Hugging Face's logo

mmousavi660
/
cls_headline_llama3_v1

trl
sft
Generated from Trainer
Model card Files Community
cls_headline_llama3_v1 / README.md
mmousavi660's picture
mmousavi660
Model save
c910d39 verified
preview code
|
raw
history blame contribute delete
6.57 kB
metadata
license: llama3
base_model: meta-llama/Meta-Llama-3-8B-Instruct
tags:
  - trl
  - sft
  - generated_from_trainer
datasets:
  - generator
model-index:
  - name: cls_headline_llama3_v1
    results: []

cls_headline_llama3_v1

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the generator dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2376

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: 2
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 4
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • lr_scheduler_warmup_ratio: 0.03
  • num_epochs: 2
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
No log 0.02 5 0.5809
0.7814 0.04 10 0.4209
0.7814 0.06 15 0.3950
0.3932 0.08 20 0.3830
0.3932 0.1 25 0.3727
0.3726 0.12 30 0.3658
0.3726 0.14 35 0.3597
0.3572 0.16 40 0.3560
0.3572 0.18 45 0.3524
0.3437 0.2 50 0.3493
0.3437 0.22 55 0.3452
0.3404 0.24 60 0.3449
0.3404 0.26 65 0.3427
0.3349 0.28 70 0.3398
0.3349 0.3 75 0.3368
0.3342 0.32 80 0.3350
0.3342 0.34 85 0.3322
0.3316 0.36 90 0.3310
0.3316 0.38 95 0.3290
0.3251 0.4 100 0.3271
0.3251 0.42 105 0.3251
0.3137 0.44 110 0.3251
0.3137 0.46 115 0.3245
0.3216 0.48 120 0.3216
0.3216 0.5 125 0.3204
0.3171 0.52 130 0.3198
0.3171 0.54 135 0.3167
0.3161 0.56 140 0.3153
0.3161 0.58 145 0.3152
0.3176 0.6 150 0.3135
0.3176 0.62 155 0.3123
0.3089 0.64 160 0.3109
0.3089 0.66 165 0.3102
0.3109 0.68 170 0.3094
0.3109 0.7 175 0.3070
0.3058 0.72 180 0.3053
0.3058 0.74 185 0.3030
0.3041 0.76 190 0.3026
0.3041 0.78 195 0.3012
0.2975 0.8 200 0.2988
0.2975 0.82 205 0.2972
0.2861 0.84 210 0.2976
0.2861 0.86 215 0.2956
0.2964 0.88 220 0.2944
0.2964 0.9 225 0.2918
0.2907 0.92 230 0.2899
0.2907 0.94 235 0.2893
0.2861 0.96 240 0.2873
0.2861 0.98 245 0.2870
0.2794 1.0 250 0.2859
0.2794 1.02 255 0.2902
0.251 1.04 260 0.2857
0.251 1.06 265 0.2876
0.2498 1.08 270 0.2837
0.2498 1.1 275 0.2852
0.2445 1.12 280 0.2807
0.2445 1.14 285 0.2809
0.251 1.16 290 0.2806
0.251 1.18 295 0.2818
0.2455 1.2 300 0.2795
0.2455 1.22 305 0.2772
0.2416 1.24 310 0.2775
0.2416 1.26 315 0.2754
0.2463 1.28 320 0.2740
0.2463 1.3 325 0.2740
0.2381 1.32 330 0.2732
0.2381 1.34 335 0.2727
0.2401 1.36 340 0.2714
0.2401 1.38 345 0.2698
0.2426 1.4 350 0.2691
0.2426 1.42 355 0.2671
0.2371 1.44 360 0.2657
0.2371 1.46 365 0.2650
0.2409 1.48 370 0.2646
0.2409 1.5 375 0.2620
0.2386 1.52 380 0.2599
0.2386 1.54 385 0.2592
0.2331 1.56 390 0.2584
0.2331 1.58 395 0.2571
0.2339 1.6 400 0.2554
0.2339 1.62 405 0.2563
0.2295 1.64 410 0.2550
0.2295 1.66 415 0.2527
0.2349 1.68 420 0.2537
0.2349 1.7 425 0.2515
0.2296 1.72 430 0.2514
0.2296 1.74 435 0.2486
0.2288 1.76 440 0.2479
0.2288 1.78 445 0.2478
0.2346 1.8 450 0.2456
0.2346 1.82 455 0.2440
0.227 1.84 460 0.2424
0.227 1.86 465 0.2427
0.2217 1.88 470 0.2410
0.2217 1.9 475 0.2402
0.2203 1.92 480 0.2398
0.2203 1.94 485 0.2406
0.2129 1.96 490 0.2388
0.2129 1.98 495 0.2358
0.2195 2.0 500 0.2376

Framework versions

  • Transformers 4.35.0.dev0
  • Pytorch 2.0.1
  • Datasets 2.19.1
  • Tokenizers 0.14.1