andrewzhang505 commited on
Commit
96e3eda
1 Parent(s): bbc9d5c

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +34 -3
README.md CHANGED
@@ -15,11 +15,42 @@ model-index:
15
  type: Ant
16
  metrics:
17
  - type: mean_reward
18
- value: 11830.10 +/- 875.26
19
  name: mean_reward
20
  verified: false
21
  ---
22
 
23
  A(n) **APPO** model trained on the **Ant** environment.
24
- This model was trained using Sample Factory 2.0: https://github.com/alex-petrenko/sample-factory
25
-
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15
  type: Ant
16
  metrics:
17
  - type: mean_reward
18
+ value: 11827.40 +/- 1185.26
19
  name: mean_reward
20
  verified: false
21
  ---
22
 
23
  A(n) **APPO** model trained on the **Ant** environment.
24
+
25
+ This model was trained using Sample-Factory 2.0: https://github.com/alex-petrenko/sample-factory.
26
+ Documentation for how to use Sample-Factory can be found at https://www.samplefactory.dev/
27
+
28
+
29
+ **Downloading the model**
30
+
31
+ After installing Sample-Factory, download the model with:
32
+ ```
33
+ python -m sample_factory.huggingface.load_from_hub -r andrewzhang505/isaacgym_ant
34
+ ```
35
+
36
+
37
+ **Using the model**
38
+
39
+ To run the model after download, use the `enjoy` script corresponding to this environment:
40
+ ```
41
+ python -m sf_examples.isaacgym_examples.enjoy_isaacgym --algo=APPO --env=Ant --train_dir=./train_dir --experiment=isaacgym_ant
42
+ ```
43
+
44
+
45
+ You can also upload models to the Hugging Face Hub using the same script with the `--push_to_hub` flag.
46
+ See https://www.samplefactory.dev/10-huggingface/huggingface/ for more details
47
+
48
+ **Training with this model**
49
+
50
+ To continue training with this model, use the `train` script corresponding to this environment:
51
+ ```
52
+ python -m sf_examples.isaacgym_examples.train_isaacgym --algo=APPO --env=Ant --train_dir=./train_dir --experiment=isaacgym_ant --restart_behavior=resume --train_for_env_steps=10000000000
53
+ ```
54
+
55
+ Note, you may have to adjust `--train_for_env_steps` to a suitably high number as the experiment will resume at the number of steps it concluded at.
56
+