update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,133 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-sa-4.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
metrics:
|
| 6 |
+
- accuracy
|
| 7 |
+
model-index:
|
| 8 |
+
- name: 2024-01-03_one_stage_subgraphs_weighted_txt_vis_conc_1_4_8_12_ramp
|
| 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 |
+
# 2024-01-03_one_stage_subgraphs_weighted_txt_vis_conc_1_4_8_12_ramp
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [microsoft/layoutlmv3-base](https://huggingface.co/microsoft/layoutlmv3-base) on an unknown dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.8574
|
| 20 |
+
- Accuracy: 0.77
|
| 21 |
+
- Exit 0 Accuracy: 0.2
|
| 22 |
+
- Exit 1 Accuracy: 0.3
|
| 23 |
+
- Exit 2 Accuracy: 0.1125
|
| 24 |
+
- Exit 3 Accuracy: 0.2725
|
| 25 |
+
- Exit 4 Accuracy: 0.2675
|
| 26 |
+
- Exit 5 Accuracy: 0.51
|
| 27 |
+
- Exit 6 Accuracy: 0.55
|
| 28 |
+
- Exit 7 Accuracy: 0.63
|
| 29 |
+
- Exit 8 Accuracy: 0.525
|
| 30 |
+
- Exit 9 Accuracy: 0.3425
|
| 31 |
+
- Exit 10 Accuracy: 0.445
|
| 32 |
+
- Exit 11 Accuracy: 0.6875
|
| 33 |
+
- Exit 12 Accuracy: 0.7575
|
| 34 |
+
|
| 35 |
+
## Model description
|
| 36 |
+
|
| 37 |
+
More information needed
|
| 38 |
+
|
| 39 |
+
## Intended uses & limitations
|
| 40 |
+
|
| 41 |
+
More information needed
|
| 42 |
+
|
| 43 |
+
## Training and evaluation data
|
| 44 |
+
|
| 45 |
+
More information needed
|
| 46 |
+
|
| 47 |
+
## Training procedure
|
| 48 |
+
|
| 49 |
+
### Training hyperparameters
|
| 50 |
+
|
| 51 |
+
The following hyperparameters were used during training:
|
| 52 |
+
- learning_rate: 2e-05
|
| 53 |
+
- train_batch_size: 8
|
| 54 |
+
- eval_batch_size: 4
|
| 55 |
+
- seed: 42
|
| 56 |
+
- gradient_accumulation_steps: 24
|
| 57 |
+
- total_train_batch_size: 192
|
| 58 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 59 |
+
- lr_scheduler_type: linear
|
| 60 |
+
- num_epochs: 60
|
| 61 |
+
|
| 62 |
+
### Training results
|
| 63 |
+
|
| 64 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Exit 0 Accuracy | Exit 1 Accuracy | Exit 2 Accuracy | Exit 3 Accuracy | Exit 4 Accuracy | Exit 5 Accuracy | Exit 6 Accuracy | Exit 7 Accuracy | Exit 8 Accuracy | Exit 9 Accuracy | Exit 10 Accuracy | Exit 11 Accuracy | Exit 12 Accuracy |
|
| 65 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------------:|:---------------:|:---------------:|:---------------:|:---------------:|:---------------:|:---------------:|:---------------:|:---------------:|:---------------:|:----------------:|:----------------:|:----------------:|
|
| 66 |
+
| No log | 0.96 | 4 | 2.7477 | 0.1025 | 0.07 | 0.0625 | 0.0625 | 0.0625 | 0.06 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.06 |
|
| 67 |
+
| No log | 1.96 | 8 | 2.7020 | 0.1275 | 0.0775 | 0.0625 | 0.0625 | 0.0625 | 0.0575 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.06 |
|
| 68 |
+
| No log | 2.96 | 12 | 2.6514 | 0.19 | 0.08 | 0.0625 | 0.0625 | 0.0625 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.065 |
|
| 69 |
+
| No log | 3.96 | 16 | 2.5720 | 0.215 | 0.095 | 0.065 | 0.0625 | 0.0625 | 0.06 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0925 |
|
| 70 |
+
| No log | 4.96 | 20 | 2.5033 | 0.2375 | 0.1 | 0.0675 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.1475 |
|
| 71 |
+
| No log | 5.96 | 24 | 2.4003 | 0.275 | 0.115 | 0.08 | 0.0625 | 0.0625 | 0.0625 | 0.0775 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.1525 |
|
| 72 |
+
| No log | 6.96 | 28 | 2.3192 | 0.3 | 0.12 | 0.0875 | 0.0625 | 0.0625 | 0.0625 | 0.09 | 0.075 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.1625 |
|
| 73 |
+
| No log | 7.96 | 32 | 2.2199 | 0.3325 | 0.1325 | 0.0875 | 0.0625 | 0.0625 | 0.0625 | 0.1025 | 0.075 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.185 |
|
| 74 |
+
| No log | 8.96 | 36 | 2.1335 | 0.36 | 0.1425 | 0.1025 | 0.0625 | 0.0625 | 0.0625 | 0.1375 | 0.09 | 0.07 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.235 |
|
| 75 |
+
| No log | 9.96 | 40 | 2.0290 | 0.4 | 0.1475 | 0.1025 | 0.0625 | 0.0625 | 0.06 | 0.15 | 0.0925 | 0.0725 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.27 |
|
| 76 |
+
| No log | 10.96 | 44 | 1.9285 | 0.4525 | 0.155 | 0.1025 | 0.0625 | 0.065 | 0.0675 | 0.1725 | 0.1225 | 0.08 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.305 |
|
| 77 |
+
| No log | 11.96 | 48 | 1.8071 | 0.495 | 0.155 | 0.1075 | 0.0625 | 0.0775 | 0.0825 | 0.2025 | 0.135 | 0.0925 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.355 |
|
| 78 |
+
| No log | 12.96 | 52 | 1.7194 | 0.5375 | 0.16 | 0.1075 | 0.0625 | 0.0775 | 0.0875 | 0.22 | 0.1625 | 0.1025 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.37 |
|
| 79 |
+
| No log | 13.96 | 56 | 1.5732 | 0.6 | 0.1625 | 0.11 | 0.0625 | 0.08 | 0.0775 | 0.2425 | 0.2125 | 0.14 | 0.0625 | 0.0625 | 0.0625 | 0.0625 | 0.43 |
|
| 80 |
+
| No log | 14.96 | 60 | 1.5141 | 0.6125 | 0.16 | 0.115 | 0.0625 | 0.0825 | 0.09 | 0.29 | 0.245 | 0.18 | 0.0625 | 0.065 | 0.0625 | 0.0625 | 0.4325 |
|
| 81 |
+
| No log | 15.96 | 64 | 1.4049 | 0.65 | 0.1575 | 0.1175 | 0.0625 | 0.0925 | 0.1 | 0.3275 | 0.2725 | 0.24 | 0.0675 | 0.065 | 0.0625 | 0.0625 | 0.51 |
|
| 82 |
+
| No log | 16.96 | 68 | 1.3476 | 0.665 | 0.16 | 0.115 | 0.0625 | 0.095 | 0.1025 | 0.345 | 0.285 | 0.2625 | 0.0675 | 0.075 | 0.0625 | 0.0625 | 0.5275 |
|
| 83 |
+
| No log | 17.96 | 72 | 1.2825 | 0.6925 | 0.16 | 0.1175 | 0.0625 | 0.1025 | 0.11 | 0.35 | 0.2875 | 0.2925 | 0.0675 | 0.075 | 0.065 | 0.0625 | 0.55 |
|
| 84 |
+
| No log | 18.96 | 76 | 1.2102 | 0.71 | 0.16 | 0.1175 | 0.0625 | 0.1025 | 0.135 | 0.365 | 0.29 | 0.31 | 0.0675 | 0.0775 | 0.08 | 0.0625 | 0.5925 |
|
| 85 |
+
| No log | 19.96 | 80 | 1.1664 | 0.725 | 0.16 | 0.1175 | 0.0625 | 0.1125 | 0.1325 | 0.3675 | 0.3075 | 0.365 | 0.0775 | 0.0775 | 0.08 | 0.0625 | 0.6025 |
|
| 86 |
+
| No log | 20.96 | 84 | 1.1363 | 0.735 | 0.16 | 0.12 | 0.0625 | 0.115 | 0.145 | 0.3725 | 0.3275 | 0.3775 | 0.0775 | 0.075 | 0.07 | 0.0625 | 0.6175 |
|
| 87 |
+
| No log | 21.96 | 88 | 1.0745 | 0.74 | 0.16 | 0.1225 | 0.0625 | 0.1175 | 0.1325 | 0.375 | 0.355 | 0.4175 | 0.0775 | 0.0825 | 0.065 | 0.0625 | 0.6275 |
|
| 88 |
+
| No log | 22.96 | 92 | 1.0377 | 0.7525 | 0.16 | 0.13 | 0.0625 | 0.115 | 0.1575 | 0.38 | 0.3775 | 0.4125 | 0.0825 | 0.075 | 0.07 | 0.0625 | 0.64 |
|
| 89 |
+
| No log | 23.96 | 96 | 1.0321 | 0.74 | 0.16 | 0.1375 | 0.0625 | 0.115 | 0.165 | 0.3825 | 0.385 | 0.4375 | 0.08 | 0.0975 | 0.07 | 0.0625 | 0.66 |
|
| 90 |
+
| No log | 24.96 | 100 | 0.9702 | 0.76 | 0.1625 | 0.15 | 0.0625 | 0.11 | 0.1725 | 0.385 | 0.4075 | 0.455 | 0.0825 | 0.0975 | 0.0725 | 0.0625 | 0.68 |
|
| 91 |
+
| No log | 25.96 | 104 | 0.9861 | 0.7525 | 0.1675 | 0.1525 | 0.0625 | 0.1125 | 0.1825 | 0.395 | 0.4 | 0.4675 | 0.0825 | 0.115 | 0.0675 | 0.07 | 0.6825 |
|
| 92 |
+
| No log | 26.96 | 108 | 0.9339 | 0.7525 | 0.16 | 0.15 | 0.0625 | 0.1175 | 0.195 | 0.4075 | 0.4275 | 0.4825 | 0.0875 | 0.13 | 0.07 | 0.095 | 0.7025 |
|
| 93 |
+
| No log | 27.96 | 112 | 0.9362 | 0.7575 | 0.1625 | 0.1425 | 0.0625 | 0.1175 | 0.1925 | 0.4175 | 0.43 | 0.515 | 0.095 | 0.14 | 0.0675 | 0.105 | 0.71 |
|
| 94 |
+
| No log | 28.96 | 116 | 0.8872 | 0.755 | 0.165 | 0.15 | 0.0625 | 0.1175 | 0.2 | 0.4275 | 0.4325 | 0.53 | 0.095 | 0.1425 | 0.07 | 0.125 | 0.7425 |
|
| 95 |
+
| No log | 29.96 | 120 | 0.8939 | 0.7675 | 0.1625 | 0.15 | 0.0625 | 0.1175 | 0.2 | 0.4475 | 0.4325 | 0.55 | 0.1 | 0.1425 | 0.085 | 0.1325 | 0.7175 |
|
| 96 |
+
| No log | 30.96 | 124 | 0.8767 | 0.7475 | 0.16 | 0.1575 | 0.0625 | 0.12 | 0.195 | 0.4425 | 0.4475 | 0.545 | 0.1 | 0.1525 | 0.0925 | 0.2025 | 0.7575 |
|
| 97 |
+
| No log | 31.96 | 128 | 0.8658 | 0.76 | 0.165 | 0.17 | 0.0625 | 0.1225 | 0.195 | 0.455 | 0.455 | 0.555 | 0.1025 | 0.1375 | 0.11 | 0.245 | 0.7375 |
|
| 98 |
+
| No log | 32.96 | 132 | 0.8736 | 0.7625 | 0.165 | 0.1875 | 0.0625 | 0.125 | 0.195 | 0.465 | 0.45 | 0.5625 | 0.105 | 0.155 | 0.0975 | 0.275 | 0.7575 |
|
| 99 |
+
| No log | 33.96 | 136 | 0.8380 | 0.7625 | 0.1675 | 0.21 | 0.0625 | 0.125 | 0.195 | 0.465 | 0.4625 | 0.565 | 0.1175 | 0.13 | 0.115 | 0.3225 | 0.755 |
|
| 100 |
+
| No log | 34.96 | 140 | 0.8386 | 0.7725 | 0.1675 | 0.2325 | 0.0625 | 0.1275 | 0.1975 | 0.4675 | 0.4575 | 0.575 | 0.12 | 0.125 | 0.13 | 0.345 | 0.765 |
|
| 101 |
+
| No log | 35.96 | 144 | 0.8610 | 0.755 | 0.17 | 0.2425 | 0.0625 | 0.1275 | 0.19 | 0.4625 | 0.4725 | 0.5875 | 0.13 | 0.105 | 0.18 | 0.39 | 0.755 |
|
| 102 |
+
| No log | 36.96 | 148 | 0.8444 | 0.76 | 0.17 | 0.255 | 0.0625 | 0.1325 | 0.2 | 0.4575 | 0.4675 | 0.595 | 0.12 | 0.1475 | 0.22 | 0.445 | 0.7525 |
|
| 103 |
+
| No log | 37.96 | 152 | 0.8845 | 0.75 | 0.1725 | 0.2725 | 0.0625 | 0.1375 | 0.205 | 0.46 | 0.475 | 0.59 | 0.1425 | 0.175 | 0.255 | 0.45 | 0.7575 |
|
| 104 |
+
| No log | 38.96 | 156 | 0.8464 | 0.7625 | 0.18 | 0.275 | 0.0625 | 0.145 | 0.2075 | 0.4675 | 0.4825 | 0.5925 | 0.18 | 0.22 | 0.265 | 0.51 | 0.755 |
|
| 105 |
+
| No log | 39.96 | 160 | 0.8539 | 0.7575 | 0.1825 | 0.2825 | 0.065 | 0.1475 | 0.215 | 0.48 | 0.515 | 0.6025 | 0.2 | 0.2425 | 0.2725 | 0.5275 | 0.755 |
|
| 106 |
+
| No log | 40.96 | 164 | 0.8697 | 0.76 | 0.185 | 0.2775 | 0.0675 | 0.1525 | 0.23 | 0.485 | 0.5325 | 0.605 | 0.2175 | 0.21 | 0.29 | 0.53 | 0.7625 |
|
| 107 |
+
| No log | 41.96 | 168 | 0.8395 | 0.775 | 0.185 | 0.2825 | 0.075 | 0.16 | 0.2225 | 0.4925 | 0.54 | 0.6 | 0.225 | 0.225 | 0.2875 | 0.5725 | 0.77 |
|
| 108 |
+
| No log | 42.96 | 172 | 0.8570 | 0.7675 | 0.1875 | 0.285 | 0.08 | 0.1575 | 0.2275 | 0.485 | 0.5475 | 0.61 | 0.2325 | 0.2325 | 0.3075 | 0.6075 | 0.7525 |
|
| 109 |
+
| No log | 43.96 | 176 | 0.8462 | 0.765 | 0.195 | 0.28 | 0.08 | 0.165 | 0.2325 | 0.49 | 0.5425 | 0.6125 | 0.2475 | 0.2425 | 0.3125 | 0.6225 | 0.755 |
|
| 110 |
+
| No log | 44.96 | 180 | 0.8563 | 0.765 | 0.195 | 0.2825 | 0.085 | 0.1775 | 0.235 | 0.495 | 0.535 | 0.6075 | 0.2975 | 0.22 | 0.3175 | 0.62 | 0.75 |
|
| 111 |
+
| No log | 45.96 | 184 | 0.8670 | 0.7675 | 0.195 | 0.28 | 0.085 | 0.1825 | 0.24 | 0.4975 | 0.54 | 0.615 | 0.3525 | 0.215 | 0.325 | 0.6375 | 0.76 |
|
| 112 |
+
| No log | 46.96 | 188 | 0.8708 | 0.77 | 0.195 | 0.29 | 0.0925 | 0.185 | 0.2375 | 0.4975 | 0.535 | 0.6125 | 0.365 | 0.2275 | 0.3175 | 0.64 | 0.7575 |
|
| 113 |
+
| No log | 47.96 | 192 | 0.8535 | 0.7675 | 0.19 | 0.29 | 0.095 | 0.2075 | 0.24 | 0.4975 | 0.5375 | 0.6125 | 0.4025 | 0.24 | 0.35 | 0.6575 | 0.755 |
|
| 114 |
+
| No log | 48.96 | 196 | 0.8592 | 0.765 | 0.19 | 0.285 | 0.0975 | 0.2175 | 0.2425 | 0.495 | 0.54 | 0.615 | 0.4175 | 0.2375 | 0.365 | 0.6575 | 0.7475 |
|
| 115 |
+
| No log | 49.96 | 200 | 0.8717 | 0.765 | 0.19 | 0.2925 | 0.1 | 0.235 | 0.25 | 0.5 | 0.545 | 0.6125 | 0.4325 | 0.25 | 0.3725 | 0.66 | 0.76 |
|
| 116 |
+
| No log | 50.96 | 204 | 0.8684 | 0.765 | 0.1925 | 0.2975 | 0.105 | 0.245 | 0.2575 | 0.5025 | 0.545 | 0.61 | 0.4475 | 0.2775 | 0.3775 | 0.675 | 0.7625 |
|
| 117 |
+
| No log | 51.96 | 208 | 0.8662 | 0.76 | 0.1925 | 0.295 | 0.1025 | 0.245 | 0.2625 | 0.5025 | 0.55 | 0.6175 | 0.455 | 0.2925 | 0.39 | 0.68 | 0.76 |
|
| 118 |
+
| No log | 52.96 | 212 | 0.8718 | 0.7625 | 0.1925 | 0.295 | 0.1075 | 0.2525 | 0.2625 | 0.5025 | 0.55 | 0.6225 | 0.485 | 0.3075 | 0.4125 | 0.6825 | 0.755 |
|
| 119 |
+
| No log | 53.96 | 216 | 0.8798 | 0.76 | 0.195 | 0.295 | 0.11 | 0.265 | 0.2625 | 0.505 | 0.5475 | 0.6275 | 0.495 | 0.3175 | 0.4275 | 0.68 | 0.7475 |
|
| 120 |
+
| No log | 54.96 | 220 | 0.8703 | 0.7575 | 0.2 | 0.2975 | 0.11 | 0.2675 | 0.2675 | 0.5075 | 0.545 | 0.6225 | 0.4975 | 0.3275 | 0.435 | 0.6825 | 0.745 |
|
| 121 |
+
| No log | 55.96 | 224 | 0.8622 | 0.765 | 0.2 | 0.3 | 0.11 | 0.265 | 0.27 | 0.51 | 0.545 | 0.625 | 0.505 | 0.33 | 0.435 | 0.69 | 0.7525 |
|
| 122 |
+
| No log | 56.96 | 228 | 0.8590 | 0.77 | 0.2 | 0.3 | 0.11 | 0.27 | 0.2675 | 0.5125 | 0.5475 | 0.6325 | 0.5075 | 0.34 | 0.4375 | 0.6875 | 0.76 |
|
| 123 |
+
| No log | 57.96 | 232 | 0.8572 | 0.7725 | 0.2 | 0.3025 | 0.11 | 0.27 | 0.2675 | 0.51 | 0.5475 | 0.6325 | 0.5175 | 0.34 | 0.44 | 0.6875 | 0.7575 |
|
| 124 |
+
| No log | 58.96 | 236 | 0.8570 | 0.7725 | 0.2 | 0.3025 | 0.1125 | 0.2725 | 0.2675 | 0.51 | 0.55 | 0.6325 | 0.5225 | 0.34 | 0.445 | 0.6875 | 0.76 |
|
| 125 |
+
| No log | 59.96 | 240 | 0.8574 | 0.77 | 0.2 | 0.3 | 0.1125 | 0.2725 | 0.2675 | 0.51 | 0.55 | 0.63 | 0.525 | 0.3425 | 0.445 | 0.6875 | 0.7575 |
|
| 126 |
+
|
| 127 |
+
|
| 128 |
+
### Framework versions
|
| 129 |
+
|
| 130 |
+
- Transformers 4.26.1
|
| 131 |
+
- Pytorch 1.13.1.post200
|
| 132 |
+
- Datasets 2.9.0
|
| 133 |
+
- Tokenizers 0.13.2
|