train_boolq_1745950274

This model is a fine-tuned version of google/gemma-3-1b-it on the boolq dataset. It achieves the following results on the evaluation set:

• Loss: 2.9271
• Num Input Tokens Seen: 34633072

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: 2
• eval_batch_size: 2
• seed: 123
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• training_steps: 40000

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
2.6453	0.0943	200	3.2031	174096
3.5809	0.1886	400	3.1503	344560
3.5279	0.2829	600	3.1043	517536
2.8551	0.3772	800	3.0762	696016
2.7868	0.4715	1000	3.0509	868992
2.6766	0.5658	1200	3.0614	1040544
3.84	0.6601	1400	3.0431	1211680
2.3861	0.7544	1600	3.0519	1381792
3.4698	0.8487	1800	3.0226	1559456
2.9566	0.9430	2000	3.0275	1735840
2.8968	1.0372	2200	3.0316	1910848
3.1856	1.1315	2400	3.0049	2081696
3.4134	1.2258	2600	3.0010	2255952
2.9332	1.3201	2800	3.0122	2427152
2.9351	1.4144	3000	3.0129	2601296
2.5434	1.5087	3200	2.9970	2774672
3.1122	1.6030	3400	2.9784	2944896
2.7179	1.6973	3600	2.9600	3117216
2.8625	1.7916	3800	2.9709	3287952
2.7659	1.8859	4000	2.9735	3464640
2.9328	1.9802	4200	2.9566	3638880
3.0105	2.0745	4400	2.9613	3812624
3.4794	2.1688	4600	2.9646	3986544
2.7816	2.2631	4800	2.9783	4158272
3.1668	2.3574	5000	2.9668	4328240
3.0955	2.4517	5200	2.9730	4507760
2.8676	2.5460	5400	2.9506	4681664
3.3803	2.6403	5600	2.9642	4856928
3.1192	2.7346	5800	2.9736	5024976
3.3355	2.8289	6000	2.9883	5202368
3.3114	2.9231	6200	2.9882	5377360
2.1893	3.0174	6400	2.9834	5550480
3.533	3.1117	6600	2.9786	5724080
2.7837	3.2060	6800	2.9849	5896688
3.127	3.3003	7000	2.9557	6070544
2.7459	3.3946	7200	2.9538	6244624
3.2816	3.4889	7400	2.9563	6416176
3.3454	3.5832	7600	2.9709	6587616
2.7435	3.6775	7800	2.9540	6759696
3.1908	3.7718	8000	2.9497	6932384
2.6196	3.8661	8200	2.9479	7103328
3.1352	3.9604	8400	2.9728	7276304
2.9739	4.0547	8600	2.9485	7448112
2.5534	4.1490	8800	2.9555	7623632
2.5543	4.2433	9000	2.9271	7799248
2.5284	4.3376	9200	2.9455	7974368
2.8538	4.4319	9400	2.9379	8146384
2.7943	4.5262	9600	2.9336	8321456
2.4081	4.6205	9800	2.9336	8490096
2.7996	4.7148	10000	2.9381	8665904
2.3661	4.8091	10200	2.9455	8837712
2.966	4.9033	10400	2.9509	9010400
3.1711	4.9976	10600	2.9699	9185584
3.311	5.0919	10800	2.9491	9358160
2.4932	5.1862	11000	2.9626	9535520
3.3139	5.2805	11200	2.9806	9709232
2.7371	5.3748	11400	2.9771	9880896
2.9514	5.4691	11600	2.9830	10053056
3.3724	5.5634	11800	2.9859	10229152
3.1756	5.6577	12000	2.9848	10404384
2.9306	5.7520	12200	2.9874	10573872
2.9906	5.8463	12400	2.9814	10748304
3.0223	5.9406	12600	2.9898	10917920
3.4233	6.0349	12800	2.9858	11092736
2.554	6.1292	13000	2.9839	11269264
3.5816	6.2235	13200	2.9827	11441120
2.9904	6.3178	13400	2.9840	11614176
2.3922	6.4121	13600	2.9881	11785424
3.0193	6.5064	13800	2.9901	11960752
2.663	6.6007	14000	2.9775	12132672
3.3592	6.6950	14200	2.9821	12303424
3.1617	6.7893	14400	2.9830	12474592
3.1247	6.8835	14600	2.9735	12649424
2.4094	6.9778	14800	2.9854	12821280
2.8975	7.0721	15000	2.9798	12996208
3.3305	7.1664	15200	2.9779	13172592
2.8335	7.2607	15400	2.9754	13342864
3.2162	7.3550	15600	2.9741	13515600
3.1557	7.4493	15800	2.9858	13688640
3.123	7.5436	16000	2.9802	13863312
3.1461	7.6379	16200	2.9744	14032992
2.5753	7.7322	16400	2.9744	14205936
3.0835	7.8265	16600	2.9788	14378336
2.9754	7.9208	16800	2.9861	14551456
2.7244	8.0151	17000	2.9781	14730672
3.4109	8.1094	17200	2.9844	14904544
2.7873	8.2037	17400	2.9804	15078832
3.263	8.2980	17600	2.9827	15254544
2.6633	8.3923	17800	2.9717	15422256
2.6194	8.4866	18000	2.9880	15595776
2.8025	8.5809	18200	2.9845	15768288
3.2739	8.6752	18400	2.9862	15941776
3.0337	8.7694	18600	2.9897	16115152
3.0608	8.8637	18800	2.9865	16284384
3.5312	8.9580	19000	2.9885	16457552
2.6771	9.0523	19200	2.9896	16632272
2.8448	9.1466	19400	2.9859	16806304
3.4979	9.2409	19600	2.9860	16979072
3.4671	9.3352	19800	2.9841	17150160
3.6682	9.4295	20000	2.9850	17321280
2.8798	9.5238	20200	2.9839	17495488
3.5262	9.6181	20400	2.9818	17670576
3.3741	9.7124	20600	2.9857	17843440
2.8687	9.8067	20800	2.9815	18012496
2.7849	9.9010	21000	2.9824	18186480
2.7368	9.9953	21200	2.9828	18360368
2.6571	10.0896	21400	2.9818	18539664
2.6093	10.1839	21600	2.9837	18718016
2.8979	10.2782	21800	2.9838	18888560
2.3822	10.3725	22000	2.9835	19061328
2.8941	10.4668	22200	2.9847	19236176
2.2785	10.5611	22400	2.9793	19404288
2.7086	10.6554	22600	2.9829	19574224
3.0499	10.7496	22800	2.9829	19744496
2.5357	10.8439	23000	2.9834	19915984
2.8058	10.9382	23200	2.9827	20090944
3.2345	11.0325	23400	2.9843	20264992
2.6316	11.1268	23600	2.9810	20437952
3.344	11.2211	23800	2.9823	20611040
3.0959	11.3154	24000	2.9831	20787488
3.3262	11.4097	24200	2.9829	20958240
3.9468	11.5040	24400	2.9828	21133392
2.874	11.5983	24600	2.9810	21303360
2.9608	11.6926	24800	2.9846	21475184
2.9467	11.7869	25000	2.9840	21649744
2.8529	11.8812	25200	2.9841	21819728
3.0579	11.9755	25400	2.9836	21993120
2.9273	12.0698	25600	2.9827	22164624
3.3136	12.1641	25800	2.9837	22340064
2.507	12.2584	26000	2.9838	22515088
2.7376	12.3527	26200	2.9842	22692240
2.3293	12.4470	26400	2.9816	22864512
3.2821	12.5413	26600	2.9816	23037568
2.8383	12.6355	26800	2.9818	23207936
2.491	12.7298	27000	2.9850	23381376
2.7425	12.8241	27200	2.9821	23553008
3.0866	12.9184	27400	2.9818	23722608
3.0738	13.0127	27600	2.9836	23892928
2.6363	13.1070	27800	2.9819	24063632
3.15	13.2013	28000	2.9816	24237248
2.9501	13.2956	28200	2.9806	24411712
3.1561	13.3899	28400	2.9818	24584800
3.0268	13.4842	28600	2.9812	24759888
2.6915	13.5785	28800	2.9827	24936720
2.1768	13.6728	29000	2.9816	25110864
2.769	13.7671	29200	2.9800	25284944
3.4771	13.8614	29400	2.9812	25456816
3.0152	13.9557	29600	2.9807	25631728
2.8072	14.0500	29800	2.9815	25801056
3.4366	14.1443	30000	2.9832	25978896
3.0998	14.2386	30200	2.9835	26156672
2.6254	14.3329	30400	2.9815	26330592
3.1786	14.4272	30600	2.9817	26502800
2.803	14.5215	30800	2.9820	26671584
2.881	14.6157	31000	2.9818	26845568
3.4202	14.7100	31200	2.9797	27017952
2.9903	14.8043	31400	2.9816	27191600
2.9173	14.8986	31600	2.9821	27362144
2.5902	14.9929	31800	2.9826	27536992
3.4131	15.0872	32000	2.9853	27707728
3.1571	15.1815	32200	2.9846	27886368
4.0238	15.2758	32400	2.9850	28061984
2.7433	15.3701	32600	2.9855	28233360
2.9622	15.4644	32800	2.9850	28411200
2.9855	15.5587	33000	2.9855	28582944
2.8146	15.6530	33200	2.9850	28756240
2.1957	15.7473	33400	2.9850	28926208
3.0656	15.8416	33600	2.9850	29096816
2.781	15.9359	33800	2.9844	29267072
2.8638	16.0302	34000	2.9844	29435360
3.3154	16.1245	34200	2.9844	29610720
2.351	16.2188	34400	2.9844	29781472
2.8905	16.3131	34600	2.9844	29959568
2.4058	16.4074	34800	2.9844	30134704
3.0833	16.5017	35000	2.9844	30305200
3.0284	16.5959	35200	2.9844	30478576
2.7625	16.6902	35400	2.9844	30647744
2.5376	16.7845	35600	2.9844	30823072
3.3257	16.8788	35800	2.9844	30996032
2.8614	16.9731	36000	2.9844	31167328
2.8231	17.0674	36200	2.9844	31341392
2.4818	17.1617	36400	2.9844	31515648
3.5857	17.2560	36600	2.9844	31690208
2.8301	17.3503	36800	2.9844	31868288
2.6124	17.4446	37000	2.9844	32041536
3.5769	17.5389	37200	2.9844	32213584
3.0379	17.6332	37400	2.9844	32385920
2.76	17.7275	37600	2.9844	32555856
3.1422	17.8218	37800	2.9844	32729024
2.7946	17.9161	38000	2.9844	32902832
3.2047	18.0104	38200	2.9844	33076912
2.4714	18.1047	38400	2.9844	33248832
2.6947	18.1990	38600	2.9844	33420800
3.2644	18.2933	38800	2.9844	33594000
2.9905	18.3876	39000	2.9844	33765936
2.7909	18.4818	39200	2.9844	33936896
2.7959	18.5761	39400	2.9844	34110592
2.8925	18.6704	39600	2.9844	34284208
3.0191	18.7647	39800	2.9844	34458576
3.5334	18.8590	40000	2.9844	34633072

Framework versions

• PEFT 0.15.2.dev0
• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1