metadata

base_model: defog/sqlcoder-7b-2
library_name: peft
license: cc-by-sa-4.0
tags:
  - trl
  - sft
  - QLora
  - peft
  - SQL
  - causal-lm
model-index:
  - name: sqlcoder-7b-2_FineTuned_PEFT_QLORA_adapter
    results: []
language:
  - en

sqlcoder-7b-2_FineTuned_QLORA_Adapter

This model is a fine-tuned version of defog/sqlcoder-7b-2 on 260 MS SQL examples (Task, Schema and Answer).

Intended uses & limitations

MS Server - SQL Query Generation

Training and evaluation data

This model was re-trained using the QLoRA method, with the following configurations:

'r = 64', 'lora_alpha = 32', 'lora_dropout = 0.05', bias='none', task_type='CAUSAL_LM'

Quantization parameters:

load_in_4bit=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=torch.bfloat16

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0002
train_batch_size: 1
eval_batch_size: 8
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: linear
lr_scheduler_warmup_steps: 10
num_epochs: 5
mixed_precision_training: Native AMP

Framework versions

PEFT 0.13.2
Transformers 4.44.2
Pytorch 2.4.1+cu121
Datasets 3.0.2
Tokenizers 0.19.1