GSM8K-Binary_Llama-3.2-1B-8kwse8de

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: 1.4787
• Model Preparation Time: 0.0059
• Mdl: 5279.8389
• Accumulated Loss: 3659.7055
• Correct Preds: 1822.0
• Total Preds: 2475.0
• Accuracy: 0.7362
• Correct Gen Preds: 1743.0
• Gen Accuracy: 0.7042
• Correct Gen Preds 34192: 834.0
• Correct Preds 34192: 870.0
• Total Labels 34192: 1196.0
• Accuracy 34192: 0.7274
• Gen Accuracy 34192: 0.6973
• Correct Gen Preds 41568: 900.0
• Correct Preds 41568: 952.0
• Total Labels 41568: 1267.0
• Accuracy 41568: 0.7514
• Gen Accuracy 41568: 0.7103

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	0.0059	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
0.7404	1.0	5	0.7519	0.0059	2684.8085	1860.9674	1301.0	2475.0	0.5257	9.0	0.0036	0.0	97.0	1196.0	0.0811	0.0	1.0	1204.0	1267.0	0.9503	0.0008
1.4345	2.0	10	0.6475	0.0059	2312.1054	1602.6293	1678.0	2475.0	0.6780	8.0	0.0032	0.0	926.0	1196.0	0.7742	0.0	0.0	752.0	1267.0	0.5935	0.0
0.3056	3.0	15	0.6249	0.0059	2231.2720	1546.5999	1767.0	2475.0	0.7139	9.0	0.0036	1.0	906.0	1196.0	0.7575	0.0008	0.0	861.0	1267.0	0.6796	0.0
0.3324	4.0	20	0.6716	0.0059	2398.2346	1662.3295	1794.0	2475.0	0.7248	125.0	0.0505	9.0	831.0	1196.0	0.6948	0.0075	108.0	963.0	1267.0	0.7601	0.0852
0.7534	5.0	25	1.2676	0.0059	4526.0621	3137.2272	1499.0	2475.0	0.6057	932.0	0.3766	82.0	267.0	1196.0	0.2232	0.0686	842.0	1232.0	1267.0	0.9724	0.6646
0.2081	6.0	30	1.5980	0.0059	5705.7968	3954.9570	1505.0	2475.0	0.6081	703.0	0.2840	618.0	1175.0	1196.0	0.9824	0.5167	77.0	330.0	1267.0	0.2605	0.0608
0.082	7.0	35	1.1486	0.0059	4101.3733	2842.8553	1612.0	2475.0	0.6513	992.0	0.4008	120.0	449.0	1196.0	0.3754	0.1003	863.0	1163.0	1267.0	0.9179	0.6811
0.6616	8.0	40	1.2311	0.0059	4395.8751	3046.9884	1779.0	2475.0	0.7188	1492.0	0.6028	826.0	1015.0	1196.0	0.8487	0.6906	657.0	764.0	1267.0	0.6030	0.5185
0.0017	9.0	45	1.6432	0.0059	5867.3174	4066.9145	1756.0	2475.0	0.7095	1610.0	0.6505	923.0	1023.0	1196.0	0.8554	0.7717	678.0	733.0	1267.0	0.5785	0.5351
0.0001	10.0	50	2.1381	0.0059	7634.3190	5291.7067	1718.0	2475.0	0.6941	1546.0	0.6246	983.0	1082.0	1196.0	0.9047	0.8219	554.0	636.0	1267.0	0.5020	0.4373
0.0001	11.0	55	1.4472	0.0059	5167.3448	3581.7305	1813.0	2475.0	0.7325	1610.0	0.6505	792.0	898.0	1196.0	0.7508	0.6622	809.0	915.0	1267.0	0.7222	0.6385
0.0	12.0	60	1.4471	0.0059	5167.1333	3581.5839	1815.0	2475.0	0.7333	1670.0	0.6747	770.0	835.0	1196.0	0.6982	0.6438	891.0	980.0	1267.0	0.7735	0.7032
0.0	13.0	65	1.4645	0.0059	5229.3257	3624.6924	1820.0	2475.0	0.7354	1726.0	0.6974	812.0	852.0	1196.0	0.7124	0.6789	905.0	968.0	1267.0	0.7640	0.7143
1.3069	14.0	70	1.4787	0.0059	5279.8389	3659.7055	1822.0	2475.0	0.7362	1743.0	0.7042	834.0	870.0	1196.0	0.7274	0.6973	900.0	952.0	1267.0	0.7514	0.7103
0.6534	15.0	75	1.4931	0.0059	5331.4229	3695.4608	1820.0	2475.0	0.7354	1757.0	0.7099	859.0	888.0	1196.0	0.7425	0.7182	889.0	932.0	1267.0	0.7356	0.7017
0.6535	16.0	80	1.5030	0.0059	5366.7260	3719.9310	1818.0	2475.0	0.7345	1766.0	0.7135	869.0	893.0	1196.0	0.7467	0.7266	888.0	925.0	1267.0	0.7301	0.7009
0.0	17.0	85	1.5122	0.0059	5399.3942	3742.5749	1820.0	2475.0	0.7354	1767.0	0.7139	874.0	898.0	1196.0	0.7508	0.7308	884.0	922.0	1267.0	0.7277	0.6977
0.0	18.0	90	1.5168	0.0059	5415.9772	3754.0693	1822.0	2475.0	0.7362	1772.0	0.7160	879.0	902.0	1196.0	0.7542	0.7349	884.0	920.0	1267.0	0.7261	0.6977
0.0	19.0	95	1.5232	0.0059	5438.9175	3769.9703	1822.0	2475.0	0.7362	1774.0	0.7168	881.0	903.0	1196.0	0.7550	0.7366	884.0	919.0	1267.0	0.7253	0.6977
0.0	20.0	100	1.5241	0.0059	5442.2286	3772.2654	1819.0	2475.0	0.7349	1771.0	0.7156	884.0	905.0	1196.0	0.7567	0.7391	878.0	914.0	1267.0	0.7214	0.6930
0.0	21.0	105	1.5278	0.0059	5455.2160	3781.2676	1821.0	2475.0	0.7358	1778.0	0.7184	884.0	905.0	1196.0	0.7567	0.7391	885.0	916.0	1267.0	0.7230	0.6985
0.6535	22.0	110	1.5296	0.0059	5461.6471	3785.7253	1819.0	2475.0	0.7349	1776.0	0.7176	887.0	907.0	1196.0	0.7584	0.7416	880.0	912.0	1267.0	0.7198	0.6946
0.0	23.0	115	1.5328	0.0059	5473.0012	3793.5954	1821.0	2475.0	0.7358	1782.0	0.72	888.0	907.0	1196.0	0.7584	0.7425	885.0	914.0	1267.0	0.7214	0.6985
0.0	24.0	120	1.5339	0.0059	5477.0890	3796.4288	1821.0	2475.0	0.7358	1778.0	0.7184	889.0	910.0	1196.0	0.7609	0.7433	880.0	911.0	1267.0	0.7190	0.6946
1.3069	25.0	125	1.5357	0.0059	5483.3601	3800.7756	1818.0	2475.0	0.7345	1777.0	0.7180	886.0	907.0	1196.0	0.7584	0.7408	882.0	911.0	1267.0	0.7190	0.6961
0.0	26.0	130	1.5390	0.0059	5495.1006	3808.9135	1820.0	2475.0	0.7354	1779.0	0.7188	888.0	909.0	1196.0	0.7600	0.7425	882.0	911.0	1267.0	0.7190	0.6961
0.6534	27.0	135	1.5373	0.0059	5489.3342	3804.9165	1820.0	2475.0	0.7354	1782.0	0.72	889.0	908.0	1196.0	0.7592	0.7433	884.0	912.0	1267.0	0.7198	0.6977
0.0	28.0	140	1.5419	0.0059	5505.6494	3816.2253	1822.0	2475.0	0.7362	1780.0	0.7192	890.0	911.0	1196.0	0.7617	0.7441	881.0	911.0	1267.0	0.7190	0.6953
0.0	29.0	145	1.5433	0.0059	5510.5924	3819.6516	1821.0	2475.0	0.7358	1779.0	0.7188	889.0	910.0	1196.0	0.7609	0.7433	881.0	911.0	1267.0	0.7190	0.6953
0.0	30.0	150	1.5439	0.0059	5512.6644	3821.0878	1819.0	2475.0	0.7349	1777.0	0.7180	889.0	909.0	1196.0	0.7600	0.7433	879.0	910.0	1267.0	0.7182	0.6938
0.0	31.0	155	1.5443	0.0059	5514.1591	3822.1238	1820.0	2475.0	0.7354	1781.0	0.7196	890.0	911.0	1196.0	0.7617	0.7441	882.0	909.0	1267.0	0.7174	0.6961
0.6534	32.0	160	1.5471	0.0059	5524.2001	3829.0837	1820.0	2475.0	0.7354	1776.0	0.7176	891.0	912.0	1196.0	0.7625	0.7450	876.0	908.0	1267.0	0.7167	0.6914
0.0	33.0	165	1.5472	0.0059	5524.7178	3829.4426	1821.0	2475.0	0.7358	1778.0	0.7184	891.0	912.0	1196.0	0.7625	0.7450	878.0	909.0	1267.0	0.7174	0.6930
0.0	34.0	170	1.5496	0.0059	5533.2649	3835.3670	1817.0	2475.0	0.7341	1777.0	0.7180	890.0	911.0	1196.0	0.7617	0.7441	878.0	906.0	1267.0	0.7151	0.6930
0.0	35.0	175	1.5519	0.0059	5541.3527	3840.9730	1820.0	2475.0	0.7354	1780.0	0.7192	890.0	910.0	1196.0	0.7609	0.7441	881.0	910.0	1267.0	0.7182	0.6953
0.0	36.0	180	1.5514	0.0059	5539.5094	3839.6954	1820.0	2475.0	0.7354	1781.0	0.7196	891.0	912.0	1196.0	0.7625	0.7450	881.0	908.0	1267.0	0.7167	0.6953
0.0	37.0	185	1.5539	0.0059	5548.5974	3845.9946	1819.0	2475.0	0.7349	1780.0	0.7192	891.0	912.0	1196.0	0.7625	0.7450	880.0	907.0	1267.0	0.7159	0.6946
0.0	38.0	190	1.5534	0.0059	5546.8413	3844.7774	1819.0	2475.0	0.7349	1781.0	0.7196	892.0	912.0	1196.0	0.7625	0.7458	880.0	907.0	1267.0	0.7159	0.6946
0.0	39.0	195	1.5541	0.0059	5549.1300	3846.3638	1820.0	2475.0	0.7354	1781.0	0.7196	892.0	912.0	1196.0	0.7625	0.7458	880.0	908.0	1267.0	0.7167	0.6946
0.0	40.0	200	1.5561	0.0059	5556.3793	3851.3886	1821.0	2475.0	0.7358	1785.0	0.7212	894.0	914.0	1196.0	0.7642	0.7475	882.0	907.0	1267.0	0.7159	0.6961
0.6534	41.0	205	1.5581	0.0059	5563.5837	3856.3823	1815.0	2475.0	0.7333	1778.0	0.7184	891.0	910.0	1196.0	0.7609	0.7450	878.0	905.0	1267.0	0.7143	0.6930
0.6534	42.0	210	1.5582	0.0059	5563.8211	3856.5469	1819.0	2475.0	0.7349	1784.0	0.7208	893.0	913.0	1196.0	0.7634	0.7467	882.0	906.0	1267.0	0.7151	0.6961
0.6534	43.0	215	1.5591	0.0059	5566.9433	3858.7111	1819.0	2475.0	0.7349	1784.0	0.7208	895.0	915.0	1196.0	0.7651	0.7483	880.0	904.0	1267.0	0.7135	0.6946
0.0	44.0	220	1.5600	0.0059	5570.2078	3860.9738	1818.0	2475.0	0.7345	1779.0	0.7188	893.0	913.0	1196.0	0.7634	0.7467	877.0	905.0	1267.0	0.7143	0.6922

Framework versions

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