Hugging Face's logo

herwoww
/
mms-MGB5

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
mms-MGB5 / README.md
herwoww's picture
herwoww
Model save
033073c verified
preview code
|
raw
history blame contribute delete
No virus
10.7 kB
metadata
license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: mms-MGB5
    results: []

mms-MGB5

This model is a fine-tuned version of facebook/mms-1b-all on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.2328
  • Wer: 88.9419

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: 14
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant_with_warmup
  • lr_scheduler_warmup_steps: 50
  • num_epochs: 25

Training results

Training Loss Epoch Step Validation Loss Wer
7.0257 0.16 250 6.2113 1.0041
3.2731 0.33 500 3.2218 1.0
2.8821 0.49 750 2.8522 1.0
2.1862 0.66 1000 2.0623 0.9721
1.6714 0.82 1250 1.6496 0.9398
1.7441 0.99 1500 1.5210 0.9314
1.4705 1.15 1750 1.4589 0.9251
1.4794 1.32 2000 1.4186 0.9213
1.4261 1.48 2250 1.3921 0.9175
1.3905 1.64 2500 1.3732 0.9185
1.4405 1.81 2750 1.3548 0.9134
1.3561 1.97 3000 1.3398 0.9093
1.4154 2.14 3250 1.3280 0.9128
1.361 2.3 3500 1.3161 0.9082
1.3421 2.47 3750 1.3064 0.9037
1.3615 2.63 4000 1.2976 0.9005
1.3517 2.8 4250 1.2996 0.9056
1.4316 2.96 4500 1.3136 0.9091
1.3946 3.12 4750 1.3534 0.9143
1.4888 3.29 5000 1.3797 0.9242
1.4223 3.45 5250 1.3707 0.9289
1.4616 3.62 5500 1.3463 0.9205
1.4196 3.78 5750 1.3362 0.9158
1.4163 3.95 6000 1.3392 0.9101
1.3408 4.11 6250 1.2765 89.9911
1.333 4.28 6500 1.2633 89.3855
1.3207 4.44 6750 1.2502 89.8381
1.2747 4.61 7000 1.2434 89.3478
1.3101 4.77 7250 1.2363 88.8864
1.281 4.93 7500 1.2290 88.7311
1.3351 5.1 7750 1.2211 88.3274
1.2738 5.26 8000 1.2147 88.2076
1.253 5.43 8250 1.2092 88.0834
1.2188 5.59 8500 1.2039 87.5776
1.2393 5.76 8750 1.1976 87.7418
1.2605 5.92 9000 1.1927 87.7019
1.2279 6.09 9250 1.1878 87.7307
1.2538 6.25 9500 1.1826 87.4201
1.2139 6.41 9750 1.1771 86.8301
1.1945 6.58 10000 1.1719 86.9210
1.2752 6.74 10250 1.1673 86.7902
1.1857 6.91 10500 1.1642 86.5306
1.1545 7.07 10750 1.1626 86.4929
1.2422 7.24 11000 1.1677 87.3270
1.2511 7.4 11250 1.2051 88.0435
1.2199 7.57 11500 1.2128 88.7844
1.2514 7.73 11750 1.2097 88.8199
1.2947 7.89 12000 1.2109 88.4339
1.3974 8.06 12250 1.2198 88.5426
1.2454 8.22 12500 1.2145 88.0546
1.2317 8.39 12750 1.2137 87.6353
1.3143 8.55 13000 1.2154 87.2671
1.3269 8.72 13250 1.2247 86.7924
1.2602 8.88 13500 1.2295 87.0453
1.2919 9.05 13750 1.2416 87.2205
1.2577 9.21 14000 1.2346 88.0058
1.2439 9.38 14250 1.2330 88.8421
1.4106 9.54 14500 1.2328 88.9286
1.2935 9.7 14750 1.2328 88.9419
1.2167 9.87 15000 1.2328 88.9419
1.3232 10.03 15250 1.2328 88.9419
1.2935 10.2 15500 1.2328 88.9419
1.2459 10.36 15750 1.2328 88.9419
1.2716 10.53 16000 1.2328 88.9419
1.2941 10.69 16250 1.2328 88.9419
1.2849 10.86 16500 1.2328 88.9419
1.2813 11.02 16750 1.2328 88.9419
1.351 11.18 17000 1.2328 88.9419
1.2302 11.35 17250 1.2328 88.9419
1.2724 11.51 17500 1.2328 88.9419
1.293 11.68 17750 1.2328 88.9419
1.2799 11.84 18000 1.2328 88.9419
1.2969 12.01 18250 1.2328 88.9419
1.2824 12.17 18500 1.2328 88.9419
1.2564 12.34 18750 1.2328 88.9419
1.2867 12.5 19000 1.2328 88.9419
1.2606 12.66 19250 1.2328 88.9419
1.2988 12.83 19500 1.2328 88.9419
1.2882 12.99 19750 1.2328 88.9419
1.3214 13.16 20000 1.2328 88.9419
1.2781 13.32 20250 1.2328 88.9419
1.3764 13.49 20500 1.2328 88.9419
1.2828 13.65 20750 1.2328 88.9419
1.2571 13.82 21000 1.2328 88.9419
1.2649 13.98 21250 1.2328 88.9419
1.2571 14.14 21500 1.2328 88.9419
1.2657 14.31 21750 1.2328 88.9419
1.3216 14.47 22000 1.2328 88.9419
1.2843 14.64 22250 1.2328 88.9419
1.307 14.8 22500 1.2328 88.9419
1.3198 14.97 22750 1.2328 88.9419
1.3673 15.13 23000 1.2328 88.9419
1.3636 15.3 23250 1.2328 88.9419
1.2858 15.46 23500 1.2328 88.9419
1.2658 15.62 23750 1.2328 88.9419
1.2532 15.79 24000 1.2328 88.9419
1.2647 15.95 24250 1.2328 88.9419
1.2912 16.12 24500 1.2328 88.9419
1.2707 16.28 24750 1.2328 88.9419
1.3161 16.45 25000 1.2328 88.9419
1.3061 16.61 25250 1.2328 88.9419
1.2688 16.78 25500 1.2328 88.9419
1.2528 16.94 25750 1.2328 88.9419
1.2585 17.11 26000 1.2328 88.9419
1.2681 17.27 26250 1.2328 88.9419
1.2705 17.43 26500 1.2328 88.9419
1.2662 17.6 26750 1.2328 88.9419
1.3273 17.76 27000 1.2328 88.9419
1.2774 17.93 27250 1.2328 88.9419
1.293 18.09 27500 1.2328 88.9419
1.2305 18.26 27750 1.2328 88.9419
1.3117 18.42 28000 1.2328 88.9419
1.397 18.59 28250 1.2328 88.9419
1.2693 18.75 28500 1.2328 88.9419
1.3219 18.91 28750 1.2328 88.9419
1.3352 19.08 29000 1.2328 88.9419
1.3321 19.24 29250 1.2328 88.9419
1.2534 19.41 29500 1.2328 88.9419
1.2507 19.57 29750 1.2328 88.9419
1.273 19.74 30000 1.2328 88.9419
1.2856 19.9 30250 1.2328 88.9419
1.3163 20.07 30500 1.2328 88.9419
1.2463 20.23 30750 1.2328 88.9419
1.3017 20.39 31000 1.2328 88.9419
1.3236 20.56 31250 1.2328 88.9419
1.3012 20.72 31500 1.2328 88.9419
1.3296 20.89 31750 1.2328 88.9419
1.362 21.05 32000 1.2328 88.9419
1.2816 21.22 32250 1.2328 88.9419
1.2383 21.38 32500 1.2328 88.9419
1.2127 21.55 32750 1.2328 88.9419
1.2443 21.71 33000 1.2328 88.9419
1.2553 21.88 33250 1.2328 88.9419
1.3227 22.04 33500 1.2328 88.9419
1.2923 22.2 33750 1.2328 88.9419
1.2954 22.37 34000 1.2328 88.9419
1.3123 22.53 34250 1.2328 88.9419
1.2675 22.7 34500 1.2328 88.9419
1.2565 22.86 34750 1.2328 88.9419
1.2139 23.03 35000 1.2328 88.9419
1.3014 23.19 35250 1.2328 88.9419
1.2949 23.36 35500 1.2328 88.9419
1.2854 23.52 35750 1.2328 88.9419
1.2639 23.68 36000 1.2328 88.9419
1.2692 23.85 36250 1.2328 88.9419
1.2935 24.01 36500 1.2328 88.9419
1.4062 24.18 36750 1.2328 88.9419
1.2626 24.34 37000 1.2328 88.9419
1.2555 24.51 37250 1.2328 88.9419
1.2876 24.67 37500 1.2328 88.9419
1.3362 24.84 37750 1.2328 88.9419
1.2625 25.0 38000 1.2328 88.9419

Framework versions

  • Transformers 4.33.2
  • Pytorch 2.0.1
  • Datasets 2.19.1
  • Tokenizers 0.13.3