oscarloch commited on
Commit
050dc89
·
verified ·
1 Parent(s): 6d36f43

Upload model

Browse files
Files changed (5) hide show
  1. README.md +199 -0
  2. config.json +2252 -0
  3. generation_config.json +5 -0
  4. model.safetensors +3 -0
  5. modelling_longitudinal.py +512 -0
README.md ADDED
@@ -0,0 +1,199 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ tags: []
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+ This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
config.json ADDED
@@ -0,0 +1,2252 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LongitudinalPromptMultiCXREncoderDecoderModel"
4
+ ],
5
+ "auto_map": {
6
+ "AutoModel": "modelling_longitudinal.LongitudinalPromptMultiCXREncoderDecoderModel"
7
+ },
8
+ "decoder": {
9
+ "_name_or_path": "",
10
+ "add_cross_attention": true,
11
+ "architectures": null,
12
+ "attention_probs_dropout_prob": 0.1,
13
+ "bad_words_ids": null,
14
+ "begin_suppress_tokens": null,
15
+ "bos_token_id": null,
16
+ "chunk_size_feed_forward": 0,
17
+ "classifier_dropout": null,
18
+ "cross_attention_hidden_size": null,
19
+ "decoder_start_token_id": null,
20
+ "diversity_penalty": 0.0,
21
+ "do_sample": false,
22
+ "early_stopping": false,
23
+ "encoder_no_repeat_ngram_size": 0,
24
+ "eos_token_id": null,
25
+ "exponential_decay_length_penalty": null,
26
+ "finetuning_task": null,
27
+ "forced_bos_token_id": null,
28
+ "forced_eos_token_id": null,
29
+ "hidden_act": "gelu",
30
+ "hidden_dropout_prob": 0.1,
31
+ "hidden_size": 768,
32
+ "id2label": {
33
+ "0": "LABEL_0",
34
+ "1": "LABEL_1"
35
+ },
36
+ "initializer_range": 0.02,
37
+ "intermediate_size": 3072,
38
+ "is_decoder": true,
39
+ "is_encoder_decoder": false,
40
+ "label2id": {
41
+ "LABEL_0": 0,
42
+ "LABEL_1": 1
43
+ },
44
+ "layer_norm_eps": 1e-12,
45
+ "length_penalty": 1.0,
46
+ "max_length": 20,
47
+ "max_position_embeddings": 512,
48
+ "min_length": 0,
49
+ "model_type": "bert",
50
+ "no_repeat_ngram_size": 0,
51
+ "num_attention_heads": 12,
52
+ "num_beam_groups": 1,
53
+ "num_beams": 1,
54
+ "num_hidden_layers": 6,
55
+ "num_return_sequences": 1,
56
+ "output_attentions": false,
57
+ "output_hidden_states": false,
58
+ "output_scores": false,
59
+ "pad_token_id": 0,
60
+ "position_embedding_type": "absolute",
61
+ "prefix": null,
62
+ "problem_type": null,
63
+ "pruned_heads": {},
64
+ "remove_invalid_values": false,
65
+ "repetition_penalty": 1.0,
66
+ "return_dict": true,
67
+ "return_dict_in_generate": false,
68
+ "sep_token_id": null,
69
+ "suppress_tokens": null,
70
+ "task_specific_params": null,
71
+ "temperature": 1.0,
72
+ "tf_legacy_loss": false,
73
+ "tie_encoder_decoder": false,
74
+ "tie_word_embeddings": true,
75
+ "tokenizer_class": null,
76
+ "top_k": 50,
77
+ "top_p": 1.0,
78
+ "torch_dtype": null,
79
+ "torchscript": false,
80
+ "type_vocab_size": 2,
81
+ "typical_p": 1.0,
82
+ "use_bfloat16": false,
83
+ "use_cache": true,
84
+ "vocab_size": 30000
85
+ },
86
+ "encoder": {
87
+ "_name_or_path": "",
88
+ "add_cross_attention": false,
89
+ "architectures": [
90
+ "CvtForImageClassification"
91
+ ],
92
+ "attention_drop_rate": [
93
+ 0.0,
94
+ 0.0,
95
+ 0.0
96
+ ],
97
+ "bad_words_ids": null,
98
+ "begin_suppress_tokens": null,
99
+ "bos_token_id": null,
100
+ "chunk_size_feed_forward": 0,
101
+ "cls_token": [
102
+ false,
103
+ false,
104
+ true
105
+ ],
106
+ "cross_attention_hidden_size": null,
107
+ "decoder_start_token_id": null,
108
+ "depth": [
109
+ 1,
110
+ 4,
111
+ 16
112
+ ],
113
+ "diversity_penalty": 0.0,
114
+ "do_sample": false,
115
+ "drop_path_rate": [
116
+ 0.0,
117
+ 0.0,
118
+ 0.1
119
+ ],
120
+ "drop_rate": [
121
+ 0.0,
122
+ 0.0,
123
+ 0.0
124
+ ],
125
+ "early_stopping": false,
126
+ "embed_dim": [
127
+ 64,
128
+ 192,
129
+ 384
130
+ ],
131
+ "encoder_no_repeat_ngram_size": 0,
132
+ "eos_token_id": null,
133
+ "exponential_decay_length_penalty": null,
134
+ "finetuning_task": null,
135
+ "forced_bos_token_id": null,
136
+ "forced_eos_token_id": null,
137
+ "id2label": {
138
+ "0": "tench, Tinca tinca",
139
+ "1": "goldfish, Carassius auratus",
140
+ "2": "great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias",
141
+ "3": "tiger shark, Galeocerdo cuvieri",
142
+ "4": "hammerhead, hammerhead shark",
143
+ "5": "electric ray, crampfish, numbfish, torpedo",
144
+ "6": "stingray",
145
+ "7": "cock",
146
+ "8": "hen",
147
+ "9": "ostrich, Struthio camelus",
148
+ "10": "brambling, Fringilla montifringilla",
149
+ "11": "goldfinch, Carduelis carduelis",
150
+ "12": "house finch, linnet, Carpodacus mexicanus",
151
+ "13": "junco, snowbird",
152
+ "14": "indigo bunting, indigo finch, indigo bird, Passerina cyanea",
153
+ "15": "robin, American robin, Turdus migratorius",
154
+ "16": "bulbul",
155
+ "17": "jay",
156
+ "18": "magpie",
157
+ "19": "chickadee",
158
+ "20": "water ouzel, dipper",
159
+ "21": "kite",
160
+ "22": "bald eagle, American eagle, Haliaeetus leucocephalus",
161
+ "23": "vulture",
162
+ "24": "great grey owl, great gray owl, Strix nebulosa",
163
+ "25": "European fire salamander, Salamandra salamandra",
164
+ "26": "common newt, Triturus vulgaris",
165
+ "27": "eft",
166
+ "28": "spotted salamander, Ambystoma maculatum",
167
+ "29": "axolotl, mud puppy, Ambystoma mexicanum",
168
+ "30": "bullfrog, Rana catesbeiana",
169
+ "31": "tree frog, tree-frog",
170
+ "32": "tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui",
171
+ "33": "loggerhead, loggerhead turtle, Caretta caretta",
172
+ "34": "leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea",
173
+ "35": "mud turtle",
174
+ "36": "terrapin",
175
+ "37": "box turtle, box tortoise",
176
+ "38": "banded gecko",
177
+ "39": "common iguana, iguana, Iguana iguana",
178
+ "40": "American chameleon, anole, Anolis carolinensis",
179
+ "41": "whiptail, whiptail lizard",
180
+ "42": "agama",
181
+ "43": "frilled lizard, Chlamydosaurus kingi",
182
+ "44": "alligator lizard",
183
+ "45": "Gila monster, Heloderma suspectum",
184
+ "46": "green lizard, Lacerta viridis",
185
+ "47": "African chameleon, Chamaeleo chamaeleon",
186
+ "48": "Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis",
187
+ "49": "African crocodile, Nile crocodile, Crocodylus niloticus",
188
+ "50": "American alligator, Alligator mississipiensis",
189
+ "51": "triceratops",
190
+ "52": "thunder snake, worm snake, Carphophis amoenus",
191
+ "53": "ringneck snake, ring-necked snake, ring snake",
192
+ "54": "hognose snake, puff adder, sand viper",
193
+ "55": "green snake, grass snake",
194
+ "56": "king snake, kingsnake",
195
+ "57": "garter snake, grass snake",
196
+ "58": "water snake",
197
+ "59": "vine snake",
198
+ "60": "night snake, Hypsiglena torquata",
199
+ "61": "boa constrictor, Constrictor constrictor",
200
+ "62": "rock python, rock snake, Python sebae",
201
+ "63": "Indian cobra, Naja naja",
202
+ "64": "green mamba",
203
+ "65": "sea snake",
204
+ "66": "horned viper, cerastes, sand viper, horned asp, Cerastes cornutus",
205
+ "67": "diamondback, diamondback rattlesnake, Crotalus adamanteus",
206
+ "68": "sidewinder, horned rattlesnake, Crotalus cerastes",
207
+ "69": "trilobite",
208
+ "70": "harvestman, daddy longlegs, Phalangium opilio",
209
+ "71": "scorpion",
210
+ "72": "black and gold garden spider, Argiope aurantia",
211
+ "73": "barn spider, Araneus cavaticus",
212
+ "74": "garden spider, Aranea diademata",
213
+ "75": "black widow, Latrodectus mactans",
214
+ "76": "tarantula",
215
+ "77": "wolf spider, hunting spider",
216
+ "78": "tick",
217
+ "79": "centipede",
218
+ "80": "black grouse",
219
+ "81": "ptarmigan",
220
+ "82": "ruffed grouse, partridge, Bonasa umbellus",
221
+ "83": "prairie chicken, prairie grouse, prairie fowl",
222
+ "84": "peacock",
223
+ "85": "quail",
224
+ "86": "partridge",
225
+ "87": "African grey, African gray, Psittacus erithacus",
226
+ "88": "macaw",
227
+ "89": "sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita",
228
+ "90": "lorikeet",
229
+ "91": "coucal",
230
+ "92": "bee eater",
231
+ "93": "hornbill",
232
+ "94": "hummingbird",
233
+ "95": "jacamar",
234
+ "96": "toucan",
235
+ "97": "drake",
236
+ "98": "red-breasted merganser, Mergus serrator",
237
+ "99": "goose",
238
+ "100": "black swan, Cygnus atratus",
239
+ "101": "tusker",
240
+ "102": "echidna, spiny anteater, anteater",
241
+ "103": "platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus",
242
+ "104": "wallaby, brush kangaroo",
243
+ "105": "koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus",
244
+ "106": "wombat",
245
+ "107": "jellyfish",
246
+ "108": "sea anemone, anemone",
247
+ "109": "brain coral",
248
+ "110": "flatworm, platyhelminth",
249
+ "111": "nematode, nematode worm, roundworm",
250
+ "112": "conch",
251
+ "113": "snail",
252
+ "114": "slug",
253
+ "115": "sea slug, nudibranch",
254
+ "116": "chiton, coat-of-mail shell, sea cradle, polyplacophore",
255
+ "117": "chambered nautilus, pearly nautilus, nautilus",
256
+ "118": "Dungeness crab, Cancer magister",
257
+ "119": "rock crab, Cancer irroratus",
258
+ "120": "fiddler crab",
259
+ "121": "king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica",
260
+ "122": "American lobster, Northern lobster, Maine lobster, Homarus americanus",
261
+ "123": "spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish",
262
+ "124": "crayfish, crawfish, crawdad, crawdaddy",
263
+ "125": "hermit crab",
264
+ "126": "isopod",
265
+ "127": "white stork, Ciconia ciconia",
266
+ "128": "black stork, Ciconia nigra",
267
+ "129": "spoonbill",
268
+ "130": "flamingo",
269
+ "131": "little blue heron, Egretta caerulea",
270
+ "132": "American egret, great white heron, Egretta albus",
271
+ "133": "bittern",
272
+ "134": "crane",
273
+ "135": "limpkin, Aramus pictus",
274
+ "136": "European gallinule, Porphyrio porphyrio",
275
+ "137": "American coot, marsh hen, mud hen, water hen, Fulica americana",
276
+ "138": "bustard",
277
+ "139": "ruddy turnstone, Arenaria interpres",
278
+ "140": "red-backed sandpiper, dunlin, Erolia alpina",
279
+ "141": "redshank, Tringa totanus",
280
+ "142": "dowitcher",
281
+ "143": "oystercatcher, oyster catcher",
282
+ "144": "pelican",
283
+ "145": "king penguin, Aptenodytes patagonica",
284
+ "146": "albatross, mollymawk",
285
+ "147": "grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus",
286
+ "148": "killer whale, killer, orca, grampus, sea wolf, Orcinus orca",
287
+ "149": "dugong, Dugong dugon",
288
+ "150": "sea lion",
289
+ "151": "Chihuahua",
290
+ "152": "Japanese spaniel",
291
+ "153": "Maltese dog, Maltese terrier, Maltese",
292
+ "154": "Pekinese, Pekingese, Peke",
293
+ "155": "Shih-Tzu",
294
+ "156": "Blenheim spaniel",
295
+ "157": "papillon",
296
+ "158": "toy terrier",
297
+ "159": "Rhodesian ridgeback",
298
+ "160": "Afghan hound, Afghan",
299
+ "161": "basset, basset hound",
300
+ "162": "beagle",
301
+ "163": "bloodhound, sleuthhound",
302
+ "164": "bluetick",
303
+ "165": "black-and-tan coonhound",
304
+ "166": "Walker hound, Walker foxhound",
305
+ "167": "English foxhound",
306
+ "168": "redbone",
307
+ "169": "borzoi, Russian wolfhound",
308
+ "170": "Irish wolfhound",
309
+ "171": "Italian greyhound",
310
+ "172": "whippet",
311
+ "173": "Ibizan hound, Ibizan Podenco",
312
+ "174": "Norwegian elkhound, elkhound",
313
+ "175": "otterhound, otter hound",
314
+ "176": "Saluki, gazelle hound",
315
+ "177": "Scottish deerhound, deerhound",
316
+ "178": "Weimaraner",
317
+ "179": "Staffordshire bullterrier, Staffordshire bull terrier",
318
+ "180": "American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier",
319
+ "181": "Bedlington terrier",
320
+ "182": "Border terrier",
321
+ "183": "Kerry blue terrier",
322
+ "184": "Irish terrier",
323
+ "185": "Norfolk terrier",
324
+ "186": "Norwich terrier",
325
+ "187": "Yorkshire terrier",
326
+ "188": "wire-haired fox terrier",
327
+ "189": "Lakeland terrier",
328
+ "190": "Sealyham terrier, Sealyham",
329
+ "191": "Airedale, Airedale terrier",
330
+ "192": "cairn, cairn terrier",
331
+ "193": "Australian terrier",
332
+ "194": "Dandie Dinmont, Dandie Dinmont terrier",
333
+ "195": "Boston bull, Boston terrier",
334
+ "196": "miniature schnauzer",
335
+ "197": "giant schnauzer",
336
+ "198": "standard schnauzer",
337
+ "199": "Scotch terrier, Scottish terrier, Scottie",
338
+ "200": "Tibetan terrier, chrysanthemum dog",
339
+ "201": "silky terrier, Sydney silky",
340
+ "202": "soft-coated wheaten terrier",
341
+ "203": "West Highland white terrier",
342
+ "204": "Lhasa, Lhasa apso",
343
+ "205": "flat-coated retriever",
344
+ "206": "curly-coated retriever",
345
+ "207": "golden retriever",
346
+ "208": "Labrador retriever",
347
+ "209": "Chesapeake Bay retriever",
348
+ "210": "German short-haired pointer",
349
+ "211": "vizsla, Hungarian pointer",
350
+ "212": "English setter",
351
+ "213": "Irish setter, red setter",
352
+ "214": "Gordon setter",
353
+ "215": "Brittany spaniel",
354
+ "216": "clumber, clumber spaniel",
355
+ "217": "English springer, English springer spaniel",
356
+ "218": "Welsh springer spaniel",
357
+ "219": "cocker spaniel, English cocker spaniel, cocker",
358
+ "220": "Sussex spaniel",
359
+ "221": "Irish water spaniel",
360
+ "222": "kuvasz",
361
+ "223": "schipperke",
362
+ "224": "groenendael",
363
+ "225": "malinois",
364
+ "226": "briard",
365
+ "227": "kelpie",
366
+ "228": "komondor",
367
+ "229": "Old English sheepdog, bobtail",
368
+ "230": "Shetland sheepdog, Shetland sheep dog, Shetland",
369
+ "231": "collie",
370
+ "232": "Border collie",
371
+ "233": "Bouvier des Flandres, Bouviers des Flandres",
372
+ "234": "Rottweiler",
373
+ "235": "German shepherd, German shepherd dog, German police dog, alsatian",
374
+ "236": "Doberman, Doberman pinscher",
375
+ "237": "miniature pinscher",
376
+ "238": "Greater Swiss Mountain dog",
377
+ "239": "Bernese mountain dog",
378
+ "240": "Appenzeller",
379
+ "241": "EntleBucher",
380
+ "242": "boxer",
381
+ "243": "bull mastiff",
382
+ "244": "Tibetan mastiff",
383
+ "245": "French bulldog",
384
+ "246": "Great Dane",
385
+ "247": "Saint Bernard, St Bernard",
386
+ "248": "Eskimo dog, husky",
387
+ "249": "malamute, malemute, Alaskan malamute",
388
+ "250": "Siberian husky",
389
+ "251": "dalmatian, coach dog, carriage dog",
390
+ "252": "affenpinscher, monkey pinscher, monkey dog",
391
+ "253": "basenji",
392
+ "254": "pug, pug-dog",
393
+ "255": "Leonberg",
394
+ "256": "Newfoundland, Newfoundland dog",
395
+ "257": "Great Pyrenees",
396
+ "258": "Samoyed, Samoyede",
397
+ "259": "Pomeranian",
398
+ "260": "chow, chow chow",
399
+ "261": "keeshond",
400
+ "262": "Brabancon griffon",
401
+ "263": "Pembroke, Pembroke Welsh corgi",
402
+ "264": "Cardigan, Cardigan Welsh corgi",
403
+ "265": "toy poodle",
404
+ "266": "miniature poodle",
405
+ "267": "standard poodle",
406
+ "268": "Mexican hairless",
407
+ "269": "timber wolf, grey wolf, gray wolf, Canis lupus",
408
+ "270": "white wolf, Arctic wolf, Canis lupus tundrarum",
409
+ "271": "red wolf, maned wolf, Canis rufus, Canis niger",
410
+ "272": "coyote, prairie wolf, brush wolf, Canis latrans",
411
+ "273": "dingo, warrigal, warragal, Canis dingo",
412
+ "274": "dhole, Cuon alpinus",
413
+ "275": "African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus",
414
+ "276": "hyena, hyaena",
415
+ "277": "red fox, Vulpes vulpes",
416
+ "278": "kit fox, Vulpes macrotis",
417
+ "279": "Arctic fox, white fox, Alopex lagopus",
418
+ "280": "grey fox, gray fox, Urocyon cinereoargenteus",
419
+ "281": "tabby, tabby cat",
420
+ "282": "tiger cat",
421
+ "283": "Persian cat",
422
+ "284": "Siamese cat, Siamese",
423
+ "285": "Egyptian cat",
424
+ "286": "cougar, puma, catamount, mountain lion, painter, panther, Felis concolor",
425
+ "287": "lynx, catamount",
426
+ "288": "leopard, Panthera pardus",
427
+ "289": "snow leopard, ounce, Panthera uncia",
428
+ "290": "jaguar, panther, Panthera onca, Felis onca",
429
+ "291": "lion, king of beasts, Panthera leo",
430
+ "292": "tiger, Panthera tigris",
431
+ "293": "cheetah, chetah, Acinonyx jubatus",
432
+ "294": "brown bear, bruin, Ursus arctos",
433
+ "295": "American black bear, black bear, Ursus americanus, Euarctos americanus",
434
+ "296": "ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus",
435
+ "297": "sloth bear, Melursus ursinus, Ursus ursinus",
436
+ "298": "mongoose",
437
+ "299": "meerkat, mierkat",
438
+ "300": "tiger beetle",
439
+ "301": "ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle",
440
+ "302": "ground beetle, carabid beetle",
441
+ "303": "long-horned beetle, longicorn, longicorn beetle",
442
+ "304": "leaf beetle, chrysomelid",
443
+ "305": "dung beetle",
444
+ "306": "rhinoceros beetle",
445
+ "307": "weevil",
446
+ "308": "fly",
447
+ "309": "bee",
448
+ "310": "ant, emmet, pismire",
449
+ "311": "grasshopper, hopper",
450
+ "312": "cricket",
451
+ "313": "walking stick, walkingstick, stick insect",
452
+ "314": "cockroach, roach",
453
+ "315": "mantis, mantid",
454
+ "316": "cicada, cicala",
455
+ "317": "leafhopper",
456
+ "318": "lacewing, lacewing fly",
457
+ "319": "dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk",
458
+ "320": "damselfly",
459
+ "321": "admiral",
460
+ "322": "ringlet, ringlet butterfly",
461
+ "323": "monarch, monarch butterfly, milkweed butterfly, Danaus plexippus",
462
+ "324": "cabbage butterfly",
463
+ "325": "sulphur butterfly, sulfur butterfly",
464
+ "326": "lycaenid, lycaenid butterfly",
465
+ "327": "starfish, sea star",
466
+ "328": "sea urchin",
467
+ "329": "sea cucumber, holothurian",
468
+ "330": "wood rabbit, cottontail, cottontail rabbit",
469
+ "331": "hare",
470
+ "332": "Angora, Angora rabbit",
471
+ "333": "hamster",
472
+ "334": "porcupine, hedgehog",
473
+ "335": "fox squirrel, eastern fox squirrel, Sciurus niger",
474
+ "336": "marmot",
475
+ "337": "beaver",
476
+ "338": "guinea pig, Cavia cobaya",
477
+ "339": "sorrel",
478
+ "340": "zebra",
479
+ "341": "hog, pig, grunter, squealer, Sus scrofa",
480
+ "342": "wild boar, boar, Sus scrofa",
481
+ "343": "warthog",
482
+ "344": "hippopotamus, hippo, river horse, Hippopotamus amphibius",
483
+ "345": "ox",
484
+ "346": "water buffalo, water ox, Asiatic buffalo, Bubalus bubalis",
485
+ "347": "bison",
486
+ "348": "ram, tup",
487
+ "349": "bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis",
488
+ "350": "ibex, Capra ibex",
489
+ "351": "hartebeest",
490
+ "352": "impala, Aepyceros melampus",
491
+ "353": "gazelle",
492
+ "354": "Arabian camel, dromedary, Camelus dromedarius",
493
+ "355": "llama",
494
+ "356": "weasel",
495
+ "357": "mink",
496
+ "358": "polecat, fitch, foulmart, foumart, Mustela putorius",
497
+ "359": "black-footed ferret, ferret, Mustela nigripes",
498
+ "360": "otter",
499
+ "361": "skunk, polecat, wood pussy",
500
+ "362": "badger",
501
+ "363": "armadillo",
502
+ "364": "three-toed sloth, ai, Bradypus tridactylus",
503
+ "365": "orangutan, orang, orangutang, Pongo pygmaeus",
504
+ "366": "gorilla, Gorilla gorilla",
505
+ "367": "chimpanzee, chimp, Pan troglodytes",
506
+ "368": "gibbon, Hylobates lar",
507
+ "369": "siamang, Hylobates syndactylus, Symphalangus syndactylus",
508
+ "370": "guenon, guenon monkey",
509
+ "371": "patas, hussar monkey, Erythrocebus patas",
510
+ "372": "baboon",
511
+ "373": "macaque",
512
+ "374": "langur",
513
+ "375": "colobus, colobus monkey",
514
+ "376": "proboscis monkey, Nasalis larvatus",
515
+ "377": "marmoset",
516
+ "378": "capuchin, ringtail, Cebus capucinus",
517
+ "379": "howler monkey, howler",
518
+ "380": "titi, titi monkey",
519
+ "381": "spider monkey, Ateles geoffroyi",
520
+ "382": "squirrel monkey, Saimiri sciureus",
521
+ "383": "Madagascar cat, ring-tailed lemur, Lemur catta",
522
+ "384": "indri, indris, Indri indri, Indri brevicaudatus",
523
+ "385": "Indian elephant, Elephas maximus",
524
+ "386": "African elephant, Loxodonta africana",
525
+ "387": "lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens",
526
+ "388": "giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca",
527
+ "389": "barracouta, snoek",
528
+ "390": "eel",
529
+ "391": "coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch",
530
+ "392": "rock beauty, Holocanthus tricolor",
531
+ "393": "anemone fish",
532
+ "394": "sturgeon",
533
+ "395": "gar, garfish, garpike, billfish, Lepisosteus osseus",
534
+ "396": "lionfish",
535
+ "397": "puffer, pufferfish, blowfish, globefish",
536
+ "398": "abacus",
537
+ "399": "abaya",
538
+ "400": "academic gown, academic robe, judge's robe",
539
+ "401": "accordion, piano accordion, squeeze box",
540
+ "402": "acoustic guitar",
541
+ "403": "aircraft carrier, carrier, flattop, attack aircraft carrier",
542
+ "404": "airliner",
543
+ "405": "airship, dirigible",
544
+ "406": "altar",
545
+ "407": "ambulance",
546
+ "408": "amphibian, amphibious vehicle",
547
+ "409": "analog clock",
548
+ "410": "apiary, bee house",
549
+ "411": "apron",
550
+ "412": "ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin",
551
+ "413": "assault rifle, assault gun",
552
+ "414": "backpack, back pack, knapsack, packsack, rucksack, haversack",
553
+ "415": "bakery, bakeshop, bakehouse",
554
+ "416": "balance beam, beam",
555
+ "417": "balloon",
556
+ "418": "ballpoint, ballpoint pen, ballpen, Biro",
557
+ "419": "Band Aid",
558
+ "420": "banjo",
559
+ "421": "bannister, banister, balustrade, balusters, handrail",
560
+ "422": "barbell",
561
+ "423": "barber chair",
562
+ "424": "barbershop",
563
+ "425": "barn",
564
+ "426": "barometer",
565
+ "427": "barrel, cask",
566
+ "428": "barrow, garden cart, lawn cart, wheelbarrow",
567
+ "429": "baseball",
568
+ "430": "basketball",
569
+ "431": "bassinet",
570
+ "432": "bassoon",
571
+ "433": "bathing cap, swimming cap",
572
+ "434": "bath towel",
573
+ "435": "bathtub, bathing tub, bath, tub",
574
+ "436": "beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon",
575
+ "437": "beacon, lighthouse, beacon light, pharos",
576
+ "438": "beaker",
577
+ "439": "bearskin, busby, shako",
578
+ "440": "beer bottle",
579
+ "441": "beer glass",
580
+ "442": "bell cote, bell cot",
581
+ "443": "bib",
582
+ "444": "bicycle-built-for-two, tandem bicycle, tandem",
583
+ "445": "bikini, two-piece",
584
+ "446": "binder, ring-binder",
585
+ "447": "binoculars, field glasses, opera glasses",
586
+ "448": "birdhouse",
587
+ "449": "boathouse",
588
+ "450": "bobsled, bobsleigh, bob",
589
+ "451": "bolo tie, bolo, bola tie, bola",
590
+ "452": "bonnet, poke bonnet",
591
+ "453": "bookcase",
592
+ "454": "bookshop, bookstore, bookstall",
593
+ "455": "bottlecap",
594
+ "456": "bow",
595
+ "457": "bow tie, bow-tie, bowtie",
596
+ "458": "brass, memorial tablet, plaque",
597
+ "459": "brassiere, bra, bandeau",
598
+ "460": "breakwater, groin, groyne, mole, bulwark, seawall, jetty",
599
+ "461": "breastplate, aegis, egis",
600
+ "462": "broom",
601
+ "463": "bucket, pail",
602
+ "464": "buckle",
603
+ "465": "bulletproof vest",
604
+ "466": "bullet train, bullet",
605
+ "467": "butcher shop, meat market",
606
+ "468": "cab, hack, taxi, taxicab",
607
+ "469": "caldron, cauldron",
608
+ "470": "candle, taper, wax light",
609
+ "471": "cannon",
610
+ "472": "canoe",
611
+ "473": "can opener, tin opener",
612
+ "474": "cardigan",
613
+ "475": "car mirror",
614
+ "476": "carousel, carrousel, merry-go-round, roundabout, whirligig",
615
+ "477": "carpenter's kit, tool kit",
616
+ "478": "carton",
617
+ "479": "car wheel",
618
+ "480": "cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM",
619
+ "481": "cassette",
620
+ "482": "cassette player",
621
+ "483": "castle",
622
+ "484": "catamaran",
623
+ "485": "CD player",
624
+ "486": "cello, violoncello",
625
+ "487": "cellular telephone, cellular phone, cellphone, cell, mobile phone",
626
+ "488": "chain",
627
+ "489": "chainlink fence",
628
+ "490": "chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour",
629
+ "491": "chain saw, chainsaw",
630
+ "492": "chest",
631
+ "493": "chiffonier, commode",
632
+ "494": "chime, bell, gong",
633
+ "495": "china cabinet, china closet",
634
+ "496": "Christmas stocking",
635
+ "497": "church, church building",
636
+ "498": "cinema, movie theater, movie theatre, movie house, picture palace",
637
+ "499": "cleaver, meat cleaver, chopper",
638
+ "500": "cliff dwelling",
639
+ "501": "cloak",
640
+ "502": "clog, geta, patten, sabot",
641
+ "503": "cocktail shaker",
642
+ "504": "coffee mug",
643
+ "505": "coffeepot",
644
+ "506": "coil, spiral, volute, whorl, helix",
645
+ "507": "combination lock",
646
+ "508": "computer keyboard, keypad",
647
+ "509": "confectionery, confectionary, candy store",
648
+ "510": "container ship, containership, container vessel",
649
+ "511": "convertible",
650
+ "512": "corkscrew, bottle screw",
651
+ "513": "cornet, horn, trumpet, trump",
652
+ "514": "cowboy boot",
653
+ "515": "cowboy hat, ten-gallon hat",
654
+ "516": "cradle",
655
+ "517": "crane",
656
+ "518": "crash helmet",
657
+ "519": "crate",
658
+ "520": "crib, cot",
659
+ "521": "Crock Pot",
660
+ "522": "croquet ball",
661
+ "523": "crutch",
662
+ "524": "cuirass",
663
+ "525": "dam, dike, dyke",
664
+ "526": "desk",
665
+ "527": "desktop computer",
666
+ "528": "dial telephone, dial phone",
667
+ "529": "diaper, nappy, napkin",
668
+ "530": "digital clock",
669
+ "531": "digital watch",
670
+ "532": "dining table, board",
671
+ "533": "dishrag, dishcloth",
672
+ "534": "dishwasher, dish washer, dishwashing machine",
673
+ "535": "disk brake, disc brake",
674
+ "536": "dock, dockage, docking facility",
675
+ "537": "dogsled, dog sled, dog sleigh",
676
+ "538": "dome",
677
+ "539": "doormat, welcome mat",
678
+ "540": "drilling platform, offshore rig",
679
+ "541": "drum, membranophone, tympan",
680
+ "542": "drumstick",
681
+ "543": "dumbbell",
682
+ "544": "Dutch oven",
683
+ "545": "electric fan, blower",
684
+ "546": "electric guitar",
685
+ "547": "electric locomotive",
686
+ "548": "entertainment center",
687
+ "549": "envelope",
688
+ "550": "espresso maker",
689
+ "551": "face powder",
690
+ "552": "feather boa, boa",
691
+ "553": "file, file cabinet, filing cabinet",
692
+ "554": "fireboat",
693
+ "555": "fire engine, fire truck",
694
+ "556": "fire screen, fireguard",
695
+ "557": "flagpole, flagstaff",
696
+ "558": "flute, transverse flute",
697
+ "559": "folding chair",
698
+ "560": "football helmet",
699
+ "561": "forklift",
700
+ "562": "fountain",
701
+ "563": "fountain pen",
702
+ "564": "four-poster",
703
+ "565": "freight car",
704
+ "566": "French horn, horn",
705
+ "567": "frying pan, frypan, skillet",
706
+ "568": "fur coat",
707
+ "569": "garbage truck, dustcart",
708
+ "570": "gasmask, respirator, gas helmet",
709
+ "571": "gas pump, gasoline pump, petrol pump, island dispenser",
710
+ "572": "goblet",
711
+ "573": "go-kart",
712
+ "574": "golf ball",
713
+ "575": "golfcart, golf cart",
714
+ "576": "gondola",
715
+ "577": "gong, tam-tam",
716
+ "578": "gown",
717
+ "579": "grand piano, grand",
718
+ "580": "greenhouse, nursery, glasshouse",
719
+ "581": "grille, radiator grille",
720
+ "582": "grocery store, grocery, food market, market",
721
+ "583": "guillotine",
722
+ "584": "hair slide",
723
+ "585": "hair spray",
724
+ "586": "half track",
725
+ "587": "hammer",
726
+ "588": "hamper",
727
+ "589": "hand blower, blow dryer, blow drier, hair dryer, hair drier",
728
+ "590": "hand-held computer, hand-held microcomputer",
729
+ "591": "handkerchief, hankie, hanky, hankey",
730
+ "592": "hard disc, hard disk, fixed disk",
731
+ "593": "harmonica, mouth organ, harp, mouth harp",
732
+ "594": "harp",
733
+ "595": "harvester, reaper",
734
+ "596": "hatchet",
735
+ "597": "holster",
736
+ "598": "home theater, home theatre",
737
+ "599": "honeycomb",
738
+ "600": "hook, claw",
739
+ "601": "hoopskirt, crinoline",
740
+ "602": "horizontal bar, high bar",
741
+ "603": "horse cart, horse-cart",
742
+ "604": "hourglass",
743
+ "605": "iPod",
744
+ "606": "iron, smoothing iron",
745
+ "607": "jack-o'-lantern",
746
+ "608": "jean, blue jean, denim",
747
+ "609": "jeep, landrover",
748
+ "610": "jersey, T-shirt, tee shirt",
749
+ "611": "jigsaw puzzle",
750
+ "612": "jinrikisha, ricksha, rickshaw",
751
+ "613": "joystick",
752
+ "614": "kimono",
753
+ "615": "knee pad",
754
+ "616": "knot",
755
+ "617": "lab coat, laboratory coat",
756
+ "618": "ladle",
757
+ "619": "lampshade, lamp shade",
758
+ "620": "laptop, laptop computer",
759
+ "621": "lawn mower, mower",
760
+ "622": "lens cap, lens cover",
761
+ "623": "letter opener, paper knife, paperknife",
762
+ "624": "library",
763
+ "625": "lifeboat",
764
+ "626": "lighter, light, igniter, ignitor",
765
+ "627": "limousine, limo",
766
+ "628": "liner, ocean liner",
767
+ "629": "lipstick, lip rouge",
768
+ "630": "Loafer",
769
+ "631": "lotion",
770
+ "632": "loudspeaker, speaker, speaker unit, loudspeaker system, speaker system",
771
+ "633": "loupe, jeweler's loupe",
772
+ "634": "lumbermill, sawmill",
773
+ "635": "magnetic compass",
774
+ "636": "mailbag, postbag",
775
+ "637": "mailbox, letter box",
776
+ "638": "maillot",
777
+ "639": "maillot, tank suit",
778
+ "640": "manhole cover",
779
+ "641": "maraca",
780
+ "642": "marimba, xylophone",
781
+ "643": "mask",
782
+ "644": "matchstick",
783
+ "645": "maypole",
784
+ "646": "maze, labyrinth",
785
+ "647": "measuring cup",
786
+ "648": "medicine chest, medicine cabinet",
787
+ "649": "megalith, megalithic structure",
788
+ "650": "microphone, mike",
789
+ "651": "microwave, microwave oven",
790
+ "652": "military uniform",
791
+ "653": "milk can",
792
+ "654": "minibus",
793
+ "655": "miniskirt, mini",
794
+ "656": "minivan",
795
+ "657": "missile",
796
+ "658": "mitten",
797
+ "659": "mixing bowl",
798
+ "660": "mobile home, manufactured home",
799
+ "661": "Model T",
800
+ "662": "modem",
801
+ "663": "monastery",
802
+ "664": "monitor",
803
+ "665": "moped",
804
+ "666": "mortar",
805
+ "667": "mortarboard",
806
+ "668": "mosque",
807
+ "669": "mosquito net",
808
+ "670": "motor scooter, scooter",
809
+ "671": "mountain bike, all-terrain bike, off-roader",
810
+ "672": "mountain tent",
811
+ "673": "mouse, computer mouse",
812
+ "674": "mousetrap",
813
+ "675": "moving van",
814
+ "676": "muzzle",
815
+ "677": "nail",
816
+ "678": "neck brace",
817
+ "679": "necklace",
818
+ "680": "nipple",
819
+ "681": "notebook, notebook computer",
820
+ "682": "obelisk",
821
+ "683": "oboe, hautboy, hautbois",
822
+ "684": "ocarina, sweet potato",
823
+ "685": "odometer, hodometer, mileometer, milometer",
824
+ "686": "oil filter",
825
+ "687": "organ, pipe organ",
826
+ "688": "oscilloscope, scope, cathode-ray oscilloscope, CRO",
827
+ "689": "overskirt",
828
+ "690": "oxcart",
829
+ "691": "oxygen mask",
830
+ "692": "packet",
831
+ "693": "paddle, boat paddle",
832
+ "694": "paddlewheel, paddle wheel",
833
+ "695": "padlock",
834
+ "696": "paintbrush",
835
+ "697": "pajama, pyjama, pj's, jammies",
836
+ "698": "palace",
837
+ "699": "panpipe, pandean pipe, syrinx",
838
+ "700": "paper towel",
839
+ "701": "parachute, chute",
840
+ "702": "parallel bars, bars",
841
+ "703": "park bench",
842
+ "704": "parking meter",
843
+ "705": "passenger car, coach, carriage",
844
+ "706": "patio, terrace",
845
+ "707": "pay-phone, pay-station",
846
+ "708": "pedestal, plinth, footstall",
847
+ "709": "pencil box, pencil case",
848
+ "710": "pencil sharpener",
849
+ "711": "perfume, essence",
850
+ "712": "Petri dish",
851
+ "713": "photocopier",
852
+ "714": "pick, plectrum, plectron",
853
+ "715": "pickelhaube",
854
+ "716": "picket fence, paling",
855
+ "717": "pickup, pickup truck",
856
+ "718": "pier",
857
+ "719": "piggy bank, penny bank",
858
+ "720": "pill bottle",
859
+ "721": "pillow",
860
+ "722": "ping-pong ball",
861
+ "723": "pinwheel",
862
+ "724": "pirate, pirate ship",
863
+ "725": "pitcher, ewer",
864
+ "726": "plane, carpenter's plane, woodworking plane",
865
+ "727": "planetarium",
866
+ "728": "plastic bag",
867
+ "729": "plate rack",
868
+ "730": "plow, plough",
869
+ "731": "plunger, plumber's helper",
870
+ "732": "Polaroid camera, Polaroid Land camera",
871
+ "733": "pole",
872
+ "734": "police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria",
873
+ "735": "poncho",
874
+ "736": "pool table, billiard table, snooker table",
875
+ "737": "pop bottle, soda bottle",
876
+ "738": "pot, flowerpot",
877
+ "739": "potter's wheel",
878
+ "740": "power drill",
879
+ "741": "prayer rug, prayer mat",
880
+ "742": "printer",
881
+ "743": "prison, prison house",
882
+ "744": "projectile, missile",
883
+ "745": "projector",
884
+ "746": "puck, hockey puck",
885
+ "747": "punching bag, punch bag, punching ball, punchball",
886
+ "748": "purse",
887
+ "749": "quill, quill pen",
888
+ "750": "quilt, comforter, comfort, puff",
889
+ "751": "racer, race car, racing car",
890
+ "752": "racket, racquet",
891
+ "753": "radiator",
892
+ "754": "radio, wireless",
893
+ "755": "radio telescope, radio reflector",
894
+ "756": "rain barrel",
895
+ "757": "recreational vehicle, RV, R.V.",
896
+ "758": "reel",
897
+ "759": "reflex camera",
898
+ "760": "refrigerator, icebox",
899
+ "761": "remote control, remote",
900
+ "762": "restaurant, eating house, eating place, eatery",
901
+ "763": "revolver, six-gun, six-shooter",
902
+ "764": "rifle",
903
+ "765": "rocking chair, rocker",
904
+ "766": "rotisserie",
905
+ "767": "rubber eraser, rubber, pencil eraser",
906
+ "768": "rugby ball",
907
+ "769": "rule, ruler",
908
+ "770": "running shoe",
909
+ "771": "safe",
910
+ "772": "safety pin",
911
+ "773": "saltshaker, salt shaker",
912
+ "774": "sandal",
913
+ "775": "sarong",
914
+ "776": "sax, saxophone",
915
+ "777": "scabbard",
916
+ "778": "scale, weighing machine",
917
+ "779": "school bus",
918
+ "780": "schooner",
919
+ "781": "scoreboard",
920
+ "782": "screen, CRT screen",
921
+ "783": "screw",
922
+ "784": "screwdriver",
923
+ "785": "seat belt, seatbelt",
924
+ "786": "sewing machine",
925
+ "787": "shield, buckler",
926
+ "788": "shoe shop, shoe-shop, shoe store",
927
+ "789": "shoji",
928
+ "790": "shopping basket",
929
+ "791": "shopping cart",
930
+ "792": "shovel",
931
+ "793": "shower cap",
932
+ "794": "shower curtain",
933
+ "795": "ski",
934
+ "796": "ski mask",
935
+ "797": "sleeping bag",
936
+ "798": "slide rule, slipstick",
937
+ "799": "sliding door",
938
+ "800": "slot, one-armed bandit",
939
+ "801": "snorkel",
940
+ "802": "snowmobile",
941
+ "803": "snowplow, snowplough",
942
+ "804": "soap dispenser",
943
+ "805": "soccer ball",
944
+ "806": "sock",
945
+ "807": "solar dish, solar collector, solar furnace",
946
+ "808": "sombrero",
947
+ "809": "soup bowl",
948
+ "810": "space bar",
949
+ "811": "space heater",
950
+ "812": "space shuttle",
951
+ "813": "spatula",
952
+ "814": "speedboat",
953
+ "815": "spider web, spider's web",
954
+ "816": "spindle",
955
+ "817": "sports car, sport car",
956
+ "818": "spotlight, spot",
957
+ "819": "stage",
958
+ "820": "steam locomotive",
959
+ "821": "steel arch bridge",
960
+ "822": "steel drum",
961
+ "823": "stethoscope",
962
+ "824": "stole",
963
+ "825": "stone wall",
964
+ "826": "stopwatch, stop watch",
965
+ "827": "stove",
966
+ "828": "strainer",
967
+ "829": "streetcar, tram, tramcar, trolley, trolley car",
968
+ "830": "stretcher",
969
+ "831": "studio couch, day bed",
970
+ "832": "stupa, tope",
971
+ "833": "submarine, pigboat, sub, U-boat",
972
+ "834": "suit, suit of clothes",
973
+ "835": "sundial",
974
+ "836": "sunglass",
975
+ "837": "sunglasses, dark glasses, shades",
976
+ "838": "sunscreen, sunblock, sun blocker",
977
+ "839": "suspension bridge",
978
+ "840": "swab, swob, mop",
979
+ "841": "sweatshirt",
980
+ "842": "swimming trunks, bathing trunks",
981
+ "843": "swing",
982
+ "844": "switch, electric switch, electrical switch",
983
+ "845": "syringe",
984
+ "846": "table lamp",
985
+ "847": "tank, army tank, armored combat vehicle, armoured combat vehicle",
986
+ "848": "tape player",
987
+ "849": "teapot",
988
+ "850": "teddy, teddy bear",
989
+ "851": "television, television system",
990
+ "852": "tennis ball",
991
+ "853": "thatch, thatched roof",
992
+ "854": "theater curtain, theatre curtain",
993
+ "855": "thimble",
994
+ "856": "thresher, thrasher, threshing machine",
995
+ "857": "throne",
996
+ "858": "tile roof",
997
+ "859": "toaster",
998
+ "860": "tobacco shop, tobacconist shop, tobacconist",
999
+ "861": "toilet seat",
1000
+ "862": "torch",
1001
+ "863": "totem pole",
1002
+ "864": "tow truck, tow car, wrecker",
1003
+ "865": "toyshop",
1004
+ "866": "tractor",
1005
+ "867": "trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi",
1006
+ "868": "tray",
1007
+ "869": "trench coat",
1008
+ "870": "tricycle, trike, velocipede",
1009
+ "871": "trimaran",
1010
+ "872": "tripod",
1011
+ "873": "triumphal arch",
1012
+ "874": "trolleybus, trolley coach, trackless trolley",
1013
+ "875": "trombone",
1014
+ "876": "tub, vat",
1015
+ "877": "turnstile",
1016
+ "878": "typewriter keyboard",
1017
+ "879": "umbrella",
1018
+ "880": "unicycle, monocycle",
1019
+ "881": "upright, upright piano",
1020
+ "882": "vacuum, vacuum cleaner",
1021
+ "883": "vase",
1022
+ "884": "vault",
1023
+ "885": "velvet",
1024
+ "886": "vending machine",
1025
+ "887": "vestment",
1026
+ "888": "viaduct",
1027
+ "889": "violin, fiddle",
1028
+ "890": "volleyball",
1029
+ "891": "waffle iron",
1030
+ "892": "wall clock",
1031
+ "893": "wallet, billfold, notecase, pocketbook",
1032
+ "894": "wardrobe, closet, press",
1033
+ "895": "warplane, military plane",
1034
+ "896": "washbasin, handbasin, washbowl, lavabo, wash-hand basin",
1035
+ "897": "washer, automatic washer, washing machine",
1036
+ "898": "water bottle",
1037
+ "899": "water jug",
1038
+ "900": "water tower",
1039
+ "901": "whiskey jug",
1040
+ "902": "whistle",
1041
+ "903": "wig",
1042
+ "904": "window screen",
1043
+ "905": "window shade",
1044
+ "906": "Windsor tie",
1045
+ "907": "wine bottle",
1046
+ "908": "wing",
1047
+ "909": "wok",
1048
+ "910": "wooden spoon",
1049
+ "911": "wool, woolen, woollen",
1050
+ "912": "worm fence, snake fence, snake-rail fence, Virginia fence",
1051
+ "913": "wreck",
1052
+ "914": "yawl",
1053
+ "915": "yurt",
1054
+ "916": "web site, website, internet site, site",
1055
+ "917": "comic book",
1056
+ "918": "crossword puzzle, crossword",
1057
+ "919": "street sign",
1058
+ "920": "traffic light, traffic signal, stoplight",
1059
+ "921": "book jacket, dust cover, dust jacket, dust wrapper",
1060
+ "922": "menu",
1061
+ "923": "plate",
1062
+ "924": "guacamole",
1063
+ "925": "consomme",
1064
+ "926": "hot pot, hotpot",
1065
+ "927": "trifle",
1066
+ "928": "ice cream, icecream",
1067
+ "929": "ice lolly, lolly, lollipop, popsicle",
1068
+ "930": "French loaf",
1069
+ "931": "bagel, beigel",
1070
+ "932": "pretzel",
1071
+ "933": "cheeseburger",
1072
+ "934": "hotdog, hot dog, red hot",
1073
+ "935": "mashed potato",
1074
+ "936": "head cabbage",
1075
+ "937": "broccoli",
1076
+ "938": "cauliflower",
1077
+ "939": "zucchini, courgette",
1078
+ "940": "spaghetti squash",
1079
+ "941": "acorn squash",
1080
+ "942": "butternut squash",
1081
+ "943": "cucumber, cuke",
1082
+ "944": "artichoke, globe artichoke",
1083
+ "945": "bell pepper",
1084
+ "946": "cardoon",
1085
+ "947": "mushroom",
1086
+ "948": "Granny Smith",
1087
+ "949": "strawberry",
1088
+ "950": "orange",
1089
+ "951": "lemon",
1090
+ "952": "fig",
1091
+ "953": "pineapple, ananas",
1092
+ "954": "banana",
1093
+ "955": "jackfruit, jak, jack",
1094
+ "956": "custard apple",
1095
+ "957": "pomegranate",
1096
+ "958": "hay",
1097
+ "959": "carbonara",
1098
+ "960": "chocolate sauce, chocolate syrup",
1099
+ "961": "dough",
1100
+ "962": "meat loaf, meatloaf",
1101
+ "963": "pizza, pizza pie",
1102
+ "964": "potpie",
1103
+ "965": "burrito",
1104
+ "966": "red wine",
1105
+ "967": "espresso",
1106
+ "968": "cup",
1107
+ "969": "eggnog",
1108
+ "970": "alp",
1109
+ "971": "bubble",
1110
+ "972": "cliff, drop, drop-off",
1111
+ "973": "coral reef",
1112
+ "974": "geyser",
1113
+ "975": "lakeside, lakeshore",
1114
+ "976": "promontory, headland, head, foreland",
1115
+ "977": "sandbar, sand bar",
1116
+ "978": "seashore, coast, seacoast, sea-coast",
1117
+ "979": "valley, vale",
1118
+ "980": "volcano",
1119
+ "981": "ballplayer, baseball player",
1120
+ "982": "groom, bridegroom",
1121
+ "983": "scuba diver",
1122
+ "984": "rapeseed",
1123
+ "985": "daisy",
1124
+ "986": "yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum",
1125
+ "987": "corn",
1126
+ "988": "acorn",
1127
+ "989": "hip, rose hip, rosehip",
1128
+ "990": "buckeye, horse chestnut, conker",
1129
+ "991": "coral fungus",
1130
+ "992": "agaric",
1131
+ "993": "gyromitra",
1132
+ "994": "stinkhorn, carrion fungus",
1133
+ "995": "earthstar",
1134
+ "996": "hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa",
1135
+ "997": "bolete",
1136
+ "998": "ear, spike, capitulum",
1137
+ "999": "toilet tissue, toilet paper, bathroom tissue"
1138
+ },
1139
+ "image_size": 384,
1140
+ "initializer_range": 0.02,
1141
+ "is_decoder": false,
1142
+ "is_encoder_decoder": false,
1143
+ "kernel_qkv": [
1144
+ 3,
1145
+ 3,
1146
+ 3
1147
+ ],
1148
+ "label2id": {
1149
+ "Afghan hound, Afghan": 160,
1150
+ "African chameleon, Chamaeleo chamaeleon": 47,
1151
+ "African crocodile, Nile crocodile, Crocodylus niloticus": 49,
1152
+ "African elephant, Loxodonta africana": 386,
1153
+ "African grey, African gray, Psittacus erithacus": 87,
1154
+ "African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus": 275,
1155
+ "Airedale, Airedale terrier": 191,
1156
+ "American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier": 180,
1157
+ "American alligator, Alligator mississipiensis": 50,
1158
+ "American black bear, black bear, Ursus americanus, Euarctos americanus": 295,
1159
+ "American chameleon, anole, Anolis carolinensis": 40,
1160
+ "American coot, marsh hen, mud hen, water hen, Fulica americana": 137,
1161
+ "American egret, great white heron, Egretta albus": 132,
1162
+ "American lobster, Northern lobster, Maine lobster, Homarus americanus": 122,
1163
+ "Angora, Angora rabbit": 332,
1164
+ "Appenzeller": 240,
1165
+ "Arabian camel, dromedary, Camelus dromedarius": 354,
1166
+ "Arctic fox, white fox, Alopex lagopus": 279,
1167
+ "Australian terrier": 193,
1168
+ "Band Aid": 419,
1169
+ "Bedlington terrier": 181,
1170
+ "Bernese mountain dog": 239,
1171
+ "Blenheim spaniel": 156,
1172
+ "Border collie": 232,
1173
+ "Border terrier": 182,
1174
+ "Boston bull, Boston terrier": 195,
1175
+ "Bouvier des Flandres, Bouviers des Flandres": 233,
1176
+ "Brabancon griffon": 262,
1177
+ "Brittany spaniel": 215,
1178
+ "CD player": 485,
1179
+ "Cardigan, Cardigan Welsh corgi": 264,
1180
+ "Chesapeake Bay retriever": 209,
1181
+ "Chihuahua": 151,
1182
+ "Christmas stocking": 496,
1183
+ "Crock Pot": 521,
1184
+ "Dandie Dinmont, Dandie Dinmont terrier": 194,
1185
+ "Doberman, Doberman pinscher": 236,
1186
+ "Dungeness crab, Cancer magister": 118,
1187
+ "Dutch oven": 544,
1188
+ "Egyptian cat": 285,
1189
+ "English foxhound": 167,
1190
+ "English setter": 212,
1191
+ "English springer, English springer spaniel": 217,
1192
+ "EntleBucher": 241,
1193
+ "Eskimo dog, husky": 248,
1194
+ "European fire salamander, Salamandra salamandra": 25,
1195
+ "European gallinule, Porphyrio porphyrio": 136,
1196
+ "French bulldog": 245,
1197
+ "French horn, horn": 566,
1198
+ "French loaf": 930,
1199
+ "German shepherd, German shepherd dog, German police dog, alsatian": 235,
1200
+ "German short-haired pointer": 210,
1201
+ "Gila monster, Heloderma suspectum": 45,
1202
+ "Gordon setter": 214,
1203
+ "Granny Smith": 948,
1204
+ "Great Dane": 246,
1205
+ "Great Pyrenees": 257,
1206
+ "Greater Swiss Mountain dog": 238,
1207
+ "Ibizan hound, Ibizan Podenco": 173,
1208
+ "Indian cobra, Naja naja": 63,
1209
+ "Indian elephant, Elephas maximus": 385,
1210
+ "Irish setter, red setter": 213,
1211
+ "Irish terrier": 184,
1212
+ "Irish water spaniel": 221,
1213
+ "Irish wolfhound": 170,
1214
+ "Italian greyhound": 171,
1215
+ "Japanese spaniel": 152,
1216
+ "Kerry blue terrier": 183,
1217
+ "Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis": 48,
1218
+ "Labrador retriever": 208,
1219
+ "Lakeland terrier": 189,
1220
+ "Leonberg": 255,
1221
+ "Lhasa, Lhasa apso": 204,
1222
+ "Loafer": 630,
1223
+ "Madagascar cat, ring-tailed lemur, Lemur catta": 383,
1224
+ "Maltese dog, Maltese terrier, Maltese": 153,
1225
+ "Mexican hairless": 268,
1226
+ "Model T": 661,
1227
+ "Newfoundland, Newfoundland dog": 256,
1228
+ "Norfolk terrier": 185,
1229
+ "Norwegian elkhound, elkhound": 174,
1230
+ "Norwich terrier": 186,
1231
+ "Old English sheepdog, bobtail": 229,
1232
+ "Pekinese, Pekingese, Peke": 154,
1233
+ "Pembroke, Pembroke Welsh corgi": 263,
1234
+ "Persian cat": 283,
1235
+ "Petri dish": 712,
1236
+ "Polaroid camera, Polaroid Land camera": 732,
1237
+ "Pomeranian": 259,
1238
+ "Rhodesian ridgeback": 159,
1239
+ "Rottweiler": 234,
1240
+ "Saint Bernard, St Bernard": 247,
1241
+ "Saluki, gazelle hound": 176,
1242
+ "Samoyed, Samoyede": 258,
1243
+ "Scotch terrier, Scottish terrier, Scottie": 199,
1244
+ "Scottish deerhound, deerhound": 177,
1245
+ "Sealyham terrier, Sealyham": 190,
1246
+ "Shetland sheepdog, Shetland sheep dog, Shetland": 230,
1247
+ "Shih-Tzu": 155,
1248
+ "Siamese cat, Siamese": 284,
1249
+ "Siberian husky": 250,
1250
+ "Staffordshire bullterrier, Staffordshire bull terrier": 179,
1251
+ "Sussex spaniel": 220,
1252
+ "Tibetan mastiff": 244,
1253
+ "Tibetan terrier, chrysanthemum dog": 200,
1254
+ "Walker hound, Walker foxhound": 166,
1255
+ "Weimaraner": 178,
1256
+ "Welsh springer spaniel": 218,
1257
+ "West Highland white terrier": 203,
1258
+ "Windsor tie": 906,
1259
+ "Yorkshire terrier": 187,
1260
+ "abacus": 398,
1261
+ "abaya": 399,
1262
+ "academic gown, academic robe, judge's robe": 400,
1263
+ "accordion, piano accordion, squeeze box": 401,
1264
+ "acorn": 988,
1265
+ "acorn squash": 941,
1266
+ "acoustic guitar": 402,
1267
+ "admiral": 321,
1268
+ "affenpinscher, monkey pinscher, monkey dog": 252,
1269
+ "agama": 42,
1270
+ "agaric": 992,
1271
+ "aircraft carrier, carrier, flattop, attack aircraft carrier": 403,
1272
+ "airliner": 404,
1273
+ "airship, dirigible": 405,
1274
+ "albatross, mollymawk": 146,
1275
+ "alligator lizard": 44,
1276
+ "alp": 970,
1277
+ "altar": 406,
1278
+ "ambulance": 407,
1279
+ "amphibian, amphibious vehicle": 408,
1280
+ "analog clock": 409,
1281
+ "anemone fish": 393,
1282
+ "ant, emmet, pismire": 310,
1283
+ "apiary, bee house": 410,
1284
+ "apron": 411,
1285
+ "armadillo": 363,
1286
+ "artichoke, globe artichoke": 944,
1287
+ "ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin": 412,
1288
+ "assault rifle, assault gun": 413,
1289
+ "axolotl, mud puppy, Ambystoma mexicanum": 29,
1290
+ "baboon": 372,
1291
+ "backpack, back pack, knapsack, packsack, rucksack, haversack": 414,
1292
+ "badger": 362,
1293
+ "bagel, beigel": 931,
1294
+ "bakery, bakeshop, bakehouse": 415,
1295
+ "balance beam, beam": 416,
1296
+ "bald eagle, American eagle, Haliaeetus leucocephalus": 22,
1297
+ "balloon": 417,
1298
+ "ballplayer, baseball player": 981,
1299
+ "ballpoint, ballpoint pen, ballpen, Biro": 418,
1300
+ "banana": 954,
1301
+ "banded gecko": 38,
1302
+ "banjo": 420,
1303
+ "bannister, banister, balustrade, balusters, handrail": 421,
1304
+ "barbell": 422,
1305
+ "barber chair": 423,
1306
+ "barbershop": 424,
1307
+ "barn": 425,
1308
+ "barn spider, Araneus cavaticus": 73,
1309
+ "barometer": 426,
1310
+ "barracouta, snoek": 389,
1311
+ "barrel, cask": 427,
1312
+ "barrow, garden cart, lawn cart, wheelbarrow": 428,
1313
+ "baseball": 429,
1314
+ "basenji": 253,
1315
+ "basketball": 430,
1316
+ "basset, basset hound": 161,
1317
+ "bassinet": 431,
1318
+ "bassoon": 432,
1319
+ "bath towel": 434,
1320
+ "bathing cap, swimming cap": 433,
1321
+ "bathtub, bathing tub, bath, tub": 435,
1322
+ "beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon": 436,
1323
+ "beacon, lighthouse, beacon light, pharos": 437,
1324
+ "beagle": 162,
1325
+ "beaker": 438,
1326
+ "bearskin, busby, shako": 439,
1327
+ "beaver": 337,
1328
+ "bee": 309,
1329
+ "bee eater": 92,
1330
+ "beer bottle": 440,
1331
+ "beer glass": 441,
1332
+ "bell cote, bell cot": 442,
1333
+ "bell pepper": 945,
1334
+ "bib": 443,
1335
+ "bicycle-built-for-two, tandem bicycle, tandem": 444,
1336
+ "bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis": 349,
1337
+ "bikini, two-piece": 445,
1338
+ "binder, ring-binder": 446,
1339
+ "binoculars, field glasses, opera glasses": 447,
1340
+ "birdhouse": 448,
1341
+ "bison": 347,
1342
+ "bittern": 133,
1343
+ "black and gold garden spider, Argiope aurantia": 72,
1344
+ "black grouse": 80,
1345
+ "black stork, Ciconia nigra": 128,
1346
+ "black swan, Cygnus atratus": 100,
1347
+ "black widow, Latrodectus mactans": 75,
1348
+ "black-and-tan coonhound": 165,
1349
+ "black-footed ferret, ferret, Mustela nigripes": 359,
1350
+ "bloodhound, sleuthhound": 163,
1351
+ "bluetick": 164,
1352
+ "boa constrictor, Constrictor constrictor": 61,
1353
+ "boathouse": 449,
1354
+ "bobsled, bobsleigh, bob": 450,
1355
+ "bolete": 997,
1356
+ "bolo tie, bolo, bola tie, bola": 451,
1357
+ "bonnet, poke bonnet": 452,
1358
+ "book jacket, dust cover, dust jacket, dust wrapper": 921,
1359
+ "bookcase": 453,
1360
+ "bookshop, bookstore, bookstall": 454,
1361
+ "borzoi, Russian wolfhound": 169,
1362
+ "bottlecap": 455,
1363
+ "bow": 456,
1364
+ "bow tie, bow-tie, bowtie": 457,
1365
+ "box turtle, box tortoise": 37,
1366
+ "boxer": 242,
1367
+ "brain coral": 109,
1368
+ "brambling, Fringilla montifringilla": 10,
1369
+ "brass, memorial tablet, plaque": 458,
1370
+ "brassiere, bra, bandeau": 459,
1371
+ "breakwater, groin, groyne, mole, bulwark, seawall, jetty": 460,
1372
+ "breastplate, aegis, egis": 461,
1373
+ "briard": 226,
1374
+ "broccoli": 937,
1375
+ "broom": 462,
1376
+ "brown bear, bruin, Ursus arctos": 294,
1377
+ "bubble": 971,
1378
+ "bucket, pail": 463,
1379
+ "buckeye, horse chestnut, conker": 990,
1380
+ "buckle": 464,
1381
+ "bulbul": 16,
1382
+ "bull mastiff": 243,
1383
+ "bullet train, bullet": 466,
1384
+ "bulletproof vest": 465,
1385
+ "bullfrog, Rana catesbeiana": 30,
1386
+ "burrito": 965,
1387
+ "bustard": 138,
1388
+ "butcher shop, meat market": 467,
1389
+ "butternut squash": 942,
1390
+ "cab, hack, taxi, taxicab": 468,
1391
+ "cabbage butterfly": 324,
1392
+ "cairn, cairn terrier": 192,
1393
+ "caldron, cauldron": 469,
1394
+ "can opener, tin opener": 473,
1395
+ "candle, taper, wax light": 470,
1396
+ "cannon": 471,
1397
+ "canoe": 472,
1398
+ "capuchin, ringtail, Cebus capucinus": 378,
1399
+ "car mirror": 475,
1400
+ "car wheel": 479,
1401
+ "carbonara": 959,
1402
+ "cardigan": 474,
1403
+ "cardoon": 946,
1404
+ "carousel, carrousel, merry-go-round, roundabout, whirligig": 476,
1405
+ "carpenter's kit, tool kit": 477,
1406
+ "carton": 478,
1407
+ "cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM": 480,
1408
+ "cassette": 481,
1409
+ "cassette player": 482,
1410
+ "castle": 483,
1411
+ "catamaran": 484,
1412
+ "cauliflower": 938,
1413
+ "cello, violoncello": 486,
1414
+ "cellular telephone, cellular phone, cellphone, cell, mobile phone": 487,
1415
+ "centipede": 79,
1416
+ "chain": 488,
1417
+ "chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour": 490,
1418
+ "chain saw, chainsaw": 491,
1419
+ "chainlink fence": 489,
1420
+ "chambered nautilus, pearly nautilus, nautilus": 117,
1421
+ "cheeseburger": 933,
1422
+ "cheetah, chetah, Acinonyx jubatus": 293,
1423
+ "chest": 492,
1424
+ "chickadee": 19,
1425
+ "chiffonier, commode": 493,
1426
+ "chime, bell, gong": 494,
1427
+ "chimpanzee, chimp, Pan troglodytes": 367,
1428
+ "china cabinet, china closet": 495,
1429
+ "chiton, coat-of-mail shell, sea cradle, polyplacophore": 116,
1430
+ "chocolate sauce, chocolate syrup": 960,
1431
+ "chow, chow chow": 260,
1432
+ "church, church building": 497,
1433
+ "cicada, cicala": 316,
1434
+ "cinema, movie theater, movie theatre, movie house, picture palace": 498,
1435
+ "cleaver, meat cleaver, chopper": 499,
1436
+ "cliff dwelling": 500,
1437
+ "cliff, drop, drop-off": 972,
1438
+ "cloak": 501,
1439
+ "clog, geta, patten, sabot": 502,
1440
+ "clumber, clumber spaniel": 216,
1441
+ "cock": 7,
1442
+ "cocker spaniel, English cocker spaniel, cocker": 219,
1443
+ "cockroach, roach": 314,
1444
+ "cocktail shaker": 503,
1445
+ "coffee mug": 504,
1446
+ "coffeepot": 505,
1447
+ "coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch": 391,
1448
+ "coil, spiral, volute, whorl, helix": 506,
1449
+ "collie": 231,
1450
+ "colobus, colobus monkey": 375,
1451
+ "combination lock": 507,
1452
+ "comic book": 917,
1453
+ "common iguana, iguana, Iguana iguana": 39,
1454
+ "common newt, Triturus vulgaris": 26,
1455
+ "computer keyboard, keypad": 508,
1456
+ "conch": 112,
1457
+ "confectionery, confectionary, candy store": 509,
1458
+ "consomme": 925,
1459
+ "container ship, containership, container vessel": 510,
1460
+ "convertible": 511,
1461
+ "coral fungus": 991,
1462
+ "coral reef": 973,
1463
+ "corkscrew, bottle screw": 512,
1464
+ "corn": 987,
1465
+ "cornet, horn, trumpet, trump": 513,
1466
+ "coucal": 91,
1467
+ "cougar, puma, catamount, mountain lion, painter, panther, Felis concolor": 286,
1468
+ "cowboy boot": 514,
1469
+ "cowboy hat, ten-gallon hat": 515,
1470
+ "coyote, prairie wolf, brush wolf, Canis latrans": 272,
1471
+ "cradle": 516,
1472
+ "crane": 517,
1473
+ "crash helmet": 518,
1474
+ "crate": 519,
1475
+ "crayfish, crawfish, crawdad, crawdaddy": 124,
1476
+ "crib, cot": 520,
1477
+ "cricket": 312,
1478
+ "croquet ball": 522,
1479
+ "crossword puzzle, crossword": 918,
1480
+ "crutch": 523,
1481
+ "cucumber, cuke": 943,
1482
+ "cuirass": 524,
1483
+ "cup": 968,
1484
+ "curly-coated retriever": 206,
1485
+ "custard apple": 956,
1486
+ "daisy": 985,
1487
+ "dalmatian, coach dog, carriage dog": 251,
1488
+ "dam, dike, dyke": 525,
1489
+ "damselfly": 320,
1490
+ "desk": 526,
1491
+ "desktop computer": 527,
1492
+ "dhole, Cuon alpinus": 274,
1493
+ "dial telephone, dial phone": 528,
1494
+ "diamondback, diamondback rattlesnake, Crotalus adamanteus": 67,
1495
+ "diaper, nappy, napkin": 529,
1496
+ "digital clock": 530,
1497
+ "digital watch": 531,
1498
+ "dingo, warrigal, warragal, Canis dingo": 273,
1499
+ "dining table, board": 532,
1500
+ "dishrag, dishcloth": 533,
1501
+ "dishwasher, dish washer, dishwashing machine": 534,
1502
+ "disk brake, disc brake": 535,
1503
+ "dock, dockage, docking facility": 536,
1504
+ "dogsled, dog sled, dog sleigh": 537,
1505
+ "dome": 538,
1506
+ "doormat, welcome mat": 539,
1507
+ "dough": 961,
1508
+ "dowitcher": 142,
1509
+ "dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk": 319,
1510
+ "drake": 97,
1511
+ "drilling platform, offshore rig": 540,
1512
+ "drum, membranophone, tympan": 541,
1513
+ "drumstick": 542,
1514
+ "dugong, Dugong dugon": 149,
1515
+ "dumbbell": 543,
1516
+ "dung beetle": 305,
1517
+ "ear, spike, capitulum": 998,
1518
+ "earthstar": 995,
1519
+ "echidna, spiny anteater, anteater": 102,
1520
+ "eel": 390,
1521
+ "eft": 27,
1522
+ "eggnog": 969,
1523
+ "electric fan, blower": 545,
1524
+ "electric guitar": 546,
1525
+ "electric locomotive": 547,
1526
+ "electric ray, crampfish, numbfish, torpedo": 5,
1527
+ "entertainment center": 548,
1528
+ "envelope": 549,
1529
+ "espresso": 967,
1530
+ "espresso maker": 550,
1531
+ "face powder": 551,
1532
+ "feather boa, boa": 552,
1533
+ "fiddler crab": 120,
1534
+ "fig": 952,
1535
+ "file, file cabinet, filing cabinet": 553,
1536
+ "fire engine, fire truck": 555,
1537
+ "fire screen, fireguard": 556,
1538
+ "fireboat": 554,
1539
+ "flagpole, flagstaff": 557,
1540
+ "flamingo": 130,
1541
+ "flat-coated retriever": 205,
1542
+ "flatworm, platyhelminth": 110,
1543
+ "flute, transverse flute": 558,
1544
+ "fly": 308,
1545
+ "folding chair": 559,
1546
+ "football helmet": 560,
1547
+ "forklift": 561,
1548
+ "fountain": 562,
1549
+ "fountain pen": 563,
1550
+ "four-poster": 564,
1551
+ "fox squirrel, eastern fox squirrel, Sciurus niger": 335,
1552
+ "freight car": 565,
1553
+ "frilled lizard, Chlamydosaurus kingi": 43,
1554
+ "frying pan, frypan, skillet": 567,
1555
+ "fur coat": 568,
1556
+ "gar, garfish, garpike, billfish, Lepisosteus osseus": 395,
1557
+ "garbage truck, dustcart": 569,
1558
+ "garden spider, Aranea diademata": 74,
1559
+ "garter snake, grass snake": 57,
1560
+ "gas pump, gasoline pump, petrol pump, island dispenser": 571,
1561
+ "gasmask, respirator, gas helmet": 570,
1562
+ "gazelle": 353,
1563
+ "geyser": 974,
1564
+ "giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca": 388,
1565
+ "giant schnauzer": 197,
1566
+ "gibbon, Hylobates lar": 368,
1567
+ "go-kart": 573,
1568
+ "goblet": 572,
1569
+ "golden retriever": 207,
1570
+ "goldfinch, Carduelis carduelis": 11,
1571
+ "goldfish, Carassius auratus": 1,
1572
+ "golf ball": 574,
1573
+ "golfcart, golf cart": 575,
1574
+ "gondola": 576,
1575
+ "gong, tam-tam": 577,
1576
+ "goose": 99,
1577
+ "gorilla, Gorilla gorilla": 366,
1578
+ "gown": 578,
1579
+ "grand piano, grand": 579,
1580
+ "grasshopper, hopper": 311,
1581
+ "great grey owl, great gray owl, Strix nebulosa": 24,
1582
+ "great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias": 2,
1583
+ "green lizard, Lacerta viridis": 46,
1584
+ "green mamba": 64,
1585
+ "green snake, grass snake": 55,
1586
+ "greenhouse, nursery, glasshouse": 580,
1587
+ "grey fox, gray fox, Urocyon cinereoargenteus": 280,
1588
+ "grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus": 147,
1589
+ "grille, radiator grille": 581,
1590
+ "grocery store, grocery, food market, market": 582,
1591
+ "groenendael": 224,
1592
+ "groom, bridegroom": 982,
1593
+ "ground beetle, carabid beetle": 302,
1594
+ "guacamole": 924,
1595
+ "guenon, guenon monkey": 370,
1596
+ "guillotine": 583,
1597
+ "guinea pig, Cavia cobaya": 338,
1598
+ "gyromitra": 993,
1599
+ "hair slide": 584,
1600
+ "hair spray": 585,
1601
+ "half track": 586,
1602
+ "hammer": 587,
1603
+ "hammerhead, hammerhead shark": 4,
1604
+ "hamper": 588,
1605
+ "hamster": 333,
1606
+ "hand blower, blow dryer, blow drier, hair dryer, hair drier": 589,
1607
+ "hand-held computer, hand-held microcomputer": 590,
1608
+ "handkerchief, hankie, hanky, hankey": 591,
1609
+ "hard disc, hard disk, fixed disk": 592,
1610
+ "hare": 331,
1611
+ "harmonica, mouth organ, harp, mouth harp": 593,
1612
+ "harp": 594,
1613
+ "hartebeest": 351,
1614
+ "harvester, reaper": 595,
1615
+ "harvestman, daddy longlegs, Phalangium opilio": 70,
1616
+ "hatchet": 596,
1617
+ "hay": 958,
1618
+ "head cabbage": 936,
1619
+ "hen": 8,
1620
+ "hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa": 996,
1621
+ "hermit crab": 125,
1622
+ "hip, rose hip, rosehip": 989,
1623
+ "hippopotamus, hippo, river horse, Hippopotamus amphibius": 344,
1624
+ "hog, pig, grunter, squealer, Sus scrofa": 341,
1625
+ "hognose snake, puff adder, sand viper": 54,
1626
+ "holster": 597,
1627
+ "home theater, home theatre": 598,
1628
+ "honeycomb": 599,
1629
+ "hook, claw": 600,
1630
+ "hoopskirt, crinoline": 601,
1631
+ "horizontal bar, high bar": 602,
1632
+ "hornbill": 93,
1633
+ "horned viper, cerastes, sand viper, horned asp, Cerastes cornutus": 66,
1634
+ "horse cart, horse-cart": 603,
1635
+ "hot pot, hotpot": 926,
1636
+ "hotdog, hot dog, red hot": 934,
1637
+ "hourglass": 604,
1638
+ "house finch, linnet, Carpodacus mexicanus": 12,
1639
+ "howler monkey, howler": 379,
1640
+ "hummingbird": 94,
1641
+ "hyena, hyaena": 276,
1642
+ "iPod": 605,
1643
+ "ibex, Capra ibex": 350,
1644
+ "ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus": 296,
1645
+ "ice cream, icecream": 928,
1646
+ "ice lolly, lolly, lollipop, popsicle": 929,
1647
+ "impala, Aepyceros melampus": 352,
1648
+ "indigo bunting, indigo finch, indigo bird, Passerina cyanea": 14,
1649
+ "indri, indris, Indri indri, Indri brevicaudatus": 384,
1650
+ "iron, smoothing iron": 606,
1651
+ "isopod": 126,
1652
+ "jacamar": 95,
1653
+ "jack-o'-lantern": 607,
1654
+ "jackfruit, jak, jack": 955,
1655
+ "jaguar, panther, Panthera onca, Felis onca": 290,
1656
+ "jay": 17,
1657
+ "jean, blue jean, denim": 608,
1658
+ "jeep, landrover": 609,
1659
+ "jellyfish": 107,
1660
+ "jersey, T-shirt, tee shirt": 610,
1661
+ "jigsaw puzzle": 611,
1662
+ "jinrikisha, ricksha, rickshaw": 612,
1663
+ "joystick": 613,
1664
+ "junco, snowbird": 13,
1665
+ "keeshond": 261,
1666
+ "kelpie": 227,
1667
+ "killer whale, killer, orca, grampus, sea wolf, Orcinus orca": 148,
1668
+ "kimono": 614,
1669
+ "king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica": 121,
1670
+ "king penguin, Aptenodytes patagonica": 145,
1671
+ "king snake, kingsnake": 56,
1672
+ "kit fox, Vulpes macrotis": 278,
1673
+ "kite": 21,
1674
+ "knee pad": 615,
1675
+ "knot": 616,
1676
+ "koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus": 105,
1677
+ "komondor": 228,
1678
+ "kuvasz": 222,
1679
+ "lab coat, laboratory coat": 617,
1680
+ "lacewing, lacewing fly": 318,
1681
+ "ladle": 618,
1682
+ "ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle": 301,
1683
+ "lakeside, lakeshore": 975,
1684
+ "lampshade, lamp shade": 619,
1685
+ "langur": 374,
1686
+ "laptop, laptop computer": 620,
1687
+ "lawn mower, mower": 621,
1688
+ "leaf beetle, chrysomelid": 304,
1689
+ "leafhopper": 317,
1690
+ "leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea": 34,
1691
+ "lemon": 951,
1692
+ "lens cap, lens cover": 622,
1693
+ "leopard, Panthera pardus": 288,
1694
+ "lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens": 387,
1695
+ "letter opener, paper knife, paperknife": 623,
1696
+ "library": 624,
1697
+ "lifeboat": 625,
1698
+ "lighter, light, igniter, ignitor": 626,
1699
+ "limousine, limo": 627,
1700
+ "limpkin, Aramus pictus": 135,
1701
+ "liner, ocean liner": 628,
1702
+ "lion, king of beasts, Panthera leo": 291,
1703
+ "lionfish": 396,
1704
+ "lipstick, lip rouge": 629,
1705
+ "little blue heron, Egretta caerulea": 131,
1706
+ "llama": 355,
1707
+ "loggerhead, loggerhead turtle, Caretta caretta": 33,
1708
+ "long-horned beetle, longicorn, longicorn beetle": 303,
1709
+ "lorikeet": 90,
1710
+ "lotion": 631,
1711
+ "loudspeaker, speaker, speaker unit, loudspeaker system, speaker system": 632,
1712
+ "loupe, jeweler's loupe": 633,
1713
+ "lumbermill, sawmill": 634,
1714
+ "lycaenid, lycaenid butterfly": 326,
1715
+ "lynx, catamount": 287,
1716
+ "macaque": 373,
1717
+ "macaw": 88,
1718
+ "magnetic compass": 635,
1719
+ "magpie": 18,
1720
+ "mailbag, postbag": 636,
1721
+ "mailbox, letter box": 637,
1722
+ "maillot": 638,
1723
+ "maillot, tank suit": 639,
1724
+ "malamute, malemute, Alaskan malamute": 249,
1725
+ "malinois": 225,
1726
+ "manhole cover": 640,
1727
+ "mantis, mantid": 315,
1728
+ "maraca": 641,
1729
+ "marimba, xylophone": 642,
1730
+ "marmoset": 377,
1731
+ "marmot": 336,
1732
+ "mashed potato": 935,
1733
+ "mask": 643,
1734
+ "matchstick": 644,
1735
+ "maypole": 645,
1736
+ "maze, labyrinth": 646,
1737
+ "measuring cup": 647,
1738
+ "meat loaf, meatloaf": 962,
1739
+ "medicine chest, medicine cabinet": 648,
1740
+ "meerkat, mierkat": 299,
1741
+ "megalith, megalithic structure": 649,
1742
+ "menu": 922,
1743
+ "microphone, mike": 650,
1744
+ "microwave, microwave oven": 651,
1745
+ "military uniform": 652,
1746
+ "milk can": 653,
1747
+ "miniature pinscher": 237,
1748
+ "miniature poodle": 266,
1749
+ "miniature schnauzer": 196,
1750
+ "minibus": 654,
1751
+ "miniskirt, mini": 655,
1752
+ "minivan": 656,
1753
+ "mink": 357,
1754
+ "missile": 657,
1755
+ "mitten": 658,
1756
+ "mixing bowl": 659,
1757
+ "mobile home, manufactured home": 660,
1758
+ "modem": 662,
1759
+ "monarch, monarch butterfly, milkweed butterfly, Danaus plexippus": 323,
1760
+ "monastery": 663,
1761
+ "mongoose": 298,
1762
+ "monitor": 664,
1763
+ "moped": 665,
1764
+ "mortar": 666,
1765
+ "mortarboard": 667,
1766
+ "mosque": 668,
1767
+ "mosquito net": 669,
1768
+ "motor scooter, scooter": 670,
1769
+ "mountain bike, all-terrain bike, off-roader": 671,
1770
+ "mountain tent": 672,
1771
+ "mouse, computer mouse": 673,
1772
+ "mousetrap": 674,
1773
+ "moving van": 675,
1774
+ "mud turtle": 35,
1775
+ "mushroom": 947,
1776
+ "muzzle": 676,
1777
+ "nail": 677,
1778
+ "neck brace": 678,
1779
+ "necklace": 679,
1780
+ "nematode, nematode worm, roundworm": 111,
1781
+ "night snake, Hypsiglena torquata": 60,
1782
+ "nipple": 680,
1783
+ "notebook, notebook computer": 681,
1784
+ "obelisk": 682,
1785
+ "oboe, hautboy, hautbois": 683,
1786
+ "ocarina, sweet potato": 684,
1787
+ "odometer, hodometer, mileometer, milometer": 685,
1788
+ "oil filter": 686,
1789
+ "orange": 950,
1790
+ "orangutan, orang, orangutang, Pongo pygmaeus": 365,
1791
+ "organ, pipe organ": 687,
1792
+ "oscilloscope, scope, cathode-ray oscilloscope, CRO": 688,
1793
+ "ostrich, Struthio camelus": 9,
1794
+ "otter": 360,
1795
+ "otterhound, otter hound": 175,
1796
+ "overskirt": 689,
1797
+ "ox": 345,
1798
+ "oxcart": 690,
1799
+ "oxygen mask": 691,
1800
+ "oystercatcher, oyster catcher": 143,
1801
+ "packet": 692,
1802
+ "paddle, boat paddle": 693,
1803
+ "paddlewheel, paddle wheel": 694,
1804
+ "padlock": 695,
1805
+ "paintbrush": 696,
1806
+ "pajama, pyjama, pj's, jammies": 697,
1807
+ "palace": 698,
1808
+ "panpipe, pandean pipe, syrinx": 699,
1809
+ "paper towel": 700,
1810
+ "papillon": 157,
1811
+ "parachute, chute": 701,
1812
+ "parallel bars, bars": 702,
1813
+ "park bench": 703,
1814
+ "parking meter": 704,
1815
+ "partridge": 86,
1816
+ "passenger car, coach, carriage": 705,
1817
+ "patas, hussar monkey, Erythrocebus patas": 371,
1818
+ "patio, terrace": 706,
1819
+ "pay-phone, pay-station": 707,
1820
+ "peacock": 84,
1821
+ "pedestal, plinth, footstall": 708,
1822
+ "pelican": 144,
1823
+ "pencil box, pencil case": 709,
1824
+ "pencil sharpener": 710,
1825
+ "perfume, essence": 711,
1826
+ "photocopier": 713,
1827
+ "pick, plectrum, plectron": 714,
1828
+ "pickelhaube": 715,
1829
+ "picket fence, paling": 716,
1830
+ "pickup, pickup truck": 717,
1831
+ "pier": 718,
1832
+ "piggy bank, penny bank": 719,
1833
+ "pill bottle": 720,
1834
+ "pillow": 721,
1835
+ "pineapple, ananas": 953,
1836
+ "ping-pong ball": 722,
1837
+ "pinwheel": 723,
1838
+ "pirate, pirate ship": 724,
1839
+ "pitcher, ewer": 725,
1840
+ "pizza, pizza pie": 963,
1841
+ "plane, carpenter's plane, woodworking plane": 726,
1842
+ "planetarium": 727,
1843
+ "plastic bag": 728,
1844
+ "plate": 923,
1845
+ "plate rack": 729,
1846
+ "platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus": 103,
1847
+ "plow, plough": 730,
1848
+ "plunger, plumber's helper": 731,
1849
+ "pole": 733,
1850
+ "polecat, fitch, foulmart, foumart, Mustela putorius": 358,
1851
+ "police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria": 734,
1852
+ "pomegranate": 957,
1853
+ "poncho": 735,
1854
+ "pool table, billiard table, snooker table": 736,
1855
+ "pop bottle, soda bottle": 737,
1856
+ "porcupine, hedgehog": 334,
1857
+ "pot, flowerpot": 738,
1858
+ "potpie": 964,
1859
+ "potter's wheel": 739,
1860
+ "power drill": 740,
1861
+ "prairie chicken, prairie grouse, prairie fowl": 83,
1862
+ "prayer rug, prayer mat": 741,
1863
+ "pretzel": 932,
1864
+ "printer": 742,
1865
+ "prison, prison house": 743,
1866
+ "proboscis monkey, Nasalis larvatus": 376,
1867
+ "projectile, missile": 744,
1868
+ "projector": 745,
1869
+ "promontory, headland, head, foreland": 976,
1870
+ "ptarmigan": 81,
1871
+ "puck, hockey puck": 746,
1872
+ "puffer, pufferfish, blowfish, globefish": 397,
1873
+ "pug, pug-dog": 254,
1874
+ "punching bag, punch bag, punching ball, punchball": 747,
1875
+ "purse": 748,
1876
+ "quail": 85,
1877
+ "quill, quill pen": 749,
1878
+ "quilt, comforter, comfort, puff": 750,
1879
+ "racer, race car, racing car": 751,
1880
+ "racket, racquet": 752,
1881
+ "radiator": 753,
1882
+ "radio telescope, radio reflector": 755,
1883
+ "radio, wireless": 754,
1884
+ "rain barrel": 756,
1885
+ "ram, tup": 348,
1886
+ "rapeseed": 984,
1887
+ "recreational vehicle, RV, R.V.": 757,
1888
+ "red fox, Vulpes vulpes": 277,
1889
+ "red wine": 966,
1890
+ "red wolf, maned wolf, Canis rufus, Canis niger": 271,
1891
+ "red-backed sandpiper, dunlin, Erolia alpina": 140,
1892
+ "red-breasted merganser, Mergus serrator": 98,
1893
+ "redbone": 168,
1894
+ "redshank, Tringa totanus": 141,
1895
+ "reel": 758,
1896
+ "reflex camera": 759,
1897
+ "refrigerator, icebox": 760,
1898
+ "remote control, remote": 761,
1899
+ "restaurant, eating house, eating place, eatery": 762,
1900
+ "revolver, six-gun, six-shooter": 763,
1901
+ "rhinoceros beetle": 306,
1902
+ "rifle": 764,
1903
+ "ringlet, ringlet butterfly": 322,
1904
+ "ringneck snake, ring-necked snake, ring snake": 53,
1905
+ "robin, American robin, Turdus migratorius": 15,
1906
+ "rock beauty, Holocanthus tricolor": 392,
1907
+ "rock crab, Cancer irroratus": 119,
1908
+ "rock python, rock snake, Python sebae": 62,
1909
+ "rocking chair, rocker": 765,
1910
+ "rotisserie": 766,
1911
+ "rubber eraser, rubber, pencil eraser": 767,
1912
+ "ruddy turnstone, Arenaria interpres": 139,
1913
+ "ruffed grouse, partridge, Bonasa umbellus": 82,
1914
+ "rugby ball": 768,
1915
+ "rule, ruler": 769,
1916
+ "running shoe": 770,
1917
+ "safe": 771,
1918
+ "safety pin": 772,
1919
+ "saltshaker, salt shaker": 773,
1920
+ "sandal": 774,
1921
+ "sandbar, sand bar": 977,
1922
+ "sarong": 775,
1923
+ "sax, saxophone": 776,
1924
+ "scabbard": 777,
1925
+ "scale, weighing machine": 778,
1926
+ "schipperke": 223,
1927
+ "school bus": 779,
1928
+ "schooner": 780,
1929
+ "scoreboard": 781,
1930
+ "scorpion": 71,
1931
+ "screen, CRT screen": 782,
1932
+ "screw": 783,
1933
+ "screwdriver": 784,
1934
+ "scuba diver": 983,
1935
+ "sea anemone, anemone": 108,
1936
+ "sea cucumber, holothurian": 329,
1937
+ "sea lion": 150,
1938
+ "sea slug, nudibranch": 115,
1939
+ "sea snake": 65,
1940
+ "sea urchin": 328,
1941
+ "seashore, coast, seacoast, sea-coast": 978,
1942
+ "seat belt, seatbelt": 785,
1943
+ "sewing machine": 786,
1944
+ "shield, buckler": 787,
1945
+ "shoe shop, shoe-shop, shoe store": 788,
1946
+ "shoji": 789,
1947
+ "shopping basket": 790,
1948
+ "shopping cart": 791,
1949
+ "shovel": 792,
1950
+ "shower cap": 793,
1951
+ "shower curtain": 794,
1952
+ "siamang, Hylobates syndactylus, Symphalangus syndactylus": 369,
1953
+ "sidewinder, horned rattlesnake, Crotalus cerastes": 68,
1954
+ "silky terrier, Sydney silky": 201,
1955
+ "ski": 795,
1956
+ "ski mask": 796,
1957
+ "skunk, polecat, wood pussy": 361,
1958
+ "sleeping bag": 797,
1959
+ "slide rule, slipstick": 798,
1960
+ "sliding door": 799,
1961
+ "slot, one-armed bandit": 800,
1962
+ "sloth bear, Melursus ursinus, Ursus ursinus": 297,
1963
+ "slug": 114,
1964
+ "snail": 113,
1965
+ "snorkel": 801,
1966
+ "snow leopard, ounce, Panthera uncia": 289,
1967
+ "snowmobile": 802,
1968
+ "snowplow, snowplough": 803,
1969
+ "soap dispenser": 804,
1970
+ "soccer ball": 805,
1971
+ "sock": 806,
1972
+ "soft-coated wheaten terrier": 202,
1973
+ "solar dish, solar collector, solar furnace": 807,
1974
+ "sombrero": 808,
1975
+ "sorrel": 339,
1976
+ "soup bowl": 809,
1977
+ "space bar": 810,
1978
+ "space heater": 811,
1979
+ "space shuttle": 812,
1980
+ "spaghetti squash": 940,
1981
+ "spatula": 813,
1982
+ "speedboat": 814,
1983
+ "spider monkey, Ateles geoffroyi": 381,
1984
+ "spider web, spider's web": 815,
1985
+ "spindle": 816,
1986
+ "spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish": 123,
1987
+ "spoonbill": 129,
1988
+ "sports car, sport car": 817,
1989
+ "spotlight, spot": 818,
1990
+ "spotted salamander, Ambystoma maculatum": 28,
1991
+ "squirrel monkey, Saimiri sciureus": 382,
1992
+ "stage": 819,
1993
+ "standard poodle": 267,
1994
+ "standard schnauzer": 198,
1995
+ "starfish, sea star": 327,
1996
+ "steam locomotive": 820,
1997
+ "steel arch bridge": 821,
1998
+ "steel drum": 822,
1999
+ "stethoscope": 823,
2000
+ "stingray": 6,
2001
+ "stinkhorn, carrion fungus": 994,
2002
+ "stole": 824,
2003
+ "stone wall": 825,
2004
+ "stopwatch, stop watch": 826,
2005
+ "stove": 827,
2006
+ "strainer": 828,
2007
+ "strawberry": 949,
2008
+ "street sign": 919,
2009
+ "streetcar, tram, tramcar, trolley, trolley car": 829,
2010
+ "stretcher": 830,
2011
+ "studio couch, day bed": 831,
2012
+ "stupa, tope": 832,
2013
+ "sturgeon": 394,
2014
+ "submarine, pigboat, sub, U-boat": 833,
2015
+ "suit, suit of clothes": 834,
2016
+ "sulphur butterfly, sulfur butterfly": 325,
2017
+ "sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita": 89,
2018
+ "sundial": 835,
2019
+ "sunglass": 836,
2020
+ "sunglasses, dark glasses, shades": 837,
2021
+ "sunscreen, sunblock, sun blocker": 838,
2022
+ "suspension bridge": 839,
2023
+ "swab, swob, mop": 840,
2024
+ "sweatshirt": 841,
2025
+ "swimming trunks, bathing trunks": 842,
2026
+ "swing": 843,
2027
+ "switch, electric switch, electrical switch": 844,
2028
+ "syringe": 845,
2029
+ "tabby, tabby cat": 281,
2030
+ "table lamp": 846,
2031
+ "tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui": 32,
2032
+ "tank, army tank, armored combat vehicle, armoured combat vehicle": 847,
2033
+ "tape player": 848,
2034
+ "tarantula": 76,
2035
+ "teapot": 849,
2036
+ "teddy, teddy bear": 850,
2037
+ "television, television system": 851,
2038
+ "tench, Tinca tinca": 0,
2039
+ "tennis ball": 852,
2040
+ "terrapin": 36,
2041
+ "thatch, thatched roof": 853,
2042
+ "theater curtain, theatre curtain": 854,
2043
+ "thimble": 855,
2044
+ "three-toed sloth, ai, Bradypus tridactylus": 364,
2045
+ "thresher, thrasher, threshing machine": 856,
2046
+ "throne": 857,
2047
+ "thunder snake, worm snake, Carphophis amoenus": 52,
2048
+ "tick": 78,
2049
+ "tiger beetle": 300,
2050
+ "tiger cat": 282,
2051
+ "tiger shark, Galeocerdo cuvieri": 3,
2052
+ "tiger, Panthera tigris": 292,
2053
+ "tile roof": 858,
2054
+ "timber wolf, grey wolf, gray wolf, Canis lupus": 269,
2055
+ "titi, titi monkey": 380,
2056
+ "toaster": 859,
2057
+ "tobacco shop, tobacconist shop, tobacconist": 860,
2058
+ "toilet seat": 861,
2059
+ "toilet tissue, toilet paper, bathroom tissue": 999,
2060
+ "torch": 862,
2061
+ "totem pole": 863,
2062
+ "toucan": 96,
2063
+ "tow truck, tow car, wrecker": 864,
2064
+ "toy poodle": 265,
2065
+ "toy terrier": 158,
2066
+ "toyshop": 865,
2067
+ "tractor": 866,
2068
+ "traffic light, traffic signal, stoplight": 920,
2069
+ "trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi": 867,
2070
+ "tray": 868,
2071
+ "tree frog, tree-frog": 31,
2072
+ "trench coat": 869,
2073
+ "triceratops": 51,
2074
+ "tricycle, trike, velocipede": 870,
2075
+ "trifle": 927,
2076
+ "trilobite": 69,
2077
+ "trimaran": 871,
2078
+ "tripod": 872,
2079
+ "triumphal arch": 873,
2080
+ "trolleybus, trolley coach, trackless trolley": 874,
2081
+ "trombone": 875,
2082
+ "tub, vat": 876,
2083
+ "turnstile": 877,
2084
+ "tusker": 101,
2085
+ "typewriter keyboard": 878,
2086
+ "umbrella": 879,
2087
+ "unicycle, monocycle": 880,
2088
+ "upright, upright piano": 881,
2089
+ "vacuum, vacuum cleaner": 882,
2090
+ "valley, vale": 979,
2091
+ "vase": 883,
2092
+ "vault": 884,
2093
+ "velvet": 885,
2094
+ "vending machine": 886,
2095
+ "vestment": 887,
2096
+ "viaduct": 888,
2097
+ "vine snake": 59,
2098
+ "violin, fiddle": 889,
2099
+ "vizsla, Hungarian pointer": 211,
2100
+ "volcano": 980,
2101
+ "volleyball": 890,
2102
+ "vulture": 23,
2103
+ "waffle iron": 891,
2104
+ "walking stick, walkingstick, stick insect": 313,
2105
+ "wall clock": 892,
2106
+ "wallaby, brush kangaroo": 104,
2107
+ "wallet, billfold, notecase, pocketbook": 893,
2108
+ "wardrobe, closet, press": 894,
2109
+ "warplane, military plane": 895,
2110
+ "warthog": 343,
2111
+ "washbasin, handbasin, washbowl, lavabo, wash-hand basin": 896,
2112
+ "washer, automatic washer, washing machine": 897,
2113
+ "water bottle": 898,
2114
+ "water buffalo, water ox, Asiatic buffalo, Bubalus bubalis": 346,
2115
+ "water jug": 899,
2116
+ "water ouzel, dipper": 20,
2117
+ "water snake": 58,
2118
+ "water tower": 900,
2119
+ "weasel": 356,
2120
+ "web site, website, internet site, site": 916,
2121
+ "weevil": 307,
2122
+ "whippet": 172,
2123
+ "whiptail, whiptail lizard": 41,
2124
+ "whiskey jug": 901,
2125
+ "whistle": 902,
2126
+ "white stork, Ciconia ciconia": 127,
2127
+ "white wolf, Arctic wolf, Canis lupus tundrarum": 270,
2128
+ "wig": 903,
2129
+ "wild boar, boar, Sus scrofa": 342,
2130
+ "window screen": 904,
2131
+ "window shade": 905,
2132
+ "wine bottle": 907,
2133
+ "wing": 908,
2134
+ "wire-haired fox terrier": 188,
2135
+ "wok": 909,
2136
+ "wolf spider, hunting spider": 77,
2137
+ "wombat": 106,
2138
+ "wood rabbit, cottontail, cottontail rabbit": 330,
2139
+ "wooden spoon": 910,
2140
+ "wool, woolen, woollen": 911,
2141
+ "worm fence, snake fence, snake-rail fence, Virginia fence": 912,
2142
+ "wreck": 913,
2143
+ "yawl": 914,
2144
+ "yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum": 986,
2145
+ "yurt": 915,
2146
+ "zebra": 340,
2147
+ "zucchini, courgette": 939
2148
+ },
2149
+ "layer_norm_eps": 1e-12,
2150
+ "length_penalty": 1.0,
2151
+ "max_length": 20,
2152
+ "min_length": 0,
2153
+ "mlp_ratio": [
2154
+ 4.0,
2155
+ 4.0,
2156
+ 4.0
2157
+ ],
2158
+ "model_type": "cvt",
2159
+ "no_repeat_ngram_size": 0,
2160
+ "num_beam_groups": 1,
2161
+ "num_beams": 1,
2162
+ "num_channels": 3,
2163
+ "num_heads": [
2164
+ 1,
2165
+ 3,
2166
+ 6
2167
+ ],
2168
+ "num_return_sequences": 1,
2169
+ "num_stages": 3,
2170
+ "output_attentions": false,
2171
+ "output_hidden_states": false,
2172
+ "output_scores": false,
2173
+ "pad_token_id": null,
2174
+ "padding_kv": [
2175
+ 1,
2176
+ 1,
2177
+ 1
2178
+ ],
2179
+ "padding_q": [
2180
+ 1,
2181
+ 1,
2182
+ 1
2183
+ ],
2184
+ "patch_padding": [
2185
+ 2,
2186
+ 1,
2187
+ 1
2188
+ ],
2189
+ "patch_sizes": [
2190
+ 7,
2191
+ 3,
2192
+ 3
2193
+ ],
2194
+ "patch_stride": [
2195
+ 4,
2196
+ 2,
2197
+ 2
2198
+ ],
2199
+ "pos_embed": [
2200
+ false,
2201
+ false,
2202
+ false
2203
+ ],
2204
+ "prefix": null,
2205
+ "problem_type": null,
2206
+ "projection_size": 768,
2207
+ "pruned_heads": {},
2208
+ "qkv_bias": [
2209
+ true,
2210
+ true,
2211
+ true
2212
+ ],
2213
+ "qkv_projection_method": [
2214
+ "dw_bn",
2215
+ "dw_bn",
2216
+ "dw_bn"
2217
+ ],
2218
+ "remove_invalid_values": false,
2219
+ "repetition_penalty": 1.0,
2220
+ "return_dict": true,
2221
+ "return_dict_in_generate": false,
2222
+ "sep_token_id": null,
2223
+ "stride_kv": [
2224
+ 2,
2225
+ 2,
2226
+ 2
2227
+ ],
2228
+ "stride_q": [
2229
+ 1,
2230
+ 1,
2231
+ 1
2232
+ ],
2233
+ "suppress_tokens": null,
2234
+ "task_specific_params": null,
2235
+ "temperature": 1.0,
2236
+ "tf_legacy_loss": false,
2237
+ "tie_encoder_decoder": false,
2238
+ "tie_word_embeddings": true,
2239
+ "tokenizer_class": null,
2240
+ "top_k": 50,
2241
+ "top_p": 1.0,
2242
+ "torch_dtype": "float32",
2243
+ "torchscript": false,
2244
+ "typical_p": 1.0,
2245
+ "use_bfloat16": false
2246
+ },
2247
+ "is_encoder_decoder": true,
2248
+ "model_type": "vision-encoder-decoder",
2249
+ "tie_word_embeddings": false,
2250
+ "torch_dtype": "float32",
2251
+ "transformers_version": "4.45.2"
2252
+ }
generation_config.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "pad_token_id": 0,
4
+ "transformers_version": "4.45.2"
5
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06a1e94ff55bee9d5189c814cc6c9ce0ea63ff3e56c5f00bd6cc3396f54d4cf7
3
+ size 450117792
modelling_longitudinal.py ADDED
@@ -0,0 +1,512 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import warnings
3
+ from dataclasses import dataclass
4
+ from typing import Any, Optional, Tuple, Union
5
+
6
+ import torch
7
+ import transformers
8
+ from peft import LoraConfig, TaskType, get_peft_config, get_peft_model
9
+ from torch.nn import CrossEntropyLoss
10
+ from transformers import AutoModel, PreTrainedTokenizerFast, VisionEncoderDecoderModel
11
+ from transformers.configuration_utils import PretrainedConfig
12
+ from transformers.modeling_outputs import BaseModelOutput, ModelOutput, Seq2SeqLMOutput
13
+ from transformers.modeling_utils import PreTrainedModel
14
+ from transformers.models.vision_encoder_decoder.configuration_vision_encoder_decoder import (
15
+ VisionEncoderDecoderConfig,
16
+ )
17
+ from transformers.utils import logging
18
+
19
+ logger = logging.get_logger(__name__)
20
+
21
+
22
+ class CvtWithProjectionHeadConfig(transformers.CvtConfig):
23
+ def __init__(self, projection_size: int = None, **kwargs: Any) -> None:
24
+ super().__init__(**kwargs)
25
+ self.projection_size = projection_size
26
+
27
+
28
+ class CvtProjectionHead(torch.nn.Module):
29
+
30
+ def __init__(self, config) -> None:
31
+ super().__init__()
32
+
33
+ # https://github.com/huggingface/transformers/blob/68287689f2f0d8b7063c400230b3766987abf18d/src/transformers/models/cvt/modeling_cvt.py#L657
34
+ self.layer_norm = torch.nn.LayerNorm(config.embed_dim[-1], eps=config.layer_norm_eps)
35
+
36
+ # No bias as following layer normalisation with bias:
37
+ self.projection = torch.nn.Linear(config.embed_dim[-1], config.projection_size, bias=False)
38
+
39
+
40
+ def forward(self, x: torch.Tensor) -> torch.Tensor:
41
+ x = self.layer_norm(x)
42
+ x = self.projection(x)
43
+ return x
44
+
45
+
46
+ class MultiCvtWithProjectionHead(transformers.CvtPreTrainedModel):
47
+ def __init__(self, config):
48
+ super().__init__(config)
49
+
50
+ self.cvt = transformers.CvtModel(config, add_pooling_layer=False)
51
+ self.projection_head = CvtProjectionHead(config)
52
+
53
+ # Initialize weights and apply final processing:
54
+ self.post_init()
55
+
56
+ def forward(
57
+ self,
58
+ pixel_values: Optional[torch.Tensor] = None,
59
+ output_hidden_states: Optional[bool] = None,
60
+ return_dict: Optional[bool] = None,
61
+ output_attentions: Optional[bool] = None,
62
+ ) -> Union[Tuple, ModelOutput]:
63
+
64
+ return_dict = return_dict if return_dict is not None else self.config.use_return_dict
65
+
66
+ # Flatten the batch and study_id dimensions:
67
+ outputs = self.cvt(
68
+ pixel_values.view(-1, *pixel_values.shape[2:]),
69
+ output_hidden_states=output_hidden_states,
70
+ return_dict=return_dict,
71
+ )
72
+
73
+ # Flatten h x w:
74
+ last_hidden_state = torch.flatten(outputs.last_hidden_state, 2)
75
+
76
+ # Project the features for each spatial position to the decoder's hidden size:
77
+ projection = self.projection_head(torch.permute(last_hidden_state, [0, 2, 1]))
78
+
79
+ # Concatenate the features for each chest X-ray:
80
+ projection = projection.view(pixel_values.shape[0], -1, projection.shape[-1])
81
+
82
+ # Derive the attention mask from the pixel values:
83
+ attention_mask = (pixel_values[:, :, 0, 0, 0] != 0.0).repeat_interleave(last_hidden_state.shape[-1], dim=1)
84
+
85
+ if not return_dict:
86
+ return projection
87
+
88
+ return ModelOutput(
89
+ last_hidden_state=projection, attention_mask=attention_mask,
90
+ )
91
+
92
+
93
+ class LongitudinalPromptMultiCXREncoderDecoderModel(VisionEncoderDecoderModel):
94
+
95
+ config_class = VisionEncoderDecoderConfig
96
+ base_model_prefix = "vision_encoder_decoder"
97
+ main_input_name = "pixel_values"
98
+ supports_gradient_checkpointing = True
99
+
100
+ def __init__(
101
+ self,
102
+ config: Optional[PretrainedConfig] = None,
103
+ encoder: Optional[PreTrainedModel] = None,
104
+ decoder: Optional[PreTrainedModel] = None,
105
+ encoder_decoder_ckpt_name: Optional[str] = None,
106
+ ):
107
+
108
+ if decoder:
109
+ assert decoder.config.add_cross_attention, '"add_cross_attention" must be True for the given decoder'
110
+ assert decoder.config.is_decoder, '"is_decoder" must be True for the given decoder'
111
+
112
+ if config is None and (encoder is None or decoder is None):
113
+ raise ValueError("Either a configuration or an encoder and a decoder has to be provided.")
114
+ if config is None:
115
+ config = VisionEncoderDecoderConfig.from_encoder_decoder_configs(encoder.config, decoder.config)
116
+ else:
117
+ if not isinstance(config, self.config_class):
118
+ raise ValueError(f"Config: {config} has to be of type {self.config_class}")
119
+
120
+ config.tie_word_embeddings = False
121
+
122
+ # initialize with config
123
+ PreTrainedModel.__init__(self, config)
124
+
125
+ # Encoder:
126
+ if encoder is None:
127
+ encoder = MultiCvtWithProjectionHead(config=config.encoder)
128
+
129
+ # Decoder:
130
+ if decoder is None:
131
+ decoder = transformers.BertLMHeadModel(config=config.decoder)
132
+
133
+ self.encoder = encoder
134
+ self.decoder = decoder
135
+
136
+ if self.encoder.config.to_dict() != self.config.encoder.to_dict():
137
+ logger.warning(
138
+ f"Config of the encoder: {self.encoder.__class__} is overwritten by shared encoder config:"
139
+ f" {self.config.encoder}"
140
+ )
141
+ if self.decoder.config.to_dict() != self.config.decoder.to_dict():
142
+ logger.warning(
143
+ f"Config of the decoder: {self.decoder.__class__} is overwritten by shared decoder config:"
144
+ f" {self.config.decoder}"
145
+ )
146
+
147
+ self.encoder.config = self.config.encoder
148
+ self.decoder.config = self.config.decoder
149
+
150
+ # Load multi checkpoint:
151
+ if encoder_decoder_ckpt_name:
152
+ encoder_decoder = AutoModel.from_pretrained(encoder_decoder_ckpt_name, trust_remote_code=True)
153
+ self.load_state_dict(encoder_decoder.state_dict())
154
+ else:
155
+ warnings.warn('The encoder-to-decoder model was not warm-started before applying low-rank approximation.')
156
+
157
+ # Freeze the encoder:
158
+ for p in self.encoder.parameters():
159
+ p.requires_grad = False
160
+
161
+ # Freeze the decoder and add LoRA:
162
+ peft_config = LoraConfig(
163
+ inference_mode=False,
164
+ r=8,
165
+ lora_alpha=32,
166
+ lora_dropout=0.1,
167
+ target_modules='bert.encoder.layer.[0-9]+.attention.self.(query|key)',
168
+ )
169
+ self.decoder = get_peft_model(self.decoder, peft_config)
170
+ self.decoder.print_trainable_parameters()
171
+
172
+ def forward(
173
+ self,
174
+ pixel_values: Optional[torch.FloatTensor] = None,
175
+ decoder_input_ids: Optional[torch.LongTensor] = None,
176
+ decoder_attention_mask: Optional[torch.BoolTensor] = None,
177
+ encoder_outputs: Optional[Tuple[torch.FloatTensor]] = None,
178
+ past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
179
+ decoder_inputs_embeds: Optional[torch.FloatTensor] = None,
180
+ labels: Optional[torch.LongTensor] = None,
181
+ use_cache: Optional[bool] = None,
182
+ output_attentions: Optional[bool] = None,
183
+ output_hidden_states: Optional[bool] = None,
184
+ return_dict: Optional[bool] = None,
185
+ **kwargs,
186
+ ) -> Union[Tuple[torch.FloatTensor], Seq2SeqLMOutput]:
187
+
188
+ return_dict = return_dict if return_dict is not None else self.config.use_return_dict
189
+
190
+ kwargs_encoder = {argument: value for argument, value in kwargs.items() if not argument.startswith("decoder_")}
191
+
192
+ kwargs_decoder = {
193
+ argument[len("decoder_") :]: value for argument, value in kwargs.items() if argument.startswith("decoder_")
194
+ }
195
+
196
+ if encoder_outputs is None:
197
+ if pixel_values is None:
198
+ raise ValueError("You have to specify pixel_values")
199
+
200
+ encoder_outputs = self.encoder(
201
+ pixel_values,
202
+ output_hidden_states=output_hidden_states,
203
+ return_dict=return_dict,
204
+ **kwargs_encoder,
205
+ ) # CvT does not support output_attentions.
206
+ elif isinstance(encoder_outputs, tuple):
207
+ encoder_outputs = BaseModelOutput(*encoder_outputs)
208
+
209
+ encoder_hidden_states = encoder_outputs[0]
210
+
211
+ decoder_outputs = self.decoder(
212
+ input_ids=decoder_input_ids,
213
+ attention_mask=decoder_attention_mask,
214
+ encoder_hidden_states=encoder_hidden_states,
215
+ encoder_attention_mask=encoder_outputs.attention_mask,
216
+ inputs_embeds=decoder_inputs_embeds,
217
+ output_attentions=output_attentions,
218
+ output_hidden_states=output_hidden_states,
219
+ use_cache=use_cache,
220
+ past_key_values=past_key_values,
221
+ return_dict=return_dict,
222
+ **kwargs_decoder,
223
+ )
224
+
225
+ # Loss:
226
+ loss = None
227
+ if labels is not None:
228
+ logits = decoder_outputs.logits if return_dict else decoder_outputs[0]
229
+ loss_fct = CrossEntropyLoss()
230
+ loss = loss_fct(logits.reshape(-1, self.decoder.config.vocab_size), labels.reshape(-1))
231
+
232
+ if not return_dict:
233
+ if loss is not None:
234
+ return (loss,) + decoder_outputs + encoder_outputs
235
+ else:
236
+ return decoder_outputs + encoder_outputs
237
+
238
+ return Seq2SeqLMOutput(
239
+ loss=loss,
240
+ logits=decoder_outputs.logits,
241
+ past_key_values=decoder_outputs.past_key_values,
242
+ decoder_hidden_states=decoder_outputs.hidden_states,
243
+ decoder_attentions=decoder_outputs.attentions,
244
+ cross_attentions=decoder_outputs.cross_attentions,
245
+ encoder_last_hidden_state=encoder_outputs.last_hidden_state,
246
+ # encoder_hidden_states=encoder_outputs.hidden_states,
247
+ # encoder_attentions=encoder_outputs.attentions,
248
+ )
249
+
250
+ def prepare_inputs_for_generation(
251
+ self,
252
+ input_ids,
253
+ special_token_ids,
254
+ mask_token_id,
255
+ past_key_values=None,
256
+ attention_mask=None,
257
+ use_cache=None,
258
+ encoder_outputs=None,
259
+ **kwargs,
260
+ ):
261
+ """
262
+ Modification of:
263
+ https://github.com/huggingface/transformers/blob/main/src/transformers/models/encoder_decoder/modeling_encoder_decoder.py#L660
264
+ """
265
+
266
+ # An update to generate() now prepends bos_token_id to each sequence if it does not exist at the start of the input:
267
+ # https://github.com/huggingface/transformers/blob/d533465150532b0c5de167b574e59f64c68b1154/src/transformers/generation/utils.py#L699C13-L699C30
268
+ # Hence, we remove the prepended bos_token_id from each sequence if it is there:
269
+ if torch.all(input_ids[:, 0] == 1):
270
+ input_ids = input_ids[:, 1:]
271
+
272
+ decoder_inputs = self.decoder.prepare_inputs_for_generation(input_ids, past_key_values=past_key_values)
273
+ decoder_attention_mask = (input_ids != mask_token_id).int()
274
+ decoder_position_ids = torch.nn.functional.relu(
275
+ torch.cumsum(decoder_attention_mask, dim=1, dtype=torch.int64) - 1
276
+ )
277
+
278
+ if not past_key_values:
279
+ token_type_ids = self.token_ids_to_token_type_ids(input_ids, special_token_ids, [0, 1, 0, 1])
280
+ else:
281
+ token_type_ids = self.token_ids_to_token_type_ids_past(input_ids, special_token_ids, [0, 1, 0, 1])
282
+ decoder_position_ids = decoder_position_ids[:, -1:]
283
+
284
+ input_dict = {
285
+ 'attention_mask': attention_mask,
286
+ 'decoder_attention_mask': decoder_attention_mask,
287
+ 'decoder_input_ids': decoder_inputs['input_ids'],
288
+ 'decoder_token_type_ids': token_type_ids,
289
+ 'decoder_position_ids': decoder_position_ids,
290
+ 'encoder_outputs': encoder_outputs,
291
+ 'past_key_values': decoder_inputs['past_key_values'],
292
+ 'use_cache': use_cache,
293
+ }
294
+ return input_dict
295
+
296
+ def token_ids_to_token_type_ids(self, token_ids, special_token_ids, token_type_id_sections=None):
297
+ """
298
+ Extract token type identifiers from the token identifiers.
299
+
300
+ Argument/s:
301
+ token_ids - token identifiers.
302
+ special_token_ids - special token identifiers that indicate the separation between sections.
303
+ token_type_id_section - token type identifier for each section.
304
+
305
+ Returns:
306
+ token_type_ids - token type identifiers.
307
+ """
308
+
309
+ token_type_id_sections = token_type_id_sections if token_type_id_sections is not None else list(range(len(special_token_ids) + 1))
310
+
311
+ mbatch_size, seq_len = token_ids.shape
312
+ token_type_ids = torch.full_like(token_ids, token_type_id_sections[0], dtype=torch.long, device=token_ids.device)
313
+
314
+ for i, j in enumerate(special_token_ids):
315
+ # Find first occurrence of special tokens that indicate the boundary between sections:
316
+ cols = (token_ids == j).int().argmax(dim=1)
317
+ rows = torch.arange(mbatch_size, device=token_ids.device)
318
+
319
+ # https://huggingface.co/docs/transformers/model_doc/bert#transformers.BertTokenizer.create_token_type_ids_from_sequences.example
320
+ cols += 1
321
+
322
+ # Ensure that the column index is not out of bounds. If 0, then token_id not present.
323
+ # This is safe as index 0 is always a special token (now equal to 1 due to +1):
324
+ rows = rows[torch.logical_and(cols != 1, cols < seq_len)]
325
+ cols = cols[torch.logical_and(cols != 1, cols < seq_len)]
326
+
327
+ # Indices to that correspond to the second sequence:
328
+ if rows.nelement() != 0:
329
+ ids = torch.stack([
330
+ torch.stack([x, z]) for (x, y) in zip(rows, cols) for z in torch.arange(
331
+ y, seq_len, device=token_ids.device,
332
+ )
333
+ ])
334
+
335
+ token_type_ids[ids[:, 0], ids[:, 1]] = token_type_id_sections[i + 1]
336
+
337
+ return token_type_ids
338
+
339
+ def token_ids_to_token_type_ids_past(self, token_ids, special_token_ids, token_type_id_sections=None):
340
+ """
341
+ Extract token type identifiers from the token identifiers if past != None.
342
+
343
+ Argument/s:
344
+ token_ids - token identifiers.
345
+ special_token_ids - special token identifiers that indicate the separation between sections.
346
+
347
+ Returns:
348
+ token_type_ids - token type identifiers.
349
+ """
350
+
351
+ token_type_id_sections = token_type_id_sections if token_type_id_sections is not None else list(range(len(special_token_ids) + 1))
352
+ token_type_ids = torch.full([token_ids.shape[0], 1], token_type_id_sections[0], dtype=torch.long, device=token_ids.device)
353
+
354
+ # https://huggingface.co/docs/transformers/model_doc/bert#transformers.BertTokenizer.create_token_type_ids_from_sequences.example
355
+ token_ids = token_ids[:, :-1]
356
+
357
+ for i, j in enumerate(special_token_ids):
358
+
359
+ # Find first occurrence of special token, which indicates the boundary between sections:
360
+ exists = torch.any(token_ids == j, dim=1, keepdim=True)
361
+ token_type_ids[exists] = token_type_id_sections[i + 1]
362
+
363
+ return token_type_ids
364
+
365
+ def tokenize_report_teacher_forcing(self, findings: str, impression: str, tokenizer: PreTrainedTokenizerFast, max_len: int):
366
+ """
367
+ Tokenize the reports and creates the inputs and targets for teacher forcing.
368
+
369
+ Argument/s:
370
+ findings - findings section.
371
+ impression - impression section.
372
+ return_token_type_ids - return the token type identifiers.
373
+ tokenizer - Hugging Face tokenizer.
374
+ max_len - maximum number of tokens.
375
+
376
+ Returns:
377
+ decoder_input_ids - the token identifiers for the input of the decoder.
378
+ decoder_attention_mask - the attention mask for the decoder_input_ids.
379
+ label_ids - the label token identifiers for the decoder.
380
+ """
381
+
382
+ # Prepare the sections for the tokenizer by placing special tokens between each section:
383
+ report = [f'{tokenizer.bos_token}{i}{tokenizer.sep_token}{j}{tokenizer.eos_token}' for i, j in
384
+ zip(findings, impression)]
385
+
386
+ # Tokenize the report:
387
+ tokenized = tokenizer(
388
+ report,
389
+ padding='longest',
390
+ truncation=True,
391
+ max_length=max_len + 1, # +1 to account for the bias between input and target.
392
+ return_tensors='pt',
393
+ return_token_type_ids=False,
394
+ add_special_tokens=False,
395
+ ).to(self.device)
396
+
397
+ # Modify for language modelling:
398
+ batch_dict = {
399
+
400
+ # Labels for the decoder (shifted right by one for autoregression):
401
+ 'label_ids': tokenized['input_ids'][:, 1:].detach().clone(),
402
+
403
+ # Remove last token identifier to match the sequence length of the labels:
404
+ 'decoder_input_ids': tokenized['input_ids'][:, :-1],
405
+
406
+ # Attention mask for the decoder_input_ids (remove first token so that the eos_token_id is not considered):
407
+ 'decoder_attention_mask': tokenized['attention_mask'][:, 1:],
408
+ }
409
+
410
+ return batch_dict
411
+
412
+ def split_and_decode_sections(self, token_ids, special_token_ids, tokenizer: PreTrainedTokenizerFast):
413
+ """
414
+ Split the token identifiers into sections, then convert the token identifiers into strings.
415
+
416
+ Argument/s:
417
+ token_ids - token identifiers.
418
+ special_token_ids - special token identifiers that indicate the end of each section.
419
+ tokenizer - Hugging Face tokenizer.
420
+
421
+ Returns:
422
+ token_type_ids - token type identifiers.
423
+ """
424
+
425
+ _, seq_len = token_ids.shape
426
+
427
+ # The number of sections is the same as the number of special_token_ids:
428
+ num_sections = len(special_token_ids)
429
+
430
+ sections = {k: [] for k in range(num_sections)}
431
+
432
+ for i in token_ids:
433
+ prev_col = 0
434
+ for j, k in enumerate(special_token_ids):
435
+
436
+ # The maximum sequence length was exceeded, thus no more tokens:
437
+ if prev_col >= seq_len:
438
+ sections[j].append('')
439
+ continue
440
+
441
+ # Find first occurrence of special tokens that indicate the boundary between sections:
442
+ col = (i == k).int().argmax().item()
443
+
444
+ # If equal to 0, token was not found, set the column to the sequence length (as the decoder exceeded
445
+ # the maximum sequence length):
446
+ if col == 0:
447
+ col = seq_len
448
+
449
+ # Extract section token identifiers:
450
+ section_token_ids = i[prev_col:col]
451
+ prev_col = col
452
+ section_string = tokenizer.decode(section_token_ids, skip_special_tokens=True)
453
+
454
+ sections[j].append(section_string)
455
+
456
+ return tuple(sections.values())
457
+
458
+ def tokenize_prompt(
459
+ self,
460
+ previous_findings: str,
461
+ previous_impression: str,
462
+ tokenizer: PreTrainedTokenizerFast,
463
+ max_len: int,
464
+ add_bos_token_id: bool = False,
465
+ ):
466
+ """
467
+ Tokenize the sections of the previous report to be used as a prompt.
468
+
469
+ Argument/s:
470
+ previous_findings - previous findings section.
471
+ previous_impression - previous impression section.
472
+ tokenizer - Hugging Face tokenizer.
473
+ max_len - maximum number of tokens.
474
+ add_bos_token_id - whether to add the BOS token identifier to the prompt.
475
+
476
+ Returns:
477
+ input_ids - the input identifiers for the previous impression.
478
+ attention_mask - the attention mask for the previous impression
479
+ """
480
+
481
+ # Use [NPF]/[NPI] special token if no previous findings/impression:
482
+ previous_findings = ['[NPF]' if not i else i for i in previous_findings]
483
+ previous_impression = ['[NPI]' if not i else i for i in previous_impression]
484
+
485
+ # Prepare the sections for the tokenizer by placing special tokens:
486
+ previous_sections = [
487
+ f'[PMT]{i}[PMT-SEP]{j}{tokenizer.bos_token}' if add_bos_token_id else f'[PMT]{i}[PMT-SEP]{j}' \
488
+ for i, j in zip(previous_findings, previous_impression)
489
+ ]
490
+
491
+ # Tokenize:
492
+ previous_sections = tokenizer(
493
+ previous_sections,
494
+ padding='longest',
495
+ truncation=True,
496
+ max_length=max_len,
497
+ return_tensors='pt',
498
+ return_token_type_ids=False,
499
+ add_special_tokens=False,
500
+ ).to(self.device)
501
+
502
+ # Ensure BOS token identifier is at the end of the input_ids:
503
+ if previous_sections.input_ids.shape[1] == max_len:
504
+ previous_sections.input_ids[:, -1] = torch.where(
505
+ previous_sections.attention_mask[:, -1] == 1,
506
+ tokenizer.bos_token_id,
507
+ previous_sections.input_ids[:, -1],
508
+ )
509
+
510
+ assert previous_sections.input_ids.shape[1] <= max_len
511
+
512
+ return {'input_ids': previous_sections.input_ids, 'attention_mask': previous_sections.attention_mask}