train_sft.py

from datasets import load_dataset
from transformers import AutoModelForCausalLM, AutoTokenizer
from trl import SFTTrainer, SFTTrainingArguments

MODEL_ID = "Salesforce/codegen-350M-multi"
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
model     = AutoModelForCausalLM.from_pretrained(MODEL_ID)

# 1) load your local JSONL
ds = load_dataset("json", data_files="data/train_dataset.jsonl", split="train")

# 2) tokenize & format
def tokenize(example):
    prompt = f"DIFF:\n{example['diff']}\n\nOUTPUT FORMAT:\n"
    output = example['comments']
    text = prompt + tokenizer.decode(tokenizer.encode(str(output), add_special_tokens=False))
    tokens = tokenizer(text, truncation=True, max_length=512)
    tokens["labels"] = tokens["input_ids"].copy()
    return tokens

ds = ds.map(tokenize, remove_columns=ds.column_names, batched=False)

# 3) SFT arguments
training_args = SFTTrainingArguments(
    output_dir="sft-model",
    per_device_train_batch_size=2,
    gradient_accumulation_steps=8,
    learning_rate=2e-5,
    max_train_steps=500,
    logging_steps=50,
    save_steps=200
)

# 4) kick off the trainer
trainer = SFTTrainer(model, tokenizer, args=training_args, train_dataset=ds)
trainer.train()
trainer.save_model("sft-model")