openllama-3b-finance

This model is a fine-tuned version of openlm-research/open_llama_3b_v2 on the financial_phrasebank dataset. It achieves the following results on the evaluation set:

  • Loss: 4.0296
  • Accuracy: 0.4143

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: 0.0002
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • num_epochs: 1

Training results

Training Loss Epoch Step Validation Loss Accuracy
21.9655 0.01 20 8.1663 0.0816
2.231 0.01 40 6.3007 0.4143
2.7452 0.02 60 4.0892 0.4143
2.4561 0.02 80 5.0314 0.4143
2.337 0.03 100 5.6176 0.4143
3.2226 0.03 120 4.4963 0.4143
2.5633 0.04 140 6.1800 0.4143
2.4764 0.04 160 4.7059 0.4143
2.45 0.05 180 5.0602 0.4143
1.4232 0.05 200 5.3418 0.4143
2.7684 0.06 220 5.1805 0.4143
1.7065 0.06 240 4.7568 0.4143
2.3417 0.07 260 6.1062 0.4143
1.907 0.07 280 12.0988 0.5041
14.6043 0.08 300 3.0283 0.0816
1.337 0.08 320 12.7786 0.4143
4.182 0.09 340 7.5619 0.4143
3.7365 0.09 360 7.8581 0.4143
3.209 0.1 380 3.2547 0.4143
3.4836 0.1 400 89.8525 0.0816
4.5805 0.11 420 103.0762 0.4143
4.6351 0.11 440 91.4501 0.4143
11.0873 0.12 460 88.0469 0.4143
1.1274 0.12 480 86.7130 0.4143
2.0398 0.13 500 86.4186 0.4143
18.6924 0.13 520 80.1491 0.4143
1.2216 0.14 540 76.8429 0.4143
1.1179 0.14 560 78.0159 0.4143
10.0981 0.15 580 71.1114 0.4143
9.0123 0.15 600 66.2945 0.4143
1.9539 0.16 620 65.6854 0.4143
8.4729 0.17 640 62.1595 0.4143
7.816 0.17 660 52.0763 0.4143
6.0443 0.18 680 41.1500 0.4143
3.1804 0.18 700 42.8007 0.4143
1.6122 0.19 720 44.0976 0.4143
9.8927 0.19 740 31.6381 0.4143
6.828 0.2 760 12.7483 0.4143
3.1457 0.2 780 13.2981 0.4143
1.9991 0.21 800 12.4846 0.4143
2.5539 0.21 820 13.7669 0.4143
1.3898 0.22 840 12.8919 0.0816
2.9251 0.22 860 15.9149 0.0816
4.0874 0.23 880 10.5282 0.4143
2.4763 0.23 900 3.0281 0.4143
2.2865 0.24 920 12.2460 0.4143
4.2438 0.24 940 10.1961 0.4143
2.547 0.25 960 1.4099 0.4143
0.8659 0.25 980 8.3217 0.4143
3.5331 0.26 1000 6.3990 0.4143
2.4704 0.26 1020 2.2337 0.0816
2.1381 0.27 1040 10.6263 0.4143
1.5927 0.27 1060 11.1989 0.4143
2.485 0.28 1080 8.8174 0.4143
2.8074 0.28 1100 5.5971 0.4143
0.8622 0.29 1120 5.5089 0.4143
2.8085 0.29 1140 5.4300 0.4143
1.2405 0.3 1160 7.5657 0.4143
3.9374 0.3 1180 2.7180 0.4143
1.7494 0.31 1200 4.9639 0.0816
2.6094 0.32 1220 2.1980 0.4143
2.2072 0.32 1240 7.3392 0.4143
0.9978 0.33 1260 7.9127 0.4143
2.3872 0.33 1280 7.0613 0.4143
3.3129 0.34 1300 4.4202 0.4143
1.776 0.34 1320 6.1467 0.4143
3.1179 0.35 1340 6.0607 0.4143
1.272 0.35 1360 5.0484 0.4143
3.0694 0.36 1380 3.1665 0.4143
1.9452 0.36 1400 4.8692 0.4143
2.3689 0.37 1420 4.9375 0.4143
2.7082 0.37 1440 3.2108 0.4143
0.8244 0.38 1460 7.0151 0.4143
2.6032 0.38 1480 5.5645 0.4143
2.8745 0.39 1500 4.2408 0.4143
2.625 0.39 1520 6.8800 0.4143
2.5335 0.4 1540 6.3109 0.4143
2.5495 0.4 1560 4.4017 0.4143
1.7234 0.41 1580 5.1739 0.4143
2.1066 0.41 1600 6.0769 0.4143
2.5541 0.42 1620 3.7539 0.4143
2.4598 0.42 1640 4.2075 0.4143
1.7211 0.43 1660 5.3975 0.4143
2.3993 0.43 1680 4.1427 0.4143
1.6161 0.44 1700 5.0871 0.4143
2.2361 0.44 1720 4.3375 0.4143
2.0841 0.45 1740 4.7357 0.4143
2.137 0.45 1760 5.2737 0.4143
2.3819 0.46 1780 3.1688 0.4143
2.6391 0.46 1800 5.6169 0.4143
1.276 0.47 1820 6.1945 0.4143
2.0694 0.48 1840 6.3761 0.4143
2.3715 0.48 1860 6.1666 0.4143
2.1428 0.49 1880 6.4718 0.4143
2.0409 0.49 1900 6.3259 0.4143
2.1924 0.5 1920 6.0853 0.4143
2.3511 0.5 1940 4.7199 0.4143
2.7335 0.51 1960 4.3591 0.4143
1.6784 0.51 1980 3.7488 0.1612
1.5525 0.52 2000 6.0497 0.4143
2.7457 0.52 2020 3.5952 0.4143
2.3929 0.53 2040 4.7684 0.4143
1.9522 0.53 2060 5.6394 0.4143
2.2257 0.54 2080 4.5801 0.4143
1.6753 0.54 2100 5.0521 0.4143
1.6154 0.55 2120 5.4730 0.4143
1.7723 0.55 2140 5.5251 0.4143
2.6963 0.56 2160 3.5098 0.4143
1.7274 0.56 2180 5.4262 0.4143
2.4059 0.57 2200 4.5019 0.4143
1.6505 0.57 2220 5.1107 0.4143
1.2469 0.58 2240 5.3456 0.4143
1.6702 0.58 2260 5.4103 0.4143
1.615 0.59 2280 5.8024 0.4143
1.5622 0.59 2300 5.6035 0.4143
2.3536 0.6 2320 5.3779 0.4143
2.0512 0.6 2340 5.2498 0.4143
2.1405 0.61 2360 5.2279 0.4143
2.1926 0.61 2380 4.3260 0.4143
2.3995 0.62 2400 4.4445 0.4143
1.4944 0.62 2420 4.9616 0.4143
2.6623 0.63 2440 4.9736 0.4143
1.4095 0.64 2460 4.6506 0.4143
2.4803 0.64 2480 4.0971 0.4143
1.2721 0.65 2500 4.3192 0.4143
1.8372 0.65 2520 4.4907 0.4143
1.8942 0.66 2540 4.7323 0.4143
2.1407 0.66 2560 4.9554 0.4143
2.5039 0.67 2580 5.1599 0.4143
1.7321 0.67 2600 5.6089 0.4143
2.0621 0.68 2620 4.8359 0.4143
2.1664 0.68 2640 4.5581 0.4143
1.8835 0.69 2660 5.1029 0.4143
3.0314 0.69 2680 3.9587 0.4143
1.1781 0.7 2700 4.4584 0.4143
3.3222 0.7 2720 4.7628 0.4143
2.1184 0.71 2740 4.4039 0.4143
1.9293 0.71 2760 3.8755 0.4143
2.2448 0.72 2780 4.4327 0.4143
2.4697 0.72 2800 3.3026 0.4143
1.8569 0.73 2820 3.7722 0.4143
0.8544 0.73 2840 4.9176 0.4143
2.2445 0.74 2860 4.3889 0.4143
1.3723 0.74 2880 4.3280 0.4143
2.2167 0.75 2900 4.4016 0.4143
1.98 0.75 2920 3.8661 0.4143
1.7344 0.76 2940 3.7919 0.4143
1.924 0.76 2960 4.1408 0.4143
1.3811 0.77 2980 4.3730 0.4143
1.8289 0.77 3000 4.2872 0.4143
1.9573 0.78 3020 4.6165 0.4143
2.4877 0.78 3040 4.5988 0.4143
1.1749 0.79 3060 4.7887 0.4143
2.1835 0.8 3080 4.9018 0.4143
2.3752 0.8 3100 4.6911 0.4143
1.9741 0.81 3120 4.5126 0.4143
1.7513 0.81 3140 4.6251 0.4143
3.0666 0.82 3160 4.0260 0.4143
0.5569 0.82 3180 4.0965 0.4143
2.1805 0.83 3200 4.5240 0.4143
2.4319 0.83 3220 4.3080 0.4143
2.126 0.84 3240 3.7823 0.4143
1.6993 0.84 3260 3.8093 0.4143
0.6861 0.85 3280 4.1618 0.4143
0.748 0.85 3300 4.5653 0.4143
2.5721 0.86 3320 4.6628 0.4143
2.0137 0.86 3340 4.2796 0.4143
2.1864 0.87 3360 4.1173 0.4143
2.4881 0.87 3380 3.9617 0.4143
2.6837 0.88 3400 3.7575 0.4143
1.5951 0.88 3420 3.6086 0.4143
2.504 0.89 3440 3.5919 0.4143
1.4982 0.89 3460 3.7519 0.4143
1.8994 0.9 3480 3.7120 0.4143
1.6126 0.9 3500 3.6854 0.4143
2.002 0.91 3520 3.7888 0.4143
1.0264 0.91 3540 3.7990 0.4143
1.9495 0.92 3560 3.9635 0.4143
2.0742 0.92 3580 3.9651 0.4143
1.7803 0.93 3600 3.9518 0.4143
2.0843 0.93 3620 3.9404 0.4143
1.8431 0.94 3640 3.9334 0.4143
1.4987 0.95 3660 3.9609 0.4143
1.8214 0.95 3680 4.0060 0.4143
1.0964 0.96 3700 4.0422 0.4143
0.9669 0.96 3720 4.0549 0.4143
1.6226 0.97 3740 4.0486 0.4143
1.8061 0.97 3760 4.0405 0.4143
2.8738 0.98 3780 4.0317 0.4143
1.684 0.98 3800 4.0319 0.4143
1.1158 0.99 3820 4.0303 0.4143
1.775 0.99 3840 4.0294 0.4143
2.1639 1.0 3860 4.0296 0.4143

Framework versions

  • Transformers 4.32.0
  • Pytorch 2.0.1+cu117
  • Datasets 2.14.4
  • Tokenizers 0.13.3
Downloads last month
21
Safetensors
Model size
3.32B params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for IngeniousArtist/openllama-3b-finance

Finetuned
(28)
this model

Dataset used to train IngeniousArtist/openllama-3b-finance

Evaluation results