santacoder-finetuned-robot4

This model is a fine-tuned version of bigcode/santacoder on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5116

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 1
• training_steps: 20

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.05	1	1.5720
No log	0.1	2	2.6819
No log	0.15	3	1.3777
No log	0.2	4	1.7661
No log	0.25	5	1.0597
No log	0.3	6	1.1801
No log	0.35	7	0.8989
No log	0.4	8	0.9572
No log	0.45	9	0.7408
1.48	0.5	10	0.7517
1.48	0.55	11	0.6493
1.48	0.6	12	0.6206
1.48	0.65	13	0.5872
1.48	0.7	14	0.5644
1.48	0.75	15	0.5415
1.48	0.8	16	0.5298
1.48	0.85	17	0.5186
1.48	0.9	18	0.5128
1.48	0.95	19	0.5120
0.5307	1.0	20	0.5116

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1