emirozturk/phi-3-mini-4k-t2c

This repository contains a fine-tuned model for Turkish code generation.

Checkpoint: step 10000
Dataset: Translated version of python_code_instructions_18k_alpaca

Example Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("emirozturk/phi-3-mini-4k-t2c")
tokenizer = AutoTokenizer.from_pretrained("emirozturk/phi-3-mini-4k-t2c")

prompt = "Python'da ax^2 + bx + c = 0 denkleminin x-kesişimlerini hesaplayan bir program yaz."
inputs = tokenizer(prompt, return_tensors="pt")
output = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(output[0], skip_special_tokens=True))