ThomasSimonini HF staff commited on
Commit
795b970
1 Parent(s): fc88083
3DBall.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ffc771138f23cdeed5528c579dc751fc90846ee5c31ee53c3689f48c3ade8202
3
+ size 73994
3DBall/3DBall-499833.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:709b0779ae01ff144c354ffcf1eecbbd74f70109897b833df7bb7c50c4386ba9
3
+ size 73994
3DBall/3DBall-499833.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15cce544c58941c0f8b4efb0ef2461d6e3a8533827a79a9920be67e416bc9875
3
+ size 448783
3DBall/3DBall-500833.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ffc771138f23cdeed5528c579dc751fc90846ee5c31ee53c3689f48c3ade8202
3
+ size 73994
3DBall/3DBall-500833.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee3d088854ac00a7a00b17e7a36c10b54c96c131b8f44c4bcd0112f1c9c7790c
3
+ size 448783
3DBall/checkpoint.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee3d088854ac00a7a00b17e7a36c10b54c96c131b8f44c4bcd0112f1c9c7790c
3
+ size 448783
3DBall/events.out.tfevents.1650347797.Thomas.23216.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3dd0058c8fff03d34e731f68ef8442f6b9a66c22466ef631389bec6c8e17fae2
3
+ size 51004
README.md ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ ---
3
+ tags:
4
+ - unity-ml-agents
5
+ - ml-agents
6
+ - deep-reinforcement-learning
7
+ - reinforcement-learning
8
+ - 3DBall
9
+ ---
10
+
11
+ # **ppo** Agent playing **3DBall**
12
+ This is a trained model of a **ppo** agent playing **3DBall** using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents).
13
+
14
+ ## Usage (with ML-Agents)
15
+ TODO: Add your code
16
+
17
+ ## Configuration file
18
+ {'default_settings': None, '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': None, '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': None, 'goal_conditioning_type': 'hyper', 'deterministic': False}}}, 'init_path': None, 'keep_checkpoints': 5, 'checkpoint_interval': 500000, 'max_steps': 500000, 'time_horizon': 1000, 'summary_freq': 12000, 'threaded': False, 'self_play': None, 'behavioral_cloning': None}}, 'env_settings': {'env_path': None, 'env_args': None, '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': False}, 'environment_parameters': None, 'checkpoint_settings': {'run_id': 'fullTestHF', 'initialize_from': None, 'load_model': False, 'resume': False, 'force': False, 'train_model': False, 'inference': False, 'results_dir': 'results'}, 'torch_settings': {'device': None}, 'debug': False}
19
+
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": null, "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": false}, "environment_parameters": null, "checkpoint_settings": {"run_id": "fullTestHF", "initialize_from": null, "load_model": false, "resume": false, "force": false, "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: null
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: false
63
+ environment_parameters: null
64
+ checkpoint_settings:
65
+ run_id: fullTestHF
66
+ initialize_from: null
67
+ load_model: false
68
+ resume: false
69
+ force: false
70
+ train_model: false
71
+ inference: false
72
+ results_dir: results
73
+ torch_settings:
74
+ device: null
75
+ debug: false
run_logs/timers.json ADDED
@@ -0,0 +1,327 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "root",
3
+ "gauges": {
4
+ "3DBall.Policy.Entropy.mean": {
5
+ "value": 1.2896696329116821,
6
+ "min": 1.2889900207519531,
7
+ "max": 1.4189382791519165,
8
+ "count": 41
9
+ },
10
+ "3DBall.Policy.Entropy.sum": {
11
+ "value": 16445.8671875,
12
+ "min": 12166.5849609375,
13
+ "max": 20346.564453125,
14
+ "count": 41
15
+ },
16
+ "3DBall.Environment.EpisodeLength.mean": {
17
+ "value": 999.0,
18
+ "min": 22.057581573896353,
19
+ "max": 999.0,
20
+ "count": 41
21
+ },
22
+ "3DBall.Environment.EpisodeLength.sum": {
23
+ "value": 11988.0,
24
+ "min": 11152.0,
25
+ "max": 12711.0,
26
+ "count": 41
27
+ },
28
+ "3DBall.Step.mean": {
29
+ "value": 491833.0,
30
+ "min": 11995.0,
31
+ "max": 491833.0,
32
+ "count": 41
33
+ },
34
+ "3DBall.Step.sum": {
35
+ "value": 491833.0,
36
+ "min": 11995.0,
37
+ "max": 491833.0,
38
+ "count": 41
39
+ },
40
+ "3DBall.Policy.ExtrinsicValueEstimate.mean": {
41
+ "value": 9.993976593017578,
42
+ "min": 0.2812666893005371,
43
+ "max": 10.00097370147705,
44
+ "count": 41
45
+ },
46
+ "3DBall.Policy.ExtrinsicValueEstimate.sum": {
47
+ "value": 119.92771911621094,
48
+ "min": 111.53549194335938,
49
+ "max": 243.1040802001953,
50
+ "count": 41
51
+ },
52
+ "3DBall.Environment.CumulativeReward.mean": {
53
+ "value": 100.00001525878906,
54
+ "min": 1.2067307359897173,
55
+ "max": 100.00001525878906,
56
+ "count": 41
57
+ },
58
+ "3DBall.Environment.CumulativeReward.sum": {
59
+ "value": 1200.0001831054688,
60
+ "min": 627.499982714653,
61
+ "max": 1269.3001952171326,
62
+ "count": 41
63
+ },
64
+ "3DBall.Policy.ExtrinsicReward.mean": {
65
+ "value": 100.00001525878906,
66
+ "min": 1.2067307359897173,
67
+ "max": 100.00001525878906,
68
+ "count": 41
69
+ },
70
+ "3DBall.Policy.ExtrinsicReward.sum": {
71
+ "value": 1200.0001831054688,
72
+ "min": 627.499982714653,
73
+ "max": 1269.3001952171326,
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": 0.09716191346342683,
90
+ "min": 0.07257706742175919,
91
+ "max": 0.10286889624501348,
92
+ "count": 39
93
+ },
94
+ "3DBall.Losses.PolicyLoss.sum": {
95
+ "value": 0.09716191346342683,
96
+ "min": 0.07257706742175919,
97
+ "max": 0.10286889624501348,
98
+ "count": 39
99
+ },
100
+ "3DBall.Losses.ValueLoss.mean": {
101
+ "value": 0.0016630031273597837,
102
+ "min": 0.0016630031273597837,
103
+ "max": 2.894371613886692,
104
+ "count": 39
105
+ },
106
+ "3DBall.Losses.ValueLoss.sum": {
107
+ "value": 0.0016630031273597837,
108
+ "min": 0.0016630031273597837,
109
+ "max": 2.894371613886692,
110
+ "count": 39
111
+ },
112
+ "3DBall.Policy.LearningRate.mean": {
113
+ "value": 4.900298366599988e-06,
114
+ "min": 4.900298366599988e-06,
115
+ "max": 0.0002927922024026,
116
+ "count": 39
117
+ },
118
+ "3DBall.Policy.LearningRate.sum": {
119
+ "value": 4.900298366599988e-06,
120
+ "min": 4.900298366599988e-06,
121
+ "max": 0.0002927922024026,
122
+ "count": 39
123
+ },
124
+ "3DBall.Policy.Epsilon.mean": {
125
+ "value": 0.10163340000000001,
126
+ "min": 0.10163340000000001,
127
+ "max": 0.19759739999999998,
128
+ "count": 39
129
+ },
130
+ "3DBall.Policy.Epsilon.sum": {
131
+ "value": 0.10163340000000001,
132
+ "min": 0.10163340000000001,
133
+ "max": 0.19759739999999998,
134
+ "count": 39
135
+ },
136
+ "3DBall.Policy.Beta.mean": {
137
+ "value": 2.617065999999996e-05,
138
+ "min": 2.617065999999996e-05,
139
+ "max": 0.00097621426,
140
+ "count": 39
141
+ },
142
+ "3DBall.Policy.Beta.sum": {
143
+ "value": 2.617065999999996e-05,
144
+ "min": 2.617065999999996e-05,
145
+ "max": 0.00097621426,
146
+ "count": 39
147
+ }
148
+ },
149
+ "metadata": {
150
+ "timer_format_version": "0.1.0",
151
+ "start_time_seconds": "1650347770",
152
+ "python_version": "3.8.13 (default, Mar 28 2022, 06:59:08) [MSC v.1916 64 bit (AMD64)]",
153
+ "command_line_arguments": "C:\\Users\\simon\\Anaconda3\\envs\\mlagents\\Scripts\\mlagents-learn config/ppo/3DBall.yaml --run-id=fullTestHF",
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": "1.7.1+cu110",
158
+ "numpy_version": "1.22.3",
159
+ "end_time_seconds": "1650348441"
160
+ },
161
+ "total": 670.4777134,
162
+ "count": 1,
163
+ "self": 0.011665499999821805,
164
+ "children": {
165
+ "run_training.setup": {
166
+ "total": 0.08510020000000007,
167
+ "count": 1,
168
+ "self": 0.08510020000000007
169
+ },
170
+ "TrainerController.start_learning": {
171
+ "total": 670.3809477000001,
172
+ "count": 1,
173
+ "self": 0.7038328999990426,
174
+ "children": {
175
+ "TrainerController._reset_env": {
176
+ "total": 29.2883102,
177
+ "count": 1,
178
+ "self": 29.2883102
179
+ },
180
+ "TrainerController.advance": {
181
+ "total": 640.3289582000012,
182
+ "count": 33602,
183
+ "self": 0.6747393000028978,
184
+ "children": {
185
+ "env_step": {
186
+ "total": 312.20899970000295,
187
+ "count": 33602,
188
+ "self": 227.34882499999514,
189
+ "children": {
190
+ "SubprocessEnvManager._take_step": {
191
+ "total": 84.34764660000867,
192
+ "count": 33602,
193
+ "self": 3.1455584000018604,
194
+ "children": {
195
+ "TorchPolicy.evaluate": {
196
+ "total": 81.20208820000681,
197
+ "count": 31801,
198
+ "self": 47.8277782000011,
199
+ "children": {
200
+ "TorchPolicy.sample_actions": {
201
+ "total": 33.374310000005714,
202
+ "count": 31801,
203
+ "self": 33.374310000005714
204
+ }
205
+ }
206
+ }
207
+ }
208
+ },
209
+ "workers": {
210
+ "total": 0.5125280999991482,
211
+ "count": 33602,
212
+ "self": 0.0,
213
+ "children": {
214
+ "worker_root": {
215
+ "total": 642.2971652999979,
216
+ "count": 33602,
217
+ "is_parallel": true,
218
+ "self": 465.0805939999972,
219
+ "children": {
220
+ "steps_from_proto": {
221
+ "total": 0.003397700000000725,
222
+ "count": 1,
223
+ "is_parallel": true,
224
+ "self": 0.00017640000000085365,
225
+ "children": {
226
+ "_process_rank_one_or_two_observation": {
227
+ "total": 0.0032212999999998715,
228
+ "count": 2,
229
+ "is_parallel": true,
230
+ "self": 0.0032212999999998715
231
+ }
232
+ }
233
+ },
234
+ "UnityEnvironment.step": {
235
+ "total": 177.21317360000077,
236
+ "count": 33602,
237
+ "is_parallel": true,
238
+ "self": 4.623666399989361,
239
+ "children": {
240
+ "UnityEnvironment._generate_step_input": {
241
+ "total": 12.354716499999505,
242
+ "count": 33602,
243
+ "is_parallel": true,
244
+ "self": 12.354716499999505
245
+ },
246
+ "communicator.exchange": {
247
+ "total": 150.37429890000874,
248
+ "count": 33602,
249
+ "is_parallel": true,
250
+ "self": 150.37429890000874
251
+ },
252
+ "steps_from_proto": {
253
+ "total": 9.860491800003157,
254
+ "count": 33602,
255
+ "is_parallel": true,
256
+ "self": 3.7401947999922953,
257
+ "children": {
258
+ "_process_rank_one_or_two_observation": {
259
+ "total": 6.120297000010861,
260
+ "count": 67204,
261
+ "is_parallel": true,
262
+ "self": 6.120297000010861
263
+ }
264
+ }
265
+ }
266
+ }
267
+ }
268
+ }
269
+ }
270
+ }
271
+ }
272
+ }
273
+ },
274
+ "trainer_advance": {
275
+ "total": 327.44521919999534,
276
+ "count": 33602,
277
+ "self": 1.3367100999911372,
278
+ "children": {
279
+ "process_trajectory": {
280
+ "total": 56.714018700004075,
281
+ "count": 33602,
282
+ "self": 56.59327900000402,
283
+ "children": {
284
+ "RLTrainer._checkpoint": {
285
+ "total": 0.12073970000005829,
286
+ "count": 1,
287
+ "self": 0.12073970000005829
288
+ }
289
+ }
290
+ },
291
+ "_update_policy": {
292
+ "total": 269.3944904000001,
293
+ "count": 39,
294
+ "self": 86.58347969999704,
295
+ "children": {
296
+ "TorchPPOOptimizer.update": {
297
+ "total": 182.81101070000307,
298
+ "count": 23013,
299
+ "self": 182.81101070000307
300
+ }
301
+ }
302
+ }
303
+ }
304
+ }
305
+ }
306
+ },
307
+ "trainer_threads": {
308
+ "total": 8.99999918146932e-07,
309
+ "count": 1,
310
+ "self": 8.99999918146932e-07
311
+ },
312
+ "TrainerController._save_models": {
313
+ "total": 0.0598454999999376,
314
+ "count": 1,
315
+ "self": 0.002040499999907297,
316
+ "children": {
317
+ "RLTrainer._checkpoint": {
318
+ "total": 0.057805000000030304,
319
+ "count": 1,
320
+ "self": 0.057805000000030304
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": 499833,
6
+ "file_path": "results\\fullTestHF\\3DBall\\3DBall-499833.onnx",
7
+ "reward": 100.00001525878906,
8
+ "creation_time": 1650348441.2484677,
9
+ "auxillary_file_paths": [
10
+ "results\\fullTestHF\\3DBall\\3DBall-499833.pt"
11
+ ]
12
+ },
13
+ {
14
+ "steps": 500833,
15
+ "file_path": "results\\fullTestHF\\3DBall\\3DBall-500833.onnx",
16
+ "reward": 100.00001525878906,
17
+ "creation_time": 1650348441.359496,
18
+ "auxillary_file_paths": [
19
+ "results\\fullTestHF\\3DBall\\3DBall-500833.pt"
20
+ ]
21
+ }
22
+ ],
23
+ "final_checkpoint": {
24
+ "steps": 500833,
25
+ "file_path": "results\\fullTestHF\\3DBall.onnx",
26
+ "reward": 100.00001525878906,
27
+ "creation_time": 1650348441.359496,
28
+ "auxillary_file_paths": [
29
+ "results\\fullTestHF\\3DBall\\3DBall-500833.pt"
30
+ ]
31
+ }
32
+ },
33
+ "metadata": {
34
+ "stats_format_version": "0.3.0",
35
+ "mlagents_version": "0.29.0.dev0",
36
+ "torch_version": "1.7.1+cu110"
37
+ }
38
+ }