speech_emotion_is25_2401

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 5.9534
• F1: 0.2172

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: 0.0003
• train_batch_size: 96
• eval_batch_size: 96
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 384
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 80

Training results

Training Loss	Epoch	Step	Validation Loss	F1
2.0794	1.0	25	2.0794	0.0273
1.9206	2.0	50	2.0781	0.0546
2.0578	3.0	75	2.0166	0.1312
1.8736	4.0	100	2.0161	0.1342
1.9971	5.0	125	1.9877	0.1640
1.8198	6.0	150	1.9416	0.2073
1.9686	7.0	175	1.9777	0.1657
1.8065	8.0	200	1.9370	0.1977
1.9424	9.0	225	1.9781	0.2017
1.7882	10.0	250	1.9437	0.2064
1.9057	11.0	275	1.9376	0.2189
1.7494	12.0	300	1.9299	0.2191
1.8207	13.0	325	1.9541	0.1991
1.6539	14.0	350	1.9275	0.2381
1.7783	15.0	375	1.9523	0.2520
1.6253	16.0	400	1.9718	0.2340
1.697	17.0	425	1.9726	0.2506
1.5304	18.0	450	1.9830	0.2483
1.6035	19.0	475	2.2017	0.1929
1.4398	20.0	500	2.2326	0.2127
1.5259	21.0	525	2.2705	0.2161
1.3323	22.0	550	2.2537	0.2314
1.4133	23.0	575	2.2838	0.2229
1.267	24.0	600	2.3778	0.2202
1.2922	25.0	625	2.5493	0.2104
1.1266	26.0	650	2.7201	0.2102
1.1796	27.0	675	2.7254	0.2238
1.017	28.0	700	2.9926	0.1992
1.0778	29.0	725	2.8849	0.2071
0.9195	30.0	750	2.9993	0.2189
0.9121	31.0	775	3.0823	0.2148
0.7994	32.0	800	3.1808	0.2125
0.777	33.0	825	3.3580	0.2077
0.7391	34.0	850	3.4756	0.1952
0.751	35.0	875	3.5514	0.2086
0.6774	36.0	900	3.7188	0.2068
0.6095	37.0	925	3.8295	0.1968
0.5689	38.0	950	3.8538	0.2218
0.5496	39.0	975	3.9988	0.1990
0.4789	40.0	1000	4.2451	0.2108
0.4903	41.0	1025	4.0233	0.2216
0.4433	42.0	1050	4.3824	0.2056
0.4537	43.0	1075	4.4037	0.2047
0.4196	44.0	1100	4.2913	0.2236
0.37	45.0	1125	4.6178	0.2072
0.2967	46.0	1150	4.6234	0.2132
0.352	47.0	1175	4.7141	0.2199
0.3013	48.0	1200	4.8603	0.2203
0.3731	49.0	1225	4.8180	0.2276
0.3029	50.0	1250	4.9620	0.2198
0.292	51.0	1275	4.9615	0.2233
0.2765	52.0	1300	4.9503	0.2175
0.2408	53.0	1325	4.9422	0.2244
0.2151	54.0	1350	5.0663	0.2138
0.2335	55.0	1375	5.3134	0.2147
0.2009	56.0	1400	5.4252	0.1973
0.2154	57.0	1425	5.3755	0.2019
0.2045	58.0	1450	5.3263	0.2191
0.2135	59.0	1475	5.3328	0.2109
0.1765	60.0	1500	5.6206	0.2098
0.1764	61.0	1525	5.5509	0.2200
0.1648	62.0	1550	5.5793	0.2088
0.1558	63.0	1575	5.5590	0.2229
0.1294	64.0	1600	5.7499	0.2177
0.152	65.0	1625	5.7696	0.2102
0.1133	66.0	1650	5.6495	0.2129
0.1425	67.0	1675	5.7512	0.2233
0.1171	68.0	1700	5.7398	0.2166
0.1681	69.0	1725	5.8166	0.2231
0.1217	70.0	1750	5.8027	0.2148
0.1228	71.0	1775	5.7813	0.2183
0.1055	72.0	1800	5.8321	0.2218
0.1174	73.0	1825	5.8462	0.2207
0.1193	74.0	1850	5.9226	0.2148
0.1104	75.0	1875	5.9136	0.2192
0.1115	76.0	1900	5.9308	0.2202
0.1122	76.8247	1920	5.9534	0.2172

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0