 | 0.3944 |
|
| 278 |
+
| 0.4263 | 1.8502 | 4520 | 0.3939 |
|
| 279 |
+
| 0.436 | 1.8584 | 4540 | 0.3943 |
|
| 280 |
+
| 0.432 | 1.8666 | 4560 | 0.3946 |
|
| 281 |
+
| 0.4302 | 1.8747 | 4580 | 0.3942 |
|
| 282 |
+
| 0.4333 | 1.8829 | 4600 | 0.3936 |
|
| 283 |
+
| 0.4316 | 1.8911 | 4620 | 0.3936 |
|
| 284 |
+
| 0.4294 | 1.8993 | 4640 | 0.3938 |
|
| 285 |
+
| 0.4265 | 1.9075 | 4660 | 0.3936 |
|
| 286 |
+
| 0.4294 | 1.9157 | 4680 | 0.3943 |
|
| 287 |
+
| 0.4319 | 1.9239 | 4700 | 0.3942 |
|
| 288 |
+
| 0.4391 | 1.9321 | 4720 | 0.3933 |
|
| 289 |
+
| 0.4243 | 1.9402 | 4740 | 0.3944 |
|
| 290 |
+
| 0.4325 | 1.9484 | 4760 | 0.3930 |
|
| 291 |
+
| 0.4343 | 1.9566 | 4780 | 0.3924 |
|
| 292 |
+
| 0.4287 | 1.9648 | 4800 | 0.3938 |
|
| 293 |
+
| 0.4322 | 1.9730 | 4820 | 0.3933 |
|
| 294 |
+
| 0.4283 | 1.9812 | 4840 | 0.3926 |
|
| 295 |
+
| 0.4309 | 1.9894 | 4860 | 0.3935 |
|
| 296 |
+
| 0.4238 | 1.9975 | 4880 | 0.3922 |
|
| 297 |
+
| 0.4217 | 2.0057 | 4900 | 0.3925 |
|
| 298 |
+
| 0.425 | 2.0139 | 4920 | 0.3926 |
|
| 299 |
+
| 0.4389 | 2.0221 | 4940 | 0.3925 |
|
| 300 |
+
| 0.4346 | 2.0303 | 4960 | 0.3920 |
|
| 301 |
+
| 0.4254 | 2.0385 | 4980 | 0.3931 |
|
| 302 |
+
| 0.4223 | 2.0467 | 5000 | 0.3919 |
|
| 303 |
+
| 0.4268 | 2.0549 | 5020 | 0.3930 |
|
| 304 |
+
| 0.4228 | 2.0630 | 5040 | 0.3929 |
|
| 305 |
+
| 0.4325 | 2.0712 | 5060 | 0.3928 |
|
| 306 |
+
| 0.4255 | 2.0794 | 5080 | 0.3928 |
|
| 307 |
+
| 0.4305 | 2.0876 | 5100 | 0.3922 |
|
| 308 |
+
| 0.4333 | 2.0958 | 5120 | 0.3919 |
|
| 309 |
+
| 0.4332 | 2.1040 | 5140 | 0.3927 |
|
| 310 |
+
| 0.4261 | 2.1122 | 5160 | 0.3929 |
|
| 311 |
+
| 0.429 | 2.1203 | 5180 | 0.3916 |
|
| 312 |
+
| 0.4274 | 2.1285 | 5200 | 0.3921 |
|
| 313 |
+
| 0.4277 | 2.1367 | 5220 | 0.3928 |
|
| 314 |
+
| 0.4356 | 2.1449 | 5240 | 0.3913 |
|
| 315 |
+
| 0.4268 | 2.1531 | 5260 | 0.3921 |
|
| 316 |
+
| 0.4297 | 2.1613 | 5280 | 0.3921 |
|
| 317 |
+
| 0.4272 | 2.1695 | 5300 | 0.3915 |
|
| 318 |
+
| 0.4337 | 2.1777 | 5320 | 0.3922 |
|
| 319 |
+
| 0.4312 | 2.1858 | 5340 | 0.3911 |
|
| 320 |
+
| 0.426 | 2.1940 | 5360 | 0.3917 |
|
| 321 |
+
| 0.4305 | 2.2022 | 5380 | 0.3925 |
|
| 322 |
+
| 0.4373 | 2.2104 | 5400 | 0.3919 |
|
| 323 |
+
| 0.4319 | 2.2186 | 5420 | 0.3914 |
|
| 324 |
+
| 0.43 | 2.2268 | 5440 | 0.3921 |
|
| 325 |
+
| 0.4307 | 2.2350 | 5460 | 0.3910 |
|
| 326 |
+
| 0.4352 | 2.2431 | 5480 | 0.3912 |
|
| 327 |
+
| 0.4323 | 2.2513 | 5500 | 0.3907 |
|
| 328 |
+
| 0.4255 | 2.2595 | 5520 | 0.3905 |
|
| 329 |
+
| 0.4286 | 2.2677 | 5540 | 0.3913 |
|
| 330 |
+
| 0.4271 | 2.2759 | 5560 | 0.3916 |
|
| 331 |
+
| 0.4319 | 2.2841 | 5580 | 0.3915 |
|
| 332 |
+
| 0.4175 | 2.2923 | 5600 | 0.3911 |
|
| 333 |
+
| 0.424 | 2.3005 | 5620 | 0.3914 |
|
| 334 |
+
| 0.4365 | 2.3086 | 5640 | 0.3907 |
|
| 335 |
+
| 0.4322 | 2.3168 | 5660 | 0.3906 |
|
| 336 |
+
| 0.4227 | 2.3250 | 5680 | 0.3910 |
|
| 337 |
+
| 0.4308 | 2.3332 | 5700 | 0.3909 |
|
| 338 |
+
| 0.4268 | 2.3414 | 5720 | 0.3910 |
|
| 339 |
+
| 0.4352 | 2.3496 | 5740 | 0.3911 |
|
| 340 |
+
| 0.4274 | 2.3578 | 5760 | 0.3898 |
|
| 341 |
+
| 0.4255 | 2.3659 | 5780 | 0.3901 |
|
| 342 |
+
| 0.4277 | 2.3741 | 5800 | 0.3903 |
|
| 343 |
+
| 0.4209 | 2.3823 | 5820 | 0.3905 |
|
| 344 |
+
| 0.4221 | 2.3905 | 5840 | 0.3911 |
|
| 345 |
+
| 0.4247 | 2.3987 | 5860 | 0.3911 |
|
| 346 |
+
| 0.4263 | 2.4069 | 5880 | 0.3910 |
|
| 347 |
+
| 0.4284 | 2.4151 | 5900 | 0.3912 |
|
| 348 |
+
| 0.4251 | 2.4233 | 5920 | 0.3910 |
|
| 349 |
+
| 0.4275 | 2.4314 | 5940 | 0.3908 |
|
| 350 |
+
| 0.4271 | 2.4396 | 5960 | 0.3904 |
|
| 351 |
+
| 0.4333 | 2.4478 | 5980 | 0.3904 |
|
| 352 |
+
| 0.4237 | 2.4560 | 6000 | 0.3903 |
|
| 353 |
+
| 0.4351 | 2.4642 | 6020 | 0.3903 |
|
| 354 |
+
| 0.4313 | 2.4724 | 6040 | 0.3902 |
|
| 355 |
+
| 0.4243 | 2.4806 | 6060 | 0.3910 |
|
| 356 |
+
| 0.4289 | 2.4887 | 6080 | 0.3907 |
|
| 357 |
+
| 0.4299 | 2.4969 | 6100 | 0.3909 |
|
| 358 |
+
| 0.428 | 2.5051 | 6120 | 0.3903 |
|
| 359 |
+
| 0.4202 | 2.5133 | 6140 | 0.3902 |
|
| 360 |
+
| 0.4291 | 2.5215 | 6160 | 0.3899 |
|
| 361 |
+
| 0.4344 | 2.5297 | 6180 | 0.3899 |
|
| 362 |
+
| 0.4256 | 2.5379 | 6200 | 0.3902 |
|
| 363 |
+
| 0.4227 | 2.5460 | 6220 | 0.3904 |
|
| 364 |
+
| 0.43 | 2.5542 | 6240 | 0.3907 |
|
| 365 |
+
| 0.4252 | 2.5624 | 6260 | 0.3900 |
|
| 366 |
+
| 0.4224 | 2.5706 | 6280 | 0.3909 |
|
| 367 |
+
| 0.4207 | 2.5788 | 6300 | 0.3909 |
|
| 368 |
+
| 0.4265 | 2.5870 | 6320 | 0.3906 |
|
| 369 |
+
| 0.4341 | 2.5952 | 6340 | 0.3907 |
|
| 370 |
+
| 0.4228 | 2.6034 | 6360 | 0.3903 |
|
| 371 |
+
| 0.4196 | 2.6115 | 6380 | 0.3904 |
|
| 372 |
+
| 0.4216 | 2.6197 | 6400 | 0.3897 |
|
| 373 |
+
| 0.4339 | 2.6279 | 6420 | 0.3904 |
|
| 374 |
+
| 0.4255 | 2.6361 | 6440 | 0.3903 |
|
| 375 |
+
| 0.4261 | 2.6443 | 6460 | 0.3905 |
|
| 376 |
+
| 0.43 | 2.6525 | 6480 | 0.3906 |
|
| 377 |
+
| 0.4265 | 2.6607 | 6500 | 0.3907 |
|
| 378 |
+
| 0.4279 | 2.6688 | 6520 | 0.3904 |
|
| 379 |
+
| 0.4298 | 2.6770 | 6540 | 0.3901 |
|
| 380 |
+
| 0.4312 | 2.6852 | 6560 | 0.3901 |
|
| 381 |
+
| 0.4199 | 2.6934 | 6580 | 0.3898 |
|
| 382 |
+
| 0.4288 | 2.7016 | 6600 | 0.3902 |
|
| 383 |
+
| 0.4325 | 2.7098 | 6620 | 0.3905 |
|
| 384 |
+
| 0.4246 | 2.7180 | 6640 | 0.3903 |
|
| 385 |
+
| 0.4281 | 2.7262 | 6660 | 0.3899 |
|
| 386 |
+
| 0.4296 | 2.7343 | 6680 | 0.3903 |
|
| 387 |
+
| 0.4247 | 2.7425 | 6700 | 0.3898 |
|
| 388 |
+
| 0.4252 | 2.7507 | 6720 | 0.3905 |
|
| 389 |
+
| 0.4255 | 2.7589 | 6740 | 0.3904 |
|
| 390 |
+
| 0.4282 | 2.7671 | 6760 | 0.3902 |
|
| 391 |
+
| 0.4225 | 2.7753 | 6780 | 0.3900 |
|
| 392 |
+
| 0.4251 | 2.7835 | 6800 | 0.3900 |
|
| 393 |
+
| 0.4201 | 2.7916 | 6820 | 0.3903 |
|
| 394 |
+
| 0.4252 | 2.7998 | 6840 | 0.3905 |
|
| 395 |
+
| 0.427 | 2.8080 | 6860 | 0.3907 |
|
| 396 |
+
| 0.428 | 2.8162 | 6880 | 0.3907 |
|
| 397 |
+
| 0.437 | 2.8244 | 6900 | 0.3900 |
|
| 398 |
+
| 0.4257 | 2.8326 | 6920 | 0.3901 |
|
| 399 |
+
| 0.4239 | 2.8408 | 6940 | 0.3905 |
|
| 400 |
+
| 0.4276 | 2.8490 | 6960 | 0.3902 |
|
| 401 |
+
| 0.4274 | 2.8571 | 6980 | 0.3897 |
|
| 402 |
+
| 0.4327 | 2.8653 | 7000 | 0.3902 |
|
| 403 |
+
| 0.4313 | 2.8735 | 7020 | 0.3896 |
|
| 404 |
+
| 0.4277 | 2.8817 | 7040 | 0.3904 |
|
| 405 |
+
| 0.4289 | 2.8899 | 7060 | 0.3904 |
|
| 406 |
+
| 0.4321 | 2.8981 | 7080 | 0.3900 |
|
| 407 |
+
| 0.4232 | 2.9063 | 7100 | 0.3902 |
|
| 408 |
+
| 0.4274 | 2.9144 | 7120 | 0.3901 |
|
| 409 |
+
| 0.4339 | 2.9226 | 7140 | 0.3901 |
|
| 410 |
+
| 0.4226 | 2.9308 | 7160 | 0.3904 |
|
| 411 |
+
| 0.4184 | 2.9390 | 7180 | 0.3902 |
|
| 412 |
+
| 0.4242 | 2.9472 | 7200 | 0.3901 |
|
| 413 |
+
| 0.4259 | 2.9554 | 7220 | 0.3902 |
|
| 414 |
+
| 0.4297 | 2.9636 | 7240 | 0.3897 |
|
| 415 |
+
| 0.4268 | 2.9718 | 7260 | 0.3900 |
|
| 416 |
+
| 0.4281 | 2.9799 | 7280 | 0.3900 |
|
| 417 |
+
| 0.4234 | 2.9881 | 7300 | 0.3901 |
|
| 418 |
+
| 0.4196 | 2.9963 | 7320 | 0.3900 |
|
| 419 |
+
|
| 420 |
+
|
| 421 |
+
### Framework versions
|
| 422 |
+
|
| 423 |
+
- PEFT 0.15.1
|
| 424 |
+
- Transformers 4.51.3
|
| 425 |
+
- Pytorch 2.6.0+cu118
|
| 426 |
+
- Datasets 3.5.0
|
| 427 |
+
- Tokenizers 0.21.1
|
adapter_config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "gpt2",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": true,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 32,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.05,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"r": 8,
|
| 24 |
+
"rank_pattern": {},
|
| 25 |
+
"revision": null,
|
| 26 |
+
"target_modules": [
|
| 27 |
+
"c_attn"
|
| 28 |
+
],
|
| 29 |
+
"task_type": "CAUSAL_LM",
|
| 30 |
+
"trainable_token_indices": null,
|
| 31 |
+
"use_dora": false,
|
| 32 |
+
"use_rslora": false
|
| 33 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8aaf332fb2b5a94ea406fe126d993b06ee1d77486027655733f14fa1dab70ad6
|
| 3 |
+
size 309980480
|
added_tokens.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<endofex>": 50259,
|
| 3 |
+
"<pad>": 50257,
|
| 4 |
+
"<startofex>": 50258
|
| 5 |
+
}
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 0.38987866044044495,
|
| 4 |
+
"eval_runtime": 27.8251,
|
| 5 |
+
"eval_samples_per_second": 602.01,
|
| 6 |
+
"eval_steps_per_second": 37.628,
|
| 7 |
+
"perplexity": 1.4768015885562091,
|
| 8 |
+
"total_flos": 1.4983411323027456e+16,
|
| 9 |
+
"train_loss": 0.5561708558514981,
|
| 10 |
+
"train_runtime": 11243.154,
|
| 11 |
+
"train_samples_per_second": 20.858,
|
| 12 |
+
"train_steps_per_second": 0.652
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 0.38987866044044495,
|
| 4 |
+
"eval_runtime": 27.8251,
|
| 5 |
+
"eval_samples_per_second": 602.01,
|
| 6 |
+
"eval_steps_per_second": 37.628,
|
| 7 |
+
"perplexity": 1.4768015885562091
|
| 8 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<startofex>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "<endofex>",
|
| 12 |
+
"lstrip": false,
|
| 13 |
+
"normalized": false,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"bos_token": "<|endoftext|>",
|
| 19 |
+
"eos_token": "<|endoftext|>",
|
| 20 |
+
"pad_token": {
|
| 21 |
+
"content": "<pad>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false
|
| 26 |
+
},
|
| 27 |
+
"unk_token": "<|endoftext|>"
|
| 28 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"50257": {
|
| 13 |
+
"content": "<pad>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"50258": {
|
| 21 |
+
"content": "<startofex>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"50259": {
|
| 29 |
+
"content": "<endofex>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
}
|
| 36 |
+
},
|
| 37 |
+
"additional_special_tokens": [
|
| 38 |
+
"<startofex>",
|
| 39 |
+
"<endofex>"
|
| 40 |
+
],
|
| 41 |
+
"bos_token": "<|endoftext|>",
|
| 42 |
+
"clean_up_tokenization_spaces": false,
|
| 43 |
+
"eos_token": "<|endoftext|>",
|
| 44 |
+
"extra_special_tokens": {},
|
| 45 |
+
"model_max_length": 1024,
|
| 46 |
+
"pad_token": "<pad>",
|
| 47 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 48 |
+
"unk_token": "<|endoftext|>"
|
| 49 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.4983411323027456e+16,
|
| 4 |
+
"train_loss": 0.5561708558514981,
|
| 5 |
+
"train_runtime": 11243.154,
|
| 6 |
+
"train_samples_per_second": 20.858,
|
| 7 |
+
"train_steps_per_second": 0.652
|
| 8 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53564b62c1a8e9108e0cbed2d27783c81779e3b6b17ff3068341bc2e47576db5
|
| 3 |
+
size 5432
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|