train_cola_1757340185

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

• Loss: 0.1891
• Num Input Tokens Seen: 3669168

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: 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: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.1682	0.5	962	0.2292	184192
0.2053	1.0	1924	0.2663	367320
0.3625	1.5	2886	0.2419	550840
0.3409	2.0	3848	0.1891	734600
0.242	2.5	4810	0.2103	918600
0.4443	3.0	5772	0.2274	1101216
0.2546	3.5	6734	0.1962	1284288
0.3322	4.0	7696	0.2311	1468552
0.1172	4.5	8658	0.2418	1651528
0.0905	5.0	9620	0.2520	1834816
0.2706	5.5	10582	0.2517	2018016
0.1555	6.0	11544	0.2436	2201584
0.2961	6.5	12506	0.2524	2385200
0.0537	7.0	13468	0.2558	2568288
0.2939	7.5	14430	0.2553	2751584
0.1063	8.0	15392	0.2559	2935056
0.1724	8.5	16354	0.2551	3118000
0.0641	9.0	17316	0.2560	3301760
0.05	9.5	18278	0.2536	3485344
0.4162	10.0	19240	0.2559	3669168

Framework versions

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