loubnabnl's picture
loubnabnl HF Staff
Model save
6b60e33 verified
|
raw
history blame
16 kB
metadata
base_model: bigcode/starencoder
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - accuracy
model-index:
  - name: classifier-llama3-php-500k
    results: []

classifier-llama3-php-500k

This model is a fine-tuned version of bigcode/starencoder on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3237
  • Precision: 0.4988
  • Recall: 0.3690
  • F1 Macro: 0.3961
  • Accuracy: 0.6229
  • F1 Binary Minimum3: 0.6382
  • F1 Binary Minimum2: 0.9386

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: 0.0001
  • train_batch_size: 16
  • eval_batch_size: 256
  • seed: 0
  • distributed_type: multi-GPU
  • num_devices: 8
  • total_train_batch_size: 128
  • total_eval_batch_size: 2048
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 200
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Macro Accuracy F1 Binary Minimum3 F1 Binary Minimum2
No log 0 0 6.4099 0.0294 0.2 0.0513 0.1471 0 0
0.3844 0.2955 1000 0.3684 0.4654 0.3097 0.3196 0.5828 0.6038 0.9300
0.3739 0.5910 2000 0.3584 0.4903 0.3176 0.3308 0.5921 0.5791 0.9321
0.3648 0.8865 3000 0.3598 0.4776 0.3252 0.3393 0.5900 0.6469 0.9307
0.3471 1.1820 4000 0.3505 0.4862 0.3290 0.3457 0.5980 0.6182 0.9328
0.3481 1.4775 5000 0.3498 0.4805 0.3314 0.3506 0.5972 0.6235 0.9323
0.3558 1.7730 6000 0.3480 0.4905 0.3258 0.3423 0.5969 0.6219 0.9324
0.3523 2.0686 7000 0.3485 0.4931 0.3315 0.3492 0.5995 0.5795 0.9337
0.3714 2.3641 8000 0.3461 0.4874 0.3316 0.3507 0.5995 0.6196 0.9326
0.3584 2.6596 9000 0.3455 0.4910 0.3335 0.3522 0.6006 0.6346 0.9331
0.3595 2.9551 10000 0.3491 0.4885 0.3282 0.3445 0.5966 0.6446 0.9316
0.3464 3.2506 11000 0.3423 0.4917 0.3398 0.3598 0.6052 0.6108 0.9349
0.3572 3.5461 12000 0.3426 0.4822 0.3419 0.3620 0.6054 0.6354 0.9341
0.3598 3.8416 13000 0.3468 0.4823 0.3367 0.3559 0.5975 0.6475 0.9327
0.3553 4.1371 14000 0.3483 0.4941 0.3394 0.3579 0.6033 0.5638 0.9350
0.3626 4.4326 15000 0.3405 0.4971 0.3345 0.3538 0.6040 0.6172 0.9343
0.3499 4.7281 16000 0.3401 0.4912 0.3408 0.3613 0.6052 0.6330 0.9345
0.3468 5.0236 17000 0.3391 0.4881 0.3415 0.3621 0.6064 0.6278 0.9346
0.333 5.3191 18000 0.3408 0.4977 0.3400 0.3584 0.6066 0.5884 0.9360
0.3512 5.6147 19000 0.3388 0.5002 0.3443 0.3645 0.6091 0.6127 0.9355
0.3583 5.9102 20000 0.3396 0.4946 0.3375 0.3580 0.6039 0.6352 0.9339
0.3391 6.2057 21000 0.3465 0.4947 0.3380 0.3550 0.6037 0.5646 0.9347
0.3396 6.5012 22000 0.3371 0.4962 0.3453 0.3658 0.6089 0.6147 0.9361
0.3476 6.7967 23000 0.3381 0.4890 0.3458 0.3672 0.6071 0.6411 0.9349
0.3517 7.0922 24000 0.3377 0.4958 0.3454 0.3654 0.6100 0.5965 0.9359
0.3501 7.3877 25000 0.3462 0.4840 0.3390 0.3580 0.5963 0.6580 0.9322
0.3421 7.6832 26000 0.3407 0.4875 0.3394 0.3600 0.6035 0.6388 0.9332
0.3392 7.9787 27000 0.3360 0.4951 0.3446 0.3666 0.6090 0.6150 0.9354
0.3444 8.2742 28000 0.3369 0.4895 0.3485 0.3707 0.6082 0.6457 0.9350
0.3566 8.5697 29000 0.3385 0.4846 0.3485 0.3708 0.6069 0.6518 0.9345
0.3395 8.8652 30000 0.3347 0.4931 0.3503 0.3719 0.6129 0.6252 0.9363
0.3451 9.1608 31000 0.3332 0.5022 0.3522 0.3744 0.6148 0.6200 0.9367
0.3383 9.4563 32000 0.3419 0.4969 0.3481 0.3641 0.6080 0.5704 0.9364
0.3361 9.7518 33000 0.3334 0.4912 0.3511 0.3740 0.6113 0.6353 0.9358
0.3427 10.0473 34000 0.3342 0.4988 0.3495 0.3721 0.6111 0.6444 0.9354
0.339 10.3428 35000 0.3337 0.4940 0.3602 0.3823 0.6155 0.6148 0.9374
0.3373 10.6383 36000 0.3322 0.4979 0.3512 0.3743 0.6134 0.6179 0.9365
0.3368 10.9338 37000 0.3328 0.4951 0.3527 0.3757 0.6131 0.6438 0.9359
0.3324 11.2293 38000 0.3336 0.4908 0.3537 0.3772 0.6107 0.6461 0.9360
0.3334 11.5248 39000 0.3322 0.4955 0.3585 0.3819 0.6163 0.6476 0.9369
0.3334 11.8203 40000 0.3309 0.4975 0.3570 0.3806 0.6162 0.6358 0.9370
0.3386 12.1158 41000 0.3305 0.4969 0.3592 0.3823 0.6183 0.6327 0.9373
0.3349 12.4113 42000 0.3320 0.4908 0.3575 0.3813 0.6148 0.6500 0.9363
0.3287 12.7069 43000 0.3313 0.4941 0.3588 0.3841 0.6144 0.6390 0.9367
0.3401 13.0024 44000 0.3311 0.4955 0.3572 0.3793 0.6161 0.6080 0.9377
0.3385 13.2979 45000 0.3308 0.4912 0.3586 0.3828 0.6162 0.6478 0.9367
0.3338 13.5934 46000 0.3300 0.4925 0.3612 0.3862 0.6167 0.6440 0.9370
0.3403 13.8889 47000 0.3421 0.4886 0.3540 0.3764 0.6011 0.6671 0.9342
0.3382 14.1844 48000 0.3335 0.4998 0.3556 0.3762 0.6143 0.5815 0.9377
0.3397 14.4799 49000 0.3289 0.4986 0.3539 0.3771 0.6167 0.6117 0.9373
0.327 14.7754 50000 0.3312 0.4942 0.3599 0.3815 0.6170 0.5997 0.9377
0.3388 15.0709 51000 0.3295 0.4965 0.3546 0.3795 0.6156 0.6222 0.9366
0.3375 15.3664 52000 0.3285 0.4960 0.3598 0.3833 0.6196 0.6211 0.9380
0.3316 15.6619 53000 0.3288 0.4977 0.3585 0.3833 0.6172 0.6431 0.9370
0.3306 15.9574 54000 0.3307 0.4947 0.3582 0.3838 0.6139 0.6407 0.9365
0.3449 16.2530 55000 0.3313 0.4947 0.3580 0.3828 0.6138 0.6562 0.9359
0.339 16.5485 56000 0.3306 0.4933 0.3623 0.3877 0.6150 0.6571 0.9364
0.3389 16.8440 57000 0.3286 0.4956 0.3654 0.3909 0.6177 0.6494 0.9376
0.3306 17.1395 58000 0.3269 0.4947 0.3676 0.3936 0.6201 0.6368 0.9379
0.3257 17.4350 59000 0.3313 0.4968 0.3590 0.3845 0.6133 0.6561 0.9357
0.3325 17.7305 60000 0.3284 0.4976 0.3595 0.3850 0.6166 0.6461 0.9368
0.3413 18.0260 61000 0.3271 0.4969 0.3628 0.3884 0.6190 0.6466 0.9374
0.3257 18.3215 62000 0.3276 0.4921 0.3676 0.3942 0.6187 0.6459 0.9376
0.3367 18.6170 63000 0.3266 0.4969 0.3644 0.3888 0.6214 0.6190 0.9385
0.3428 18.9125 64000 0.3261 0.4974 0.3618 0.3862 0.6203 0.6256 0.9381
0.3405 19.2080 65000 0.3260 0.4962 0.3651 0.3909 0.6210 0.6394 0.9377
0.3284 19.5035 66000 0.3263 0.4954 0.3627 0.3887 0.6192 0.6364 0.9374
0.3247 19.7991 67000 0.3263 0.4968 0.3624 0.3882 0.6195 0.6305 0.9376
0.3205 20.0946 68000 0.3261 0.4968 0.3660 0.3907 0.6212 0.6159 0.9386
0.3349 20.3901 69000 0.3265 0.4978 0.3666 0.3934 0.6198 0.6477 0.9376
0.3246 20.6856 70000 0.3262 0.4999 0.3626 0.3890 0.6195 0.6404 0.9376
0.3355 20.9811 71000 0.3274 0.4944 0.3636 0.3874 0.6187 0.6006 0.9383
0.3421 21.2766 72000 0.3253 0.4980 0.3636 0.3892 0.6211 0.6272 0.9382
0.3345 21.5721 73000 0.3258 0.5012 0.3625 0.3892 0.6199 0.6383 0.9373
0.3227 21.8676 74000 0.3248 0.4977 0.3672 0.3931 0.6228 0.6303 0.9385
0.3284 22.1631 75000 0.3248 0.4986 0.3658 0.3914 0.6224 0.6244 0.9387
0.3394 22.4586 76000 0.3255 0.4985 0.3661 0.3932 0.6196 0.6409 0.9379
0.314 22.7541 77000 0.3255 0.4969 0.3683 0.3950 0.6209 0.6438 0.9380
0.3268 23.0496 78000 0.3268 0.4987 0.3650 0.3919 0.6186 0.6503 0.9371
0.3285 23.3452 79000 0.3252 0.4995 0.3653 0.3913 0.6221 0.6216 0.9384
0.3304 23.6407 80000 0.3245 0.4984 0.3659 0.3918 0.6224 0.6325 0.9386
0.3298 23.9362 81000 0.3243 0.4975 0.3691 0.3955 0.6228 0.6370 0.9385
0.3206 24.2317 82000 0.3249 0.4964 0.3670 0.3939 0.6203 0.6417 0.9379
0.3326 24.5272 83000 0.3265 0.4959 0.3668 0.3933 0.6191 0.6535 0.9376
0.3265 24.8227 84000 0.3244 0.4988 0.3690 0.3958 0.6227 0.6387 0.9386
0.3406 25.1182 85000 0.3242 0.4998 0.3669 0.3939 0.6220 0.6308 0.9384
0.3276 25.4137 86000 0.3241 0.5000 0.3686 0.3955 0.6227 0.6342 0.9387
0.3277 25.7092 87000 0.3240 0.4983 0.3707 0.3976 0.6235 0.6373 0.9387
0.315 26.0047 88000 0.3244 0.4969 0.3697 0.3966 0.6225 0.6424 0.9383
0.3306 26.3002 89000 0.3254 0.5631 0.3684 0.3968 0.6198 0.6480 0.9379
0.3341 26.5957 90000 0.3237 0.4991 0.3697 0.3968 0.6239 0.6317 0.9389
0.3344 26.8913 91000 0.3241 0.4996 0.3649 0.3920 0.6217 0.6324 0.9380
0.3286 27.1868 92000 0.3242 0.4972 0.3680 0.3951 0.6214 0.6400 0.9382
0.3217 27.4823 93000 0.3238 0.4988 0.3688 0.3959 0.6223 0.6332 0.9386
0.3241 27.7778 94000 0.3242 0.4972 0.3673 0.3944 0.6215 0.6394 0.9381
0.3287 28.0733 95000 0.3236 0.5010 0.3693 0.3964 0.6235 0.6297 0.9390
0.3381 28.3688 96000 0.3239 0.4992 0.3682 0.3954 0.6222 0.6397 0.9383
0.3287 28.6643 97000 0.3235 0.4989 0.3695 0.3964 0.6234 0.6348 0.9388
0.3185 28.9598 98000 0.3239 0.4988 0.3689 0.3961 0.6225 0.6398 0.9384
0.3289 29.2553 99000 0.3238 0.4986 0.3691 0.3963 0.6225 0.6399 0.9385
0.3301 29.5508 100000 0.3235 0.4997 0.3682 0.3951 0.6231 0.6292 0.9388
0.3261 29.8463 101000 0.3237 0.4988 0.3690 0.3961 0.6229 0.6382 0.9386

Framework versions

  • Transformers 4.43.4
  • Pytorch 2.4.0+cu121
  • Datasets 2.21.0
  • Tokenizers 0.19.1