zhou777 commited on
Commit
6f3c5ae
·
verified ·
1 Parent(s): 2de6616

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -3
README.md CHANGED
@@ -10,7 +10,7 @@ tags:
10
 
11
  **If you use this model, please pay attention to our work: LandGPT: A Multimodal Large Language Model for Parcel-Level Land Use Classification with Multi-Source Data (under submission).**
12
 
13
- To reproduce the main contributions of the paper, please download this model and configure the model runtime library according to `https://lmdeploy.readthedocs.io/en/latest/`.
14
 
15
  In addition, a proper JSONL file is required as an image index to feed into the model. The code for creating this JSONL file has been open-sourced at `https://figshare.com/s/1eb38bac0cc970698379`
16
 
@@ -22,13 +22,15 @@ python true_pre_2level.py
22
 
23
  This will launch the model inference.
24
 
 
 
25
  # LandGPT
26
 
27
  **此仓库为LandGPT-8B和相应的模型推理代码**
28
 
29
  **如使用此模型,请关注我们的工作:LandGPT: A Multimodal Large Language Model for Parcel-Level Land Use Classification with Multi-Source Data(投稿中)**
30
 
31
- 如复现文章主要工作,请下载此模型,并根据`https://lmdeploy.readthedocs.io/en/latest/`配置模型运行库。
32
 
33
  此外,还需要合适的jsonl文件作为图像索引送入模型,制作此jsonl文件的代码已经开源在`https://figshare.com/s/1eb38bac0cc970698379`。
34
 
@@ -38,4 +40,6 @@ This will launch the model inference.
38
  python true_pre_2level.py
39
  ```
40
 
41
- 即可启动模型推理。
 
 
 
10
 
11
  **If you use this model, please pay attention to our work: LandGPT: A Multimodal Large Language Model for Parcel-Level Land Use Classification with Multi-Source Data (under submission).**
12
 
13
+ To reproduce the main contributions of the paper, please download this model and configure the model runtime library according to `pip install lmdeploy>=0.5.3`.
14
 
15
  In addition, a proper JSONL file is required as an image index to feed into the model. The code for creating this JSONL file has been open-sourced at `https://figshare.com/s/1eb38bac0cc970698379`
16
 
 
22
 
23
  This will launch the model inference.
24
 
25
+ We recommend using a computing card with at least 24GB of VRAM for model inference, as insufficient VRAM may affect the inference quality.
26
+
27
  # LandGPT
28
 
29
  **此仓库为LandGPT-8B和相应的模型推理代码**
30
 
31
  **如使用此模型,请关注我们的工作:LandGPT: A Multimodal Large Language Model for Parcel-Level Land Use Classification with Multi-Source Data(投稿中)**
32
 
33
+ 如复现文章主要工作,请下载此模型,并根据` pip install lmdeploy>=0.5.3`配置模型运行库。
34
 
35
  此外,还需要合适的jsonl文件作为图像索引送入模型,制作此jsonl文件的代码已经开源在`https://figshare.com/s/1eb38bac0cc970698379`。
36
 
 
40
  python true_pre_2level.py
41
  ```
42
 
43
+ 即可启动模型推理。
44
+
45
+ 我们建议使用至少24GB显存的计算卡来进行模型推理,过低的显存可能会影响推理质量。