Llama-3.3-70B-Instruct-v2-3d-2M-200K-0.1-reverse-padzero-99-128D-1L-2H-512I

This model is a fine-tuned version of meta-llama/Llama-3.3-70B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.4417

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: 0.001
• train_batch_size: 128
• eval_batch_size: 128
• 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: 0.05
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0	0	3.1621
1.9685	0.032	500	1.9346
1.8076	0.064	1000	1.8022
1.6821	0.096	1500	1.6713
1.6435	0.128	2000	1.6414
1.5828	0.16	2500	1.5671
1.5491	0.192	3000	1.5456
1.5287	0.224	3500	1.5313
1.5266	0.256	4000	1.5242
1.5182	0.288	4500	1.5195
1.5165	0.32	5000	1.5157
1.5136	0.352	5500	1.5126
1.5108	0.384	6000	1.5100
1.482	0.416	6500	1.4837
1.4831	0.448	7000	1.4801
1.4795	0.48	7500	1.4779
1.475	0.512	8000	1.4763
1.4759	0.544	8500	1.4732
1.4714	0.576	9000	1.4713
1.4696	0.608	9500	1.4714
1.4719	0.64	10000	1.4696
1.4679	0.672	10500	1.4692
1.4707	0.704	11000	1.4682
1.4692	0.736	11500	1.4675
1.4699	0.768	12000	1.4673
1.4619	0.8	12500	1.4657
1.4648	0.832	13000	1.4659
1.462	0.864	13500	1.4642
1.4627	0.896	14000	1.4630
1.4634	0.928	14500	1.4624
1.4606	0.96	15000	1.4617
1.4626	0.992	15500	1.4628
1.4625	1.024	16000	1.4626
1.4594	1.056	16500	1.4603
1.4626	1.088	17000	1.4613
1.4601	1.12	17500	1.4631
1.4595	1.152	18000	1.4598
1.459	1.184	18500	1.4594
1.459	1.216	19000	1.4587
1.4576	1.248	19500	1.4583
1.4584	1.28	20000	1.4563
1.4568	1.312	20500	1.4567
1.4575	1.3440	21000	1.4566
1.4563	1.376	21500	1.4565
1.4588	1.408	22000	1.4553
1.4545	1.44	22500	1.4565
1.4558	1.472	23000	1.4578
1.4573	1.504	23500	1.4559
1.454	1.536	24000	1.4543
1.4535	1.568	24500	1.4558
1.4559	1.6	25000	1.4528
1.4511	1.6320	25500	1.4531
1.4522	1.6640	26000	1.4526
1.4523	1.696	26500	1.4530
1.4512	1.728	27000	1.4517
1.4549	1.76	27500	1.4528
1.4558	1.792	28000	1.4505
1.4536	1.8240	28500	1.4523
1.4515	1.8560	29000	1.4506
1.4502	1.888	29500	1.4494
1.4495	1.92	30000	1.4508
1.4484	1.952	30500	1.4505
1.4509	1.984	31000	1.4493
1.4498	2.016	31500	1.4505
1.4489	2.048	32000	1.4487
1.4487	2.08	32500	1.4505
1.4486	2.112	33000	1.4478
1.4502	2.144	33500	1.4484
1.4492	2.176	34000	1.4488
1.4489	2.208	34500	1.4478
1.4487	2.24	35000	1.4480
1.4483	2.2720	35500	1.4474
1.4468	2.304	36000	1.4471
1.4491	2.336	36500	1.4472
1.4448	2.368	37000	1.4466
1.4459	2.4	37500	1.4467
1.4481	2.432	38000	1.4460
1.4478	2.464	38500	1.4492
1.4456	2.496	39000	1.4470
1.4475	2.528	39500	1.4457
1.4465	2.56	40000	1.4456
1.445	2.592	40500	1.4454
1.4449	2.624	41000	1.4457
1.4465	2.656	41500	1.4456
1.4459	2.6880	42000	1.4450
1.4472	2.7200	42500	1.4449
1.4458	2.752	43000	1.4453
1.4463	2.784	43500	1.4446
1.4426	2.816	44000	1.4451
1.4454	2.848	44500	1.4448
1.4447	2.88	45000	1.4444
1.4439	2.912	45500	1.4445
1.4443	2.944	46000	1.4442
1.445	2.976	46500	1.4438
1.4419	3.008	47000	1.4437
1.4421	3.04	47500	1.4440
1.4408	3.072	48000	1.4435
1.4426	3.104	48500	1.4434
1.4422	3.136	49000	1.4434
1.4431	3.168	49500	1.4432
1.4458	3.2	50000	1.4432
1.4421	3.232	50500	1.4430
1.4455	3.2640	51000	1.4431
1.4425	3.296	51500	1.4427
1.4435	3.328	52000	1.4431
1.4427	3.36	52500	1.4431
1.4427	3.392	53000	1.4429
1.4414	3.424	53500	1.4429
1.4439	3.456	54000	1.4429
1.443	3.488	54500	1.4426
1.4413	3.52	55000	1.4426
1.4423	3.552	55500	1.4425
1.4404	3.584	56000	1.4423
1.4433	3.616	56500	1.4424
1.4418	3.648	57000	1.4423
1.4444	3.68	57500	1.4422
1.4396	3.7120	58000	1.4422
1.4427	3.7440	58500	1.4422
1.4423	3.776	59000	1.4421
1.4432	3.808	59500	1.4420
1.4421	3.84	60000	1.4420
1.4437	3.872	60500	1.4420
1.4412	3.904	61000	1.4419
1.4405	3.936	61500	1.4419
1.4418	3.968	62000	1.4419
1.4414	4.0	62500	1.4418
1.4418	4.032	63000	1.4418
1.442	4.064	63500	1.4419
1.439	4.096	64000	1.4418
1.4431	4.128	64500	1.4418
1.4391	4.16	65000	1.4418
1.4413	4.192	65500	1.4418
1.4408	4.224	66000	1.4418
1.443	4.256	66500	1.4417
1.4441	4.288	67000	1.4417
1.4431	4.32	67500	1.4417
1.4417	4.352	68000	1.4417
1.4419	4.384	68500	1.4417
1.4404	4.416	69000	1.4417
1.4418	4.448	69500	1.4417
1.4402	4.48	70000	1.4417
1.4397	4.5120	70500	1.4417
1.4395	4.5440	71000	1.4417
1.4398	4.576	71500	1.4417
1.4395	4.608	72000	1.4417
1.4413	4.64	72500	1.4417
1.4388	4.672	73000	1.4417
1.4436	4.704	73500	1.4417
1.4427	4.736	74000	1.4417
1.4413	4.768	74500	1.4417
1.4393	4.8	75000	1.4417
1.4417	4.832	75500	1.4417
1.445	4.864	76000	1.4417
1.4417	4.896	76500	1.4417
1.4407	4.928	77000	1.4417
1.4416	4.96	77500	1.4417
1.4414	4.992	78000	1.4417

Framework versions

• Transformers 4.57.1
• Pytorch 2.9.0+cu128
• Datasets 4.5.0
• Tokenizers 0.22.1