learning_rate: 5e-05
train_batch_size: 8
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Use 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae
8.5061	0.3429	300	1.7282	1.0945	0.8522
8.1139	0.6857	600	1.6352	1.0425	0.8249
7.7043	1.0286	900	1.5126	0.9853	0.8051
6.9404	1.3714	1200	1.3621	0.8932	0.7625
6.6003	1.7143	1500	1.2822	0.8509	0.7413
5.8303	2.0571	1800	1.1538	0.7647	0.7005
5.4533	2.4	2100	1.1136	0.7337	0.6812
5.0087	2.7429	2400	1.0895	0.7149	0.6684
4.8301	3.0857	2700	1.0786	0.7060	0.6610
4.676	3.4286	3000	1.0654	0.6971	0.6550
4.6187	3.7714	3300	1.0473	0.6842	0.6447
4.3955	4.1143	3600	1.0129	0.6608	0.6308
4.4388	4.4571	3900	1.0255	0.6688	0.6355
4.3574	4.8	4200	0.9843	0.6405	0.6154
4.384	5.1429	4500	0.9835	0.6403	0.6153
4.4066	5.4857	4800	1.0027	0.6535	0.6256
4.4052	5.8286	5100	0.9643	0.6275	0.6050
4.3846	6.1714	5400	0.9507	0.6175	0.5982
4.31	6.5143	5700	0.9536	0.6193	0.5989
4.2701	6.8571	6000	0.9481	0.6159	0.5943
4.2284	7.2	6300	0.9181	0.5951	0.5815
4.1039	7.5429	6600	0.9036	0.5857	0.5731
4.2027	7.8857	6900	0.8906	0.5758	0.5655
4.1092	8.2286	7200	0.8899	0.5761	0.5653
4.213	8.5714	7500	0.8903	0.5778	0.5658
4.1775	8.9143	7800	0.8766	0.5663	0.5582
4.1293	9.2571	8100	0.8891	0.5739	0.5637
4.135	9.6	8400	0.8823	0.5701	0.5599
4.1324	9.9429	8700	0.8633	0.5590	0.5546
4.0383	10.2857	9000	0.8562	0.5524	0.5462
4.0424	10.6286	9300	0.8390	0.5403	0.5376
4.0851	10.9714	9600	0.8455	0.5446	0.5421
3.9225	11.3143	9900	0.8190	0.5283	0.5305
4.0771	11.6571	10200	0.8195	0.5261	0.5268
3.9188	12.0	10500	0.8112	0.5218	0.5233
3.9961	12.3429	10800	0.8157	0.5247	0.5241
3.9793	12.6857	11100	0.7877	0.5059	0.5097
4.0327	13.0286	11400	0.8180	0.5256	0.5242
3.9755	13.3714	11700	0.8043	0.5173	0.5183
3.9531	13.7143	12000	0.7860	0.5035	0.5060
3.9182	14.0571	12300	0.7914	0.5070	0.5123
3.8201	14.4	12600	0.7873	0.5033	0.5095
3.8887	14.7429	12900	0.7868	0.5039	0.5077
3.9021	15.0857	13200	0.7736	0.4964	0.5016
3.8241	15.4286	13500	0.7677	0.4903	0.5002
3.8387	15.7714	13800	0.7751	0.4953	0.5025
3.8344	16.1143	14100	0.7802	0.5004	0.5023
3.8173	16.4571	14400	0.7588	0.4853	0.4958
3.9998	16.8	14700	0.7556	0.4828	0.4915
3.838	17.1429	15000	0.7690	0.4916	0.4988
3.7549	17.4857	15300	0.7545	0.4819	0.4920
3.8553	17.8286	15600	0.7901	0.5058	0.5077
3.7462	18.1714	15900	0.7558	0.4829	0.4947
3.7608	18.5143	16200	0.7518	0.4796	0.4890
3.9487	18.8571	16500	0.7590	0.4836	0.4915
3.9489	19.2	16800	0.7278	0.4639	0.4758
3.8653	19.5429	17100	0.7498	0.4786	0.4914
3.7533	19.8857	17400	0.7641	0.4879	0.4952

Framework versions

Transformers 4.56.1
Pytorch 2.7.0
Datasets 4.0.0
Tokenizers 0.22.0

eternis/eternis_router_encoder_sft_9Sep

eternis_router_encoder_sft_9Sep

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

eternis/
eternis_router_encoder_sft_9Sep