Upload LlavaBaichuanForCausalLM
Browse files- config.json +3 -2
- llava_baichuan.py +145 -0
config.json
CHANGED
@@ -5,8 +5,9 @@
|
|
5 |
"LlavaBaichuanForCausalLM"
|
6 |
],
|
7 |
"auto_map": {
|
8 |
-
"AutoConfig": "
|
9 |
-
"AutoModelForCausalLM": "baichuan-inc/Baichuan2-7B-Chat--modeling_baichuan.BaichuanForCausalLM"
|
|
|
10 |
},
|
11 |
"bos_token_id": 1,
|
12 |
"eos_token_id": 2,
|
|
|
5 |
"LlavaBaichuanForCausalLM"
|
6 |
],
|
7 |
"auto_map": {
|
8 |
+
"AutoConfig": "llava_baichuan.LlavaBaichuanConfig",
|
9 |
+
"AutoModelForCausalLM": "baichuan-inc/Baichuan2-7B-Chat--modeling_baichuan.BaichuanForCausalLM",
|
10 |
+
"AutoModelForVisualQuestionAnswering": "llava_baichuan.LlavaBaichuanForCausalLM"
|
11 |
},
|
12 |
"bos_token_id": 1,
|
13 |
"eos_token_id": 2,
|
llava_baichuan.py
ADDED
@@ -0,0 +1,145 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import List, Optional, Tuple, Union
|
2 |
+
|
3 |
+
import torch
|
4 |
+
import torch.nn as nn
|
5 |
+
|
6 |
+
from transformers import AutoConfig, AutoModelForCausalLM
|
7 |
+
|
8 |
+
from transformers.modeling_outputs import CausalLMOutputWithPast
|
9 |
+
from transformers.generation.utils import GenerateOutput
|
10 |
+
|
11 |
+
from llava_arch import LlavaMetaModel, LlavaMetaForCausalLM
|
12 |
+
|
13 |
+
from configuration_baichuan import BaichuanConfig
|
14 |
+
from modeling_baichuan import BaichuanModel, BaichuanForCausalLM
|
15 |
+
|
16 |
+
|
17 |
+
class LlavaBaichuanConfig(BaichuanConfig):
|
18 |
+
model_type = "llava_baichuan"
|
19 |
+
|
20 |
+
|
21 |
+
class LlavaBaichuanModel(LlavaMetaModel, BaichuanModel):
|
22 |
+
config_class = LlavaBaichuanConfig
|
23 |
+
|
24 |
+
def __init__(self, config: BaichuanConfig):
|
25 |
+
super(LlavaBaichuanModel, self).__init__(config)
|
26 |
+
|
27 |
+
|
28 |
+
class LlavaBaichuanForCausalLM(BaichuanForCausalLM, LlavaMetaForCausalLM):
|
29 |
+
config_class = LlavaBaichuanConfig
|
30 |
+
|
31 |
+
def __init__(self, config):
|
32 |
+
super(BaichuanForCausalLM, self).__init__(config)
|
33 |
+
self.model = LlavaBaichuanModel(config)
|
34 |
+
|
35 |
+
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
|
36 |
+
|
37 |
+
# Initialize weights and apply final processing
|
38 |
+
self.post_init()
|
39 |
+
|
40 |
+
def get_model(self):
|
41 |
+
return self.model
|
42 |
+
|
43 |
+
def forward(
|
44 |
+
self,
|
45 |
+
input_ids: torch.LongTensor = None,
|
46 |
+
attention_mask: Optional[torch.Tensor] = None,
|
47 |
+
position_ids: Optional[torch.LongTensor] = None,
|
48 |
+
past_key_values: Optional[List[torch.FloatTensor]] = None,
|
49 |
+
inputs_embeds: Optional[torch.FloatTensor] = None,
|
50 |
+
labels: Optional[torch.LongTensor] = None,
|
51 |
+
use_cache: Optional[bool] = None,
|
52 |
+
output_attentions: Optional[bool] = None,
|
53 |
+
output_hidden_states: Optional[bool] = None,
|
54 |
+
images: Optional[torch.FloatTensor] = None,
|
55 |
+
image_sizes: Optional[List[List[int]]] = None,
|
56 |
+
return_dict: Optional[bool] = None,
|
57 |
+
) -> Union[Tuple, CausalLMOutputWithPast]:
|
58 |
+
|
59 |
+
if inputs_embeds is None:
|
60 |
+
(
|
61 |
+
input_ids,
|
62 |
+
position_ids,
|
63 |
+
attention_mask,
|
64 |
+
past_key_values,
|
65 |
+
inputs_embeds,
|
66 |
+
labels
|
67 |
+
) = self.prepare_inputs_labels_for_multimodal(
|
68 |
+
input_ids,
|
69 |
+
position_ids,
|
70 |
+
attention_mask,
|
71 |
+
past_key_values,
|
72 |
+
labels,
|
73 |
+
images,
|
74 |
+
image_sizes
|
75 |
+
)
|
76 |
+
|
77 |
+
return super().forward(
|
78 |
+
input_ids=input_ids,
|
79 |
+
attention_mask=attention_mask,
|
80 |
+
position_ids=position_ids,
|
81 |
+
past_key_values=past_key_values,
|
82 |
+
inputs_embeds=inputs_embeds,
|
83 |
+
labels=labels,
|
84 |
+
use_cache=use_cache,
|
85 |
+
output_attentions=output_attentions,
|
86 |
+
output_hidden_states=output_hidden_states,
|
87 |
+
return_dict=return_dict
|
88 |
+
)
|
89 |
+
|
90 |
+
@torch.no_grad()
|
91 |
+
def generate(
|
92 |
+
self,
|
93 |
+
inputs: Optional[torch.Tensor] = None,
|
94 |
+
images: Optional[torch.Tensor] = None,
|
95 |
+
image_sizes: Optional[torch.Tensor] = None,
|
96 |
+
**kwargs,
|
97 |
+
) -> Union[GenerateOutput, torch.LongTensor]:
|
98 |
+
position_ids = kwargs.pop("position_ids", None)
|
99 |
+
attention_mask = kwargs.pop("attention_mask", None)
|
100 |
+
if "inputs_embeds" in kwargs:
|
101 |
+
raise NotImplementedError("`inputs_embeds` is not supported")
|
102 |
+
|
103 |
+
if images is not None:
|
104 |
+
(
|
105 |
+
inputs,
|
106 |
+
position_ids,
|
107 |
+
attention_mask,
|
108 |
+
_,
|
109 |
+
inputs_embeds,
|
110 |
+
_
|
111 |
+
) = self.prepare_inputs_labels_for_multimodal(
|
112 |
+
inputs,
|
113 |
+
position_ids,
|
114 |
+
attention_mask,
|
115 |
+
None,
|
116 |
+
None,
|
117 |
+
images,
|
118 |
+
image_sizes=image_sizes
|
119 |
+
)
|
120 |
+
else:
|
121 |
+
inputs_embeds = self.get_model().embed_tokens(inputs)
|
122 |
+
|
123 |
+
return super().generate(
|
124 |
+
position_ids=position_ids,
|
125 |
+
attention_mask=attention_mask,
|
126 |
+
inputs_embeds=inputs_embeds,
|
127 |
+
**kwargs
|
128 |
+
)
|
129 |
+
|
130 |
+
def prepare_inputs_for_generation(self, input_ids, past_key_values=None,
|
131 |
+
inputs_embeds=None, **kwargs):
|
132 |
+
images = kwargs.pop("images", None)
|
133 |
+
image_sizes = kwargs.pop("image_sizes", None)
|
134 |
+
inputs = super().prepare_inputs_for_generation(
|
135 |
+
input_ids, past_key_values=past_key_values, inputs_embeds=inputs_embeds, **kwargs
|
136 |
+
)
|
137 |
+
if images is not None:
|
138 |
+
inputs['images'] = images
|
139 |
+
if image_sizes is not None:
|
140 |
+
inputs['image_sizes'] = image_sizes
|
141 |
+
return inputs
|
142 |
+
|
143 |
+
|
144 |
+
AutoConfig.register("llava_baichuan", LlavaBaichuanConfig)
|
145 |
+
AutoModelForCausalLM.register(LlavaBaichuanConfig, LlavaBaichuanForCausalLM)
|