OpenCodeInterpreter-DS-6.7B-SQL2

This model is a fine-tuned version of m-a-p/OpenCodeInterpreter-DS-6.7B on a downsampled (20,000 samples) b-mc2/sql-create-context dataset. It achieves the following results on the evaluation set:

• Loss: 0.3180

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: 0.0002
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 30
• training_steps: 1600

Training results

Training Loss	Epoch	Step	Validation Loss
0.3643	1.44	200	0.3582
0.337	2.88	400	0.3357
0.3197	4.32	600	0.3271
0.3126	5.76	800	0.3222
0.3034	7.19	1000	0.3198
0.2998	8.63	1200	0.3185
0.297	10.07	1400	0.3180
0.2964	11.51	1600	0.3180

Framework versions

• PEFT 0.10.0
• Transformers 4.40.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2