bertweet-large-sentiment-tuned-p

This model is a fine-tuned version of vinai/bertweet-large on the EPFL CS-433 Text Classification dataset. It achieves the following results on the evaluation set:

• Loss: 0.2088
• Accuracy: 0.9178
• F1: 0.9178
• Precision: 0.9178
• Recall: 0.9178

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 4e-06
• train_batch_size: 64
• eval_batch_size: 64
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 2.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.5966	0.0400	1413	0.3083	0.8754	0.8754	0.8756	0.8754
0.2814	0.0800	2826	0.2545	0.8913	0.8913	0.8914	0.8913
0.2577	0.1200	4239	0.2521	0.896	0.8959	0.8973	0.896
0.2474	0.1600	5652	0.2368	0.9019	0.9019	0.9019	0.9019
0.2415	0.2000	7065	0.2336	0.9023	0.9023	0.9026	0.9023
0.2367	0.2400	8478	0.2423	0.9033	0.9032	0.9044	0.9033
0.2338	0.2800	9891	0.2406	0.9019	0.9018	0.9036	0.9019
0.2292	0.3200	11304	0.2278	0.9073	0.9073	0.9075	0.9073
0.2285	0.3600	12717	0.2335	0.9083	0.9083	0.9083	0.9083
0.224	0.4000	14130	0.2255	0.9106	0.9106	0.9108	0.9106
0.2249	0.4400	15543	0.2317	0.91	0.9100	0.9100	0.91
0.2227	0.4801	16956	0.2208	0.9107	0.9107	0.9113	0.9107
0.224	0.5201	18369	0.2189	0.9108	0.9108	0.9108	0.9108
0.2228	0.5601	19782	0.2177	0.9121	0.9121	0.9124	0.9121
0.2216	0.6001	21195	0.2185	0.9103	0.9103	0.9103	0.9103
0.2201	0.6401	22608	0.2176	0.9114	0.9114	0.9116	0.9114
0.2187	0.6801	24021	0.2138	0.9126	0.9126	0.9126	0.9126
0.2168	0.7201	25434	0.2135	0.9127	0.9127	0.9131	0.9127
0.2185	0.7601	26847	0.2146	0.9142	0.9142	0.9142	0.9142
0.2171	0.8001	28260	0.2099	0.9142	0.9142	0.9142	0.9142
0.216	0.8401	29673	0.2103	0.9142	0.9142	0.9142	0.9142
0.218	0.8801	31086	0.2097	0.9139	0.9139	0.9139	0.9139
0.2154	0.9201	32499	0.2098	0.912	0.9120	0.9120	0.912
0.212	0.9601	33912	0.2079	0.9147	0.9147	0.9147	0.9147
0.2102	1.0001	35325	0.2105	0.9135	0.9135	0.9136	0.9135
0.1938	1.0401	36738	0.2114	0.9165	0.9165	0.9166	0.9165
0.1934	1.0801	38151	0.2118	0.9135	0.9135	0.9135	0.9135
0.1965	1.1201	39564	0.2121	0.9142	0.9142	0.9143	0.9142
0.1952	1.1601	40977	0.2156	0.9126	0.9126	0.9128	0.9126
0.1911	1.2001	42390	0.2142	0.9141	0.9141	0.9143	0.9141
0.1928	1.2401	43803	0.2143	0.9148	0.9148	0.9149	0.9148
0.1957	1.2801	45216	0.2116	0.9133	0.9133	0.9134	0.9133
0.1964	1.3201	46629	0.2108	0.9155	0.9155	0.9155	0.9155
0.1916	1.3602	48042	0.2125	0.9149	0.9149	0.9149	0.9149
0.1917	1.4002	49455	0.2158	0.9154	0.9154	0.9154	0.9154
0.1942	1.4402	50868	0.2124	0.9143	0.9143	0.9144	0.9143
0.1918	1.4802	52281	0.2075	0.9147	0.9147	0.9147	0.9147
0.1931	1.5202	53694	0.2154	0.9158	0.9158	0.9158	0.9158
0.1895	1.5602	55107	0.2084	0.916	0.9160	0.9160	0.916
0.192	1.6002	56520	0.2099	0.917	0.9170	0.9170	0.917
0.1948	1.6402	57933	0.2080	0.9163	0.9163	0.9163	0.9163
0.1907	1.6802	59346	0.2084	0.9166	0.9166	0.9166	0.9166
0.191	1.7202	60759	0.2094	0.9159	0.9159	0.9160	0.9159
0.1878	1.7602	62172	0.2104	0.9157	0.9157	0.9157	0.9157
0.192	1.8002	63585	0.2061	0.9179	0.9179	0.9179	0.9179
0.1894	1.8402	64998	0.2092	0.916	0.9160	0.9160	0.916
0.1922	1.8802	66411	0.2076	0.9165	0.9165	0.9165	0.9165
0.1908	1.9202	67824	0.2071	0.917	0.9170	0.9170	0.917
0.1883	1.9602	69237	0.2088	0.9178	0.9178	0.9178	0.9178

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0