train_multirc_1745950266

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the multirc dataset. It achieves the following results on the evaluation set:

• Loss: 0.9281
• Num Input Tokens Seen: 75778784

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
0.8429	0.0326	200	1.0602	378944
0.8908	0.0653	400	0.9846	758192
1.0758	0.0979	600	0.9602	1141408
1.1887	0.1305	800	0.9494	1518336
0.4649	0.1631	1000	0.9558	1901264
1.1186	0.1958	1200	0.9523	2279552
1.1403	0.2284	1400	0.9474	2668256
0.6437	0.2610	1600	0.9507	3047328
0.6645	0.2937	1800	0.9448	3429984
0.951	0.3263	2000	0.9496	3814576
0.9459	0.3589	2200	0.9451	4190352
0.6611	0.3915	2400	0.9471	4567440
1.9537	0.4242	2600	0.9527	4944384
0.9478	0.4568	2800	0.9585	5325216
0.9183	0.4894	3000	0.9481	5698896
1.5122	0.5221	3200	0.9483	6074432
0.51	0.5547	3400	0.9481	6454208
0.4708	0.5873	3600	0.9469	6831056
0.6829	0.6200	3800	0.9577	7209536
0.9297	0.6526	4000	0.9484	7593024
0.8081	0.6852	4200	0.9388	7977072
0.9532	0.7178	4400	0.9327	8353296
0.7015	0.7505	4600	0.9427	8733232
1.0696	0.7831	4800	0.9324	9113632
0.6209	0.8157	5000	0.9560	9487952
0.8961	0.8484	5200	0.9395	9861104
1.0547	0.8810	5400	0.9412	10239088
1.2457	0.9136	5600	0.9445	10619840
1.3315	0.9462	5800	0.9514	10994720
1.0866	0.9789	6000	0.9464	11376976
0.9159	1.0114	6200	0.9383	11758656
1.2179	1.0440	6400	0.9451	12144016
0.8416	1.0767	6600	0.9362	12531776
1.1413	1.1093	6800	0.9466	12905136
0.9718	1.1419	7000	0.9424	13278096
1.295	1.1746	7200	0.9421	13651520
0.6159	1.2072	7400	0.9364	14034784
1.0405	1.2398	7600	0.9419	14415120
0.7077	1.2725	7800	0.9339	14794784
0.6286	1.3051	8000	0.9354	15176240
1.015	1.3377	8200	0.9345	15548080
0.5649	1.3703	8400	0.9363	15926832
1.288	1.4030	8600	0.9464	16305344
0.6352	1.4356	8800	0.9344	16686528
1.0296	1.4682	9000	0.9397	17073648
0.6318	1.5009	9200	0.9450	17457952
1.2265	1.5335	9400	0.9558	17831104
0.752	1.5661	9600	0.9367	18215168
0.6444	1.5987	9800	0.9360	18592816
1.2319	1.6314	10000	0.9408	18972864
0.6575	1.6640	10200	0.9439	19350160
1.3234	1.6966	10400	0.9475	19735024
1.2022	1.7293	10600	0.9402	20108768
1.3244	1.7619	10800	0.9433	20489424
1.2097	1.7945	11000	0.9463	20870832
0.9912	1.8271	11200	0.9415	21240960
1.1266	1.8598	11400	0.9293	21615744
0.9204	1.8924	11600	0.9569	21991984
0.8482	1.9250	11800	0.9444	22366624
1.301	1.9577	12000	0.9329	22746000
1.1116	1.9903	12200	0.9432	23122688
0.7897	2.0228	12400	0.9398	23494112
1.1651	2.0555	12600	0.9495	23876160
1.0717	2.0881	12800	0.9446	24261904
0.8885	2.1207	13000	0.9414	24643776
0.6751	2.1534	13200	0.9347	25020496
0.9654	2.1860	13400	0.9370	25391072
0.7338	2.2186	13600	0.9449	25762416
1.0438	2.2512	13800	0.9505	26139456
0.7634	2.2839	14000	0.9568	26511344
1.0118	2.3165	14200	0.9423	26891616
1.0028	2.3491	14400	0.9470	27274960
0.9065	2.3818	14600	0.9416	27652224
1.2445	2.4144	14800	0.9423	28033168
1.0826	2.4470	15000	0.9462	28414784
1.1692	2.4796	15200	0.9433	28787168
0.9041	2.5123	15400	0.9309	29164512
0.666	2.5449	15600	0.9459	29545056
0.6744	2.5775	15800	0.9435	29922176
0.9566	2.6102	16000	0.9458	30304336
0.9229	2.6428	16200	0.9431	30688608
0.981	2.6754	16400	0.9449	31067744
0.9639	2.7081	16600	0.9513	31455328
0.688	2.7407	16800	0.9445	31833136
1.2851	2.7733	17000	0.9450	32213296
0.9594	2.8059	17200	0.9332	32588128
1.596	2.8386	17400	0.9458	32971552
1.2732	2.8712	17600	0.9381	33356064
0.8654	2.9038	17800	0.9527	33739984
1.1866	2.9365	18000	0.9534	34121824
1.0331	2.9691	18200	0.9447	34498368
0.9694	3.0016	18400	0.9325	34866272
0.8546	3.0343	18600	0.9470	35258768
1.1217	3.0669	18800	0.9438	35644416
1.0455	3.0995	19000	0.9385	36017808
0.5291	3.1321	19200	0.9523	36393536
1.4627	3.1648	19400	0.9393	36770432
0.7338	3.1974	19600	0.9378	37152448
0.9475	3.2300	19800	0.9320	37532496
1.041	3.2627	20000	0.9395	37910480
0.8279	3.2953	20200	0.9451	38286080
1.0918	3.3279	20400	0.9556	38664512
1.0382	3.3606	20600	0.9494	39053472
1.3256	3.3932	20800	0.9338	39432032
1.2405	3.4258	21000	0.9404	39812704
0.7672	3.4584	21200	0.9427	40191088
0.973	3.4911	21400	0.9413	40567216
0.8321	3.5237	21600	0.9382	40947696
0.9788	3.5563	21800	0.9526	41330624
1.235	3.5890	22000	0.9441	41708800
0.4985	3.6216	22200	0.9452	42087824
1.1512	3.6542	22400	0.9353	42461936
0.7916	3.6868	22600	0.9488	42843696
0.8338	3.7195	22800	0.9428	43221120
1.1238	3.7521	23000	0.9366	43597776
0.6936	3.7847	23200	0.9418	43979312
0.6829	3.8174	23400	0.9284	44354480
0.9888	3.8500	23600	0.9381	44727696
1.0987	3.8826	23800	0.9352	45108608
1.1806	3.9152	24000	0.9479	45482928
0.6657	3.9479	24200	0.9359	45861584
1.0493	3.9805	24400	0.9422	46243072
1.0646	4.0131	24600	0.9422	46619680
0.9162	4.0457	24800	0.9401	47007360
1.3567	4.0783	25000	0.9361	47391600
0.7256	4.1109	25200	0.9404	47768320
1.2478	4.1436	25400	0.9429	48143424
1.087	4.1762	25600	0.9353	48524368
0.8097	4.2088	25800	0.9346	48899856
0.6918	4.2415	26000	0.9357	49280208
0.9613	4.2741	26200	0.9350	49658080
1.1259	4.3067	26400	0.9395	50034848
0.7122	4.3393	26600	0.9460	50413376
1.0615	4.3720	26800	0.9428	50793248
1.1671	4.4046	27000	0.9421	51170976
1.1558	4.4372	27200	0.9426	51559504
0.6195	4.4699	27400	0.9444	51928704
0.731	4.5025	27600	0.9370	52297776
0.7066	4.5351	27800	0.9397	52669472
0.6422	4.5677	28000	0.9472	53045856
1.0387	4.6004	28200	0.9486	53429232
1.1628	4.6330	28400	0.9281	53810560
1.1705	4.6656	28600	0.9332	54191536
0.9436	4.6983	28800	0.9376	54572176
0.867	4.7309	29000	0.9443	54952896
1.0229	4.7635	29200	0.9492	55327776
0.445	4.7961	29400	0.9445	55708896
0.7952	4.8288	29600	0.9441	56085712
0.6884	4.8614	29800	0.9549	56467376
0.691	4.8940	30000	0.9420	56841328
0.8571	4.9267	30200	0.9397	57227184
0.9701	4.9593	30400	0.9355	57605632
0.8541	4.9919	30600	0.9416	57987472
0.4922	5.0245	30800	0.9544	58367056
0.9541	5.0571	31000	0.9414	58746720
0.5754	5.0897	31200	0.9393	59124272
1.1303	5.1224	31400	0.9290	59504688
0.7824	5.1550	31600	0.9336	59875840
0.859	5.1876	31800	0.9425	60247360
0.8638	5.2202	32000	0.9414	60622464
0.8986	5.2529	32200	0.9392	61006768
0.7302	5.2855	32400	0.9455	61386992
1.3051	5.3181	32600	0.9346	61770000
1.1335	5.3508	32800	0.9374	62154640
0.9185	5.3834	33000	0.9444	62541664
0.5254	5.4160	33200	0.9312	62912976
0.9978	5.4486	33400	0.9312	63289520
1.6788	5.4813	33600	0.9449	63668416
0.807	5.5139	33800	0.9414	64043792
0.6642	5.5465	34000	0.9414	64433840
1.287	5.5792	34200	0.9414	64808624
0.8623	5.6118	34400	0.9414	65182704
0.7179	5.6444	34600	0.9414	65562192
1.1127	5.6771	34800	0.9414	65940816
0.9003	5.7097	35000	0.9414	66326768
0.6414	5.7423	35200	0.9414	66705744
1.3465	5.7749	35400	0.9414	67084928
0.68	5.8076	35600	0.9414	67462064
0.8508	5.8402	35800	0.9414	67846112
0.8323	5.8728	36000	0.9414	68221552
0.9347	5.9055	36200	0.9414	68606416
0.9616	5.9381	36400	0.9414	68980176
0.7618	5.9707	36600	0.9414	69349984
1.3682	6.0033	36800	0.9414	69729984
0.6724	6.0359	37000	0.9414	70107936
1.0439	6.0685	37200	0.9414	70487856
0.5013	6.1012	37400	0.9414	70865792
0.6052	6.1338	37600	0.9414	71244784
1.1087	6.1664	37800	0.9414	71630704
0.5706	6.1990	38000	0.9414	72002688
1.8746	6.2317	38200	0.9414	72385776
0.9978	6.2643	38400	0.9414	72773152
0.8948	6.2969	38600	0.9414	73149584
1.2147	6.3296	38800	0.9414	73519536
1.0777	6.3622	39000	0.9414	73902896
1.1562	6.3948	39200	0.9414	74278960
1.6357	6.4274	39400	0.9414	74655728
0.9998	6.4601	39600	0.9414	75025808
0.8366	6.4927	39800	0.9414	75402576
0.8005	6.5253	40000	0.9414	75778784

Framework versions

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