GSM8K-Binary_Llama-3.2-1B-bpzb2ktm

This model is a fine-tuned version of meta-llama/Llama-3.2-1B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.3525
• Model Preparation Time: 0.0056
• Mdl: 4829.4638
• Accumulated Loss: 3347.5292
• Correct Preds: 2001.0
• Total Preds: 2475.0
• Accuracy: 0.8085
• Correct Gen Preds: 2001.0
• Gen Accuracy: 0.8085
• Correct Gen Preds 34192: 1028.0
• Correct Preds 34192: 1033.0
• Total Labels 34192: 1196.0
• Accuracy 34192: 0.8637
• Gen Accuracy 34192: 0.8595
• Correct Gen Preds 41568: 965.0
• Correct Preds 41568: 968.0
• Total Labels 41568: 1267.0
• Accuracy 41568: 0.7640
• Gen Accuracy 41568: 0.7616

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 64
• 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: 0.01
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Mdl	Accumulated Loss	Correct Preds	Total Preds	Accuracy	Correct Gen Preds	Gen Accuracy	Correct Gen Preds 34192	Correct Preds 34192	Total Labels 34192	Accuracy 34192	Gen Accuracy 34192	Correct Gen Preds 41568	Correct Preds 41568	Total Labels 41568	Accuracy 41568	Gen Accuracy 41568
No log	0	0	1.4656	0.0056	5233.1723	3627.3586	1196.0	2475.0	0.4832	1204.0	0.4865	1196.0	1196.0	1196.0	1.0	1.0	0.0	0.0	1267.0	0.0	0.0
0.4643	1.0	32	0.6079	0.0056	2170.4901	1504.4691	1770.0	2475.0	0.7152	1776.0	0.7176	821.0	821.0	1196.0	0.6865	0.6865	947.0	949.0	1267.0	0.7490	0.7474
0.5059	2.0	64	0.5626	0.0056	2008.7864	1392.3846	1839.0	2475.0	0.7430	10.0	0.0040	0.0	1091.0	1196.0	0.9122	0.0	1.0	748.0	1267.0	0.5904	0.0008
0.4259	3.0	96	0.6197	0.0056	2212.8014	1533.7970	1859.0	2475.0	0.7511	176.0	0.0711	0.0	1134.0	1196.0	0.9482	0.0	168.0	725.0	1267.0	0.5722	0.1326
0.5362	4.0	128	0.6927	0.0056	2473.5441	1714.5301	1919.0	2475.0	0.7754	508.0	0.2053	79.0	832.0	1196.0	0.6957	0.0661	420.0	1087.0	1267.0	0.8579	0.3315
0.1605	5.0	160	0.8818	0.0056	3148.5150	2182.3843	1925.0	2475.0	0.7778	1554.0	0.6279	667.0	892.0	1196.0	0.7458	0.5577	878.0	1033.0	1267.0	0.8153	0.6930
0.0332	6.0	192	1.1003	0.0056	3928.6770	2723.1514	1914.0	2475.0	0.7733	1550.0	0.6263	870.0	1085.0	1196.0	0.9072	0.7274	670.0	829.0	1267.0	0.6543	0.5288
0.1398	7.0	224	1.2557	0.0056	4483.5623	3107.7686	1896.0	2475.0	0.7661	1728.0	0.6982	675.0	777.0	1196.0	0.6497	0.5644	1043.0	1119.0	1267.0	0.8832	0.8232
0.6464	8.0	256	1.4174	0.0056	5061.1064	3508.0916	1859.0	2475.0	0.7511	1796.0	0.7257	690.0	734.0	1196.0	0.6137	0.5769	1097.0	1125.0	1267.0	0.8879	0.8658
0.0	9.0	288	1.4300	0.0056	5105.9013	3539.1411	1987.0	2475.0	0.8028	1989.0	0.8036	1058.0	1062.0	1196.0	0.8880	0.8846	922.0	925.0	1267.0	0.7301	0.7277
0.0003	10.0	320	1.2444	0.0056	4443.1854	3079.7814	1983.0	2475.0	0.8012	1968.0	0.7952	1022.0	1036.0	1196.0	0.8662	0.8545	938.0	947.0	1267.0	0.7474	0.7403
0.941	11.0	352	1.3525	0.0056	4829.4638	3347.5292	2001.0	2475.0	0.8085	2001.0	0.8085	1028.0	1033.0	1196.0	0.8637	0.8595	965.0	968.0	1267.0	0.7640	0.7616
0.0	12.0	384	1.3653	0.0056	4874.9553	3379.0615	2000.0	2475.0	0.8081	2001.0	0.8085	1030.0	1035.0	1196.0	0.8654	0.8612	963.0	965.0	1267.0	0.7616	0.7601
0.0	13.0	416	1.3662	0.0056	4878.3502	3381.4147	1999.0	2475.0	0.8077	2001.0	0.8085	1030.0	1034.0	1196.0	0.8645	0.8612	963.0	965.0	1267.0	0.7616	0.7601
0.0	14.0	448	1.3674	0.0056	4882.3632	3384.1963	1998.0	2475.0	0.8073	1998.0	0.8073	1029.0	1034.0	1196.0	0.8645	0.8604	961.0	964.0	1267.0	0.7609	0.7585
0.0	15.0	480	1.3707	0.0056	4894.2185	3392.4138	1997.0	2475.0	0.8069	1999.0	0.8077	1030.0	1034.0	1196.0	0.8645	0.8612	961.0	963.0	1267.0	0.7601	0.7585
0.4705	16.0	512	1.3717	0.0056	4898.0271	3395.0537	1998.0	2475.0	0.8073	2001.0	0.8085	1029.0	1033.0	1196.0	0.8637	0.8604	964.0	965.0	1267.0	0.7616	0.7609
0.0	17.0	544	1.3733	0.0056	4903.4386	3398.8046	1996.0	2475.0	0.8065	1999.0	0.8077	1031.0	1034.0	1196.0	0.8645	0.8620	960.0	962.0	1267.0	0.7593	0.7577
0.0	18.0	576	1.3761	0.0056	4913.6611	3405.8903	1996.0	2475.0	0.8065	1997.0	0.8069	1030.0	1035.0	1196.0	0.8654	0.8612	959.0	961.0	1267.0	0.7585	0.7569
0.0	19.0	608	1.3771	0.0056	4917.0501	3408.2394	1998.0	2475.0	0.8073	2000.0	0.8081	1030.0	1034.0	1196.0	0.8645	0.8612	962.0	964.0	1267.0	0.7609	0.7593
0.4705	20.0	640	1.3790	0.0056	4923.8825	3412.9753	2001.0	2475.0	0.8085	2003.0	0.8093	1033.0	1037.0	1196.0	0.8671	0.8637	962.0	964.0	1267.0	0.7609	0.7593
0.0	21.0	672	1.3786	0.0056	4922.3614	3411.9210	1996.0	2475.0	0.8065	1999.0	0.8077	1032.0	1035.0	1196.0	0.8654	0.8629	959.0	961.0	1267.0	0.7585	0.7569
0.0	22.0	704	1.3813	0.0056	4932.1419	3418.7003	1997.0	2475.0	0.8069	2000.0	0.8081	1033.0	1036.0	1196.0	0.8662	0.8637	959.0	961.0	1267.0	0.7585	0.7569
0.0	23.0	736	1.3823	0.0056	4935.9056	3421.3091	1995.0	2475.0	0.8061	1999.0	0.8077	1033.0	1035.0	1196.0	0.8654	0.8637	958.0	960.0	1267.0	0.7577	0.7561
0.0	24.0	768	1.3830	0.0056	4938.2231	3422.9154	1997.0	2475.0	0.8069	1999.0	0.8077	1032.0	1036.0	1196.0	0.8662	0.8629	959.0	961.0	1267.0	0.7585	0.7569
0.0	25.0	800	1.3842	0.0056	4942.4284	3425.8303	1998.0	2475.0	0.8073	2000.0	0.8081	1032.0	1036.0	1196.0	0.8662	0.8629	960.0	962.0	1267.0	0.7593	0.7577
0.0	26.0	832	1.3851	0.0056	4945.7223	3428.1134	1996.0	2475.0	0.8065	1998.0	0.8073	1032.0	1036.0	1196.0	0.8662	0.8629	958.0	960.0	1267.0	0.7577	0.7561
0.4705	27.0	864	1.3860	0.0056	4948.9212	3430.3308	1998.0	2475.0	0.8073	1999.0	0.8077	1032.0	1037.0	1196.0	0.8671	0.8629	959.0	961.0	1267.0	0.7585	0.7569

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1