train_record_123_1763024605

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

• Loss: 6.4921
• Num Input Tokens Seen: 928969984

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: 4
• eval_batch_size: 4
• seed: 123
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.3045	1.0	31242	0.3144	46454112
0.2037	2.0	62484	0.2961	92908288
0.2339	3.0	93726	0.2803	139351808
0.2753	4.0	124968	0.2709	185790304
0.2165	5.0	156210	0.2694	232243968
0.2011	6.0	187452	0.2638	278686752
0.1549	7.0	218694	0.2635	325137568
0.2504	8.0	249936	0.2607	371592704
0.1917	9.0	281178	0.2620	418033696
0.3011	10.0	312420	0.2620	464483424
0.2058	11.0	343662	0.2632	510926720
0.1434	12.0	374904	0.2629	557369088
0.2686	13.0	406146	0.2662	603816992
0.204	14.0	437388	0.2736	650269248
0.1561	15.0	468630	0.2768	696727936
0.2222	16.0	499872	0.2874	743174112
0.2384	17.0	531114	0.2887	789614720
0.1332	18.0	562356	0.2930	836057280
0.171	19.0	593598	0.2942	882504192
0.191	20.0	624840	0.2959	928969984

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1