<!-- 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. -->
margin-element-detector-fm-still-firefly-29
This model is a fine-tuned version of toobiza/margin-element-detector-fm-likely-serenity-27 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1889
- Loss Ce: 0.0075
- Loss Bbox: 0.0068
- Cardinality Error: 0.0446
- Giou: 92.6060
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: 1e-05
- train_batch_size: 4
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 60
Training results
Training Loss | Epoch | Step | Validation Loss | Loss Ce | Loss Bbox | Cardinality Error | Giou |
---|---|---|---|---|---|---|---|
0.3498 | 0.89 | 2000 | 0.3246 | 0.0183 | 0.0116 | 0.1049 | 87.5875 |
0.3372 | 1.77 | 4000 | 0.3313 | 0.0186 | 0.0120 | 0.125 | 87.3635 |
0.3329 | 2.66 | 6000 | 0.3013 | 0.0162 | 0.0109 | 0.1116 | 88.4716 |
0.322 | 3.54 | 8000 | 0.3135 | 0.0180 | 0.0111 | 0.0971 | 87.9892 |
0.3265 | 4.43 | 10000 | 0.3032 | 0.0161 | 0.0108 | 0.0837 | 88.3363 |
0.3301 | 5.31 | 12000 | 0.3031 | 0.0172 | 0.0112 | 0.1150 | 88.4994 |
0.3175 | 6.2 | 14000 | 0.2764 | 0.0154 | 0.0099 | 0.0904 | 89.4291 |
0.3171 | 7.08 | 16000 | 0.3009 | 0.0144 | 0.0106 | 0.0871 | 88.3145 |
0.3121 | 7.97 | 18000 | 0.2920 | 0.0174 | 0.0103 | 0.1295 | 88.8467 |
0.3125 | 8.85 | 20000 | 0.2759 | 0.0172 | 0.0098 | 0.1071 | 89.4993 |
0.321 | 9.74 | 22000 | 0.2867 | 0.0165 | 0.0099 | 0.0893 | 88.9708 |
0.3121 | 10.62 | 24000 | 0.2805 | 0.0172 | 0.0098 | 0.1105 | 89.2836 |
0.292 | 11.51 | 26000 | 0.2646 | 0.0139 | 0.0093 | 0.0915 | 89.7797 |
0.3039 | 12.39 | 28000 | 0.2804 | 0.0160 | 0.0099 | 0.1004 | 89.2424 |
0.3024 | 13.28 | 30000 | 0.2739 | 0.0136 | 0.0096 | 0.0915 | 89.3642 |
0.2902 | 14.17 | 32000 | 0.2689 | 0.0150 | 0.0092 | 0.0770 | 89.6054 |
0.2958 | 15.05 | 34000 | 0.2668 | 0.0144 | 0.0095 | 0.0904 | 89.7392 |
0.3014 | 15.94 | 36000 | 0.2640 | 0.0147 | 0.0092 | 0.0781 | 89.8237 |
0.2942 | 16.82 | 38000 | 0.2549 | 0.0133 | 0.0090 | 0.0837 | 90.1652 |
0.2821 | 17.71 | 40000 | 0.2548 | 0.0131 | 0.0090 | 0.0748 | 90.1510 |
0.2909 | 18.59 | 42000 | 0.2510 | 0.0140 | 0.0088 | 0.0859 | 90.3495 |
0.2852 | 19.48 | 44000 | 0.2389 | 0.0132 | 0.0086 | 0.0926 | 90.8433 |
0.2799 | 20.36 | 46000 | 0.2651 | 0.0124 | 0.0093 | 0.0547 | 89.6861 |
0.2637 | 21.25 | 48000 | 0.2393 | 0.0114 | 0.0087 | 0.0580 | 90.7806 |
0.2767 | 22.13 | 50000 | 0.2501 | 0.0125 | 0.0088 | 0.0737 | 90.3082 |
0.2684 | 23.02 | 52000 | 0.2366 | 0.0120 | 0.0083 | 0.0647 | 90.8476 |
0.2749 | 23.9 | 54000 | 0.2424 | 0.0117 | 0.0083 | 0.0580 | 90.5364 |
0.2592 | 24.79 | 56000 | 0.2404 | 0.0112 | 0.0086 | 0.0603 | 90.6891 |
0.273 | 25.68 | 58000 | 0.2330 | 0.0114 | 0.0084 | 0.0536 | 91.0108 |
0.2618 | 26.56 | 60000 | 0.2464 | 0.0125 | 0.0088 | 0.0703 | 90.4816 |
0.2748 | 27.45 | 62000 | 0.2339 | 0.0120 | 0.0081 | 0.0647 | 90.9212 |
0.257 | 28.33 | 64000 | 0.2308 | 0.0123 | 0.0082 | 0.0670 | 91.1289 |
0.2513 | 29.22 | 66000 | 0.2545 | 0.0110 | 0.0088 | 0.0603 | 90.0094 |
0.2704 | 30.1 | 68000 | 0.2263 | 0.0118 | 0.0080 | 0.0625 | 91.2595 |
0.2651 | 30.99 | 70000 | 0.2222 | 0.0110 | 0.0079 | 0.0592 | 91.4144 |
0.2485 | 31.87 | 72000 | 0.2243 | 0.0119 | 0.0079 | 0.0647 | 91.335 |
0.244 | 32.76 | 74000 | 0.2265 | 0.0103 | 0.0079 | 0.0491 | 91.1472 |
0.2383 | 33.64 | 76000 | 0.2327 | 0.0110 | 0.0081 | 0.0614 | 90.9362 |
0.2502 | 34.53 | 78000 | 0.2293 | 0.0109 | 0.0080 | 0.0658 | 91.0688 |
0.2581 | 35.41 | 80000 | 0.2255 | 0.0111 | 0.0078 | 0.0625 | 91.2117 |
0.25 | 36.3 | 82000 | 0.2168 | 0.0102 | 0.0076 | 0.0603 | 91.5687 |
0.2564 | 37.18 | 84000 | 0.2152 | 0.0095 | 0.0077 | 0.0614 | 91.6154 |
0.2587 | 38.07 | 86000 | 0.2090 | 0.0103 | 0.0075 | 0.0569 | 91.9276 |
0.2366 | 38.96 | 88000 | 0.2151 | 0.0086 | 0.0075 | 0.0458 | 91.5430 |
0.2435 | 39.84 | 90000 | 0.2110 | 0.0089 | 0.0075 | 0.0502 | 91.7639 |
0.2504 | 40.73 | 92000 | 0.2104 | 0.0080 | 0.0074 | 0.0592 | 91.7350 |
0.2421 | 41.61 | 94000 | 0.2084 | 0.0084 | 0.0074 | 0.0513 | 91.8353 |
0.2451 | 42.5 | 96000 | 0.2080 | 0.0090 | 0.0072 | 0.0547 | 91.8259 |
0.2318 | 43.38 | 98000 | 0.2133 | 0.0106 | 0.0074 | 0.0703 | 91.6963 |
0.2339 | 44.27 | 100000 | 0.2056 | 0.0096 | 0.0075 | 0.0491 | 92.0649 |
0.2448 | 45.15 | 102000 | 0.2026 | 0.0082 | 0.0073 | 0.0525 | 92.0915 |
0.2408 | 46.04 | 104000 | 0.1980 | 0.0095 | 0.0071 | 0.0480 | 92.3465 |
0.2427 | 46.92 | 106000 | 0.1959 | 0.0094 | 0.0070 | 0.0558 | 92.4126 |
0.2299 | 47.81 | 108000 | 0.2051 | 0.0091 | 0.0072 | 0.0435 | 91.9999 |
0.2434 | 48.69 | 110000 | 0.2056 | 0.0080 | 0.0073 | 0.0424 | 91.9368 |
0.2359 | 49.58 | 112000 | 0.1934 | 0.0079 | 0.0069 | 0.0435 | 92.4476 |
0.2347 | 50.46 | 114000 | 0.1969 | 0.0084 | 0.0070 | 0.0480 | 92.3191 |
0.2311 | 51.35 | 116000 | 0.1958 | 0.0084 | 0.0070 | 0.0502 | 92.3792 |
0.2283 | 52.24 | 118000 | 0.1889 | 0.0075 | 0.0068 | 0.0402 | 92.6285 |
0.2341 | 53.12 | 120000 | 0.1989 | 0.0086 | 0.0069 | 0.0603 | 92.1915 |
0.2338 | 54.01 | 122000 | 0.1952 | 0.0084 | 0.0068 | 0.0603 | 92.3580 |
0.2262 | 54.89 | 124000 | 0.1956 | 0.0084 | 0.0068 | 0.0435 | 92.3380 |
0.215 | 55.78 | 126000 | 0.1971 | 0.0086 | 0.0069 | 0.0536 | 92.2867 |
0.2116 | 56.66 | 128000 | 0.1890 | 0.0077 | 0.0067 | 0.0446 | 92.6109 |
0.2202 | 57.55 | 130000 | 0.1913 | 0.0078 | 0.0068 | 0.0480 | 92.5177 |
0.2391 | 58.43 | 132000 | 0.1904 | 0.0084 | 0.0068 | 0.0513 | 92.5878 |
0.2234 | 59.32 | 134000 | 0.1889 | 0.0075 | 0.0068 | 0.0446 | 92.6060 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.13.3