ATL-Machine/affine-test-01

out_affine

This model is a fine-tuned version of ineso22/affine-vscode using LoRA (Low-Rank Adaptation).

Model Details

• Base Model: ineso22/affine-vscode
• Training Method: LoRA (Low-Rank Adaptation)
• LoRA Rank: 16
• LoRA Alpha: 32
• LoRA Dropout: 0.05
• Quantization: 4-bit QLoRA
• Precision: bfloat16

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained(
    "out_affine/merged_model",
    torch_dtype=torch.bfloat16,
    trust_remote_code=True,
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("out_affine/merged_model")

# Your inference code here