Hugging Face's logo

radar-generalist
/
RADAR

Safetensors
English
vision-language pretraining
medical
ct diagnosis
Model card Files
xet
 Community
changwx commited on
Commit
2bf227e
·
verified ·
1 Parent(s): e504a3e

first upload

Browse files
Files changed (20) hide show
  1. .gitattributes +1 -0
  2. README.md +31 -1
  3. bert-base-chinese/config.json +26 -0
  4. bert-base-chinese/config_decoder.json +28 -0
  5. bert-base-chinese/pytorch_model.bin +3 -0
  6. bert-base-chinese/tokenizer.json +0 -0
  7. bert-base-chinese/tokenizer_config.json +1 -0
  8. bert-base-chinese/vocab.txt +0 -0
  9. bert-base-uncased/config.json +24 -0
  10. bert-base-uncased/config_decoder.json +26 -0
  11. bert-base-uncased/model.safetensors +3 -0
  12. bert-base-uncased/pytorch_model.bin +3 -0
  13. bert-base-uncased/tokenizer.json +0 -0
  14. bert-base-uncased/tokenizer_config.json +3 -0
  15. bert-base-uncased/vocab.txt +0 -0
  16. checkpoint_radar_plus.pth +3 -0
  17. checkpoint_radar_plus_finetuned_on_merlin.pth +3 -0
  18. checkpoint_radar_pretrain.pth +3 -0
  19. checkpoint_unet.pth +3 -0
  20. radar_fig0.png +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ radar_fig0.png filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -1,3 +1,33 @@
1
  ---
2
- license: cc-by-sa-4.0
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: cc-by-nc-sa-4.0
3
+ language:
4
+ - en
5
+ tags:
6
+ - vision-language pretraining
7
+ - medical
8
+ - ct diagnosis
9
  ---
10
+
11
+ # RADAR: An Expert-Level Generalist AI for Abdominal CT Diagnosis
12
+ [![GitHub](https://img.shields.io/badge/GitHub-Code-blue?logo=github)](https://github.com/alibaba-damo-academy/damo-radar)
13
+ [![Zenodo](https://img.shields.io/badge/Zenodo-Code-7B61FF?logo=zenodo&logoColor=white)](https://zenodo.org/records/21271172)
14
+ [![License: CC BY-NC-SA 4.0](https://img.shields.io/badge/License-CC_BY--NC--SA_4.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc-sa/4.0/)
15
+
16
+ RADAR is a generalist vision-language model trained on over 400,000 contrast-enhanced abdominal CT examinations with 15 million anatomy-aware image–text pairs, learning directly from clinical reports without manual annotation. RADAR provides a scalable and versatile framework for radiology AI, demonstrating expert-level performance across both routine and complex clinical tasks.
17
+
18
+ <p align="center">
19
+ <img src="radar_fig0.png" alt="RADAR Overview" width="90%">
20
+ </p>
21
+
22
+ ## Model Preparation
23
+
24
+ Pre-trained checkpoints are available on [HuggingFace](https://huggingface.co/datasets/radar-generalist/RADAR).
25
+
26
+ | File | Description | Destination |
27
+ | --- | --- | --- |
28
+ | `checkpoint_radar_pretrain.pth` | RADAR pre-trained on RAD-CT | `radar/ckpt/checkpoint_radar_pretrain.pth` |
29
+ | `bert-base-chinese` | BERT tokenizer and model (Chinese) | `radar/ckpt/bert-base-chinese/` |
30
+ | `bert-base-uncased` | BERT tokenizer and model (English) | `radar/ckpt/bert-base-uncased/` |
31
+ | `checkpoint_unet.pth` | Pretrained VisionBranch (UNet) checkpoint | `radar/ckpt/checkpoint_unet.pth` |
32
+ | `checkpoint_radar_plus.pth` | RADAR+ checkpoint trained from scratch on Merlin-CT-Train | `radar/ckpt/checkpoint_radar_plus.pth` |
33
+ | `checkpoint_radar_plus_finetuned_on_merlin.pth` | RADAR+ checkpoint pretrained on RAD-CT and finetuned on Merlin-CT-Train | `radar/ckpt/checkpoint_radar_plus_finetuned_on_merlin.pth` |
bert-base-chinese/config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForMaskedLM"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "directionality": "bidi",
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 3072,
12
+ "layer_norm_eps": 1e-12,
13
+ "max_position_embeddings": 512,
14
+ "model_type": "bert",
15
+ "num_attention_heads": 12,
16
+ "num_hidden_layers": 12,
17
+ "pad_token_id": 0,
18
+ "pooler_fc_size": 768,
19
+ "pooler_num_attention_heads": 12,
20
+ "pooler_num_fc_layers": 3,
21
+ "pooler_size_per_head": 128,
22
+ "pooler_type": "first_token_transform",
23
+ "type_vocab_size": 2,
24
+ "vocab_size": 21128,
25
+ "add_type_embeddings": false
26
+ }
bert-base-chinese/config_decoder.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForMaskedLM"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "directionality": "bidi",
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 3072,
12
+ "layer_norm_eps": 1e-12,
13
+ "max_position_embeddings": 512,
14
+ "model_type": "bert",
15
+ "num_attention_heads": 12,
16
+ "num_hidden_layers": 12,
17
+ "pad_token_id": 0,
18
+ "pooler_fc_size": 768,
19
+ "pooler_num_attention_heads": 12,
20
+ "pooler_num_fc_layers": 3,
21
+ "pooler_size_per_head": 128,
22
+ "pooler_type": "first_token_transform",
23
+ "type_vocab_size": 2,
24
+ "vocab_size": 21128,
25
+ "add_type_embeddings": false,
26
+ "encoder_width": 768,
27
+ "add_cross_attention": true
28
+ }
bert-base-chinese/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a693db616eaf647ed2bfe531e1fa446637358fc108a8bf04e8d4db17e837ee9
3
+ size 411577189
bert-base-chinese/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
bert-base-chinese/tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": false, "model_max_length": 512}
bert-base-chinese/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
bert-base-uncased/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForMaskedLM"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "gradient_checkpointing": false,
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 3072,
12
+ "layer_norm_eps": 1e-12,
13
+ "max_position_embeddings": 512,
14
+ "model_type": "bert",
15
+ "num_attention_heads": 12,
16
+ "num_hidden_layers": 12,
17
+ "pad_token_id": 0,
18
+ "position_embedding_type": "absolute",
19
+ "transformers_version": "4.6.0.dev0",
20
+ "type_vocab_size": 2,
21
+ "use_cache": true,
22
+ "vocab_size": 30522,
23
+ "add_type_embeddings": false
24
+ }
bert-base-uncased/config_decoder.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForMaskedLM"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "gradient_checkpointing": false,
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 3072,
12
+ "layer_norm_eps": 1e-12,
13
+ "max_position_embeddings": 512,
14
+ "model_type": "bert",
15
+ "num_attention_heads": 12,
16
+ "num_hidden_layers": 12,
17
+ "pad_token_id": 0,
18
+ "position_embedding_type": "absolute",
19
+ "transformers_version": "4.6.0.dev0",
20
+ "type_vocab_size": 2,
21
+ "use_cache": true,
22
+ "vocab_size": 30522,
23
+ "add_type_embeddings": false,
24
+ "encoder_width": 768,
25
+ "add_cross_attention": true
26
+ }
bert-base-uncased/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68d45e234eb4a928074dfd868cead0219ab85354cc53d20e772753c6bb9169d3
3
+ size 440449768
bert-base-uncased/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:097417381d6c7230bd9e3557456d726de6e83245ec8b24f529f60198a67b203a
3
+ size 440473133
bert-base-uncased/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
bert-base-uncased/tokenizer_config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "do_lower_case": true
3
+ }
bert-base-uncased/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint_radar_plus.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4dd201111000fa11085d49eddf287520a37cc76288eaca93eab8f8510480f15a
3
+ size 1651336682
checkpoint_radar_plus_finetuned_on_merlin.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd5aae5635de3712312c067dd82114cd4c215e7dcfc11eb3619dab73fcceaebc
3
+ size 1651335230
checkpoint_radar_pretrain.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e8b1b50b92162fcafcf8231e51ffbe9d1be2be52ccff4f967e4beb71e4499f4
3
+ size 1566049482
checkpoint_unet.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:052bbe8cbbc5d23859cb8fe1935249e94c1ef733e8e4a79153e9fc8988e84985
3
+ size 208341958
radar_fig0.png ADDED

Git LFS Details

  • SHA256: 6022f13fa7bde6fd5402b0f2cf4876c39d89f2f8428e47a522751b0385ba38b5
  • Pointer size: 132 Bytes
  • Size of remote file: 1.53 MB