santacoder-finetuned-robot3

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

• Loss: 0.5689

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	1.6250
No log	0.1	2	2.1620
No log	0.15	3	1.7060
No log	0.2	4	1.7167
No log	0.25	5	1.0462
No log	0.3	6	1.2153
No log	0.35	7	1.0301
No log	0.4	8	0.9399
No log	0.45	9	1.0030
1.4139	0.5	10	0.8322
1.4139	0.55	11	0.7111
1.4139	0.6	12	0.7151
1.4139	0.65	13	0.6482
1.4139	0.7	14	0.6228
1.4139	0.75	15	0.6105
1.4139	0.8	16	0.5827
1.4139	0.85	17	0.5791
1.4139	0.9	18	0.5726
1.4139	0.95	19	0.5696
0.5282	1.0	20	0.5689

Framework versions

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