# Copyright (c) 2025 Baidu
# Licensed under the MIT License. See LICENSE file in the project root for full license information.
import warnings
from typing import List, Optional, Tuple, Union
from transformers  import AutoTokenizer
import torch.distributed as dist
import torch.utils.checkpoint
import transformers
from .conversation import get_conv_template
from torch import nn
from torch.nn import CrossEntropyLoss
from transformers import (AutoModel, GenerationConfig, Qwen2ForCausalLM)
from transformers.modeling_outputs import CausalLMOutputWithPast
from transformers.modeling_utils import PreTrainedModel
from transformers.utils import ModelOutput, logging

from .configuration_qianfanvl_chat import QianfanVLChatConfig
from .modeling_intern_vit import InternVisionModel, has_flash_attn
import difflib

logger = logging.get_logger(__name__)


def version_cmp(v1, v2, op='eq'):
    import operator

    from packaging import version
    op_func = getattr(operator, op)
    return op_func(version.parse(v1), version.parse(v2))


class QianfanVLChatModel(PreTrainedModel):
    config_class = QianfanVLChatConfig
    main_input_name = 'pixel_values'
    base_model_prefix = 'language_model'
    _no_split_modules = ['InternVisionModel', 'Qwen2DecoderLayer']
    _supports_flash_attn_2 = True
    supports_gradient_checkpointing = True

    def __init__(self, config: QianfanVLChatConfig, vision_model=None, language_model=None, use_flash_attn=True):
        super().__init__(config)

        assert version_cmp(transformers.__version__, '4.37.0', 'ge')
        image_size = config.force_image_size or config.vision_config.image_size
        patch_size = config.vision_config.patch_size
        self.patch_size = patch_size
        self.select_layer = config.select_layer
        self.template = config.template
        self.num_image_token = int((image_size // patch_size) ** 2 * (config.downsample_ratio ** 2))
        self.downsample_ratio = config.downsample_ratio
        self.ps_version = config.ps_version
        self.llm_arch_name = config.llm_config.architectures[0]
        # Enable Flash Attention if supported, otherwise fall back to eager attention.
        use_flash_attn = False #  use_flash_attn if has_flash_attn else False
        config.vision_config.use_flash_attn = True if use_flash_attn else False
        config.llm_config.attn_implementation = 'flash_attention_2' if use_flash_attn else 'eager'

        if vision_model is not None:
            self.vision_model = vision_model
        else:
            self.vision_model = InternVisionModel(config.vision_config)
        if language_model is not None:
            self.language_model = language_model
        else:
            self.language_model = Qwen2ForCausalLM(config.llm_config)

        vit_hidden_size = config.vision_config.hidden_size
        llm_hidden_size = config.llm_config.hidden_size
        
        self.mlp1 = nn.Sequential(
            nn.LayerNorm(vit_hidden_size * int(1 / self.downsample_ratio) ** 2),
            nn.Linear(vit_hidden_size * int(1 / self.downsample_ratio) ** 2, llm_hidden_size),
            nn.GELU(),
            nn.Linear(llm_hidden_size, llm_hidden_size)
        )

        self.img_context_token_id = None
        self.conv_template = get_conv_template(self.template)
        if hasattr(config, 'system_message'):
            self.system_message = config.system_message
        else:
            self.system_message = self.conv_template.system_message
        self.num_samples = 0

    def forward(
            self,
            pixel_values: torch.FloatTensor,
            input_ids: torch.LongTensor = None,
            attention_mask: Optional[torch.Tensor] = None,
            position_ids: Optional[torch.LongTensor] = None,
            image_flags: Optional[torch.LongTensor] = None,
            past_key_values: Optional[List[torch.FloatTensor]] = None,
            labels: Optional[torch.LongTensor] = None,
            use_cache: Optional[bool] = None,
            output_attentions: Optional[bool] = None,
            output_hidden_states: Optional[bool] = None,
            return_dict: Optional[bool] = None,
            statistics: Optional[torch.LongTensor] = None,
            loss_weight: Optional[List] = None,
            loss_reduction_all_gather: Optional[bool] = False,
    ) -> Union[Tuple, CausalLMOutputWithPast]:
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        image_flags = image_flags.squeeze(-1)
        input_embeds = self.language_model.get_input_embeddings()(input_ids).clone()
        vit_embeds = self.extract_feature(pixel_values)
        vit_embeds = vit_embeds[image_flags == 1]
        vit_batch_size = pixel_values.shape[0]
        B, N, C = input_embeds.shape
        input_embeds = input_embeds.reshape(B * N, C)

        if torch.distributed.is_initialized() and torch.distributed.get_rank() == 0:
            print(f'dynamic ViT batch size: {vit_batch_size}, images per sample: {vit_batch_size / B}, dynamic token length: {N}')
            if statistics is not None:
                num_samples, num_padding_tokens, num_padding_images = statistics.tolist()
                self.num_samples += num_samples
                print(f'total_samples={self.num_samples}, {num_samples=}, {num_padding_tokens=}, {num_padding_images=}')
        
        input_ids = input_ids.reshape(B * N)
        selected = (input_ids == self.img_context_token_id)
        try:
            input_embeds[selected] = input_embeds[selected] * 0.0 + vit_embeds.reshape(-1, C)
            ignore_flag = False
        except Exception as e:
            vit_embeds = vit_embeds.reshape(-1, C)
            print(f'warning: {e}, input_embeds[selected].shape={input_embeds[selected].shape}, '
                  f'vit_embeds.shape={vit_embeds.shape}')
            n_token = selected.sum()
            input_embeds[selected] = input_embeds[selected] * 0.0 + vit_embeds[:n_token]
            ignore_flag = True

        input_embeds = input_embeds.reshape(B, N, C)

        outputs = self.language_model(
            inputs_embeds=input_embeds,
            attention_mask=attention_mask,
            position_ids=position_ids,
            past_key_values=past_key_values,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )
        logits = outputs.logits

        loss = None
        if labels is not None and loss_weight is not None:
            loss_weight = torch.tensor(loss_weight, dtype=torch.float32, device=labels.device)
            # Shift so that tokens < n predict n
            shift_logits = logits[..., :-1, :].contiguous()
            shift_labels = labels[..., 1:].contiguous()
            shift_weights = loss_weight[..., 1:].contiguous()
            # Flatten the tokens
            loss_fct = CrossEntropyLoss(reduction='none')
            shift_logits = shift_logits.view(-1, self.language_model.config.vocab_size)
            shift_labels = shift_labels.view(-1)
            shift_weights = shift_weights.view(-1)
            # Enable model parallelism
            shift_labels = shift_labels.to(shift_logits.device)
            shift_weights = shift_weights.to(shift_logits.device)
            loss = loss_fct(shift_logits, shift_labels)

            shift_weights_sum = shift_weights.sum()
            if loss_reduction_all_gather:
                dist.all_reduce(shift_weights_sum, op=dist.ReduceOp.AVG)

            loss = loss * shift_weights
            loss = loss.sum() / shift_weights_sum
            if ignore_flag:
                loss = loss * 0.0
        elif labels is not None:
            # Shift so that tokens < n predict n
            shift_logits = logits[..., :-1, :].contiguous()
            shift_labels = labels[..., 1:].contiguous()
            # Flatten the tokens
            loss_fct = CrossEntropyLoss()
            shift_logits = shift_logits.view(-1, self.language_model.config.vocab_size)
            shift_labels = shift_labels.view(-1)
            # Enable model parallelism
            shift_labels = shift_labels.to(shift_logits.device)
            loss = loss_fct(shift_logits, shift_labels)
            if ignore_flag:
                loss = loss * 0.0

        if not return_dict:
            output = (logits,) + outputs[1:]
            return (loss,) + output if loss is not None else output

        return CausalLMOutputWithPast(
            loss=loss,
            logits=logits,
            past_key_values=outputs.past_key_values,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
        )

    def pixel_shuffle(self, x, scale_factor=0.5):
        n, w, h, c = x.size()
        # N, W, H, C --> N, W, H * scale, C // scale
        x = x.view(n, w, int(h * scale_factor), int(c / scale_factor))
        # N, W, H * scale, C // scale --> N, H * scale, W, C // scale
        x = x.permute(0, 2, 1, 3).contiguous()
        # N, H * scale, W, C // scale --> N, H * scale, W * scale, C // (scale ** 2)
        x = x.view(n, int(h * scale_factor), int(w * scale_factor),
                   int(c / (scale_factor * scale_factor)))
        if self.ps_version == 'v1':
            warnings.warn("In ps_version 'v1', the height and width have not been swapped back, "
                          'which results in a transposed image.')
        else:
            x = x.permute(0, 2, 1, 3).contiguous()
        return x

    def extract_feature(self, pixel_values):
        if self.select_layer == -1:
            vit_embeds = self.vision_model(
                pixel_values=pixel_values,
                output_hidden_states=False,
                return_dict=True).last_hidden_state
        else:
            vit_embeds = self.vision_model(
                pixel_values=pixel_values,
                output_hidden_states=True,
                return_dict=True).hidden_states[self.select_layer]
        vit_embeds = vit_embeds[:, 1:, :]
        h = w = int(vit_embeds.shape[1] ** 0.5)
        vit_embeds = vit_embeds.reshape(vit_embeds.shape[0], h, w, -1)
        vit_embeds = self.pixel_shuffle(vit_embeds, scale_factor=self.downsample_ratio)
        vit_embeds = vit_embeds.reshape(vit_embeds.shape[0], -1, vit_embeds.shape[-1])
        vit_embeds = self.mlp1(vit_embeds)
        return vit_embeds
    
    def batch_chat(self, tokenizer, pixel_values, questions, generation_config, num_patches_list=None,
                   history=None, return_history=False, IMG_START_TOKEN='<img>', IMG_END_TOKEN='</img>',
                   IMG_CONTEXT_TOKEN='<IMG_CONTEXT>', verbose=False, image_counts=None):
        if history is not None or return_history:
            print('Now multi-turn chat is not supported in batch_chat.')
            raise NotImplementedError

        if image_counts is not None:
            num_patches_list = image_counts
            print('Warning: `image_counts` is deprecated. Please use `num_patches_list` instead.')

        img_context_token_id = tokenizer.convert_tokens_to_ids(IMG_CONTEXT_TOKEN)
        self.img_context_token_id = img_context_token_id

        if verbose and pixel_values is not None:
            image_bs = pixel_values.shape[0]
            print(f'dynamic ViT batch size: {image_bs}')

        queries = []
        for idx, num_patches in enumerate(num_patches_list):
            question = questions[idx]
            if pixel_values is not None and '<image>' not in question:
                question = '<image>' + question
            template = get_conv_template(self.template)
            template.system_message = self.system_message
            template.append_message(template.roles[0], question)
            template.append_message(template.roles[1], None)
            query = template.get_prompt()

            image_tokens = IMG_START_TOKEN + IMG_CONTEXT_TOKEN * self.num_image_token * num_patches + IMG_END_TOKEN
            query = query.replace('<image>', image_tokens, 1)
            queries.append(query)

        tokenizer.padding_side = 'left'
        model_inputs = tokenizer(queries, return_tensors='pt', padding=True)
        device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
        input_ids = model_inputs['input_ids'].to(device)
        attention_mask = model_inputs['attention_mask'].to(device)
        eos_token_id = tokenizer.convert_tokens_to_ids(template.sep.strip())
        generation_config['eos_token_id'] = eos_token_id
        #generation_config['eos_token_id'] = 181896
        generation_output = self.generate(
            pixel_values=pixel_values,
            input_ids=input_ids,
            attention_mask=attention_mask,
            **generation_config
        )
        responses = tokenizer.batch_decode(generation_output, skip_special_tokens=False)
        responses = [response.split(template.sep.strip())[0].strip() for response in responses]
        return responses

    def chat(self, tokenizer, pixel_values, question, generation_config, history=None, return_history=False,
             num_patches_list=None, IMG_START_TOKEN='<img>', IMG_END_TOKEN='</img>', IMG_CONTEXT_TOKEN='<IMG_CONTEXT>',
             verbose=False, debug=False):

        if history is None and pixel_values is not None and '<image>' not in question:
          question = '<image>' + question

        if num_patches_list is None:
            num_patches_list = [pixel_values.shape[0]] if pixel_values is not None else []
        assert pixel_values is None or len(pixel_values) == sum(num_patches_list)

        img_context_token_id = tokenizer.convert_tokens_to_ids(IMG_CONTEXT_TOKEN)
        self.img_context_token_id = img_context_token_id

        template = get_conv_template(self.template)
        eos_token_id = tokenizer.convert_tokens_to_ids(template.sep.strip())

        history = [] if history is None else history
        for (old_question, old_answer) in history:
            template.append_message(template.roles[0], old_question)
            template.append_message(template.roles[1], old_answer)
        template.append_message(template.roles[0], question)
        template.append_message(template.roles[1], None)
        query = template.get_prompt()

        if verbose and pixel_values is not None:
            image_bs = pixel_values.shape[0]
            print(f'dynamic ViT batch size: {image_bs}')

        for num_patches in num_patches_list:
            image_tokens = IMG_START_TOKEN + IMG_CONTEXT_TOKEN * self.num_image_token * num_patches + IMG_END_TOKEN
            query = query.replace('<image>', image_tokens, 1)

        model_inputs = tokenizer(query, return_tensors='pt')
        device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
        input_ids = model_inputs['input_ids'].to(device)
        attention_mask = model_inputs['attention_mask'].to(device)
        generation_config['eos_token_id'] = eos_token_id

        generation_output = self.generate(
            pixel_values=pixel_values,
            input_ids=input_ids,
            attention_mask=attention_mask,
            **generation_config
        )
        
        if debug:
            return input_ids, generation_output
        response = tokenizer.batch_decode(generation_output, skip_special_tokens=False)[0]
        response = response.split(template.sep.strip())[0].strip()
        history.append((question, response))
        if return_history:
            return response, history
        else:
            query_to_print = query.replace(IMG_CONTEXT_TOKEN, '')
            query_to_print = query_to_print.replace(f'{IMG_START_TOKEN}{IMG_END_TOKEN}', '<image>')
            if verbose:
                print(query_to_print, response)
            return response

    @torch.no_grad()
    def generate(
            self,
            pixel_values: Optional[torch.FloatTensor] = None,
            input_ids: Optional[torch.FloatTensor] = None,
            attention_mask: Optional[torch.LongTensor] = None,
            visual_features: Optional[torch.FloatTensor] = None,
            generation_config: Optional[GenerationConfig] = None,
            output_hidden_states: Optional[bool] = None,
            **generate_kwargs,
    ):

        assert self.img_context_token_id is not None
        if pixel_values is not None:
            if visual_features is not None:
                vit_embeds = visual_features
            else:
                vit_embeds = self.extract_feature(pixel_values)
            #print("ViTEmbedding shape:", vit_embeds.shape)
            input_embeds = self.language_model.get_input_embeddings()(input_ids)
            B, N, C = input_embeds.shape
            input_embeds = input_embeds.reshape(B * N, C)
            input_ids = input_ids.reshape(B * N)
            selected = (input_ids == self.img_context_token_id)
            assert selected.sum() != 0
            input_embeds[selected] = vit_embeds.reshape(-1, C).to(input_embeds.device)

            input_embeds = input_embeds.reshape(B, N, C)
        else:
            input_embeds = self.language_model.get_input_embeddings()(input_ids)

        outputs = self.language_model.generate(
            inputs_embeds=input_embeds,
            attention_mask=attention_mask,
            generation_config=generation_config,
            output_hidden_states=output_hidden_states,
            use_cache=True,
            **generate_kwargs,
        )

        return outputs

    @property
    def lm_head(self):
        return self.language_model.get_output_embeddings()

    def get_input_embeddings(self):
        return self.language_model.get_input_embeddings()

    def get_output_embeddings(self):
        return self.language_model.get_output_embeddings()