snsd0805/vlnbert
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update readme and names

Browse Source
This commit is contained in:
Yicong Hong 2021-01-14 16:44:01 +11:00
parent aca9e7180a
commit fe995bd0e7
3 changed files with 3 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@ -47,8 +47,6 @@ Please refer to [vlnbert_init.py](https://github.com/YicongHong/Recurrent-VLN-BE
Please read Peter Anderson's VLN paper for the [R2R Navigation task](https://arxiv.org/abs/1711.07280).
Our code is based on the code structure of the [EnvDrop](https://github.com/airsplay/R2R-EnvDrop).
### Reproduce Testing Results
To replicate the performance reported in our paper, load the trained network weights and run validation:
@ -56,7 +54,7 @@ To replicate the performance reported in our paper, load the trained network wei
bash run/test_agent.bash
```
You can simply switch between the OSCAR-based and the PREVALENT-based VLN models by changing the arguments `vlnmodel` (oscar or prevalent) and `load` (trained model paths).
You can simply switch between the OSCAR-based and the PREVALENT-based VLN models by changing the arguments `vlnbert` (oscar or prevalent) and `load` (trained model paths).
### Training

2
run/test_agent.bash
View File

@ -1,4 +1,4 @@
name=VLNBERT-test
name=VLNBERT-test-Prevalent
flag="--vlnbert prevalent

2
run/train_agent.bash
View File

@ -1,4 +1,4 @@
name=VLNBERT-train
name=VLNBERT-train-Prevalent
flag="--vlnbert prevalent