#!/bin/bash
# Test script for CodeLlama inference

cd /workspace/ftt/codellama-migration

MODEL_PATH="training-outputs/codellama-fifo-v1"
BASE_MODEL_PATH="models/base-models/CodeLlama-7B-Instruct"

echo "======================================================================"
echo "🧪 Testing CodeLlama Inference"
echo "======================================================================"
echo "Fine-tuned Model: $MODEL_PATH"
echo "Base Model: $BASE_MODEL_PATH"
echo "======================================================================"

# Activate virtual environment if needed
if [ -f /venv/main/bin/activate ]; then
    source /venv/main/bin/activate
fi

# Test prompt
TEST_PROMPT="You are Elinnos RTL Code Generator v1.0, a specialized Verilog/SystemVerilog code generation agent. Your role: Generate clean, synthesizable RTL code for hardware design tasks. Output ONLY functional RTL code with no \$display, assertions, comments, or debug statements.

Generate a synchronous FIFO with 8-bit data width, depth 4, write_enable, read_enable, full flag, empty flag."

echo ""
echo "📝 Test Prompt:"
echo "$TEST_PROMPT"
echo ""
echo "======================================================================"
echo "🚀 Running Inference..."
echo "======================================================================"
echo ""

# Run inference
python3 scripts/inference/inference_codellama.py \
    --mode local \
    --model-path "$MODEL_PATH" \
    --base-model-path "$BASE_MODEL_PATH" \
    --prompt "$TEST_PROMPT" \
    --max-new-tokens 800 \
    --temperature 0.3

echo ""
echo "======================================================================"
echo "✅ Inference Complete!"
echo "======================================================================"