First Push
Browse files- README.md +31 -0
- SoccerTwos.onnx +3 -0
- SoccerTwos/SoccerTwos-48499680.onnx +3 -0
- SoccerTwos/SoccerTwos-48499680.pt +3 -0
- SoccerTwos/SoccerTwos-48999460.onnx +3 -0
- SoccerTwos/SoccerTwos-48999460.pt +3 -0
- SoccerTwos/SoccerTwos-49499964.onnx +3 -0
- SoccerTwos/SoccerTwos-49499964.pt +3 -0
- SoccerTwos/SoccerTwos-49999430.onnx +3 -0
- SoccerTwos/SoccerTwos-49999430.pt +3 -0
- SoccerTwos/SoccerTwos-50000890.onnx +3 -0
- SoccerTwos/SoccerTwos-50000890.pt +3 -0
- SoccerTwos/checkpoint.pt +3 -0
- SoccerTwos/events.out.tfevents.1685525223.fingerdl3.379268.0 +3 -0
- config.json +1 -0
- configuration.yaml +82 -0
- run_logs/Player-0.log +0 -0
- run_logs/timers.json +424 -0
- run_logs/training_status.json +66 -0
README.md
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
---
|
3 |
+
tags:
|
4 |
+
- unity-ml-agents
|
5 |
+
- ml-agents
|
6 |
+
- deep-reinforcement-learning
|
7 |
+
- reinforcement-learning
|
8 |
+
- ML-Agents-SoccerTwos
|
9 |
+
library_name: ml-agents
|
10 |
+
---
|
11 |
+
|
12 |
+
# **poca** Agent playing **SoccerTwos**
|
13 |
+
This is a trained model of a **poca** agent playing **SoccerTwos** using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents).
|
14 |
+
|
15 |
+
## Usage (with ML-Agents)
|
16 |
+
The Documentation: https://github.com/huggingface/ml-agents#get-started
|
17 |
+
We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub:
|
18 |
+
|
19 |
+
|
20 |
+
### Resume the training
|
21 |
+
```
|
22 |
+
mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume
|
23 |
+
```
|
24 |
+
### Watch your Agent play
|
25 |
+
You can watch your agent **playing directly in your browser:**.
|
26 |
+
|
27 |
+
1. Go to https://huggingface.co/spaces/unity/ML-Agents-SoccerTwos
|
28 |
+
2. Step 1: Write your model_id: Xoyo/poca-SoccerTwos
|
29 |
+
3. Step 2: Select your *.nn /*.onnx file
|
30 |
+
4. Click on Watch the agent play 👀
|
31 |
+
|
SoccerTwos.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6e35749cd79129ab2c135f06df3e0b46c5ed340035cafc3e55cb2a981a6b76d7
|
3 |
+
size 1764633
|
SoccerTwos/SoccerTwos-48499680.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:297147efff5775ca301efb8a1ffacea3ab82adc6980ed192d30a508be3ba3f73
|
3 |
+
size 1764633
|
SoccerTwos/SoccerTwos-48499680.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:02ce25db65425d49e1e0a2b68e343b74ea0bbfb9ca3147f5c241979571893558
|
3 |
+
size 28422481
|
SoccerTwos/SoccerTwos-48999460.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:844cbd920d83f36e935ca68fe64d21c2db4f78346272ff7f391a5fba6bb58aba
|
3 |
+
size 1764633
|
SoccerTwos/SoccerTwos-48999460.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9c6413fc69aa7b7e2dbd0bf6bc28c94136bd2c6fd4dbc70a9fd8a36d730da64b
|
3 |
+
size 28422481
|
SoccerTwos/SoccerTwos-49499964.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a4d417d793a2464e0d75c32edbcd348dad6ca4b979f00a3bf5467f4cd181c1f6
|
3 |
+
size 1764633
|
SoccerTwos/SoccerTwos-49499964.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0b7fc6779cf0f8110742690935d256f678f73ea0e2777085c90a6373a9671d9e
|
3 |
+
size 28422481
|
SoccerTwos/SoccerTwos-49999430.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6e35749cd79129ab2c135f06df3e0b46c5ed340035cafc3e55cb2a981a6b76d7
|
3 |
+
size 1764633
|
SoccerTwos/SoccerTwos-49999430.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a2a75a66ac1781097eed7963d15d17da4256468ed455f1edf94daecc9d0eb057
|
3 |
+
size 28422481
|
SoccerTwos/SoccerTwos-50000890.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6e35749cd79129ab2c135f06df3e0b46c5ed340035cafc3e55cb2a981a6b76d7
|
3 |
+
size 1764633
|
SoccerTwos/SoccerTwos-50000890.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9ebd6a82b31bd98d2125299eb402370bf98753ca9b5a36f629feda9cc3b44a4f
|
3 |
+
size 28422481
|
SoccerTwos/checkpoint.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9ebd6a82b31bd98d2125299eb402370bf98753ca9b5a36f629feda9cc3b44a4f
|
3 |
+
size 28422481
|
SoccerTwos/events.out.tfevents.1685525223.fingerdl3.379268.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e6198b18a4f86be88518fed581dbef180c2731313f6ce1d857e9b32591d7d51f
|
3 |
+
size 7282268
|
config.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"default_settings": null, "behaviors": {"SoccerTwos": {"trainer_type": "poca", "hyperparameters": {"batch_size": 2048, "buffer_size": 20480, "learning_rate": 0.0003, "beta": 0.005, "epsilon": 0.2, "lambd": 0.95, "num_epoch": 3, "learning_rate_schedule": "constant", "beta_schedule": "constant", "epsilon_schedule": "constant"}, "checkpoint_interval": 500000, "network_settings": {"normalize": false, "hidden_units": 512, "num_layers": 2, "vis_encode_type": "simple", "memory": null, "goal_conditioning_type": "hyper", "deterministic": false}, "reward_signals": {"extrinsic": {"gamma": 0.99, "strength": 1.0, "network_settings": {"normalize": false, "hidden_units": 128, "num_layers": 2, "vis_encode_type": "simple", "memory": null, "goal_conditioning_type": "hyper", "deterministic": false}}}, "init_path": null, "keep_checkpoints": 5, "even_checkpoints": false, "max_steps": 50000000, "time_horizon": 1000, "summary_freq": 10000, "threaded": false, "self_play": {"save_steps": 50000, "team_change": 200000, "swap_steps": 2000, "window": 10, "play_against_latest_model_ratio": 0.5, "initial_elo": 1200.0}, "behavioral_cloning": null}}, "env_settings": {"env_path": "./training-envs-executables/SoccerTwos/SoccerTwos.x86_64", "env_args": null, "base_port": 5005, "num_envs": 1, "num_areas": 1, "seed": -1, "max_lifetime_restarts": 10, "restarts_rate_limit_n": 1, "restarts_rate_limit_period_s": 60}, "engine_settings": {"width": 84, "height": 84, "quality_level": 5, "time_scale": 20, "target_frame_rate": -1, "capture_frame_rate": 60, "no_graphics": true}, "environment_parameters": null, "checkpoint_settings": {"run_id": "SoccerTwos", "initialize_from": null, "load_model": false, "resume": false, "force": true, "train_model": false, "inference": false, "results_dir": "results"}, "torch_settings": {"device": null}, "debug": false}
|
configuration.yaml
ADDED
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
default_settings: null
|
2 |
+
behaviors:
|
3 |
+
SoccerTwos:
|
4 |
+
trainer_type: poca
|
5 |
+
hyperparameters:
|
6 |
+
batch_size: 2048
|
7 |
+
buffer_size: 20480
|
8 |
+
learning_rate: 0.0003
|
9 |
+
beta: 0.005
|
10 |
+
epsilon: 0.2
|
11 |
+
lambd: 0.95
|
12 |
+
num_epoch: 3
|
13 |
+
learning_rate_schedule: constant
|
14 |
+
beta_schedule: constant
|
15 |
+
epsilon_schedule: constant
|
16 |
+
checkpoint_interval: 500000
|
17 |
+
network_settings:
|
18 |
+
normalize: false
|
19 |
+
hidden_units: 512
|
20 |
+
num_layers: 2
|
21 |
+
vis_encode_type: simple
|
22 |
+
memory: null
|
23 |
+
goal_conditioning_type: hyper
|
24 |
+
deterministic: false
|
25 |
+
reward_signals:
|
26 |
+
extrinsic:
|
27 |
+
gamma: 0.99
|
28 |
+
strength: 1.0
|
29 |
+
network_settings:
|
30 |
+
normalize: false
|
31 |
+
hidden_units: 128
|
32 |
+
num_layers: 2
|
33 |
+
vis_encode_type: simple
|
34 |
+
memory: null
|
35 |
+
goal_conditioning_type: hyper
|
36 |
+
deterministic: false
|
37 |
+
init_path: null
|
38 |
+
keep_checkpoints: 5
|
39 |
+
even_checkpoints: false
|
40 |
+
max_steps: 50000000
|
41 |
+
time_horizon: 1000
|
42 |
+
summary_freq: 10000
|
43 |
+
threaded: false
|
44 |
+
self_play:
|
45 |
+
save_steps: 50000
|
46 |
+
team_change: 200000
|
47 |
+
swap_steps: 2000
|
48 |
+
window: 10
|
49 |
+
play_against_latest_model_ratio: 0.5
|
50 |
+
initial_elo: 1200.0
|
51 |
+
behavioral_cloning: null
|
52 |
+
env_settings:
|
53 |
+
env_path: ./training-envs-executables/SoccerTwos/SoccerTwos.x86_64
|
54 |
+
env_args: null
|
55 |
+
base_port: 5005
|
56 |
+
num_envs: 1
|
57 |
+
num_areas: 1
|
58 |
+
seed: -1
|
59 |
+
max_lifetime_restarts: 10
|
60 |
+
restarts_rate_limit_n: 1
|
61 |
+
restarts_rate_limit_period_s: 60
|
62 |
+
engine_settings:
|
63 |
+
width: 84
|
64 |
+
height: 84
|
65 |
+
quality_level: 5
|
66 |
+
time_scale: 20
|
67 |
+
target_frame_rate: -1
|
68 |
+
capture_frame_rate: 60
|
69 |
+
no_graphics: true
|
70 |
+
environment_parameters: null
|
71 |
+
checkpoint_settings:
|
72 |
+
run_id: SoccerTwos
|
73 |
+
initialize_from: null
|
74 |
+
load_model: false
|
75 |
+
resume: false
|
76 |
+
force: true
|
77 |
+
train_model: false
|
78 |
+
inference: false
|
79 |
+
results_dir: results
|
80 |
+
torch_settings:
|
81 |
+
device: null
|
82 |
+
debug: false
|
run_logs/Player-0.log
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run_logs/timers.json
ADDED
@@ -0,0 +1,424 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"name": "root",
|
3 |
+
"gauges": {
|
4 |
+
"SoccerTwos.Policy.Entropy.mean": {
|
5 |
+
"value": 2.556478977203369,
|
6 |
+
"min": 0.8628996014595032,
|
7 |
+
"max": 3.295734405517578,
|
8 |
+
"count": 5000
|
9 |
+
},
|
10 |
+
"SoccerTwos.Policy.Entropy.sum": {
|
11 |
+
"value": 58819.46875,
|
12 |
+
"min": 9834.373046875,
|
13 |
+
"max": 155859.171875,
|
14 |
+
"count": 5000
|
15 |
+
},
|
16 |
+
"SoccerTwos.Environment.EpisodeLength.mean": {
|
17 |
+
"value": 320.0625,
|
18 |
+
"min": 256.94444444444446,
|
19 |
+
"max": 999.0,
|
20 |
+
"count": 5000
|
21 |
+
},
|
22 |
+
"SoccerTwos.Environment.EpisodeLength.sum": {
|
23 |
+
"value": 20484.0,
|
24 |
+
"min": 15984.0,
|
25 |
+
"max": 25532.0,
|
26 |
+
"count": 5000
|
27 |
+
},
|
28 |
+
"SoccerTwos.Self-play.ELO.mean": {
|
29 |
+
"value": 1246.1294126353869,
|
30 |
+
"min": 1187.3139795928823,
|
31 |
+
"max": 1246.4191325003276,
|
32 |
+
"count": 444
|
33 |
+
},
|
34 |
+
"SoccerTwos.Self-play.ELO.sum": {
|
35 |
+
"value": 37383.88237906161,
|
36 |
+
"min": 2374.6279591857647,
|
37 |
+
"max": 37383.88237906161,
|
38 |
+
"count": 444
|
39 |
+
},
|
40 |
+
"SoccerTwos.Step.mean": {
|
41 |
+
"value": 49999430.0,
|
42 |
+
"min": 9784.0,
|
43 |
+
"max": 49999430.0,
|
44 |
+
"count": 5000
|
45 |
+
},
|
46 |
+
"SoccerTwos.Step.sum": {
|
47 |
+
"value": 49999430.0,
|
48 |
+
"min": 9784.0,
|
49 |
+
"max": 49999430.0,
|
50 |
+
"count": 5000
|
51 |
+
},
|
52 |
+
"SoccerTwos.Policy.ExtrinsicBaselineEstimate.mean": {
|
53 |
+
"value": 0.012056336738169193,
|
54 |
+
"min": -0.10242440551519394,
|
55 |
+
"max": 0.042053915560245514,
|
56 |
+
"count": 5000
|
57 |
+
},
|
58 |
+
"SoccerTwos.Policy.ExtrinsicBaselineEstimate.sum": {
|
59 |
+
"value": 0.37374642491340637,
|
60 |
+
"min": -1.536352276802063,
|
61 |
+
"max": 1.0224354267120361,
|
62 |
+
"count": 5000
|
63 |
+
},
|
64 |
+
"SoccerTwos.Policy.ExtrinsicValueEstimate.mean": {
|
65 |
+
"value": 0.012648879550397396,
|
66 |
+
"min": -0.10245412588119507,
|
67 |
+
"max": 0.04426921531558037,
|
68 |
+
"count": 5000
|
69 |
+
},
|
70 |
+
"SoccerTwos.Policy.ExtrinsicValueEstimate.sum": {
|
71 |
+
"value": 0.3921152651309967,
|
72 |
+
"min": -1.5367718935012817,
|
73 |
+
"max": 1.1084165573120117,
|
74 |
+
"count": 5000
|
75 |
+
},
|
76 |
+
"SoccerTwos.Environment.CumulativeReward.mean": {
|
77 |
+
"value": 0.0,
|
78 |
+
"min": 0.0,
|
79 |
+
"max": 0.0,
|
80 |
+
"count": 5000
|
81 |
+
},
|
82 |
+
"SoccerTwos.Environment.CumulativeReward.sum": {
|
83 |
+
"value": 0.0,
|
84 |
+
"min": 0.0,
|
85 |
+
"max": 0.0,
|
86 |
+
"count": 5000
|
87 |
+
},
|
88 |
+
"SoccerTwos.Policy.ExtrinsicReward.mean": {
|
89 |
+
"value": -0.1501677425638322,
|
90 |
+
"min": -0.7717600007851918,
|
91 |
+
"max": 0.43691111273235744,
|
92 |
+
"count": 5000
|
93 |
+
},
|
94 |
+
"SoccerTwos.Policy.ExtrinsicReward.sum": {
|
95 |
+
"value": -4.655200019478798,
|
96 |
+
"min": -16.417999982833862,
|
97 |
+
"max": 8.89359986782074,
|
98 |
+
"count": 5000
|
99 |
+
},
|
100 |
+
"SoccerTwos.Environment.GroupCumulativeReward.mean": {
|
101 |
+
"value": -0.1501677425638322,
|
102 |
+
"min": -0.7717600007851918,
|
103 |
+
"max": 0.43691111273235744,
|
104 |
+
"count": 5000
|
105 |
+
},
|
106 |
+
"SoccerTwos.Environment.GroupCumulativeReward.sum": {
|
107 |
+
"value": -4.655200019478798,
|
108 |
+
"min": -16.417999982833862,
|
109 |
+
"max": 8.89359986782074,
|
110 |
+
"count": 5000
|
111 |
+
},
|
112 |
+
"SoccerTwos.IsTraining.mean": {
|
113 |
+
"value": 1.0,
|
114 |
+
"min": 1.0,
|
115 |
+
"max": 1.0,
|
116 |
+
"count": 5000
|
117 |
+
},
|
118 |
+
"SoccerTwos.IsTraining.sum": {
|
119 |
+
"value": 1.0,
|
120 |
+
"min": 1.0,
|
121 |
+
"max": 1.0,
|
122 |
+
"count": 5000
|
123 |
+
},
|
124 |
+
"SoccerTwos.Losses.PolicyLoss.mean": {
|
125 |
+
"value": 0.01427030615353336,
|
126 |
+
"min": 0.009354956900763985,
|
127 |
+
"max": 0.025422956169738123,
|
128 |
+
"count": 2286
|
129 |
+
},
|
130 |
+
"SoccerTwos.Losses.PolicyLoss.sum": {
|
131 |
+
"value": 0.01427030615353336,
|
132 |
+
"min": 0.009354956900763985,
|
133 |
+
"max": 0.025422956169738123,
|
134 |
+
"count": 2286
|
135 |
+
},
|
136 |
+
"SoccerTwos.Losses.ValueLoss.mean": {
|
137 |
+
"value": 0.009385776737083991,
|
138 |
+
"min": 2.688042489695323e-12,
|
139 |
+
"max": 0.02662869206542382,
|
140 |
+
"count": 2286
|
141 |
+
},
|
142 |
+
"SoccerTwos.Losses.ValueLoss.sum": {
|
143 |
+
"value": 0.009385776737083991,
|
144 |
+
"min": 2.688042489695323e-12,
|
145 |
+
"max": 0.02662869206542382,
|
146 |
+
"count": 2286
|
147 |
+
},
|
148 |
+
"SoccerTwos.Losses.BaselineLoss.mean": {
|
149 |
+
"value": 0.00949315377511084,
|
150 |
+
"min": 4.768202434453361e-12,
|
151 |
+
"max": 0.027282530008233153,
|
152 |
+
"count": 2286
|
153 |
+
},
|
154 |
+
"SoccerTwos.Losses.BaselineLoss.sum": {
|
155 |
+
"value": 0.00949315377511084,
|
156 |
+
"min": 4.768202434453361e-12,
|
157 |
+
"max": 0.027282530008233153,
|
158 |
+
"count": 2286
|
159 |
+
},
|
160 |
+
"SoccerTwos.Policy.LearningRate.mean": {
|
161 |
+
"value": 0.0003,
|
162 |
+
"min": 0.0003,
|
163 |
+
"max": 0.0003,
|
164 |
+
"count": 2286
|
165 |
+
},
|
166 |
+
"SoccerTwos.Policy.LearningRate.sum": {
|
167 |
+
"value": 0.0003,
|
168 |
+
"min": 0.0003,
|
169 |
+
"max": 0.0003,
|
170 |
+
"count": 2286
|
171 |
+
},
|
172 |
+
"SoccerTwos.Policy.Epsilon.mean": {
|
173 |
+
"value": 0.20000000000000007,
|
174 |
+
"min": 0.20000000000000007,
|
175 |
+
"max": 0.20000000000000007,
|
176 |
+
"count": 2286
|
177 |
+
},
|
178 |
+
"SoccerTwos.Policy.Epsilon.sum": {
|
179 |
+
"value": 0.20000000000000007,
|
180 |
+
"min": 0.20000000000000007,
|
181 |
+
"max": 0.20000000000000007,
|
182 |
+
"count": 2286
|
183 |
+
},
|
184 |
+
"SoccerTwos.Policy.Beta.mean": {
|
185 |
+
"value": 0.005000000000000001,
|
186 |
+
"min": 0.005000000000000001,
|
187 |
+
"max": 0.005000000000000001,
|
188 |
+
"count": 2286
|
189 |
+
},
|
190 |
+
"SoccerTwos.Policy.Beta.sum": {
|
191 |
+
"value": 0.005000000000000001,
|
192 |
+
"min": 0.005000000000000001,
|
193 |
+
"max": 0.005000000000000001,
|
194 |
+
"count": 2286
|
195 |
+
}
|
196 |
+
},
|
197 |
+
"metadata": {
|
198 |
+
"timer_format_version": "0.1.0",
|
199 |
+
"start_time_seconds": "1685525223",
|
200 |
+
"python_version": "3.9.16 (main, Mar 8 2023, 14:00:05) \n[GCC 11.2.0]",
|
201 |
+
"command_line_arguments": "/home/wyt/anaconda3/envs/rl/bin/mlagents-learn ./config/poca/SoccerTwos.yaml --env=./training-envs-executables/SoccerTwos/SoccerTwos.x86_64 --run-id=SoccerTwos --no-graphics --force",
|
202 |
+
"mlagents_version": "0.31.0.dev0",
|
203 |
+
"mlagents_envs_version": "0.31.0.dev0",
|
204 |
+
"communication_protocol_version": "1.5.0",
|
205 |
+
"pytorch_version": "1.11.0+cu102",
|
206 |
+
"numpy_version": "1.21.2",
|
207 |
+
"end_time_seconds": "1685584266"
|
208 |
+
},
|
209 |
+
"total": 59042.89769419585,
|
210 |
+
"count": 1,
|
211 |
+
"self": 0.21814238373190165,
|
212 |
+
"children": {
|
213 |
+
"run_training.setup": {
|
214 |
+
"total": 0.0063062182161957026,
|
215 |
+
"count": 1,
|
216 |
+
"self": 0.0063062182161957026
|
217 |
+
},
|
218 |
+
"TrainerController.start_learning": {
|
219 |
+
"total": 59042.673245593905,
|
220 |
+
"count": 1,
|
221 |
+
"self": 54.184389336034656,
|
222 |
+
"children": {
|
223 |
+
"TrainerController._reset_env": {
|
224 |
+
"total": 18.66194161004387,
|
225 |
+
"count": 249,
|
226 |
+
"self": 18.66194161004387
|
227 |
+
},
|
228 |
+
"TrainerController.advance": {
|
229 |
+
"total": 58969.67677217396,
|
230 |
+
"count": 3249605,
|
231 |
+
"self": 52.1698747985065,
|
232 |
+
"children": {
|
233 |
+
"env_step": {
|
234 |
+
"total": 44698.29575628159,
|
235 |
+
"count": 3249605,
|
236 |
+
"self": 28578.731281354558,
|
237 |
+
"children": {
|
238 |
+
"SubprocessEnvManager._take_step": {
|
239 |
+
"total": 16084.75626065908,
|
240 |
+
"count": 3249605,
|
241 |
+
"self": 298.7151764656883,
|
242 |
+
"children": {
|
243 |
+
"TorchPolicy.evaluate": {
|
244 |
+
"total": 15786.041084193392,
|
245 |
+
"count": 6463122,
|
246 |
+
"self": 15786.041084193392
|
247 |
+
}
|
248 |
+
}
|
249 |
+
},
|
250 |
+
"workers": {
|
251 |
+
"total": 34.80821426794864,
|
252 |
+
"count": 3249605,
|
253 |
+
"self": 0.0,
|
254 |
+
"children": {
|
255 |
+
"worker_root": {
|
256 |
+
"total": 58964.13316858723,
|
257 |
+
"count": 3249605,
|
258 |
+
"is_parallel": true,
|
259 |
+
"self": 36394.802758344915,
|
260 |
+
"children": {
|
261 |
+
"run_training.setup": {
|
262 |
+
"total": 0.0,
|
263 |
+
"count": 0,
|
264 |
+
"is_parallel": true,
|
265 |
+
"self": 0.0,
|
266 |
+
"children": {
|
267 |
+
"steps_from_proto": {
|
268 |
+
"total": 0.0016105701215565205,
|
269 |
+
"count": 2,
|
270 |
+
"is_parallel": true,
|
271 |
+
"self": 0.00037895189598202705,
|
272 |
+
"children": {
|
273 |
+
"_process_rank_one_or_two_observation": {
|
274 |
+
"total": 0.0012316182255744934,
|
275 |
+
"count": 8,
|
276 |
+
"is_parallel": true,
|
277 |
+
"self": 0.0012316182255744934
|
278 |
+
}
|
279 |
+
}
|
280 |
+
},
|
281 |
+
"UnityEnvironment.step": {
|
282 |
+
"total": 0.016228169901296496,
|
283 |
+
"count": 1,
|
284 |
+
"is_parallel": true,
|
285 |
+
"self": 0.00037772185169160366,
|
286 |
+
"children": {
|
287 |
+
"UnityEnvironment._generate_step_input": {
|
288 |
+
"total": 0.00032189092598855495,
|
289 |
+
"count": 1,
|
290 |
+
"is_parallel": true,
|
291 |
+
"self": 0.00032189092598855495
|
292 |
+
},
|
293 |
+
"communicator.exchange": {
|
294 |
+
"total": 0.01430563093163073,
|
295 |
+
"count": 1,
|
296 |
+
"is_parallel": true,
|
297 |
+
"self": 0.01430563093163073
|
298 |
+
},
|
299 |
+
"steps_from_proto": {
|
300 |
+
"total": 0.0012229261919856071,
|
301 |
+
"count": 2,
|
302 |
+
"is_parallel": true,
|
303 |
+
"self": 0.0002754807937890291,
|
304 |
+
"children": {
|
305 |
+
"_process_rank_one_or_two_observation": {
|
306 |
+
"total": 0.000947445398196578,
|
307 |
+
"count": 8,
|
308 |
+
"is_parallel": true,
|
309 |
+
"self": 0.000947445398196578
|
310 |
+
}
|
311 |
+
}
|
312 |
+
}
|
313 |
+
}
|
314 |
+
}
|
315 |
+
}
|
316 |
+
},
|
317 |
+
"UnityEnvironment.step": {
|
318 |
+
"total": 22569.036006505834,
|
319 |
+
"count": 3249604,
|
320 |
+
"is_parallel": true,
|
321 |
+
"self": 1224.9911103511695,
|
322 |
+
"children": {
|
323 |
+
"UnityEnvironment._generate_step_input": {
|
324 |
+
"total": 875.6204554305878,
|
325 |
+
"count": 3249604,
|
326 |
+
"is_parallel": true,
|
327 |
+
"self": 875.6204554305878
|
328 |
+
},
|
329 |
+
"communicator.exchange": {
|
330 |
+
"total": 16670.457968858536,
|
331 |
+
"count": 3249604,
|
332 |
+
"is_parallel": true,
|
333 |
+
"self": 16670.457968858536
|
334 |
+
},
|
335 |
+
"steps_from_proto": {
|
336 |
+
"total": 3797.9664718655404,
|
337 |
+
"count": 6499208,
|
338 |
+
"is_parallel": true,
|
339 |
+
"self": 776.744841856882,
|
340 |
+
"children": {
|
341 |
+
"_process_rank_one_or_two_observation": {
|
342 |
+
"total": 3021.2216300086584,
|
343 |
+
"count": 25996832,
|
344 |
+
"is_parallel": true,
|
345 |
+
"self": 3021.2216300086584
|
346 |
+
}
|
347 |
+
}
|
348 |
+
}
|
349 |
+
}
|
350 |
+
},
|
351 |
+
"steps_from_proto": {
|
352 |
+
"total": 0.2944037364795804,
|
353 |
+
"count": 496,
|
354 |
+
"is_parallel": true,
|
355 |
+
"self": 0.0608798258472234,
|
356 |
+
"children": {
|
357 |
+
"_process_rank_one_or_two_observation": {
|
358 |
+
"total": 0.233523910632357,
|
359 |
+
"count": 1984,
|
360 |
+
"is_parallel": true,
|
361 |
+
"self": 0.233523910632357
|
362 |
+
}
|
363 |
+
}
|
364 |
+
}
|
365 |
+
}
|
366 |
+
}
|
367 |
+
}
|
368 |
+
}
|
369 |
+
}
|
370 |
+
},
|
371 |
+
"trainer_advance": {
|
372 |
+
"total": 14219.211141093867,
|
373 |
+
"count": 3249605,
|
374 |
+
"self": 439.10235753725283,
|
375 |
+
"children": {
|
376 |
+
"process_trajectory": {
|
377 |
+
"total": 5498.635246976977,
|
378 |
+
"count": 3249605,
|
379 |
+
"self": 5482.912984529277,
|
380 |
+
"children": {
|
381 |
+
"RLTrainer._checkpoint": {
|
382 |
+
"total": 15.722262447699904,
|
383 |
+
"count": 100,
|
384 |
+
"self": 15.722262447699904
|
385 |
+
}
|
386 |
+
}
|
387 |
+
},
|
388 |
+
"_update_policy": {
|
389 |
+
"total": 8281.473536579637,
|
390 |
+
"count": 2286,
|
391 |
+
"self": 5230.548123164568,
|
392 |
+
"children": {
|
393 |
+
"TorchPOCAOptimizer.update": {
|
394 |
+
"total": 3050.9254134150688,
|
395 |
+
"count": 68580,
|
396 |
+
"self": 3050.9254134150688
|
397 |
+
}
|
398 |
+
}
|
399 |
+
}
|
400 |
+
}
|
401 |
+
}
|
402 |
+
}
|
403 |
+
},
|
404 |
+
"trainer_threads": {
|
405 |
+
"total": 5.359761416912079e-07,
|
406 |
+
"count": 1,
|
407 |
+
"self": 5.359761416912079e-07
|
408 |
+
},
|
409 |
+
"TrainerController._save_models": {
|
410 |
+
"total": 0.15014193789102137,
|
411 |
+
"count": 1,
|
412 |
+
"self": 0.000920719001442194,
|
413 |
+
"children": {
|
414 |
+
"RLTrainer._checkpoint": {
|
415 |
+
"total": 0.14922121888957918,
|
416 |
+
"count": 1,
|
417 |
+
"self": 0.14922121888957918
|
418 |
+
}
|
419 |
+
}
|
420 |
+
}
|
421 |
+
}
|
422 |
+
}
|
423 |
+
}
|
424 |
+
}
|
run_logs/training_status.json
ADDED
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"SoccerTwos": {
|
3 |
+
"checkpoints": [
|
4 |
+
{
|
5 |
+
"steps": 48499680,
|
6 |
+
"file_path": "results/SoccerTwos/SoccerTwos/SoccerTwos-48499680.onnx",
|
7 |
+
"reward": 0.0,
|
8 |
+
"creation_time": 1685582717.6973877,
|
9 |
+
"auxillary_file_paths": [
|
10 |
+
"results/SoccerTwos/SoccerTwos/SoccerTwos-48499680.pt"
|
11 |
+
]
|
12 |
+
},
|
13 |
+
{
|
14 |
+
"steps": 48999460,
|
15 |
+
"file_path": "results/SoccerTwos/SoccerTwos/SoccerTwos-48999460.onnx",
|
16 |
+
"reward": null,
|
17 |
+
"creation_time": 1685583235.9581857,
|
18 |
+
"auxillary_file_paths": [
|
19 |
+
"results/SoccerTwos/SoccerTwos/SoccerTwos-48999460.pt"
|
20 |
+
]
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"steps": 49499964,
|
24 |
+
"file_path": "results/SoccerTwos/SoccerTwos/SoccerTwos-49499964.onnx",
|
25 |
+
"reward": 0.0,
|
26 |
+
"creation_time": 1685583750.5695598,
|
27 |
+
"auxillary_file_paths": [
|
28 |
+
"results/SoccerTwos/SoccerTwos/SoccerTwos-49499964.pt"
|
29 |
+
]
|
30 |
+
},
|
31 |
+
{
|
32 |
+
"steps": 49999430,
|
33 |
+
"file_path": "results/SoccerTwos/SoccerTwos/SoccerTwos-49999430.onnx",
|
34 |
+
"reward": 0.0,
|
35 |
+
"creation_time": 1685584265.6484303,
|
36 |
+
"auxillary_file_paths": [
|
37 |
+
"results/SoccerTwos/SoccerTwos/SoccerTwos-49999430.pt"
|
38 |
+
]
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"steps": 50000890,
|
42 |
+
"file_path": "results/SoccerTwos/SoccerTwos/SoccerTwos-50000890.onnx",
|
43 |
+
"reward": 0.0,
|
44 |
+
"creation_time": 1685584265.870237,
|
45 |
+
"auxillary_file_paths": [
|
46 |
+
"results/SoccerTwos/SoccerTwos/SoccerTwos-50000890.pt"
|
47 |
+
]
|
48 |
+
}
|
49 |
+
],
|
50 |
+
"elo": 1244.212996620815,
|
51 |
+
"final_checkpoint": {
|
52 |
+
"steps": 50000890,
|
53 |
+
"file_path": "results/SoccerTwos/SoccerTwos.onnx",
|
54 |
+
"reward": 0.0,
|
55 |
+
"creation_time": 1685584265.870237,
|
56 |
+
"auxillary_file_paths": [
|
57 |
+
"results/SoccerTwos/SoccerTwos/SoccerTwos-50000890.pt"
|
58 |
+
]
|
59 |
+
}
|
60 |
+
},
|
61 |
+
"metadata": {
|
62 |
+
"stats_format_version": "0.3.0",
|
63 |
+
"mlagents_version": "0.31.0.dev0",
|
64 |
+
"torch_version": "1.11.0+cu102"
|
65 |
+
}
|
66 |
+
}
|