bge_large_cn_new_prompt_llama3_70

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9517
• Precision: 0.4483
• Recall: 0.3656
• F1 Macro: 0.3577
• Accuracy: 0.4617

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: 0.0003
• train_batch_size: 256
• eval_batch_size: 128
• seed: 0
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1 Macro	Accuracy
No log	0	0	10.4077	0.1217	0.1667	0.0201	0.0638
1.01	0.3672	1000	1.0195	0.4334	0.3321	0.3218	0.4338
1.0013	0.7345	2000	0.9928	0.4524	0.3418	0.3300	0.4429
0.9922	1.1017	3000	0.9758	0.4591	0.3556	0.3460	0.4574
0.9353	1.4690	4000	0.9716	0.4607	0.3557	0.3431	0.4597
0.9501	1.8362	5000	0.9638	0.4765	0.3546	0.3402	0.4627
0.9574	2.2035	6000	0.9628	0.4630	0.3502	0.3320	0.4527
0.9324	2.5707	7000	0.9755	0.4565	0.3592	0.3448	0.4471
0.9373	2.9379	8000	0.9504	0.4749	0.3635	0.3503	0.4694
0.9303	3.3052	9000	0.9510	0.4809	0.3638	0.3516	0.4702
0.9453	3.6724	10000	0.9519	0.4622	0.3619	0.3472	0.4588
0.9278	4.0397	11000	0.9469	0.4678	0.3604	0.3453	0.4591
0.9246	4.4069	12000	0.9440	0.4731	0.3634	0.3476	0.4628
0.9229	4.7741	13000	0.9423	0.4778	0.3666	0.3539	0.4704
0.9105	5.1414	14000	0.9388	0.4717	0.3655	0.3511	0.4647
0.8994	5.5086	15000	0.9471	0.4807	0.3638	0.3517	0.4707
0.9138	5.8759	16000	0.9412	0.4768	0.3620	0.3489	0.4692
0.9157	6.2431	17000	0.9393	0.4703	0.3638	0.3495	0.4631
0.9074	6.6104	18000	0.9372	0.4699	0.3676	0.3553	0.4679
0.9048	6.9776	19000	0.9378	0.4671	0.3688	0.3575	0.4690
0.8827	7.3448	20000	0.9421	0.4603	0.3673	0.3550	0.4628
0.888	7.7121	21000	0.9384	0.4632	0.3675	0.3584	0.4655
0.8701	8.0793	22000	0.9449	0.4617	0.3665	0.3535	0.4614
0.8808	8.4466	23000	0.9382	0.4595	0.3642	0.3517	0.4668
0.901	8.8138	24000	0.9436	0.4693	0.3654	0.3554	0.4682
0.8729	9.1811	25000	0.9459	0.4624	0.3625	0.3483	0.4580
0.8798	9.5483	26000	0.9459	0.4546	0.3665	0.3541	0.4581
0.8745	9.9155	27000	0.9518	0.4633	0.3688	0.3623	0.4700
0.8484	10.2828	28000	0.9426	0.4568	0.3655	0.3543	0.4623
0.8654	10.6500	29000	0.9448	0.4689	0.3659	0.3560	0.4680
0.8543	11.0173	30000	0.9421	0.4614	0.3625	0.3507	0.4645
0.8446	11.3845	31000	0.9451	0.4600	0.3684	0.3606	0.4671
0.8589	11.7517	32000	0.9414	0.4597	0.3672	0.3578	0.4644
0.8201	12.1190	33000	0.9469	0.4568	0.3641	0.3533	0.4587
0.8303	12.4862	34000	0.9527	0.4603	0.3652	0.3578	0.4663
0.8391	12.8535	35000	0.9466	0.4568	0.3647	0.3532	0.4601
0.8393	13.2207	36000	0.9489	0.4518	0.3612	0.3488	0.4555
0.8388	13.5880	37000	0.9463	0.4575	0.3639	0.3536	0.4628
0.8396	13.9552	38000	0.9466	0.4554	0.3663	0.3557	0.4635
0.8296	14.3224	39000	0.9488	0.4555	0.3665	0.3575	0.4661
0.8351	14.6897	40000	0.9500	0.4523	0.3675	0.3610	0.4635
0.8362	15.0569	41000	0.9500	0.4585	0.3645	0.3559	0.4637
0.8174	15.4242	42000	0.9510	0.4547	0.3700	0.3634	0.4646
0.8249	15.7914	43000	0.9541	0.4513	0.3670	0.3605	0.4642
0.8057	16.1586	44000	0.9504	0.4519	0.3646	0.3562	0.4604
0.8145	16.5259	45000	0.9539	0.4537	0.3667	0.3595	0.4641
0.8237	16.8931	46000	0.9515	0.4524	0.3662	0.3582	0.4622
0.8082	17.2604	47000	0.9515	0.4484	0.3646	0.3547	0.4588
0.8249	17.6276	48000	0.9512	0.4488	0.3651	0.3573	0.4608
0.8074	17.9949	49000	0.9507	0.4498	0.3662	0.3584	0.4610
0.7944	18.3621	50000	0.9512	0.4516	0.3649	0.3554	0.4605
0.8041	18.7293	51000	0.9520	0.4515	0.3660	0.3581	0.4623
0.7958	19.0966	52000	0.9528	0.4497	0.3651	0.3574	0.4617
0.8173	19.4638	53000	0.9518	0.4511	0.3644	0.3558	0.4615
0.7909	19.8311	54000	0.9517	0.4483	0.3656	0.3577	0.4617

Framework versions

• Transformers 4.43.3
• Pytorch 2.4.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1