Update README.md
Browse files
README.md
CHANGED
|
@@ -1,40 +1,77 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
license:
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
-
|
| 11 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
datasets:
|
| 13 |
-
-
|
| 14 |
metrics:
|
| 15 |
-
- accuracy
|
| 16 |
-
- bleu
|
| 17 |
-
- wer
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
|
| 36 |
|
| 37 |
-
---
|
| 38 |
|
| 39 |
### **3. 提供可下载文件**
|
| 40 |
确保以下文件已上传到仓库,便于用户下载和运行:
|
|
|
|
| 1 |
+
## 模型卡
|
| 2 |
+
---------------------------------------------------------------------
|
| 3 |
+
metadata:
|
| 4 |
+
language: multilingual # AutoModel 是一个支持多语言处理的多模态模型
|
| 5 |
+
license:
|
| 6 |
+
- apache-2.0
|
| 7 |
+
- MIT # Apache 2.0 和 MIT 是开源许可
|
| 8 |
+
library_name: pytorch # 该模型基于 PyTorch 构建
|
| 9 |
+
tags:
|
| 10 |
+
- multimodal # 该模型是多模态模型
|
| 11 |
+
- image # 处理图像任务
|
| 12 |
+
- text # 处理文本任务
|
| 13 |
+
- audio # 处理语音任务
|
| 14 |
+
- vqa # 支持视觉问答任务
|
| 15 |
+
- automatspeerecognition # 支持自动语音识别任务
|
| 16 |
+
- retrieval # 支持信息检索任务
|
| 17 |
datasets:
|
| 18 |
+
- synthetdataset # 训练和验证使用了合成的多模态数据集
|
| 19 |
metrics:
|
| 20 |
+
- accuracy # 视觉问答任务的准确率
|
| 21 |
+
- bleu # 生成式任务(如字幕生成)的 BLEU 指标
|
| 22 |
+
- wer # 语音识别任务的 WER(Word Error Rate)
|
| 23 |
+
base_model: None # 该模型为独立设计,没有基于预训练模型
|
| 24 |
+
widget:
|
| 25 |
+
- text: "A cat playing with a ball"
|
| 26 |
+
example_title: "Cat"
|
| 27 |
+
- text: "A dog jumping over a fence"
|
| 28 |
+
example_title: "Dog"
|
| 29 |
+
|
| 30 |
+
model_index:
|
| 31 |
+
- name: AutoModel
|
| 32 |
+
results:
|
| 33 |
+
- task:
|
| 34 |
+
type: vqa # 支持视觉问答任务
|
| 35 |
+
name: Visual Question Answering
|
| 36 |
+
dataset:
|
| 37 |
+
type: synthetdataset
|
| 38 |
+
name: Synthetic Multimodal Dataset
|
| 39 |
+
config: default
|
| 40 |
+
split: test
|
| 41 |
+
revision: main
|
| 42 |
+
metrics:
|
| 43 |
+
- type: accuracy
|
| 44 |
+
value: 85.0
|
| 45 |
+
name: VQA Accuracy
|
| 46 |
+
- task:
|
| 47 |
+
type: automatspeerecognition
|
| 48 |
+
name: Automatic Speech Recognition
|
| 49 |
+
dataset:
|
| 50 |
+
type: synthetdataset
|
| 51 |
+
name: Synthetic Multimodal Dataset
|
| 52 |
+
config: default
|
| 53 |
+
split: test
|
| 54 |
+
revision: main
|
| 55 |
+
metrics:
|
| 56 |
+
- type: wer
|
| 57 |
+
value: 15.3
|
| 58 |
+
name: Test WER
|
| 59 |
+
- task:
|
| 60 |
+
type: captioning
|
| 61 |
+
name: Image Captioning
|
| 62 |
+
dataset:
|
| 63 |
+
type: synthetdataset
|
| 64 |
+
name: Synthetic Multimodal Dataset
|
| 65 |
+
config: default
|
| 66 |
+
split: test
|
| 67 |
+
revision: main
|
| 68 |
+
metrics:
|
| 69 |
+
- type: bleu
|
| 70 |
+
value: 27.5
|
| 71 |
+
name: BL4
|
| 72 |
+
-----------------------------------------------------------
|
| 73 |
|
| 74 |
|
|
|
|
| 75 |
|
| 76 |
### **3. 提供可下载文件**
|
| 77 |
确保以下文件已上传到仓库,便于用户下载和运行:
|