#pragma once

#include "engine/framework/assets/tensor_source.h"
#include "engine/framework/core/execution_context.h"
#include "engine/models/qwen3_asr/assets.h"
#include "engine/models/qwen3_asr/types.h"

#include <cstddef>
#include <memory>

namespace engine::models::qwen3_asr {

class Qwen3ASRThinkerRuntime {
public:
    struct Impl;

    Qwen3ASRThinkerRuntime(
        std::shared_ptr<const Qwen3ASRAssets> assets,
        core::ExecutionContext & execution,
        size_t prefill_graph_arena_bytes,
        size_t decode_graph_arena_bytes,
        size_t weight_context_bytes,
        assets::TensorStorageType weight_storage_type);
    ~Qwen3ASRThinkerRuntime();

    Qwen3ASRGeneratedTokens generate(
        const Qwen3ASRPrompt & prompt,
        const Qwen3ASRAudioEmbeddings & audio_embeddings,
        const Qwen3ASRGenerationOptions & options);

    std::vector<int32_t> classify_prompt(
        const Qwen3ASRPrompt & prompt,
        const Qwen3ASRAudioEmbeddings & audio_embeddings);

private:
    std::unique_ptr<Impl> impl_;
};

}  // namespace engine::models::qwen3_asr