train_conala_123_1760637665

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

• Loss: 1.0658
• Num Input Tokens Seen: 3047552

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
0.8894	1.0	536	0.6439	152672
0.6937	2.0	1072	0.6180	305288
0.4875	3.0	1608	0.5986	457952
0.4151	4.0	2144	0.6029	610944
0.8716	5.0	2680	0.6069	762440
0.5753	6.0	3216	0.6013	914920
0.4714	7.0	3752	0.6204	1067520
0.549	8.0	4288	0.6045	1220200
0.6915	9.0	4824	0.6048	1372560
0.3567	10.0	5360	0.6144	1524216
0.5933	11.0	5896	0.6268	1675880
0.5259	12.0	6432	0.6402	1828344
0.489	13.0	6968	0.6632	1980376
0.2976	14.0	7504	0.6913	2132544
0.3224	15.0	8040	0.7080	2284440
0.2692	16.0	8576	0.7352	2436520
0.2695	17.0	9112	0.7561	2589096
0.3296	18.0	9648	0.7822	2741936
0.2217	19.0	10184	0.7954	2894976
0.2721	20.0	10720	0.8006	3047552

Framework versions

• PEFT 0.17.1
• Transformers 4.51.3
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4