suyashmittal/lnm-ner

model

This model is a fine-tuned version of distilbert-base-uncased on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 1.0789
• Precision: 0.2067
• Recall: 0.2760
• F1: 0.2364
• Accuracy: 0.8146

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.02	24	1.7564	0.0923	0.0081	0.0149	0.8092
No log	1.02	48	1.5458	0.0060	0.0014	0.0022	0.8020
No log	2.02	72	1.3843	0.0444	0.0135	0.0207	0.8034
No log	3.02	96	1.2963	0.08	0.0379	0.0514	0.8087
No log	4.02	120	1.2395	0.1008	0.0663	0.0800	0.8143
No log	5.02	144	1.1908	0.1270	0.0974	0.1103	0.8169
No log	6.02	168	1.1559	0.1467	0.1191	0.1314	0.8164
No log	7.02	192	1.1295	0.1553	0.1380	0.1461	0.8188
No log	8.02	216	1.1157	0.1659	0.1556	0.1606	0.8185
No log	9.02	240	1.1168	0.1808	0.1705	0.1755	0.8211
No log	10.02	264	1.1058	0.1917	0.1813	0.1864	0.8244
No log	11.02	288	1.1112	0.1865	0.1759	0.1811	0.8257
No log	12.02	312	1.1226	0.2072	0.1867	0.1964	0.8319
No log	13.02	336	1.1214	0.2196	0.1786	0.1970	0.8340
No log	14.02	360	1.0690	0.2087	0.1827	0.1948	0.8287
No log	15.02	384	1.0238	0.2154	0.2233	0.2193	0.8211
No log	16.02	408	1.0150	0.2074	0.2355	0.2205	0.8147
No log	17.02	432	1.0123	0.2082	0.2395	0.2228	0.8157
No log	18.02	456	1.0150	0.2078	0.2449	0.2248	0.8164
No log	19.02	480	1.0192	0.2050	0.2436	0.2226	0.8174
0.9784	20.02	504	1.0199	0.1980	0.2449	0.2190	0.8163
0.9784	21.02	528	1.0164	0.1987	0.2544	0.2231	0.8122
0.9784	22.02	552	1.0212	0.2012	0.2625	0.2278	0.8121
0.9784	23.02	576	1.0216	0.1998	0.2625	0.2269	0.8095
0.9784	24.02	600	1.0302	0.2010	0.2666	0.2292	0.8099
0.9784	25.02	624	1.0324	0.1958	0.2652	0.2253	0.8092
0.9784	26.02	648	1.0304	0.1984	0.2733	0.2299	0.8060
0.9784	27.02	672	1.0413	0.2016	0.2760	0.2330	0.8105
0.9784	28.02	696	1.0346	0.1993	0.2882	0.2356	0.8048
0.9784	29.02	720	1.0302	0.1902	0.2950	0.2313	0.7904
0.9784	30.02	744	1.0339	0.1967	0.3031	0.2386	0.7963
0.9784	31.02	768	1.0417	0.2028	0.2909	0.2390	0.8006
0.9784	32.02	792	1.0473	0.2029	0.2869	0.2377	0.8058
0.9784	33.02	816	1.0591	0.2041	0.2828	0.2371	0.8125
0.9784	34.02	840	1.0573	0.1992	0.2869	0.2352	0.8063
0.9784	35.02	864	1.0681	0.2037	0.2801	0.2359	0.8135
0.9784	36.02	888	1.0676	0.1990	0.2788	0.2322	0.8108
0.9784	37.02	912	1.0744	0.2036	0.2760	0.2343	0.8139
0.9784	38.02	936	1.0738	0.2017	0.2815	0.2350	0.8128
0.9784	39.02	960	1.0791	0.2083	0.2774	0.2380	0.8155
0.9784	40.02	984	1.0784	0.2051	0.2747	0.2348	0.8143
0.5145	41.02	1000	1.0789	0.2067	0.2760	0.2364	0.8146

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0