gemma-3-4b-it-code-translation

gemma-3-4b-it-code-translation

A fine-tuned version of google/gemma-3-4b-it for translating code between C++, Java, and Python.

Training

• Base model: google/gemma-3-4b-it
• Method: LoRA (Low-Rank Adaptation) via LLaMA-Factory
• Dataset: tkeskin/leetcode-solutions (instruct config) — directed C++/Java/Python translation pairs derived from LeetCode solutions
• Hardware: AMD MI210 (ROCm) / NVIDIA CUDA, flash_attn: sdpa
• LoRA target: all linear layers (lora_target: all)
• Precision: bf16

Evaluation

Evaluated with an execution-based translation benchmark: each held-out evaluation-config payload from tkeskin/leetcode-solutions is a directed source→target translation whose output is compiled and run against the problem's input/output pairs. The eval split is held out from training (no leakage). Metric is pass@1 (all test cases pass), n-weighted over 3,336 payloads.

pass@1 by language pair × difficulty (%):

Full methodology is in the llm-fine-tune repo (Stage 5).

Intended use

Given source code in one of C++, Java, or Python, the model generates a translation into the target language, following the same logic and structure.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "tkeskin/gemma-3-4b-it-code-translation"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

messages = [
    {
        "role": "user",
        "content": "Translate the following C++ code to Python:\n\nint add(int a, int b) { return a + b; }"
    }
]
inputs = tokenizer.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True)
outputs = model.generate(inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True))