Hugging Face's logo

infinitejoy
/
MLAgents-3DBall

Reinforcement Learning
ml-agents
TensorBoard
ONNX
unity-ml-agents
deep-reinforcement-learning
ML-Agents-3DBall
Model card Files Metrics Community
infinitejoy commited on
Commit
44ef782
·
1 Parent(s): 6b219ea

default parameters training

Browse files
Files changed (13) hide show
  1. 3DBall.onnx +3 -0
  2. 3DBall/3DBall-499224.onnx +3 -0
  3. 3DBall/3DBall-499224.pt +3 -0
  4. 3DBall/3DBall-500224.onnx +3 -0
  5. 3DBall/3DBall-500224.pt +3 -0
  6. 3DBall/checkpoint.pt +3 -0
  7. 3DBall/events.out.tfevents.1657904948.Joydeeps-MacBook-Pro.local.38585.0 +3 -0
  8. README.md +31 -0
  9. config.json +1 -0
  10. configuration.yaml +75 -0
  11. run_logs/Player-0.log +0 -0
  12. run_logs/timers.json +327 -0
  13. run_logs/training_status.json +38 -0
3DBall.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26821bacfaecbef69d4ea324c402bffe9b20306a1671fe0695c84305e8d7d1fc
3
+ size 73999
3DBall/3DBall-499224.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebf09630b2287d5f3541a1a9bb0573e677cb1cdc63b89373543c892729eb23d6
3
+ size 73999
3DBall/3DBall-499224.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35ad84cb60e02b8724b54cf996fdbe369f7cdf27c71d82124c2e51b8d4633db4
3
+ size 448183
3DBall/3DBall-500224.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26821bacfaecbef69d4ea324c402bffe9b20306a1671fe0695c84305e8d7d1fc
3
+ size 73999
3DBall/3DBall-500224.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15b8aaf900c9707f31fbaf4b4b2702af8a753af3be177e313e4e9949c134472f
3
+ size 448183
3DBall/checkpoint.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15b8aaf900c9707f31fbaf4b4b2702af8a753af3be177e313e4e9949c134472f
3
+ size 448183
3DBall/events.out.tfevents.1657904948.Joydeeps-MacBook-Pro.local.38585.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12f63d9d40c335986440e18d4c68f1f62518cc90a6627cc23fd1e24a614fcdfb
3
+ size 42851
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-3DBall
9
+ library_name: ml-agents
10
+ ---
11
+
12
+ # **ppo** Agent playing **3DBall**
13
+ This is a trained model of a **ppo** agent playing **3DBall** 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-3DBall
28
+ 2. Step 1: Write your model_id: infinitejoy/MLAgents-3DBall
29
+ 3. Step 2: Select your *.nn /*.onnx file
30
+ 4. Click on Watch the agent play 👀
31
+
config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"default_settings": null, "behaviors": {"3DBall": {"trainer_type": "ppo", "hyperparameters": {"batch_size": 64, "buffer_size": 12000, "learning_rate": 0.0003, "beta": 0.001, "epsilon": 0.2, "lambd": 0.99, "num_epoch": 3, "learning_rate_schedule": "linear", "beta_schedule": "linear", "epsilon_schedule": "linear"}, "network_settings": {"normalize": true, "hidden_units": 128, "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, "checkpoint_interval": 500000, "max_steps": 500000, "time_horizon": 1000, "summary_freq": 12000, "threaded": false, "self_play": null, "behavioral_cloning": null}}, "env_settings": {"env_path": "./trained-envs-executables/macos/3d_ball/3d ball", "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": "first3DBallRun3", "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 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ default_settings: null
2
+ behaviors:
3
+ 3DBall:
4
+ trainer_type: ppo
5
+ hyperparameters:
6
+ batch_size: 64
7
+ buffer_size: 12000
8
+ learning_rate: 0.0003
9
+ beta: 0.001
10
+ epsilon: 0.2
11
+ lambd: 0.99
12
+ num_epoch: 3
13
+ learning_rate_schedule: linear
14
+ beta_schedule: linear
15
+ epsilon_schedule: linear
16
+ network_settings:
17
+ normalize: true
18
+ hidden_units: 128
19
+ num_layers: 2
20
+ vis_encode_type: simple
21
+ memory: null
22
+ goal_conditioning_type: hyper
23
+ deterministic: false
24
+ reward_signals:
25
+ extrinsic:
26
+ gamma: 0.99
27
+ strength: 1.0
28
+ network_settings:
29
+ normalize: false
30
+ hidden_units: 128
31
+ num_layers: 2
32
+ vis_encode_type: simple
33
+ memory: null
34
+ goal_conditioning_type: hyper
35
+ deterministic: false
36
+ init_path: null
37
+ keep_checkpoints: 5
38
+ checkpoint_interval: 500000
39
+ max_steps: 500000
40
+ time_horizon: 1000
41
+ summary_freq: 12000
42
+ threaded: false
43
+ self_play: null
44
+ behavioral_cloning: null
45
+ env_settings:
46
+ env_path: ./trained-envs-executables/macos/3d_ball/3d ball
47
+ env_args: null
48
+ base_port: 5005
49
+ num_envs: 1
50
+ num_areas: 1
51
+ seed: -1
52
+ max_lifetime_restarts: 10
53
+ restarts_rate_limit_n: 1
54
+ restarts_rate_limit_period_s: 60
55
+ engine_settings:
56
+ width: 84
57
+ height: 84
58
+ quality_level: 5
59
+ time_scale: 20
60
+ target_frame_rate: -1
61
+ capture_frame_rate: 60
62
+ no_graphics: true
63
+ environment_parameters: null
64
+ checkpoint_settings:
65
+ run_id: first3DBallRun3
66
+ initialize_from: null
67
+ load_model: false
68
+ resume: false
69
+ force: true
70
+ train_model: false
71
+ inference: false
72
+ results_dir: results
73
+ torch_settings:
74
+ device: null
75
+ 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 +1,327 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "root",
3
+ "gauges": {
4
+ "3DBall.Policy.Entropy.mean": {
5
+ "value": 1.2530262470245361,
6
+ "min": 1.249975562095642,
7
+ "max": 1.4189382791519165,
8
+ "count": 41
9
+ },
10
+ "3DBall.Policy.Entropy.sum": {
11
+ "value": 15036.3154296875,
12
+ "min": 14999.70703125,
13
+ "max": 19505.73046875,
14
+ "count": 41
15
+ },
16
+ "3DBall.Environment.EpisodeLength.mean": {
17
+ "value": 999.0,
18
+ "min": 22.092130518234164,
19
+ "max": 999.0,
20
+ "count": 41
21
+ },
22
+ "3DBall.Environment.EpisodeLength.sum": {
23
+ "value": 11988.0,
24
+ "min": 11422.0,
25
+ "max": 12761.0,
26
+ "count": 41
27
+ },
28
+ "3DBall.Step.mean": {
29
+ "value": 491224.0,
30
+ "min": 11989.0,
31
+ "max": 491224.0,
32
+ "count": 41
33
+ },
34
+ "3DBall.Step.sum": {
35
+ "value": 491224.0,
36
+ "min": 11989.0,
37
+ "max": 491224.0,
38
+ "count": 41
39
+ },
40
+ "3DBall.Policy.ExtrinsicValueEstimate.mean": {
41
+ "value": 9.998326301574707,
42
+ "min": -0.21952161192893982,
43
+ "max": 10.004788398742676,
44
+ "count": 41
45
+ },
46
+ "3DBall.Policy.ExtrinsicValueEstimate.sum": {
47
+ "value": 119.97991180419922,
48
+ "min": -114.15123748779297,
49
+ "max": 221.3089599609375,
50
+ "count": 41
51
+ },
52
+ "3DBall.Environment.CumulativeReward.mean": {
53
+ "value": 100.00001525878906,
54
+ "min": 1.205576880161579,
55
+ "max": 100.00001525878906,
56
+ "count": 41
57
+ },
58
+ "3DBall.Environment.CumulativeReward.sum": {
59
+ "value": 1200.0001831054688,
60
+ "min": 626.899977684021,
61
+ "max": 1243.4001913070679,
62
+ "count": 41
63
+ },
64
+ "3DBall.Policy.ExtrinsicReward.mean": {
65
+ "value": 100.00001525878906,
66
+ "min": 1.205576880161579,
67
+ "max": 100.00001525878906,
68
+ "count": 41
69
+ },
70
+ "3DBall.Policy.ExtrinsicReward.sum": {
71
+ "value": 1200.0001831054688,
72
+ "min": 626.899977684021,
73
+ "max": 1243.4001913070679,
74
+ "count": 41
75
+ },
76
+ "3DBall.IsTraining.mean": {
77
+ "value": 1.0,
78
+ "min": 1.0,
79
+ "max": 1.0,
80
+ "count": 41
81
+ },
82
+ "3DBall.IsTraining.sum": {
83
+ "value": 1.0,
84
+ "min": 1.0,
85
+ "max": 1.0,
86
+ "count": 41
87
+ },
88
+ "3DBall.Losses.PolicyLoss.mean": {
89
+ "value": 0.0875553363431031,
90
+ "min": 0.0853438666777715,
91
+ "max": 0.10285655303845986,
92
+ "count": 38
93
+ },
94
+ "3DBall.Losses.PolicyLoss.sum": {
95
+ "value": 0.0875553363431031,
96
+ "min": 0.0853438666777715,
97
+ "max": 0.10285655303845986,
98
+ "count": 38
99
+ },
100
+ "3DBall.Losses.ValueLoss.mean": {
101
+ "value": 0.003986369480576965,
102
+ "min": 0.0011322325467241375,
103
+ "max": 3.11412584240732,
104
+ "count": 38
105
+ },
106
+ "3DBall.Losses.ValueLoss.sum": {
107
+ "value": 0.003986369480576965,
108
+ "min": 0.0011322325467241375,
109
+ "max": 3.11412584240732,
110
+ "count": 38
111
+ },
112
+ "3DBall.Policy.LearningRate.mean": {
113
+ "value": 1.006569664480001e-05,
114
+ "min": 1.006569664480001e-05,
115
+ "max": 0.0002927814024062,
116
+ "count": 38
117
+ },
118
+ "3DBall.Policy.LearningRate.sum": {
119
+ "value": 1.006569664480001e-05,
120
+ "min": 1.006569664480001e-05,
121
+ "max": 0.0002927814024062,
122
+ "count": 38
123
+ },
124
+ "3DBall.Policy.Epsilon.mean": {
125
+ "value": 0.10335519999999998,
126
+ "min": 0.10335519999999998,
127
+ "max": 0.19759379999999996,
128
+ "count": 38
129
+ },
130
+ "3DBall.Policy.Epsilon.sum": {
131
+ "value": 0.10335519999999998,
132
+ "min": 0.10335519999999998,
133
+ "max": 0.19759379999999996,
134
+ "count": 38
135
+ },
136
+ "3DBall.Policy.Beta.mean": {
137
+ "value": 4.321648000000003e-05,
138
+ "min": 4.321648000000003e-05,
139
+ "max": 0.0009761786200000003,
140
+ "count": 38
141
+ },
142
+ "3DBall.Policy.Beta.sum": {
143
+ "value": 4.321648000000003e-05,
144
+ "min": 4.321648000000003e-05,
145
+ "max": 0.0009761786200000003,
146
+ "count": 38
147
+ }
148
+ },
149
+ "metadata": {
150
+ "timer_format_version": "0.1.0",
151
+ "start_time_seconds": "1657904946",
152
+ "python_version": "3.8.13 (default, Mar 28 2022, 06:16:26) \n[Clang 12.0.0 ]",
153
+ "command_line_arguments": "/usr/local/Caskroom/miniconda/base/envs/ml-agents/bin/mlagents-learn ./config/ppo/3DBall.yaml --run-id=first3DBallRun3 --force --env=./trained-envs-executables/macos/3d_ball/3d ball --no-graphics",
154
+ "mlagents_version": "0.29.0.dev0",
155
+ "mlagents_envs_version": "0.29.0.dev0",
156
+ "communication_protocol_version": "1.5.0",
157
+ "pytorch_version": "1.8.1",
158
+ "numpy_version": "1.22.3",
159
+ "end_time_seconds": "1657905343"
160
+ },
161
+ "total": 397.450924047,
162
+ "count": 1,
163
+ "self": 0.34027323800000886,
164
+ "children": {
165
+ "run_training.setup": {
166
+ "total": 0.04273055499999989,
167
+ "count": 1,
168
+ "self": 0.04273055499999989
169
+ },
170
+ "TrainerController.start_learning": {
171
+ "total": 397.067920254,
172
+ "count": 1,
173
+ "self": 0.7700594520004529,
174
+ "children": {
175
+ "TrainerController._reset_env": {
176
+ "total": 2.5437515289999997,
177
+ "count": 1,
178
+ "self": 2.5437515289999997
179
+ },
180
+ "TrainerController.advance": {
181
+ "total": 393.69847658399954,
182
+ "count": 43985,
183
+ "self": 0.7830804549955701,
184
+ "children": {
185
+ "env_step": {
186
+ "total": 230.3137608409983,
187
+ "count": 43985,
188
+ "self": 202.2693348190025,
189
+ "children": {
190
+ "SubprocessEnvManager._take_step": {
191
+ "total": 27.511717070997804,
192
+ "count": 43985,
193
+ "self": 2.4751557829968895,
194
+ "children": {
195
+ "TorchPolicy.evaluate": {
196
+ "total": 25.036561288000915,
197
+ "count": 42155,
198
+ "self": 4.847578280001574,
199
+ "children": {
200
+ "TorchPolicy.sample_actions": {
201
+ "total": 20.18898300799934,
202
+ "count": 42155,
203
+ "self": 20.18898300799934
204
+ }
205
+ }
206
+ }
207
+ }
208
+ },
209
+ "workers": {
210
+ "total": 0.532708950998015,
211
+ "count": 43985,
212
+ "self": 0.0,
213
+ "children": {
214
+ "worker_root": {
215
+ "total": 393.1404451730054,
216
+ "count": 43985,
217
+ "is_parallel": true,
218
+ "self": 237.20454446299738,
219
+ "children": {
220
+ "steps_from_proto": {
221
+ "total": 0.000473571000000117,
222
+ "count": 1,
223
+ "is_parallel": true,
224
+ "self": 0.00017897400000022046,
225
+ "children": {
226
+ "_process_rank_one_or_two_observation": {
227
+ "total": 0.0002945969999998965,
228
+ "count": 2,
229
+ "is_parallel": true,
230
+ "self": 0.0002945969999998965
231
+ }
232
+ }
233
+ },
234
+ "UnityEnvironment.step": {
235
+ "total": 155.93542713900803,
236
+ "count": 43985,
237
+ "is_parallel": true,
238
+ "self": 5.6048235150036305,
239
+ "children": {
240
+ "UnityEnvironment._generate_step_input": {
241
+ "total": 7.386330972004274,
242
+ "count": 43985,
243
+ "is_parallel": true,
244
+ "self": 7.386330972004274
245
+ },
246
+ "communicator.exchange": {
247
+ "total": 132.06233277000166,
248
+ "count": 43985,
249
+ "is_parallel": true,
250
+ "self": 132.06233277000166
251
+ },
252
+ "steps_from_proto": {
253
+ "total": 10.88193988199847,
254
+ "count": 43985,
255
+ "is_parallel": true,
256
+ "self": 5.169313770000414,
257
+ "children": {
258
+ "_process_rank_one_or_two_observation": {
259
+ "total": 5.7126261119980555,
260
+ "count": 87970,
261
+ "is_parallel": true,
262
+ "self": 5.7126261119980555
263
+ }
264
+ }
265
+ }
266
+ }
267
+ }
268
+ }
269
+ }
270
+ }
271
+ }
272
+ }
273
+ },
274
+ "trainer_advance": {
275
+ "total": 162.60163528800567,
276
+ "count": 43985,
277
+ "self": 1.2461401930077045,
278
+ "children": {
279
+ "process_trajectory": {
280
+ "total": 22.86444937999798,
281
+ "count": 43985,
282
+ "self": 22.749369502998007,
283
+ "children": {
284
+ "RLTrainer._checkpoint": {
285
+ "total": 0.11507987699997102,
286
+ "count": 1,
287
+ "self": 0.11507987699997102
288
+ }
289
+ }
290
+ },
291
+ "_update_policy": {
292
+ "total": 138.49104571499998,
293
+ "count": 39,
294
+ "self": 67.82378533699823,
295
+ "children": {
296
+ "TorchPPOOptimizer.update": {
297
+ "total": 70.66726037800176,
298
+ "count": 23229,
299
+ "self": 70.66726037800176
300
+ }
301
+ }
302
+ }
303
+ }
304
+ }
305
+ }
306
+ },
307
+ "trainer_threads": {
308
+ "total": 1.0870000437535055e-06,
309
+ "count": 1,
310
+ "self": 1.0870000437535055e-06
311
+ },
312
+ "TrainerController._save_models": {
313
+ "total": 0.05563160199994854,
314
+ "count": 1,
315
+ "self": 0.0015699369999424562,
316
+ "children": {
317
+ "RLTrainer._checkpoint": {
318
+ "total": 0.05406166500000609,
319
+ "count": 1,
320
+ "self": 0.05406166500000609
321
+ }
322
+ }
323
+ }
324
+ }
325
+ }
326
+ }
327
+ }
run_logs/training_status.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "3DBall": {
3
+ "checkpoints": [
4
+ {
5
+ "steps": 499224,
6
+ "file_path": "results/first3DBallRun3/3DBall/3DBall-499224.onnx",
7
+ "reward": 100.00001525878906,
8
+ "creation_time": 1657905343.210939,
9
+ "auxillary_file_paths": [
10
+ "results/first3DBallRun3/3DBall/3DBall-499224.pt"
11
+ ]
12
+ },
13
+ {
14
+ "steps": 500224,
15
+ "file_path": "results/first3DBallRun3/3DBall/3DBall-500224.onnx",
16
+ "reward": 100.00001525878906,
17
+ "creation_time": 1657905343.3034189,
18
+ "auxillary_file_paths": [
19
+ "results/first3DBallRun3/3DBall/3DBall-500224.pt"
20
+ ]
21
+ }
22
+ ],
23
+ "final_checkpoint": {
24
+ "steps": 500224,
25
+ "file_path": "results/first3DBallRun3/3DBall.onnx",
26
+ "reward": 100.00001525878906,
27
+ "creation_time": 1657905343.3034189,
28
+ "auxillary_file_paths": [
29
+ "results/first3DBallRun3/3DBall/3DBall-500224.pt"
30
+ ]
31
+ }
32
+ },
33
+ "metadata": {
34
+ "stats_format_version": "0.3.0",
35
+ "mlagents_version": "0.29.0.dev0",
36
+ "torch_version": "1.8.1"
37
+ }
38
+ }