bge-base-en-v1.5-2024-12-05_13-34-00

This model is a fine-tuned version of BAAI/bge-base-en-v1.5 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0228
• Spearman: 0.9285
• Pearson: 0.9285
• Mse: 0.0228

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• 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: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Spearman	Pearson	Mse
0.0525	0.1000	2110	0.0402	0.8683	0.8725	0.0402
0.0351	0.1999	4220	0.0424	0.8824	0.8816	0.0424
0.0314	0.2999	6330	0.0355	0.8829	0.8860	0.0355
0.038	0.3999	8440	0.0339	0.8940	0.8932	0.0339
0.0287	0.4998	10550	0.0318	0.8959	0.8982	0.0318
0.027	0.5998	12660	0.0308	0.8986	0.9010	0.0308
0.0376	0.6998	14770	0.0309	0.9019	0.9038	0.0309
0.0264	0.7997	16880	0.0301	0.9017	0.9049	0.0301
0.0311	0.8997	18990	0.0294	0.9030	0.9070	0.0294
0.0303	0.9997	21100	0.0286	0.9052	0.9085	0.0286
0.0252	1.0996	23210	0.0290	0.9085	0.9093	0.0290
0.0312	1.1996	25320	0.0287	0.9084	0.9093	0.0287
0.0238	1.2996	27430	0.0277	0.9095	0.9132	0.0277
0.0326	1.3995	29540	0.0295	0.9089	0.9096	0.0295
0.0204	1.4995	31650	0.0272	0.9104	0.9132	0.0272
0.0237	1.5995	33760	0.0304	0.9120	0.9099	0.0304
0.0285	1.6994	35870	0.0263	0.9128	0.9168	0.0263
0.0218	1.7994	37980	0.0262	0.9152	0.9185	0.0262
0.032	1.8994	40090	0.0259	0.9149	0.9186	0.0259
0.0211	1.9993	42200	0.0256	0.9155	0.9197	0.0256
0.0209	2.0993	44310	0.0253	0.9174	0.9190	0.0253
0.016	2.1993	46420	0.0259	0.9180	0.9194	0.0259
0.0122	2.2992	48530	0.0257	0.9181	0.9211	0.0257
0.0147	2.3992	50640	0.0276	0.9205	0.9210	0.0276
0.015	2.4992	52750	0.0253	0.9196	0.9223	0.0253
0.0201	2.5991	54860	0.0243	0.9208	0.9238	0.0243
0.0137	2.6991	56970	0.0243	0.9214	0.9232	0.0243
0.0158	2.7991	59080	0.0239	0.9224	0.9250	0.0239
0.018	2.8990	61190	0.0238	0.9234	0.9258	0.0238
0.0175	2.9990	63300	0.0234	0.9231	0.9264	0.0234
0.0122	3.0990	65410	0.0234	0.9241	0.9265	0.0234
0.0107	3.1989	67520	0.0238	0.9241	0.9264	0.0238
0.0081	3.2989	69630	0.0238	0.9248	0.9264	0.0238
0.0093	3.3989	71740	0.0233	0.9251	0.9270	0.0233
0.0128	3.4988	73850	0.0229	0.9258	0.9279	0.0229
0.0106	3.5988	75960	0.0231	0.9260	0.9281	0.0231
0.0134	3.6988	78070	0.0230	0.9261	0.9284	0.0230
0.0087	3.7987	80180	0.0227	0.9269	0.9295	0.0227
0.0086	3.8987	82290	0.0228	0.9267	0.9290	0.0228
0.0101	3.9987	84400	0.0225	0.9271	0.9294	0.0225
0.0075	4.0986	86510	0.0227	0.9271	0.9290	0.0227
0.0086	4.1986	88620	0.0225	0.9273	0.9295	0.0225
0.0067	4.2986	90730	0.0227	0.9277	0.9293	0.0227
0.009	4.3985	92840	0.0224	0.9275	0.9297	0.0224
0.007	4.4985	94950	0.0225	0.9278	0.9297	0.0225
0.0066	4.5985	97060	0.0225	0.9280	0.9300	0.0225
0.0118	4.6984	99170	0.0225	0.9279	0.9298	0.0225
0.0071	4.7984	101280	0.0225	0.9279	0.9299	0.0225
0.0071	4.8984	103390	0.0225	0.9280	0.9299	0.0225
0.0087	4.9983	105500	0.0225	0.9280	0.9299	0.0225

Framework versions

• Transformers 4.46.3
• Pytorch 2.4.1+cu121
• Datasets 2.19.2
• Tokenizers 0.20.3