learning_rate: 1e-05
train_batch_size: 32
eval_batch_size: 8
seed: 42
optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.8205	0.0078	1	1.7910	97.5
1.6896	0.0156	2	1.3728	92.7419
1.8025	0.0234	3	1.3427	91.9355
1.4496	0.0312	4	1.3110	91.2097
1.594	0.0391	5	1.2896	101.8548
1.3153	0.0469	6	1.2404	99.0323
1.1479	0.0547	7	1.2017	105.2419
1.3098	0.0625	8	1.1440	124.7581
1.31	0.0703	9	1.1160	123.7097
1.1475	0.0781	10	1.0799	124.7581
1.2754	0.0859	11	1.0437	139.2742
1.0161	0.0938	12	1.0167	129.6774
1.3014	0.1016	13	0.9949	148.3871
0.966	0.1094	14	0.9597	157.0968
1.0962	0.1172	15	0.9217	130.0806
0.9943	0.125	16	0.8920	119.1129
0.8999	0.1328	17	0.8814	106.4516
0.9399	0.1406	18	0.8576	110.4032
1.0341	0.1484	19	0.8356	104.8387
0.7791	0.1562	20	0.8317	106.4516
1.0291	0.1641	21	0.8300	94.5161
0.8347	0.1719	22	0.8252	93.4677
1.0203	0.1797	23	0.8242	93.3065
0.8479	0.1875	24	0.8141	92.8226
0.8749	0.1953	25	0.8059	100.2419
0.7703	0.2031	26	0.7966	103.8710
0.7039	0.2109	27	0.7869	98.8710
0.9409	0.2188	28	0.7809	108.0645
0.8051	0.2266	29	0.7755	107.6613
0.8712	0.2344	30	0.7727	95.8065
0.9535	0.2422	31	0.7709	88.9516
0.9853	0.25	32	0.7744	94.6774
0.5933	0.2578	33	0.7788	95.9677
1.1144	0.2656	34	0.7680	83.7097
0.9498	0.2734	35	0.7619	87.8226
0.7747	0.2812	36	0.7596	88.2258
0.7927	0.2891	37	0.7583	87.7419
0.8625	0.2969	38	0.7551	96.8548
0.933	0.3047	39	0.7546	94.9194
0.8725	0.3125	40	0.7574	96.2903
0.8884	0.3203	41	0.7609	106.8548
0.8008	0.3281	42	0.7586	112.7419
0.8222	0.3359	43	0.7515	95.4032
0.7829	0.3438	44	0.7488	96.2097
0.7861	0.3516	45	0.7485	88.7097
0.6381	0.3594	46	0.7446	89.9194
0.9817	0.3672	47	0.7384	88.7097
0.6827	0.375	48	0.7277	84.3548
0.8762	0.3828	49	0.7207	80.5645
0.8126	0.3906	50	0.7216	77.4194
0.7661	0.3984	51	0.7251	79.5161
0.7805	0.4062	52	0.7229	78.4677
0.8456	0.4141	53	0.7221	81.2097
0.7459	0.4219	54	0.7239	94.2742
0.724	0.4297	55	0.7269	108.2258
0.7065	0.4375	56	0.7280	107.0968
0.7512	0.4453	57	0.7266	90.3226
0.7915	0.4531	58	0.7229	87.4194
0.7269	0.4609	59	0.7175	85.7258
0.7296	0.4688	60	0.7155	82.8226
0.6023	0.4766	61	0.7172	79.5161
0.5805	0.4844	62	0.7197	77.0968
0.7154	0.4922	63	0.7172	76.7742
0.7627	0.5	64	0.7157	77.4194
0.7749	0.5078	65	0.7097	76.6935
0.7903	0.5156	66	0.7075	77.6613
0.7966	0.5234	67	0.7069	79.5161
0.6448	0.5312	68	0.7100	85.5645
0.8416	0.5391	69	0.7160	95.2419
0.7756	0.5469	70	0.7256	98.6290
0.6671	0.5547	71	0.7319	98.2258
0.7235	0.5625	72	0.7273	92.1774
0.6617	0.5703	73	0.7160	89.9194
0.5872	0.5781	74	0.7038	84.9194
0.5833	0.5859	75	0.6956	81.6935
0.6873	0.5938	76	0.6917	87.4194
0.5932	0.6016	77	0.6899	85.6452
0.6885	0.6094	78	0.6893	85.0806
0.5302	0.6172	79	0.6882	86.1290
0.7193	0.625	80	0.6882	86.2097
0.7908	0.6328	81	0.6891	78.2258
0.6899	0.6406	82	0.6931	79.9194
0.6676	0.6484	83	0.6997	81.9355
0.6297	0.6562	84	0.7070	84.1129
0.7199	0.6641	85	0.7121	87.9032
0.5461	0.6719	86	0.7140	89.8387
0.937	0.6797	87	0.7122	91.2097
0.6397	0.6875	88	0.7063	90.0
0.5862	0.6953	89	0.6978	86.9355
0.6388	0.7031	90	0.6914	83.5484
0.5044	0.7109	91	0.6841	80.1613
0.6567	0.7188	92	0.6792	85.3226
0.7628	0.7266	93	0.6766	77.6613
0.6366	0.7344	94	0.6757	77.5
0.6173	0.7422	95	0.6737	77.9032
0.6387	0.75	96	0.6730	79.5161
0.6653	0.7578	97	0.6720	81.2903
0.7108	0.7656	98	0.6711	87.9839
0.6456	0.7734	99	0.6727	91.6129
0.5809	0.7812	100	0.6735	83.6290
0.5899	0.7891	101	0.6714	83.4677
0.6761	0.7969	102	0.6700	83.2258
0.7017	0.8047	103	0.6708	82.5
0.6117	0.8125	104	0.6729	81.5323
0.7295	0.8203	105	0.6752	80.9677
0.5875	0.8281	106	0.6770	81.2097
0.796	0.8359	107	0.6791	82.5806
0.8962	0.8438	108	0.6804	84.5161
0.7165	0.8516	109	0.6800	85.7258
0.6525	0.8594	110	0.6767	85.8065
0.5235	0.8672	111	0.6733	85.3226
0.7755	0.875	112	0.6703	84.5161
0.6451	0.8828	113	0.6685	84.1935
0.7683	0.8906	114	0.6646	82.5
0.6895	0.8984	115	0.6603	80.2419
0.6829	0.9062	116	0.6574	78.5484
0.7359	0.9141	117	0.6555	78.4677
0.606	0.9219	118	0.6540	78.9516
0.518	0.9297	119	0.6534	79.0323
0.6314	0.9375	120	0.6528	78.4677
0.6467	0.9453	121	0.6524	77.7419
0.7231	0.9531	122	0.6522	78.7903
0.7501	0.9609	123	0.6532	78.7097
0.6274	0.9688	124	0.6547	79.2742
0.5353	0.9766	125	0.6563	81.7742
0.5631	0.9844	126	0.6591	83.3871
0.6666	0.9922	127	0.6607	83.3065
0.4207	1.0	128	0.6618	82.7419
0.4789	1.0078	129	0.6611	82.3387
0.4743	1.0156	130	0.6598	83.0645
0.6108	1.0234	131	0.6600	82.2581
0.4023	1.0312	132	0.6598	80.4839
0.4907	1.0391	133	0.6594	80.0
0.4726	1.0469	134	0.6576	79.2742
0.4386	1.0547	135	0.6555	78.4677
0.4673	1.0625	136	0.6549	79.2742
0.3555	1.0703	137	0.6539	79.0323
0.4115	1.0781	138	0.6545	77.4194
0.3697	1.0859	139	0.6567	77.1774
0.3792	1.0938	140	0.6588	77.0161
0.5365	1.1016	141	0.6620	77.5
0.3783	1.1094	142	0.6640	78.4677
0.4625	1.1172	143	0.6637	79.5161
0.4137	1.125	144	0.6651	81.2097
0.4463	1.1328	145	0.6667	81.6935
0.448	1.1406	146	0.6679	81.4516
0.3829	1.1484	147	0.6679	82.0968
0.4125	1.1562	148	0.6666	81.3710
0.4642	1.1641	149	0.6665	81.2097
0.3676	1.1719	150	0.6659	80.4839
0.5189	1.1797	151	0.6654	86.6935
0.4097	1.1875	152	0.6659	86.3710
0.2964	1.1953	153	0.6664	86.6935
0.32	1.2031	154	0.6666	87.0161
0.4166	1.2109	155	0.6665	86.9355
0.4626	1.2188	156	0.6659	88.3871
0.4282	1.2266	157	0.6647	88.3065
0.4618	1.2344	158	0.6638	87.8226
0.3474	1.2422	159	0.6644	88.2258
0.3743	1.25	160	0.6637	87.9839
0.4544	1.2578	161	0.6638	87.9032
0.3572	1.2656	162	0.6646	89.0323
0.404	1.2734	163	0.6661	89.0323
0.438	1.2812	164	0.6671	89.8387
0.4599	1.2891	165	0.6673	90.8065
0.3681	1.2969	166	0.6679	91.8548
0.4114	1.3047	167	0.6667	92.0161
0.4434	1.3125	168	0.6631	90.4839
0.4012	1.3203	169	0.6597	89.7581
0.3688	1.3281	170	0.6561	88.6290
0.4731	1.3359	171	0.6531	87.4194
0.4342	1.3438	172	0.6511	87.0968
0.3771	1.3516	173	0.6504	86.2903
0.3929	1.3594	174	0.6498	85.8871
0.2813	1.3672	175	0.6499	85.6452
0.5008	1.375	176	0.6500	85.6452
0.4107	1.3828	177	0.6505	86.5323
0.3766	1.3906	178	0.6511	87.6613
0.4951	1.3984	179	0.6519	88.3871
0.4328	1.4062	180	0.6526	88.5484
0.4443	1.4141	181	0.6531	89.2742
0.282	1.4219	182	0.6537	88.9516
0.371	1.4297	183	0.6541	88.9516
0.3711	1.4375	184	0.6542	88.7903
0.3776	1.4453	185	0.6536	88.8710
0.379	1.4531	186	0.6530	88.3871
0.2809	1.4609	187	0.6523	88.0645
0.51	1.4688	188	0.6515	87.3387
0.3402	1.4766	189	0.6509	86.6935
0.3348	1.4844	190	0.6504	87.0161
0.3278	1.4922	191	0.6499	86.4516
0.3931	1.5	192	0.6497	86.3710
0.4821	1.5078	193	0.6492	86.5323
0.3666	1.5156	194	0.6483	86.9355
0.5035	1.5234	195	0.6481	86.7742
0.3547	1.5312	196	0.6478	86.7742
0.3464	1.5391	197	0.6473	87.0968
0.387	1.5469	198	0.6466	87.6613
0.4224	1.5547	199	0.6460	87.3387
0.3741	1.5625	200	0.6455	87.5806
0.3652	1.5703	201	0.6447	86.8548
0.4389	1.5781	202	0.6444	86.9355
0.3899	1.5859	203	0.6441	87.0968
0.4128	1.5938	204	0.6442	87.5806
0.3406	1.6016	205	0.6441	88.0645
0.3645	1.6094	206	0.6445	88.3065
0.3426	1.6172	207	0.6446	88.2258
0.4304	1.625	208	0.6451	88.7097
0.3869	1.6328	209	0.6458	88.7903
0.3845	1.6406	210	0.6459	88.9516
0.3839	1.6484	211	0.6462	89.0323
0.3813	1.6562	212	0.6465	89.3548
0.269	1.6641	213	0.6468	89.1935
0.4695	1.6719	214	0.6467	89.2742
0.4237	1.6797	215	0.6466	89.1935
0.2824	1.6875	216	0.6465	89.6774
0.3297	1.6953	217	0.6463	89.3548
0.4549	1.7031	218	0.6453	89.1129
0.4411	1.7109	219	0.6454	89.2742
0.3779	1.7188	220	0.6448	88.3871
0.3803	1.7266	221	0.6444	88.3871
0.4527	1.7344	222	0.6442	88.2258
0.2786	1.7422	223	0.6437	87.9839
0.4656	1.75	224	0.6440	88.2258
0.3298	1.7578	225	0.6435	88.1452
0.4767	1.7656	226	0.6430	88.0645
0.4267	1.7734	227	0.6429	87.5
0.3965	1.7812	228	0.6428	87.5806
0.4037	1.7891	229	0.6426	87.3387
0.4294	1.7969	230	0.6422	87.4194
0.3593	1.8047	231	0.6424	87.5
0.417	1.8125	232	0.6420	87.5806
0.3476	1.8203	233	0.6417	87.4194
0.321	1.8281	234	0.6416	87.7419
0.3778	1.8359	235	0.6414	87.5
0.4419	1.8438	236	0.6417	87.4194
0.3343	1.8516	237	0.6414	87.6613
0.5093	1.8594	238	0.6414	87.6613
0.4397	1.8672	239	0.6411	87.5806
0.3782	1.875	240	0.6412	87.5806
0.3143	1.8828	241	0.6410	87.9839
0.4161	1.8906	242	0.6410	87.5806
0.3421	1.8984	243	0.6410	87.6613
0.3533	1.9062	244	0.6410	87.9032
0.4659	1.9141	245	0.6411	87.9032
0.3761	1.9219	246	0.6412	87.7419
0.465	1.9297	247	0.6409	87.7419
0.4935	1.9375	248	0.6411	87.7419
0.3086	1.9453	249	0.6408	87.6613
0.466	1.9531	250	0.6409	88.0645
0.3417	1.9609	251	0.6409	87.9032
0.4786	1.9688	252	0.6407	87.9839
0.3778	1.9766	253	0.6410	87.6613
0.462	1.9844	254	0.6411	87.5806
0.4344	1.9922	255	0.6410	87.6613
0.4778	2.0	256	0.6409	87.9032

Framework versions

Transformers 4.57.3
Pytorch 2.9.1+cu128
Datasets 4.4.1
Tokenizers 0.22.1

nkkbr/whisper-large-v3-zatoichi-ja-EX-1-TRAIN_2_TO_10_EVAL_1_COSINE

Whisper Large v3 - Japanese Zatoichi ASR

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

nkkbr/
whisper-large-v3-zatoichi-ja-EX-1-TRAIN_2_TO_10_EVAL_1_COSINE