TWON-Agent-OSN-Replies-en

This model is a fine-tuned version of meta-llama/Llama-3.2-3B-Instruct on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.3637

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: 4
• eval_batch_size: 8
• seed: 42
• 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
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss
2.368	0.0522	200	2.1934
2.1134	0.1044	400	2.0532
1.9845	0.1567	600	1.9324
1.8706	0.2089	800	1.8196
1.7668	0.2611	1000	1.7153
1.6587	0.3133	1200	1.6127
1.5714	0.3655	1400	1.5284
1.5041	0.4178	1600	1.4524
1.4018	0.4700	1800	1.3912
1.3586	0.5222	2000	1.3282
1.2971	0.5744	2200	1.2665
1.2328	0.6266	2400	1.2156
1.2013	0.6789	2600	1.1665
1.1608	0.7311	2800	1.1344
1.1356	0.7833	3000	1.0867
1.0931	0.8355	3200	1.0500
1.0772	0.8877	3400	1.0196
1.0253	0.9399	3600	0.9892
1.0009	0.9922	3800	0.9622
0.9576	1.0444	4000	0.9344
0.9387	1.0966	4200	0.9117
0.9155	1.1488	4400	0.8891
0.9048	1.2010	4600	0.8720
0.891	1.2533	4800	0.8484
0.8625	1.3055	5000	0.8312
0.8397	1.3577	5200	0.8145
0.8291	1.4099	5400	0.7975
0.8244	1.4621	5600	0.7829
0.7993	1.5144	5800	0.7690
0.7817	1.5666	6000	0.7566
0.7725	1.6188	6200	0.7474
0.7402	1.6710	6400	0.7307
0.7559	1.7232	6600	0.7171
0.7314	1.7755	6800	0.7045
0.7301	1.8277	7000	0.6943
0.7164	1.8799	7200	0.6842
0.703	1.9321	7400	0.6739
0.7109	1.9843	7600	0.6638
0.6618	2.0366	7800	0.6570
0.6837	2.0888	8000	0.6474
0.658	2.1410	8200	0.6409
0.6645	2.1932	8400	0.6323
0.629	2.2454	8600	0.6237
0.6531	2.2977	8800	0.6151
0.6174	2.3499	9000	0.6121
0.6328	2.4021	9200	0.6024
0.6247	2.4543	9400	0.5949
0.6077	2.5065	9600	0.5900
0.6057	2.5587	9800	0.5817
0.6168	2.6110	10000	0.5747
0.6108	2.6632	10200	0.5711
0.5945	2.7154	10400	0.5651
0.5994	2.7676	10600	0.5578
0.5847	2.8198	10800	0.5509
0.5659	2.8721	11000	0.5495
0.5796	2.9243	11200	0.5407
0.5515	2.9765	11400	0.5350
0.551	3.0287	11600	0.5301
0.5639	3.0809	11800	0.5262
0.542	3.1332	12000	0.5214
0.5302	3.1854	12200	0.5182
0.552	3.2376	12400	0.5139
0.5433	3.2898	12600	0.5101
0.5455	3.3420	12800	0.5027
0.5229	3.3943	13000	0.5013
0.525	3.4465	13200	0.4985
0.5069	3.4987	13400	0.4926
0.5256	3.5509	13600	0.4869
0.5003	3.6031	13800	0.4858
0.5091	3.6554	14000	0.4812
0.4872	3.7076	14200	0.4781
0.5028	3.7598	14400	0.4739
0.5059	3.8120	14600	0.4708
0.481	3.8642	14800	0.4680
0.5065	3.9164	15000	0.4641
0.4764	3.9687	15200	0.4606
0.4974	4.0209	15400	0.4579
0.465	4.0731	15600	0.4561
0.4721	4.1253	15800	0.4531
0.4813	4.1775	16000	0.4490
0.4599	4.2298	16200	0.4464
0.4457	4.2820	16400	0.4438
0.464	4.3342	16600	0.4433
0.4932	4.3864	16800	0.4402
0.4885	4.4386	17000	0.4364
0.4554	4.4909	17200	0.4341
0.4573	4.5431	17400	0.4312
0.4721	4.5953	17600	0.4320
0.4472	4.6475	17800	0.4276
0.4542	4.6997	18000	0.4258
0.4455	4.7520	18200	0.4233
0.441	4.8042	18400	0.4221
0.4475	4.8564	18600	0.4203
0.4273	4.9086	18800	0.4170
0.4332	4.9608	19000	0.4149
0.4353	5.0131	19200	0.4136
0.4296	5.0653	19400	0.4119
0.424	5.1175	19600	0.4099
0.4238	5.1697	19800	0.4087
0.4354	5.2219	20000	0.4062
0.42	5.2742	20200	0.4058
0.4409	5.3264	20400	0.4039
0.4403	5.3786	20600	0.4012
0.4223	5.4308	20800	0.4004
0.4218	5.4830	21000	0.3995
0.4327	5.5352	21200	0.3994
0.4366	5.5875	21400	0.3956
0.4212	5.6397	21600	0.3942
0.4026	5.6919	21800	0.3951
0.4009	5.7441	22000	0.3927
0.4135	5.7963	22200	0.3914
0.397	5.8486	22400	0.3899
0.4191	5.9008	22600	0.3891
0.4093	5.9530	22800	0.3871
0.4135	6.0052	23000	0.3859
0.3993	6.0574	23200	0.3852
0.3874	6.1097	23400	0.3842
0.4121	6.1619	23600	0.3828
0.4175	6.2141	23800	0.3818
0.4044	6.2663	24000	0.3807
0.4004	6.3185	24200	0.3794
0.4104	6.3708	24400	0.3783
0.3962	6.4230	24600	0.3777
0.4003	6.4752	24800	0.3767
0.3887	6.5274	25000	0.3761
0.39	6.5796	25200	0.3753
0.3916	6.6319	25400	0.3745
0.4046	6.6841	25600	0.3740
0.4064	6.7363	25800	0.3725
0.3959	6.7885	26000	0.3736
0.3842	6.8407	26200	0.3718
0.3827	6.8930	26400	0.3710
0.3936	6.9452	26600	0.3707
0.3926	6.9974	26800	0.3701
0.3807	7.0496	27000	0.3696
0.3875	7.1018	27200	0.3691
0.3847	7.1540	27400	0.3685
0.3848	7.2063	27600	0.3678
0.3993	7.2585	27800	0.3678
0.3913	7.3107	28000	0.3674
0.3818	7.3629	28200	0.3668
0.3773	7.4151	28400	0.3664
0.3609	7.4674	28600	0.3658
0.3803	7.5196	28800	0.3658
0.3891	7.5718	29000	0.3652
0.3812	7.6240	29200	0.3650
0.3814	7.6762	29400	0.3648
0.3971	7.7285	29600	0.3645
0.3794	7.7807	29800	0.3641
0.3904	7.8329	30000	0.3641
0.4003	7.8851	30200	0.3638
0.3805	7.9373	30400	0.3638
0.3807	7.9896	30600	0.3637

Framework versions

• PEFT 0.12.0
• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0