Spaces:
Running
on
Zero
Running
on
Zero
File size: 10,079 Bytes
bc752b1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 |
from typing import List, Optional, Tuple, Union
import torch
import torch.nn as nn
from PIL import Image
from torch.nn import CrossEntropyLoss
from transformers import (
AutoConfig,
AutoModelForCausalLM,
MistralConfig,
MistralForCausalLM,
MistralModel,
)
from transformers.cache_utils import Cache, DynamicCache
from transformers.modeling_outputs import CausalLMOutputWithPast, MoeCausalLMOutputWithPast
from transformers.generation.utils import GenerateOutput
from ..vita_arch import VITAMetaForCausalLM, VITAMetaModel
def custom_forward(
self,
input_ids: torch.LongTensor = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.LongTensor] = None,
past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None,
inputs_embeds: Optional[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,
cache_position: Optional[torch.LongTensor] = None,
) -> Union[Tuple, CausalLMOutputWithPast]:
r"""
Args:
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
(masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
Returns:
Example:
```python
>>> from transformers import AutoTokenizer, MistralForCausalLM
>>> model = MistralForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1")
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")
>>> prompt = "Hey, are you conscious? Can you talk to me?"
>>> inputs = tokenizer(prompt, return_tensors="pt")
>>> # Generate
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
```"""
output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
output_hidden_states = (
output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
)
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
# decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
outputs = self.model(
input_ids=input_ids,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_values=past_key_values,
inputs_embeds=inputs_embeds,
use_cache=use_cache,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
cache_position=cache_position,
)
hidden_states = outputs[0]
logits = self.lm_head(hidden_states)
# logits = logits.float()
loss = None
if labels is not None:
# Shift so that tokens < n predict n
shift_logits = logits[..., :-1, :].contiguous()
shift_labels = labels[..., 1:].contiguous()
# Flatten the tokens
shift_logits = shift_logits.view(-1, self.config.vocab_size)
shift_labels = shift_labels.view(-1)
# Ensure tensors are on the same device
shift_labels = shift_labels.to(shift_logits.device)
loss_fct = CrossEntropyLoss()
loss = loss_fct(shift_logits, shift_labels)
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,
)
MistralForCausalLM.forward = custom_forward
class VITAMistralConfig(MistralConfig):
model_type = "vita-Mistral"
class VITAMistralModel(VITAMetaModel, MistralModel):
config_class = VITAMistralConfig
def __init__(self, config: MistralConfig):
super(VITAMistralModel, self).__init__(config)
class VITAMistralForCausalLM(MistralForCausalLM, VITAMetaForCausalLM):
config_class = VITAMistralConfig
def __init__(self, config):
super(MistralForCausalLM, self).__init__(config)
self.model = VITAMistralModel(config)
self.vocab_size = config.vocab_size
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
# Initialize weights and apply final processing
self.post_init()
def get_model(self):
return self.model
def forward(
self,
input_ids: torch.LongTensor = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.LongTensor] = None,
past_key_values: Optional[List[torch.FloatTensor]] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
labels: Optional[torch.LongTensor] = None,
use_cache: Optional[bool] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
images: Optional[torch.FloatTensor] = None,
audios: Optional[dict] = None,
return_dict: Optional[bool] = None,
cache_position: Optional[torch.LongTensor] = None,
) -> Union[Tuple, CausalLMOutputWithPast]:
if inputs_embeds is None:
(
input_ids,
position_ids,
attention_mask,
past_key_values,
inputs_embeds,
labels,
) = self.prepare_inputs_labels_for_multimodal(
input_ids, position_ids, attention_mask, past_key_values, labels, images, audios
)
return super().forward(
input_ids=input_ids,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_values=past_key_values,
inputs_embeds=inputs_embeds,
labels=labels,
use_cache=use_cache,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
cache_position=cache_position,
)
@torch.no_grad()
def generate(
self,
inputs: Optional[torch.Tensor] = None,
images: Optional[torch.Tensor] = None,
audios: Optional[torch.Tensor] = None,
**kwargs,
) -> Union[GenerateOutput, torch.LongTensor]:
position_ids = kwargs.pop("position_ids", None)
attention_mask = kwargs.pop("attention_mask", None)
if "inputs_embeds" in kwargs:
raise NotImplementedError("`inputs_embeds` is not supported")
if images is not None or audios is not None:
(
inputs,
position_ids,
attention_mask,
_,
inputs_embeds,
_
) = self.prepare_inputs_labels_for_multimodal(
inputs,
position_ids,
attention_mask,
None,
None,
images,
audios
)
else:
inputs_embeds = self.get_model().embed_tokens(inputs)
return super().generate(
position_ids=position_ids,
attention_mask=attention_mask,
inputs_embeds=inputs_embeds,
**kwargs
)
def prepare_inputs_for_generation(
self,
input_ids,
past_key_values=None,
inputs_embeds=None,
attention_mask=None,
**kwargs,
):
images = kwargs.pop("images", None)
audios = kwargs.pop("audios", None)
_inputs = super().prepare_inputs_for_generation(
input_ids,
past_key_values=past_key_values,
inputs_embeds=inputs_embeds,
attention_mask=attention_mask,
**kwargs,
)
if images is not None:
_inputs["images"] = images
if audios is not None:
_inputs["audios"] = audios
return _inputs
def expand2square(self, pil_img, background_color):
width, height = pil_img.size
if width == height:
return pil_img
elif width > height:
result = Image.new(pil_img.mode, (width, width), background_color)
result.paste(pil_img, (0, (width - height) // 2))
return result
else:
result = Image.new(pil_img.mode, (height, height), background_color)
result.paste(pil_img, ((height - width) // 2, 0))
return result
def process_images(self, images, model_cfg):
vision_tower = self.get_vision_tower()
if not vision_tower.is_loaded:
vision_tower.load_model()
image_processor = vision_tower.image_processor
image_aspect_ratio = getattr(model_cfg, "image_aspect_ratio", None)
new_images = []
if image_aspect_ratio == "pad":
for image in images:
image = self.expand2square(
image, tuple(int(x * 255) for x in image_processor.image_mean)
)
image = image_processor.preprocess(image, return_tensors="pt")["pixel_values"][0]
new_images.append(image)
else:
return image_processor(images, return_tensors="pt")["pixel_values"]
if all(x.shape == new_images[0].shape for x in new_images):
new_images = torch.stack(new_images, dim=0)
return new_images
AutoConfig.register("vita-Mistral", VITAMistralConfig)
AutoModelForCausalLM.register(VITAMistralConfig, VITAMistralForCausalLM)
|