scenario-kd-po-ner-half_data-univner_full44

This model is a fine-tuned version of haryoaw/scenario-TCR-NER_data-univner_full on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4036
• Precision: 0.8328
• Recall: 0.8256
• F1: 0.8292
• Accuracy: 0.9821

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 44
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.9712	0.2910	500	0.7018	0.7311	0.7712	0.7506	0.9758
0.5607	0.5821	1000	0.6182	0.7648	0.7927	0.7785	0.9780
0.5004	0.8731	1500	0.5862	0.7844	0.7895	0.7869	0.9786
0.4431	1.1641	2000	0.5566	0.7646	0.8132	0.7881	0.9787
0.4064	1.4552	2500	0.5562	0.7940	0.7979	0.7959	0.9795
0.3839	1.7462	3000	0.5438	0.7918	0.7971	0.7944	0.9791
0.3668	2.0373	3500	0.5229	0.7997	0.7928	0.7963	0.9796
0.328	2.3283	4000	0.5153	0.8032	0.7980	0.8006	0.9797
0.3247	2.6193	4500	0.5129	0.8029	0.7957	0.7993	0.9797
0.3186	2.9104	5000	0.5061	0.8012	0.8049	0.8031	0.9798
0.2965	3.2014	5500	0.4900	0.8091	0.8159	0.8125	0.9809
0.2827	3.4924	6000	0.4904	0.7916	0.8225	0.8068	0.9799
0.2788	3.7835	6500	0.4974	0.8184	0.7966	0.8073	0.9805
0.2711	4.0745	7000	0.4937	0.8129	0.8075	0.8102	0.9804
0.2557	4.3655	7500	0.4804	0.8058	0.8129	0.8093	0.9805
0.2512	4.6566	8000	0.4762	0.8092	0.8127	0.8110	0.9807
0.2493	4.9476	8500	0.4673	0.8102	0.8136	0.8119	0.9806
0.2327	5.2386	9000	0.4668	0.8113	0.8163	0.8138	0.9810
0.2268	5.5297	9500	0.4650	0.8211	0.8020	0.8115	0.9807
0.2314	5.8207	10000	0.4747	0.8086	0.8152	0.8119	0.9804
0.2238	6.1118	10500	0.4577	0.8144	0.8067	0.8105	0.9806
0.2077	6.4028	11000	0.4646	0.8157	0.8147	0.8152	0.9809
0.2114	6.6938	11500	0.4639	0.8173	0.8153	0.8163	0.9812
0.2098	6.9849	12000	0.4545	0.8230	0.8090	0.8159	0.9809
0.1965	7.2759	12500	0.4511	0.8168	0.8181	0.8174	0.9812
0.1952	7.5669	13000	0.4553	0.8153	0.8191	0.8172	0.9813
0.197	7.8580	13500	0.4441	0.8100	0.8221	0.8160	0.9811
0.1868	8.1490	14000	0.4439	0.8189	0.8214	0.8201	0.9813
0.1827	8.4400	14500	0.4500	0.8194	0.8160	0.8177	0.9810
0.1822	8.7311	15000	0.4460	0.8188	0.8139	0.8164	0.9810
0.1811	9.0221	15500	0.4449	0.8145	0.8152	0.8148	0.9809
0.1708	9.3132	16000	0.4543	0.8232	0.8110	0.8171	0.9810
0.171	9.6042	16500	0.4464	0.8182	0.8171	0.8176	0.9814
0.1726	9.8952	17000	0.4381	0.8158	0.8234	0.8196	0.9814
0.1646	10.1863	17500	0.4392	0.8183	0.8273	0.8228	0.9815
0.162	10.4773	18000	0.4351	0.8142	0.8208	0.8175	0.9811
0.1619	10.7683	18500	0.4434	0.8154	0.8191	0.8172	0.9811
0.1588	11.0594	19000	0.4441	0.8226	0.8116	0.8171	0.9811
0.155	11.3504	19500	0.4316	0.8176	0.8224	0.8200	0.9817
0.1533	11.6414	20000	0.4360	0.8249	0.8117	0.8183	0.9814
0.1528	11.9325	20500	0.4388	0.8164	0.8182	0.8173	0.9811
0.1484	12.2235	21000	0.4282	0.8258	0.8168	0.8213	0.9812
0.1468	12.5146	21500	0.4422	0.8213	0.8136	0.8174	0.9808
0.1495	12.8056	22000	0.4321	0.8156	0.8142	0.8149	0.9809
0.1452	13.0966	22500	0.4405	0.8240	0.8088	0.8164	0.9813
0.1415	13.3877	23000	0.4365	0.8268	0.8175	0.8221	0.9814
0.1412	13.6787	23500	0.4346	0.8246	0.8149	0.8197	0.9810
0.1407	13.9697	24000	0.4257	0.8191	0.8260	0.8226	0.9816
0.1359	14.2608	24500	0.4296	0.8262	0.8181	0.8221	0.9815
0.1359	14.5518	25000	0.4301	0.8180	0.8166	0.8173	0.9811
0.1356	14.8428	25500	0.4317	0.8148	0.8269	0.8208	0.9813
0.1331	15.1339	26000	0.4313	0.8266	0.8090	0.8177	0.9812
0.1308	15.4249	26500	0.4269	0.8251	0.8175	0.8213	0.9816
0.1306	15.7159	27000	0.4301	0.8217	0.8218	0.8218	0.9816
0.1319	16.0070	27500	0.4204	0.8235	0.8248	0.8242	0.9817
0.1254	16.2980	28000	0.4234	0.8257	0.8283	0.8270	0.9819
0.1265	16.5891	28500	0.4223	0.8236	0.8225	0.8231	0.9817
0.1288	16.8801	29000	0.4225	0.8332	0.8257	0.8294	0.9819
0.1259	17.1711	29500	0.4184	0.8225	0.8244	0.8235	0.9815
0.1233	17.4622	30000	0.4216	0.8310	0.8173	0.8241	0.9817
0.1243	17.7532	30500	0.4151	0.8238	0.8296	0.8267	0.9818
0.1222	18.0442	31000	0.4216	0.8290	0.8199	0.8245	0.9815
0.1207	18.3353	31500	0.4192	0.8286	0.8147	0.8216	0.9816
0.1202	18.6263	32000	0.4138	0.8308	0.8185	0.8246	0.9817
0.1209	18.9173	32500	0.4195	0.8318	0.8215	0.8267	0.9821
0.1196	19.2084	33000	0.4137	0.8356	0.8137	0.8245	0.9818
0.1184	19.4994	33500	0.4152	0.8262	0.8225	0.8244	0.9818
0.1177	19.7905	34000	0.4154	0.8333	0.8308	0.8320	0.9821
0.1155	20.0815	34500	0.4095	0.8281	0.8250	0.8265	0.9820
0.1168	20.3725	35000	0.4130	0.8326	0.8175	0.8250	0.9817
0.1147	20.6636	35500	0.4134	0.8264	0.8185	0.8224	0.9817
0.1153	20.9546	36000	0.4094	0.8306	0.8225	0.8265	0.9819
0.1144	21.2456	36500	0.4150	0.8303	0.8194	0.8248	0.9817
0.1121	21.5367	37000	0.4096	0.8283	0.8225	0.8254	0.9819
0.1135	21.8277	37500	0.4085	0.8311	0.8250	0.8280	0.9819
0.1137	22.1187	38000	0.4079	0.8351	0.8233	0.8291	0.9822
0.1098	22.4098	38500	0.4067	0.8273	0.8293	0.8283	0.9821
0.1117	22.7008	39000	0.4083	0.8278	0.8266	0.8272	0.9821
0.1105	22.9919	39500	0.4155	0.8321	0.8251	0.8286	0.9819
0.1101	23.2829	40000	0.4100	0.8301	0.8241	0.8271	0.9817
0.1087	23.5739	40500	0.4091	0.8285	0.8173	0.8229	0.9815
0.1094	23.8650	41000	0.4092	0.8292	0.8208	0.8250	0.9819
0.1081	24.1560	41500	0.4101	0.8355	0.8212	0.8283	0.9819
0.1071	24.4470	42000	0.4110	0.8319	0.8243	0.8281	0.9819
0.1089	24.7381	42500	0.4098	0.8287	0.8173	0.8230	0.9816
0.1073	25.0291	43000	0.4080	0.8301	0.8199	0.8250	0.9818
0.1068	25.3201	43500	0.4018	0.8303	0.8237	0.8270	0.9822
0.1069	25.6112	44000	0.4036	0.8328	0.8256	0.8292	0.9821

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1