jssaluja/fb-mms-1b-punjabi-ccer-train-113953-epochs-15-test-1569
This model is a fine-tuned version of facebook/mms-1b-all on the jssaluja/rajinder_singh_corrected, hustler007/training_bhai_sahib_singh_canada_sehaj_paath, hustler007/training_bhai_jagtar_singh_jachak_sehaj_paath, hustler007/training_bhai_gurpreet_singh_nalvi_sehaj_paath, hustler007/training_bhai_sarwan_singh_sehaj_paath, hustler007/training_parminder_singh_australia_sehaj_paath_kirtan, hustler007/training_akashvani_jalandhar_radio_doctor, hustler007/training_jail_chittian, hustler007/training_sachi_sakhi_sirdar_kapur_singh, hustler007/training_rangle_sajjan_akj_audiobook, hustler007/training_jsk_talks_divine_talk, hustler007/training_bhai_harpal_singh_zafarnama_katha, hustler007/training_bhai_simranjeet_singh_tohana_special_talk, hustler007/training_rangle_sajjan_gurbani_audiobook, hustler007/training_smtv_podcast_simranjot_makkar, hustler007/training_bijay_singh, hustler007/training_khalas_podcast_harsharan_kaur, hustler007/training_bhai_pinderpal_singh_salok_vaara_te_vadeek, hustler007/training_satwant_kaur, hustler007/training_sikh_siyasat_diaspora_representatives, hustler007/training_sikh_siyasat_gurmeet_singh, hustler007/training_wikimedia_punjabi_audiobooks, hustler007/training_hum_hindu_nahin_kahn_singh_nabha, hustler007/training_heer_waris_shah, hustler007/training_bhai_pinderpal_singh_jaap_sahib, hustler007/training_sarbat_media_podcast, hustler007/training_sikh_siyasat_human_rights, hustler007/training_smtv_round_table, hustler007/training_discover_punjab_exclusive_stories, hustler007/training_bhai_sewa_singh_tarmala_akath_katha, hustler007/training_jhuthe_police_mukable_di_dastan, hustler007/training_gurmat_films_guru_nanak_sakhiyan, hustler007/training_maskeen_ji_full_katha, hustler007/training_punjabi_chapters_khalsa_raj, hustler007/training_mehmi_movies_punjabi_short_films, hustler007/training_tuhituhi_pinderpal_singh_katha, hustler007/training_gurmat_elite_discover_sikhism, hustler007/training_gyandasagar_shorts, hustler007/training_chaupal_chal_mera_putt_3, hustler007/training_chaupal_pardhan, hustler007/training_lokdhun_punjabi_latest_movies, hustler007/training_rajkaregakhalsa_shorts, hustler007/training_jaanbaaz_rakha_ar_darshi, hustler007/training_safar_with_nimrat_kaur, hustler007/training_sikh_siyasat_dr_kanwaljit_singh dataset. It achieves the following results on the evaluation set:
- Loss: 0.2574
- Wer: 0.3182
- Wil: 0.5094
- Mer: 0.3137
- Cer: 0.0813
- Ncer: 0.0337
- Ccer: 0.0873
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: 0.001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
(table generated directly from trainer.state.log_history, keyed by metric name -- avoids a transformers model-card column-ordering bug seen in earlier runs)
| Training Loss | Epoch | Step | Ccer | Cer | Validation Loss | Mer | Ncer | Wer | Wil |
|---|---|---|---|---|---|---|---|---|---|
| 0.4946 | 1.0 | 7122 | 0.1301 | 0.1204 | 0.3874 | 0.4406 | 0.0499 | 0.4623 | 0.6687 |
| 0.2746 | 2.0 | 14244 | 0.1038 | 0.0986 | 0.3080 | 0.3741 | 0.0423 | 0.3877 | 0.5873 |
| 0.2515 | 3.0 | 21366 | 0.1074 | 0.0998 | 0.3296 | 0.3773 | 0.0436 | 0.3840 | 0.5928 |
| 0.2367 | 4.0 | 28488 | 0.0965 | 0.0907 | 0.2843 | 0.3488 | 0.0379 | 0.3543 | 0.5565 |
| 0.2221 | 5.0 | 35610 | 0.0966 | 0.0912 | 0.2813 | 0.3505 | 0.0382 | 0.3563 | 0.5571 |
| 0.2116 | 6.0 | 42732 | 0.0958 | 0.0922 | 0.2887 | 0.3596 | 0.0369 | 0.3641 | 0.5641 |
| 0.2020 | 7.0 | 49854 | 0.0929 | 0.0857 | 0.2767 | 0.3258 | 0.0357 | 0.3327 | 0.5280 |
| 0.1941 | 8.0 | 56976 | 0.0903 | 0.0844 | 0.2579 | 0.3317 | 0.0356 | 0.3358 | 0.5333 |
| 0.1867 | 9.0 | 64098 | 0.0915 | 0.0849 | 0.2682 | 0.3243 | 0.0346 | 0.3301 | 0.5247 |
| 0.1811 | 10.0 | 71220 | 0.0914 | 0.0846 | 0.2674 | 0.3263 | 0.0344 | 0.3307 | 0.5269 |
| 0.1744 | 11.0 | 78342 | 0.0916 | 0.0850 | 0.2674 | 0.3244 | 0.0352 | 0.3296 | 0.5244 |
| 0.1677 | 12.0 | 85464 | 0.0888 | 0.0817 | 0.2601 | 0.3132 | 0.0327 | 0.3177 | 0.5105 |
| 0.1627 | 13.0 | 92586 | 0.0902 | 0.0828 | 0.2663 | 0.3166 | 0.0348 | 0.3222 | 0.5155 |
| 0.1583 | 14.0 | 99708 | 0.0873 | 0.0813 | 0.2574 | 0.3137 | 0.0337 | 0.3182 | 0.5094 |
| 0.1542 | 15.0 | 106830 | 0.0873 | 0.0813 | 0.2574 | 0.3137 | 0.0337 | 0.3182 | 0.5094 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.4