code/support_check/model_finetune/llama32_4B.py

import ast
import json
import os
from datetime import datetime

import torch
from datasets import Dataset
from trl import SFTConfig, SFTTrainer
from unsloth import FastLanguageModel

model_name = "unsloth/Llama-3.2-3B-Instruct"
data_path = "/home/mshahidul/readctrl/data/finetuning_data/dataset_for_sft_support_check_list.json"
test_size = 0.1
seed = 3407
max_seq_length = 2048
load_in_4bit = True


def formatting_prompts_func(examples):
    convos = examples["conversations"]
    texts = [
        tokenizer.apply_chat_template(
            convo,
            tokenize=False,
            add_generation_prompt=False,
        ).removeprefix("<|begin_of_text|>")
        for convo in convos
    ]
    return {"text": texts}


def parse_label_array(raw_text):
    text = (raw_text or "").strip()
    if not text:
        return []

    if "```" in text:
        text = text.replace("```json", "").replace("```", "").strip()

    start = text.find("[")
    end = text.rfind("]")
    if start != -1 and end != -1 and end > start:
        text = text[start : end + 1]

    parsed = None
    for parser in (json.loads, ast.literal_eval):
        try:
            parsed = parser(text)
            break
        except Exception:
            continue

    if not isinstance(parsed, list):
        return []

    normalized = []
    for item in parsed:
        if not isinstance(item, str):
            normalized.append("not_supported")
            continue
        label = item.strip().lower().replace("-", "_").replace(" ", "_")
        if label not in {"supported", "not_supported"}:
            label = "not_supported"
        normalized.append(label)
    return normalized


def extract_conversation_pair(conversations):
    user_prompt = ""
    gold_response = ""
    for message in conversations:
        role = message.get("role") or message.get("from")
        content = message.get("content", "")
        if role == "user" and not user_prompt:
            user_prompt = content
        elif role == "assistant" and not gold_response:
            gold_response = content
    return user_prompt, gold_response


def generate_prediction(user_prompt):
    prompt = tokenizer.apply_chat_template(
        [{"role": "user", "content": user_prompt}],
        tokenize=False,
        add_generation_prompt=True,
    )
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
    with torch.inference_mode():
        outputs = model.generate(
            **inputs,
            max_new_tokens=128,
            do_sample=False,
            temperature=0.0,
            use_cache=True,
        )
    generated_tokens = outputs[0][inputs["input_ids"].shape[1] :]
    return tokenizer.decode(generated_tokens, skip_special_tokens=True).strip()


# 1. Load model and tokenizer
model, tokenizer = FastLanguageModel.from_pretrained(
    model_name=model_name,
    max_seq_length=max_seq_length,
    dtype=None,
    load_in_4bit=load_in_4bit,
)

# 2. Add LoRA adapters
model = FastLanguageModel.get_peft_model(
    model,
    r=16,
    target_modules=[
        "q_proj",
        "k_proj",
        "v_proj",
        "o_proj",
        "gate_proj",
        "up_proj",
        "down_proj",
    ],
    lora_alpha=16,
    lora_dropout=0,
    bias="none",
    use_gradient_checkpointing="unsloth",
    random_state=seed,
)

# 3. Data preparation
with open(data_path, "r", encoding="utf-8") as f:
    raw_data = json.load(f)

raw_dataset = Dataset.from_list(raw_data)
split_dataset = raw_dataset.train_test_split(test_size=test_size, seed=seed, shuffle=True)
train_raw = split_dataset["train"]
test_raw = split_dataset["test"]
train_dataset = train_raw.map(formatting_prompts_func, batched=True)

# 4. Save directories for this run
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
model_tag = model_name.split("/")[-1].replace(".", "_")
model_save_dir = f"/home/mshahidul/readctrl_model/support_checking_vllm/{model_tag}"
run_info_dir = os.path.join(
    "/home/mshahidul/readctrl/code/support_check/model_info",
    f"{model_tag}_{timestamp}",
)
os.makedirs(model_save_dir, exist_ok=True)
os.makedirs(run_info_dir, exist_ok=True)

# 5. Training setup
trainer = SFTTrainer(
    model=model,
    tokenizer=tokenizer,
    train_dataset=train_dataset,
    dataset_text_field="text",
    max_seq_length=max_seq_length,
    args=SFTConfig(
        per_device_train_batch_size=2,
        gradient_accumulation_steps=4,
        warmup_steps=5,
        max_steps=30,
        learning_rate=2e-4,
        fp16=not torch.cuda.is_bf16_supported(),
        bf16=torch.cuda.is_bf16_supported(),
        logging_steps=1,
        optim="adamw_8bit",
        weight_decay=0.01,
        lr_scheduler_type="linear",
        seed=seed,
        output_dir=os.path.join(run_info_dir, "trainer_outputs"),
        report_to="none",
    ),
)

# 6. Train
trainer.train()

# 7. Save merged model
model.save_pretrained_merged(model_save_dir, tokenizer, save_method="merged_16bit")
tokenizer.save_pretrained(model_save_dir)

# 8. Test-set inference + accuracy
FastLanguageModel.for_inference(model)
model.eval()

results = []
exact_match_correct = 0
label_correct = 0
label_total = 0
parsed_prediction_count = 0

for idx, sample in enumerate(test_raw):
    conversations = sample.get("conversations", [])
    user_prompt, gold_text = extract_conversation_pair(conversations)
    if not user_prompt:
        continue

    gold_labels = parse_label_array(gold_text)
    pred_text = generate_prediction(user_prompt)
    pred_labels = parse_label_array(pred_text)

    if pred_labels:
        parsed_prediction_count += 1

    exact_match = bool(gold_labels) and pred_labels == gold_labels
    if exact_match:
        exact_match_correct += 1

    sample_label_correct = 0
    for pos, gold_label in enumerate(gold_labels):
        if pos < len(pred_labels) and pred_labels[pos] == gold_label:
            sample_label_correct += 1

    label_correct += sample_label_correct
    label_total += len(gold_labels)

    results.append(
        {
            "sample_index": idx,
            "gold_labels": gold_labels,
            "predicted_labels": pred_labels,
            "raw_prediction": pred_text,
            "exact_match": exact_match,
            "label_accuracy": (
                sample_label_correct / len(gold_labels) if gold_labels else None
            ),
        }
    )

total_samples = len(results)
exact_match_accuracy = exact_match_correct / total_samples if total_samples else 0.0
label_accuracy = label_correct / label_total if label_total else 0.0

accuracy_summary = {
    "model_name": model_name,
    "model_save_dir": model_save_dir,
    "run_info_dir": run_info_dir,
    "dataset_path": data_path,
    "seed": seed,
    "test_size": test_size,
    "test_samples_evaluated": total_samples,
    "parsed_prediction_count": parsed_prediction_count,
    "exact_match_accuracy": exact_match_accuracy,
    "label_accuracy": label_accuracy,
    "exact_match_correct": exact_match_correct,
    "label_correct": label_correct,
    "label_total": label_total,
    "timestamp": timestamp,
}

predictions_path = os.path.join(run_info_dir, "test_inference.json")
accuracy_path = os.path.join(run_info_dir, "test_accuracy.json")

with open(predictions_path, "w", encoding="utf-8") as f:
    json.dump(results, f, ensure_ascii=False, indent=2)

with open(accuracy_path, "w", encoding="utf-8") as f:
    json.dump(accuracy_summary, f, ensure_ascii=False, indent=2)

print(f"Saved merged model to: {model_save_dir}")
print(f"Saved run info folder to: {run_info_dir}")
print(f"Saved test inference to: {predictions_path}")
print(f"Saved test accuracy to: {accuracy_path}")
print(f"Exact match accuracy: {exact_match_accuracy:.4f}")
print(f"Label accuracy: {label_accuracy:.4f}")