Commit
•
eab46a7
1
Parent(s):
5d66701
upload model
Browse files- .gitattributes +27 -35
- README.md +90 -0
- config.json +177 -0
- merges.txt +0 -0
- preprocessor_config.json +17 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- vocab.json +0 -0
.gitattributes
CHANGED
@@ -1,35 +1,27 @@
|
|
1 |
-
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
-
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
-
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
-
*.
|
5 |
-
*.
|
6 |
-
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
-
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
-
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
-
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
-
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
-
*.
|
12 |
-
*.
|
13 |
-
*.
|
14 |
-
*.
|
15 |
-
*.
|
16 |
-
*.
|
17 |
-
*.
|
18 |
-
*.
|
19 |
-
*.
|
20 |
-
|
21 |
-
*.
|
22 |
-
*.
|
23 |
-
*.
|
24 |
-
*.
|
25 |
-
*.
|
26 |
-
|
27 |
-
|
28 |
-
*.tar filter=lfs diff=lfs merge=lfs -text
|
29 |
-
*.tflite filter=lfs diff=lfs merge=lfs -text
|
30 |
-
*.tgz filter=lfs diff=lfs merge=lfs -text
|
31 |
-
*.wasm filter=lfs diff=lfs merge=lfs -text
|
32 |
-
*.xz filter=lfs diff=lfs merge=lfs -text
|
33 |
-
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
-
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
-
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bin.* filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
20 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
+
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
README.md
ADDED
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- image-to-text
|
4 |
+
- image-captioning
|
5 |
+
license: apache-2.0
|
6 |
+
widget:
|
7 |
+
- src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/savanna.jpg
|
8 |
+
example_title: Savanna
|
9 |
+
- src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/football-match.jpg
|
10 |
+
example_title: Football Match
|
11 |
+
- src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/airport.jpg
|
12 |
+
example_title: Airport
|
13 |
+
---
|
14 |
+
|
15 |
+
# nlpconnect/vit-gpt2-image-captioning
|
16 |
+
|
17 |
+
This is an image captioning model trained by @ydshieh in [flax ](https://github.com/huggingface/transformers/tree/main/examples/flax/image-captioning) this is pytorch version of [this](https://huggingface.co/ydshieh/vit-gpt2-coco-en-ckpts).
|
18 |
+
|
19 |
+
|
20 |
+
# The Illustrated Image Captioning using transformers
|
21 |
+
|
22 |
+
![](https://ankur3107.github.io/assets/images/vision-encoder-decoder.png)
|
23 |
+
|
24 |
+
* https://ankur3107.github.io/blogs/the-illustrated-image-captioning-using-transformers/
|
25 |
+
|
26 |
+
|
27 |
+
# Sample running code
|
28 |
+
|
29 |
+
```python
|
30 |
+
|
31 |
+
from transformers import VisionEncoderDecoderModel, ViTImageProcessor, AutoTokenizer
|
32 |
+
import torch
|
33 |
+
from PIL import Image
|
34 |
+
|
35 |
+
model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
36 |
+
feature_extractor = ViTImageProcessor.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
37 |
+
tokenizer = AutoTokenizer.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
|
38 |
+
|
39 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
40 |
+
model.to(device)
|
41 |
+
|
42 |
+
|
43 |
+
|
44 |
+
max_length = 16
|
45 |
+
num_beams = 4
|
46 |
+
gen_kwargs = {"max_length": max_length, "num_beams": num_beams}
|
47 |
+
def predict_step(image_paths):
|
48 |
+
images = []
|
49 |
+
for image_path in image_paths:
|
50 |
+
i_image = Image.open(image_path)
|
51 |
+
if i_image.mode != "RGB":
|
52 |
+
i_image = i_image.convert(mode="RGB")
|
53 |
+
|
54 |
+
images.append(i_image)
|
55 |
+
|
56 |
+
pixel_values = feature_extractor(images=images, return_tensors="pt").pixel_values
|
57 |
+
pixel_values = pixel_values.to(device)
|
58 |
+
|
59 |
+
output_ids = model.generate(pixel_values, **gen_kwargs)
|
60 |
+
|
61 |
+
preds = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
|
62 |
+
preds = [pred.strip() for pred in preds]
|
63 |
+
return preds
|
64 |
+
|
65 |
+
|
66 |
+
predict_step(['doctor.e16ba4e4.jpg']) # ['a woman in a hospital bed with a woman in a hospital bed']
|
67 |
+
|
68 |
+
```
|
69 |
+
|
70 |
+
# Sample running code using transformers pipeline
|
71 |
+
|
72 |
+
```python
|
73 |
+
|
74 |
+
from transformers import pipeline
|
75 |
+
|
76 |
+
image_to_text = pipeline("image-to-text", model="nlpconnect/vit-gpt2-image-captioning")
|
77 |
+
|
78 |
+
image_to_text("https://ankur3107.github.io/assets/images/image-captioning-example.png")
|
79 |
+
|
80 |
+
# [{'generated_text': 'a soccer game with a player jumping to catch the ball '}]
|
81 |
+
|
82 |
+
|
83 |
+
```
|
84 |
+
|
85 |
+
|
86 |
+
# Contact for any help
|
87 |
+
* https://huggingface.co/ankur310794
|
88 |
+
* https://twitter.com/ankur310794
|
89 |
+
* http://github.com/ankur3107
|
90 |
+
* https://www.linkedin.com/in/ankur310794
|
config.json
ADDED
@@ -0,0 +1,177 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "vit-gpt-pt",
|
3 |
+
"architectures": [
|
4 |
+
"VisionEncoderDecoderModel"
|
5 |
+
],
|
6 |
+
"bos_token_id": 50256,
|
7 |
+
"decoder": {
|
8 |
+
"_name_or_path": "",
|
9 |
+
"activation_function": "gelu_new",
|
10 |
+
"add_cross_attention": true,
|
11 |
+
"architectures": [
|
12 |
+
"GPT2LMHeadModel"
|
13 |
+
],
|
14 |
+
"attn_pdrop": 0.1,
|
15 |
+
"bad_words_ids": null,
|
16 |
+
"bos_token_id": 50256,
|
17 |
+
"chunk_size_feed_forward": 0,
|
18 |
+
"cross_attention_hidden_size": null,
|
19 |
+
"decoder_start_token_id": 50256,
|
20 |
+
"diversity_penalty": 0.0,
|
21 |
+
"do_sample": false,
|
22 |
+
"early_stopping": false,
|
23 |
+
"embd_pdrop": 0.1,
|
24 |
+
"encoder_no_repeat_ngram_size": 0,
|
25 |
+
"eos_token_id": 50256,
|
26 |
+
"finetuning_task": null,
|
27 |
+
"forced_bos_token_id": null,
|
28 |
+
"forced_eos_token_id": null,
|
29 |
+
"id2label": {
|
30 |
+
"0": "LABEL_0",
|
31 |
+
"1": "LABEL_1"
|
32 |
+
},
|
33 |
+
"initializer_range": 0.02,
|
34 |
+
"is_decoder": true,
|
35 |
+
"is_encoder_decoder": false,
|
36 |
+
"label2id": {
|
37 |
+
"LABEL_0": 0,
|
38 |
+
"LABEL_1": 1
|
39 |
+
},
|
40 |
+
"layer_norm_epsilon": 1e-05,
|
41 |
+
"length_penalty": 1.0,
|
42 |
+
"max_length": 20,
|
43 |
+
"min_length": 0,
|
44 |
+
"model_type": "gpt2",
|
45 |
+
"n_ctx": 1024,
|
46 |
+
"n_embd": 768,
|
47 |
+
"n_head": 12,
|
48 |
+
"n_inner": null,
|
49 |
+
"n_layer": 12,
|
50 |
+
"n_positions": 1024,
|
51 |
+
"no_repeat_ngram_size": 0,
|
52 |
+
"num_beam_groups": 1,
|
53 |
+
"num_beams": 1,
|
54 |
+
"num_return_sequences": 1,
|
55 |
+
"output_attentions": false,
|
56 |
+
"output_hidden_states": false,
|
57 |
+
"output_scores": false,
|
58 |
+
"pad_token_id": 50256,
|
59 |
+
"prefix": null,
|
60 |
+
"problem_type": null,
|
61 |
+
"pruned_heads": {},
|
62 |
+
"remove_invalid_values": false,
|
63 |
+
"reorder_and_upcast_attn": false,
|
64 |
+
"repetition_penalty": 1.0,
|
65 |
+
"resid_pdrop": 0.1,
|
66 |
+
"return_dict": true,
|
67 |
+
"return_dict_in_generate": false,
|
68 |
+
"scale_attn_by_inverse_layer_idx": false,
|
69 |
+
"scale_attn_weights": true,
|
70 |
+
"sep_token_id": null,
|
71 |
+
"summary_activation": null,
|
72 |
+
"summary_first_dropout": 0.1,
|
73 |
+
"summary_proj_to_labels": true,
|
74 |
+
"summary_type": "cls_index",
|
75 |
+
"summary_use_proj": true,
|
76 |
+
"task_specific_params": {
|
77 |
+
"text-generation": {
|
78 |
+
"do_sample": true,
|
79 |
+
"max_length": 50
|
80 |
+
}
|
81 |
+
},
|
82 |
+
"temperature": 1.0,
|
83 |
+
"tie_encoder_decoder": false,
|
84 |
+
"tie_word_embeddings": true,
|
85 |
+
"tokenizer_class": null,
|
86 |
+
"top_k": 50,
|
87 |
+
"top_p": 1.0,
|
88 |
+
"torch_dtype": null,
|
89 |
+
"torchscript": false,
|
90 |
+
"transformers_version": "4.15.0",
|
91 |
+
"use_bfloat16": false,
|
92 |
+
"use_cache": true,
|
93 |
+
"vocab_size": 50257
|
94 |
+
},
|
95 |
+
"decoder_start_token_id": 50256,
|
96 |
+
"encoder": {
|
97 |
+
"_name_or_path": "",
|
98 |
+
"add_cross_attention": false,
|
99 |
+
"architectures": [
|
100 |
+
"ViTModel"
|
101 |
+
],
|
102 |
+
"attention_probs_dropout_prob": 0.0,
|
103 |
+
"bad_words_ids": null,
|
104 |
+
"bos_token_id": null,
|
105 |
+
"chunk_size_feed_forward": 0,
|
106 |
+
"cross_attention_hidden_size": null,
|
107 |
+
"decoder_start_token_id": null,
|
108 |
+
"diversity_penalty": 0.0,
|
109 |
+
"do_sample": false,
|
110 |
+
"early_stopping": false,
|
111 |
+
"encoder_no_repeat_ngram_size": 0,
|
112 |
+
"eos_token_id": null,
|
113 |
+
"finetuning_task": null,
|
114 |
+
"forced_bos_token_id": null,
|
115 |
+
"forced_eos_token_id": null,
|
116 |
+
"hidden_act": "gelu",
|
117 |
+
"hidden_dropout_prob": 0.0,
|
118 |
+
"hidden_size": 768,
|
119 |
+
"id2label": {
|
120 |
+
"0": "LABEL_0",
|
121 |
+
"1": "LABEL_1"
|
122 |
+
},
|
123 |
+
"image_size": 224,
|
124 |
+
"initializer_range": 0.02,
|
125 |
+
"intermediate_size": 3072,
|
126 |
+
"is_decoder": false,
|
127 |
+
"is_encoder_decoder": false,
|
128 |
+
"label2id": {
|
129 |
+
"LABEL_0": 0,
|
130 |
+
"LABEL_1": 1
|
131 |
+
},
|
132 |
+
"layer_norm_eps": 1e-12,
|
133 |
+
"length_penalty": 1.0,
|
134 |
+
"max_length": 20,
|
135 |
+
"min_length": 0,
|
136 |
+
"model_type": "vit",
|
137 |
+
"no_repeat_ngram_size": 0,
|
138 |
+
"num_attention_heads": 12,
|
139 |
+
"num_beam_groups": 1,
|
140 |
+
"num_beams": 1,
|
141 |
+
"num_channels": 3,
|
142 |
+
"num_hidden_layers": 12,
|
143 |
+
"num_return_sequences": 1,
|
144 |
+
"output_attentions": false,
|
145 |
+
"output_hidden_states": false,
|
146 |
+
"output_scores": false,
|
147 |
+
"pad_token_id": null,
|
148 |
+
"patch_size": 16,
|
149 |
+
"prefix": null,
|
150 |
+
"problem_type": null,
|
151 |
+
"pruned_heads": {},
|
152 |
+
"qkv_bias": true,
|
153 |
+
"remove_invalid_values": false,
|
154 |
+
"repetition_penalty": 1.0,
|
155 |
+
"return_dict": true,
|
156 |
+
"return_dict_in_generate": false,
|
157 |
+
"sep_token_id": null,
|
158 |
+
"task_specific_params": null,
|
159 |
+
"temperature": 1.0,
|
160 |
+
"tie_encoder_decoder": false,
|
161 |
+
"tie_word_embeddings": true,
|
162 |
+
"tokenizer_class": null,
|
163 |
+
"top_k": 50,
|
164 |
+
"top_p": 1.0,
|
165 |
+
"torch_dtype": null,
|
166 |
+
"torchscript": false,
|
167 |
+
"transformers_version": "4.15.0",
|
168 |
+
"use_bfloat16": false
|
169 |
+
},
|
170 |
+
"eos_token_id": 50256,
|
171 |
+
"is_encoder_decoder": true,
|
172 |
+
"model_type": "vision-encoder-decoder",
|
173 |
+
"pad_token_id": 50256,
|
174 |
+
"tie_word_embeddings": false,
|
175 |
+
"torch_dtype": "float32",
|
176 |
+
"transformers_version": null
|
177 |
+
}
|
merges.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
preprocessor_config.json
ADDED
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": true,
|
3 |
+
"do_resize": true,
|
4 |
+
"feature_extractor_type": "ViTFeatureExtractor",
|
5 |
+
"image_mean": [
|
6 |
+
0.5,
|
7 |
+
0.5,
|
8 |
+
0.5
|
9 |
+
],
|
10 |
+
"image_std": [
|
11 |
+
0.5,
|
12 |
+
0.5,
|
13 |
+
0.5
|
14 |
+
],
|
15 |
+
"resample": 2,
|
16 |
+
"size": 224
|
17 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2605a69b760d0b218c3b5d2069ba070d28f279a16ce4bc87bf019b75a91553e6
|
3 |
+
size 982141993
|
special_tokens_map.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "unk_token": "<|endoftext|>", "pad_token": "<|endoftext|>"}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "./models/", "tokenizer_class": "GPT2Tokenizer"}
|
vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|