Retrained to 10M steps, with higher play_against_latest_model_ratio (0.75 instead of 0.25) this helped the model to learn to play defense better

SoccerTwos.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c574606e5c1839061cdb7dc9b76863444295b5b91caccfc413a1ed077ee8c3f
+oid sha256:8b6d608c4248e0b310070b2af2b4ea2275b174aeed4ae07c4b7ff7524ee580d5
 size 895526

SoccerTwos/SoccerTwos-10000046.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b6d608c4248e0b310070b2af2b4ea2275b174aeed4ae07c4b7ff7524ee580d5
+size 895526

SoccerTwos/SoccerTwos-10000046.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7da542e5ed006a647545f746b09a40c9006d9b32927ef30a3ab22b48f7c7775c
+size 10325290

SoccerTwos/SoccerTwos-8499978.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6c0ccf92cff2ea1b32b49e95f6a42b5d63affc6f5ee79e658cd79776696d523
+size 895526

SoccerTwos/SoccerTwos-8499978.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eecdb844b9847844a5be6db79e351f3711605e9c42b7cc03ab11b57b62e7522f
+size 10325138

SoccerTwos/SoccerTwos-8999988.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19f45f7b592496e18ff616e19a9a276b8269eba93055ddb4f2a1d24b6e414b20
+size 895526

SoccerTwos/SoccerTwos-8999988.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8cde9c9d152dc7cce4fe5a93171548d2a7f8103247bc932b061a6e0c7faf43c7
+size 10325138

SoccerTwos/SoccerTwos-9499993.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b6772ba57d0dda936a1b86cfe6c1d0da4bff2be72cdf1a8289ac94f43aaa77f
+size 895526

SoccerTwos/SoccerTwos-9499993.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:811ca51769e17fef7d09824811d17d0e7b4823ff3e3459cf4dcdd9fa5753c134
+size 10325138

SoccerTwos/SoccerTwos-9999982.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ae6b4d8b9bce2651e6142580cdc5b4625eb01736ba52eb9ab6a952a964b862c
+size 895526

SoccerTwos/SoccerTwos-9999982.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5a6a4d7f4080efe31a06dc37491f39b98ca4223b9ab7f31f785bfefb6ccfb1f
+size 10325138

SoccerTwos/checkpoint.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:008145e1e808608cc93b4d650ceaa479d23cc271324dee6c3b01906b96fa84f6
+oid sha256:37854fa1cf62607fbcf6e1fa3522fdc3890a2084987edc894f3955253d66549d
 size 10323922

SoccerTwos/events.out.tfevents.1710341962.stephan-ms7c02.29905.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4328e8cf67a012d3b679ca03e83f9101e4827d554072a1d9dde86d2717a0992a
+size 661583

SoccerTwos/events.out.tfevents.1710343832.stephan-ms7c02.33144.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97512bb4824eae2e636aac4106b752ef6109d681e20d6fc75355ec92b3a0529f
+size 3405815

SoccerTwos_orig/SoccerTwos-4999994.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be96c3872475cc25d67472685bc2fee156427488a3c08e79bd28cf6c98213a94
+size 895526

SoccerTwos_orig/SoccerTwos-4999994.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71e54f36b571a60d1bd3474e5ee403ddc02aa6aafcd1c007e1272b0fa33cbfd8
+size 10325138

SoccerTwos_orig/SoccerTwos-5000026.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23416b67a78fe350c105debeb7be130b546d1ca99633a6d58b7b2ab627bdac15
+size 895526

SoccerTwos_orig/SoccerTwos-5000026.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef0aa7a1f48d9f4164990361bc247cd36df0a533121533f0cdf22a25adfb7189
+size 10325138

SoccerTwos_orig/SoccerTwos-5499990.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ee5c6c1bdaa2e03dd5acac35279c5a077f98827615be82d01ca79d646f2847f
+size 895526

SoccerTwos_orig/SoccerTwos-5499990.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1b631e6e474edfbd6397a6a173d3a36a447960b4e996fe4f72efd68936a6196
+size 10325138

SoccerTwos_orig/SoccerTwos-5999976.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56acb86e98bacbe50d2c8305c10c4bcb279131ff4c1359ef6245cccfed56cd97
+size 895526

SoccerTwos_orig/SoccerTwos-5999976.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c17aec0f05d051f9f1c6ca2db48529936e5c49c6978ab68abb2f1a5a9dbc004a
+size 10325138

SoccerTwos_orig/SoccerTwos-6000104.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c574606e5c1839061cdb7dc9b76863444295b5b91caccfc413a1ed077ee8c3f
+size 895526

SoccerTwos_orig/SoccerTwos-6000104.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccaa39ff525af5de7c6578077964314fe8b52d503f9fd6cb8fa52b144355b61b
+size 10325138

SoccerTwos_orig/checkpoint.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:008145e1e808608cc93b4d650ceaa479d23cc271324dee6c3b01906b96fa84f6
+size 10323922

SoccerTwos_orig/events.out.tfevents.1710290777.stephan-ms7c02.14283.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:921364813e102c784e2b1fe10a218e4108aaf0c18241e88faad09085f782460c
+size 4854988

SoccerTwos_orig/events.out.tfevents.1710333085.stephan-ms7c02.24579.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cd4e31cb3abbb0f4426bddd18591f073816db885a8bc2560ff7951c70d47284
+size 1017522

config.json

@@ -1 +1 @@
-{"default_settings": null, "behaviors": {"SoccerTwos": {"trainer_type": "poca", "hyperparameters": {"batch_size": 4096, "buffer_size": 65536, "learning_rate": 0.0004345843188705317, "beta": 0.00019215097636640332, "epsilon": 0.15831928707038231, "lambd": 0.9467147291117369, "num_epoch": 5, "learning_rate_schedule": "linear", "beta_schedule": "linear", "epsilon_schedule": "linear"}, "checkpoint_interval": 500000, "network_settings": {"normalize": true, "hidden_units": 256, "num_layers": 3, "vis_encode_type": "fully_connected", "memory": null, "goal_conditioning_type": "hyper", "deterministic": false}, "reward_signals": {"extrinsic": {"gamma": 0.9608227414353802, "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": 6000000, "time_horizon": 32, "summary_freq": 10000, "threaded": false, "self_play": {"save_steps": 60000, "team_change": 589998, "swap_steps": 70000, "window": 25, "play_against_latest_model_ratio": 0.2427456302841774, "initial_elo": 1200.0}, "behavioral_cloning": null}}, "env_settings": {"env_path": "./ml-agents/training-envs-executables/SoccerTwos.exe", "env_args": null, "base_port": 5005, "num_envs": 1, "num_areas": 1, "timeout_wait": 60, "seed": -1, "max_lifetime_restarts": 10, "restarts_rate_limit_n": 1, "restarts_rate_limit_period_s": 60}, "engine_settings": {"width": 1280, "height": 720, "quality_level": 5, "time_scale": 20, "target_frame_rate": -1, "capture_frame_rate": 60, "no_graphics": false, "no_graphics_monitor": false}, "environment_parameters": null, "checkpoint_settings": {"run_id": "SoccerTwos-read-deal", "initialize_from": null, "load_model": false, "resume": true, "force": false, "train_model": false, "inference": false, "results_dir": "results"}, "torch_settings": {"device": null}, "debug": false}
+{"default_settings": null, "behaviors": {"SoccerTwos": {"trainer_type": "poca", "hyperparameters": {"batch_size": 4096, "buffer_size": 65536, "learning_rate": 0.0004345843188705317, "beta": 0.00019215097636640332, "epsilon": 0.15831928707038231, "lambd": 0.9467147291117369, "num_epoch": 5, "learning_rate_schedule": "linear", "beta_schedule": "linear", "epsilon_schedule": "linear"}, "checkpoint_interval": 500000, "network_settings": {"normalize": true, "hidden_units": 256, "num_layers": 3, "vis_encode_type": "fully_connected", "memory": null, "goal_conditioning_type": "hyper", "deterministic": false}, "reward_signals": {"extrinsic": {"gamma": 0.9608227414353802, "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": 10000000, "time_horizon": 32, "summary_freq": 10000, "threaded": false, "self_play": {"save_steps": 60000, "team_change": 589998, "swap_steps": 70000, "window": 25, "play_against_latest_model_ratio": 0.75, "initial_elo": 1200.0}, "behavioral_cloning": null}}, "env_settings": {"env_path": "./ml-agents/training-envs-executables/SoccerTwos.exe", "env_args": null, "base_port": 5005, "num_envs": 1, "num_areas": 1, "timeout_wait": 60, "seed": -1, "max_lifetime_restarts": 10, "restarts_rate_limit_n": 1, "restarts_rate_limit_period_s": 60}, "engine_settings": {"width": 1280, "height": 720, "quality_level": 5, "time_scale": 20, "target_frame_rate": -1, "capture_frame_rate": 60, "no_graphics": false, "no_graphics_monitor": false}, "environment_parameters": null, "checkpoint_settings": {"run_id": "SoccerTwos-read-deal", "initialize_from": null, "load_model": false, "resume": true, "force": false, "train_model": false, "inference": false, "results_dir": "results"}, "torch_settings": {"device": null}, "debug": false}

configuration.yaml

@@ -37,7 +37,7 @@ behaviors:
     init_path: null
     keep_checkpoints: 5
     even_checkpoints: false
-    max_steps: 6000000
+    max_steps: 10000000
     time_horizon: 32
     summary_freq: 10000
     threaded: false
@@ -46,7 +46,7 @@ behaviors:
       team_change: 589998
       swap_steps: 70000
       window: 25
-      play_against_latest_model_ratio: 0.2427456302841774
+      play_against_latest_model_ratio: 0.75
       initial_elo: 1200.0
     behavioral_cloning: null
 env_settings:

run_logs/Player-0.log

@@ -19,14 +19,14 @@ ns GL_ARB_stencil_texturing GL_ARB_sync GL_ARB_tessellation_shader GL_ARB_textur
 _buffer_object GL_ARB_vertex_shader GL_ARB_vertex_type_10f_11f_11f_rev GL_ARB_vertex_type_2_10_10_10_rev GL_ARB_viewport_array GL_ATI_blend_equation_separate GL_ATI_meminfo GL_ATI_texture_float GL_ATI_texture_mirror_once GL_EXT_EGL_image_storage GL_EXT_EGL_sync GL_EXT_abgr GL_EXT_blend_equation_separate GL_EXT_debug_label GL_EXT_demote_to_helper_invocation GL_EXT_depth_bounds_test GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_framebuffer_object GL_EXT_framebuffer_sRGB GL_EXT_memory_object GL_EXT_memory_object_fd GL_EXT_packed_depth_stencil GL_EXT_packed_float GL_EXT_pixel_buffer_object GL_EXT_polygon_offset_clamp GL_EXT_provoking_vertex GL_EXT_semaphore GL_EXT_semaphore_fd GL_EXT_shader_image_load_formatted GL_EXT_shader_image_load_store GL_EXT_shader_integer_mix GL_EXT_shader_samples_identical GL_EXT_texture_array GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_rgtc GL_EXT_texture_compression_s3t
 c GL_EXT_texture_filter_anisotropic GL_EXT_texture_integer GL_EXT_texture_mirror_clamp GL_EXT_texture_sRGB GL_EXT_texture_sRGB_R8 GL_EXT_texture_sRGB_decode GL_EXT_texture_shadow_lod GL_EXT_texture_shared_exponent GL_EXT_texture_snorm GL_EXT_texture_swizzle GL_EXT_timer_query GL_EXT_transform_feedback GL_EXT_vertex_array_bgra GL_EXT_vertex_attrib_64bit GL_EXT_window_rectangles GL_IBM_multimode_draw_arrays GL_INTEL_blackhole_render GL_KHR_blend_equation_advanced GL_KHR_context_flush_control GL_KHR_debug GL_KHR_no_error GL_KHR_parallel_shader_compile GL_KHR_robust_buffer_access_behavior GL_KHR_robustness GL_KHR_texture_compression_astc_ldr GL_KHR_texture_compression_astc_sliced_3d GL_MESA_framebuffer_flip_y GL_MESA_pack_invert GL_MESA_shader_integer_functions GL_MESA_texture_const_bandwidth GL_MESA_texture_signed_rgba GL_NVX_gpu_memory_info GL_NV_alpha_to_coverage_dither_control GL_NV_compute_shader_derivatives GL_NV_conditional_render GL_NV_copy_image GL_NV_depth_clamp GL_NV_packed_depth_stencil GL_NV_shader_
 atomic_int64 GL_NV_texture_barrier GL_NV_vdpau_interop GL_OES_EGL_image GL_S3_s3tc
-OPENGL LOG: Creating OpenGL 4.6 graphics device ; Context level  <OpenGL 4.5> ; Context handle 16352240
+OPENGL LOG: Creating OpenGL 4.6 graphics device ; Context level  <OpenGL 4.5> ; Context handle 22053744
 Begin MonoManager ReloadAssembly
-- Completed reload, in  0.083 seconds
+- Completed reload, in  0.092 seconds
 Default vsync count 1
 requesting resize 1280 x 720
 resizing window to 1280 x 720
 Desktop is 1920 x 1080 @ 60 Hz
-UnloadTime: 0.827430 ms
+UnloadTime: 0.807040 ms
 Registered Communicator in Agent.
 Setting up 6 worker threads for Enlighten.
 requesting resize 1280 x 720
@@ -36,7 +36,7 @@ Memory Statistics:
 [ALLOC_TEMP_TLS] TLS Allocator
   StackAllocators : 
     [ALLOC_TEMP_MAIN]
-      Peak usage frame count: [8.0 KB-16.0 KB]: 18758 frames, [2.0 MB-4.0 MB]: 1 frames
+      Peak usage frame count: [8.0 KB-16.0 KB]: 62727 frames, [2.0 MB-4.0 MB]: 1 frames
       Initial Block Size 4.0 MB
       Current Block Size 4.0 MB
       Peak Allocated Bytes 2.1 MB
@@ -211,18 +211,18 @@ Memory Statistics:
     [ALLOC_BUCKET]
       Large Block size 4.0 MB
       Used Block count 1
-      Peak Allocated bytes 2.0 MB
+      Peak Allocated bytes 1.9 MB
     [ALLOC_DEFAULT_MAIN]
-      Peak usage frame count: [8.0 MB-16.0 MB]: 18759 frames
+      Peak usage frame count: [8.0 MB-16.0 MB]: 62728 frames
       Requested Block Size 16.0 MB
       Peak Block count 1
       Peak Allocated memory 13.0 MB
       Peak Large allocation bytes 0 B
     [ALLOC_DEFAULT_THREAD]
-      Peak usage frame count: [16.0 MB-32.0 MB]: 18759 frames
+      Peak usage frame count: [16.0 MB-32.0 MB]: 62728 frames
       Requested Block Size 16.0 MB
       Peak Block count 1
-      Peak Allocated memory 29.1 MB
+      Peak Allocated memory 28.7 MB
       Peak Large allocation bytes 16.0 MB
 [ALLOC_TEMP_JOB_1_FRAME]
   Initial Block Size 2.0 MB
@@ -249,15 +249,15 @@ Memory Statistics:
     [ALLOC_BUCKET]
       Large Block size 4.0 MB
       Used Block count 1
-      Peak Allocated bytes 2.0 MB
+      Peak Allocated bytes 1.9 MB
     [ALLOC_GFX_MAIN]
-      Peak usage frame count: [32.0 KB-64.0 KB]: 18758 frames, [2.0 MB-4.0 MB]: 1 frames
+      Peak usage frame count: [32.0 KB-64.0 KB]: 62727 frames, [2.0 MB-4.0 MB]: 1 frames
       Requested Block Size 16.0 MB
       Peak Block count 1
       Peak Allocated memory 2.7 MB
       Peak Large allocation bytes 0 B
     [ALLOC_GFX_THREAD]
-      Peak usage frame count: [128.0 KB-256.0 KB]: 18759 frames
+      Peak usage frame count: [128.0 KB-256.0 KB]: 62728 frames
       Requested Block Size 16.0 MB
       Peak Block count 1
       Peak Allocated memory 174.7 KB
@@ -267,15 +267,15 @@ Memory Statistics:
     [ALLOC_BUCKET]
       Large Block size 4.0 MB
       Used Block count 1
-      Peak Allocated bytes 2.0 MB
+      Peak Allocated bytes 1.9 MB
     [ALLOC_CACHEOBJECTS_MAIN]
-      Peak usage frame count: [0.5 MB-1.0 MB]: 18758 frames, [1.0 MB-2.0 MB]: 1 frames
+      Peak usage frame count: [0.5 MB-1.0 MB]: 62727 frames, [1.0 MB-2.0 MB]: 1 frames
       Requested Block Size 4.0 MB
       Peak Block count 1
       Peak Allocated memory 1.2 MB
       Peak Large allocation bytes 0 B
     [ALLOC_CACHEOBJECTS_THREAD]
-      Peak usage frame count: [4.0 MB-8.0 MB]: 18759 frames
+      Peak usage frame count: [4.0 MB-8.0 MB]: 62728 frames
       Requested Block Size 4.0 MB
       Peak Block count 2
       Peak Allocated memory 7.8 MB
@@ -285,15 +285,15 @@ Memory Statistics:
     [ALLOC_BUCKET]
       Large Block size 4.0 MB
       Used Block count 1
-      Peak Allocated bytes 2.0 MB
+      Peak Allocated bytes 1.9 MB
     [ALLOC_TYPETREE_MAIN]
-      Peak usage frame count: [0-1.0 KB]: 18759 frames
+      Peak usage frame count: [0-1.0 KB]: 62728 frames
       Requested Block Size 2.0 MB
       Peak Block count 1
       Peak Allocated memory 1.0 KB
       Peak Large allocation bytes 0 B
     [ALLOC_TYPETREE_THREAD]
-      Peak usage frame count: [1.0 KB-2.0 KB]: 18758 frames, [2.0 KB-4.0 KB]: 1 frames
+      Peak usage frame count: [1.0 KB-2.0 KB]: 62727 frames, [2.0 KB-4.0 KB]: 1 frames
       Requested Block Size 2.0 MB
       Peak Block count 1
       Peak Allocated memory 2.2 KB

run_logs/timers.json

@@ -2,201 +2,201 @@
     "name": "root",
     "gauges": {
         "SoccerTwos.Policy.Entropy.mean": {
-            "value": 1.778828501701355,
-            "min": 1.7496646642684937,
-            "max": 1.904069185256958,
-            "count": 100
+            "value": 1.4048619270324707,
+            "min": 1.3750499486923218,
+            "max": 1.7051565647125244,
+            "count": 335
         },
         "SoccerTwos.Policy.Entropy.sum": {
-            "value": 35690.4140625,
-            "min": 34937.3046875,
-            "max": 38751.6171875,
-            "count": 100
+            "value": 28277.060546875,
+            "min": 14701.9814453125,
+            "max": 34322.671875,
+            "count": 335
         },
         "SoccerTwos.Environment.EpisodeLength.mean": {
-            "value": 43.51754385964912,
-            "min": 38.792,
-            "max": 53.82022471910113,
-            "count": 100
+            "value": 45.06422018348624,
+            "min": 35.125,
+            "max": 53.40425531914894,
+            "count": 335
         },
         "SoccerTwos.Environment.EpisodeLength.sum": {
-            "value": 19844.0,
-            "min": 18336.0,
-            "max": 20936.0,
-            "count": 100
+            "value": 19648.0,
+            "min": 8072.0,
+            "max": 20960.0,
+            "count": 335
         },
         "SoccerTwos.Self-play.ELO.mean": {
-            "value": 1561.460064375993,
-            "min": 1519.7251303553944,
-            "max": 1561.4942652168481,
-            "count": 100
+            "value": 1607.9046472284606,
+            "min": 1555.6832652825399,
+            "max": 1613.038356436412,
+            "count": 335
         },
         "SoccerTwos.Self-play.ELO.sum": {
-            "value": 356012.8946777264,
-            "min": 273814.30154407705,
-            "max": 390373.566304212,
-            "count": 100
+            "value": 350523.2130958044,
+            "min": 163654.2453609866,
+            "max": 434134.2547516843,
+            "count": 335
         },
         "SoccerTwos.Step.mean": {
-            "value": 5999976.0,
-            "min": 5009984.0,
-            "max": 5999976.0,
-            "count": 100
+            "value": 9999982.0,
+            "min": 6659988.0,
+            "max": 9999982.0,
+            "count": 335
         },
         "SoccerTwos.Step.sum": {
-            "value": 5999976.0,
-            "min": 5009984.0,
-            "max": 5999976.0,
-            "count": 100
+            "value": 9999982.0,
+            "min": 6659988.0,
+            "max": 9999982.0,
+            "count": 335
         },
         "SoccerTwos.Policy.ExtrinsicBaselineEstimate.mean": {
-            "value": 0.036884572356939316,
-            "min": -0.06633567065000534,
-            "max": 0.09457974880933762,
-            "count": 100
+            "value": 0.03223046660423279,
+            "min": -0.11106479167938232,
+            "max": 0.10332842171192169,
+            "count": 335
         },
         "SoccerTwos.Policy.ExtrinsicBaselineEstimate.sum": {
-            "value": 15.380867004394531,
-            "min": -27.06495475769043,
-            "max": 38.872276306152344,
-            "count": 100
+            "value": 13.246722221374512,
+            "min": -46.31401824951172,
+            "max": 44.017906188964844,
+            "count": 335
         },
         "SoccerTwos.Policy.ExtrinsicValueEstimate.mean": {
-            "value": 0.03724776208400726,
-            "min": -0.06787735223770142,
-            "max": 0.09511835128068924,
-            "count": 100
+            "value": 0.031953852623701096,
+            "min": -0.11250180751085281,
+            "max": 0.10477326810359955,
+            "count": 335
         },
         "SoccerTwos.Policy.ExtrinsicValueEstimate.sum": {
-            "value": 15.532317161560059,
-            "min": -27.693960189819336,
-            "max": 39.09364318847656,
-            "count": 100
+            "value": 13.133033752441406,
+            "min": -46.91325378417969,
+            "max": 44.6334114074707,
+            "count": 335
         },
         "SoccerTwos.Environment.CumulativeReward.mean": {
             "value": 0.0,
             "min": 0.0,
             "max": 0.0,
-            "count": 100
+            "count": 335
         },
         "SoccerTwos.Environment.CumulativeReward.sum": {
             "value": 0.0,
             "min": 0.0,
             "max": 0.0,
-            "count": 100
+            "count": 335
         },
         "SoccerTwos.Policy.ExtrinsicReward.mean": {
-            "value": 0.04650350934580753,
-            "min": -0.2562914263634455,
-            "max": 0.2151981968063492,
-            "count": 100
+            "value": 0.09531192604554903,
+            "min": -0.381275439994377,
+            "max": 0.26663347656634745,
+            "count": 335
         },
         "SoccerTwos.Policy.ExtrinsicReward.sum": {
-            "value": 10.602800130844116,
-            "min": -53.82119953632355,
-            "max": 47.77399969100952,
-            "count": 100
+            "value": 20.777999877929688,
+            "min": -86.93080031871796,
+            "max": 62.125600039958954,
+            "count": 335
         },
         "SoccerTwos.Environment.GroupCumulativeReward.mean": {
-            "value": 0.04650350934580753,
-            "min": -0.2562914263634455,
-            "max": 0.2151981968063492,
-            "count": 100
+            "value": 0.09531192604554903,
+            "min": -0.381275439994377,
+            "max": 0.26663347656634745,
+            "count": 335
         },
         "SoccerTwos.Environment.GroupCumulativeReward.sum": {
-            "value": 10.602800130844116,
-            "min": -53.82119953632355,
-            "max": 47.77399969100952,
-            "count": 100
+            "value": 20.777999877929688,
+            "min": -86.93080031871796,
+            "max": 62.125600039958954,
+            "count": 335
         },
         "SoccerTwos.IsTraining.mean": {
             "value": 1.0,
             "min": 1.0,
             "max": 1.0,
-            "count": 100
+            "count": 335
         },
         "SoccerTwos.IsTraining.sum": {
             "value": 1.0,
             "min": 1.0,
             "max": 1.0,
-            "count": 100
+            "count": 335
         },
         "SoccerTwos.Losses.PolicyLoss.mean": {
-            "value": 0.014634914088446749,
-            "min": 0.010371605570253451,
-            "max": 0.014634914088446749,
-            "count": 15
+            "value": 0.01046212162455049,
+            "min": 0.009879295097471185,
+            "max": 0.014559235423075734,
+            "count": 51
         },
         "SoccerTwos.Losses.PolicyLoss.sum": {
-            "value": 0.014634914088446749,
-            "min": 0.010371605570253451,
-            "max": 0.014634914088446749,
-            "count": 15
+            "value": 0.01046212162455049,
+            "min": 0.009879295097471185,
+            "max": 0.014559235423075734,
+            "count": 51
         },
         "SoccerTwos.Losses.ValueLoss.mean": {
-            "value": 0.051349658705294135,
-            "min": 0.048586570238694546,
-            "max": 0.05451308977790177,
-            "count": 15
+            "value": 0.05891079842112958,
+            "min": 0.04971672408282757,
+            "max": 0.05918900854885578,
+            "count": 51
         },
         "SoccerTwos.Losses.ValueLoss.sum": {
-            "value": 0.051349658705294135,
-            "min": 0.048586570238694546,
-            "max": 0.05451308977790177,
-            "count": 15
+            "value": 0.05891079842112958,
+            "min": 0.04971672408282757,
+            "max": 0.05918900854885578,
+            "count": 51
         },
         "SoccerTwos.Losses.BaselineLoss.mean": {
-            "value": 0.0515909023117274,
-            "min": 0.04938719081692398,
-            "max": 0.054688979778438805,
-            "count": 15
+            "value": 0.05924153439700604,
+            "min": 0.051102403877303006,
+            "max": 0.05954511137679219,
+            "count": 51
         },
         "SoccerTwos.Losses.BaselineLoss.sum": {
-            "value": 0.0515909023117274,
-            "min": 0.04938719081692398,
-            "max": 0.054688979778438805,
-            "count": 15
+            "value": 0.05924153439700604,
+            "min": 0.051102403877303006,
+            "max": 0.05954511137679219,
+            "count": 51
         },
         "SoccerTwos.Policy.LearningRate.mean": {
-            "value": 1.1853560262662174e-06,
-            "min": 1.1853560262662174e-06,
-            "max": 6.767456859236319e-05,
-            "count": 15
+            "value": 6.4449295954992225e-09,
+            "min": 6.4449295954992225e-09,
+            "max": 0.00014248644558948728,
+            "count": 51
         },
         "SoccerTwos.Policy.LearningRate.sum": {
-            "value": 1.1853560262662174e-06,
-            "min": 1.1853560262662174e-06,
-            "max": 6.767456859236319e-05,
-            "count": 15
+            "value": 6.4449295954992225e-09,
+            "min": 6.4449295954992225e-09,
+            "max": 0.00014248644558948728,
+            "count": 51
         },
         "SoccerTwos.Policy.Epsilon.mean": {
-            "value": 0.1001590561356033,
-            "min": 0.1001590561356033,
-            "max": 0.10908161546093645,
-            "count": 15
+            "value": 0.10000085146159124,
+            "min": 0.10000085146159124,
+            "max": 0.11912103967704953,
+            "count": 51
         },
         "SoccerTwos.Policy.Epsilon.sum": {
-            "value": 0.1001590561356033,
-            "min": 0.1001590561356033,
-            "max": 0.10908161546093645,
-            "count": 15
+            "value": 0.10000085146159124,
+            "min": 0.10000085146159124,
+            "max": 0.11912103967704953,
+            "count": 51
         },
         "SoccerTwos.Policy.Beta.mean": {
-            "value": 1.04967864295433e-05,
-            "min": 1.04967864295433e-05,
-            "max": 3.8364975058721184e-05,
-            "count": 15
+            "value": 1.0002659404254946e-05,
+            "min": 1.0002659404254946e-05,
+            "max": 6.972151274949521e-05,
+            "count": 51
         },
         "SoccerTwos.Policy.Beta.sum": {
-            "value": 1.04967864295433e-05,
-            "min": 1.04967864295433e-05,
-            "max": 3.8364975058721184e-05,
-            "count": 15
+            "value": 1.0002659404254946e-05,
+            "min": 1.0002659404254946e-05,
+            "max": 6.972151274949521e-05,
+            "count": 51
         }
     },
     "metadata": {
         "timer_format_version": "0.1.0",
-        "start_time_seconds": "1710333081",
+        "start_time_seconds": "1710343829",
         "python_version": "3.10.12 (main, Jul  5 2023, 18:54:27) [GCC 11.2.0]",
         "command_line_arguments": "/home/stephan/.conda/envs/rl/bin/mlagents-learn ./ml-agents/config/poca/SoccerTwos.yaml --env=./ml-agents/training-envs-executables/SoccerTwos.exe --run-id=SoccerTwos-read-deal --width=1280 --height=720 --resume",
         "mlagents_version": "1.1.0.dev0",
@@ -204,59 +204,59 @@
         "communication_protocol_version": "1.5.0",
         "pytorch_version": "2.2.1+cu121",
         "numpy_version": "1.23.5",
-        "end_time_seconds": "1710335865"
+        "end_time_seconds": "1710353568"
     },
-    "total": 2783.08172296299,
+    "total": 9739.254863360999,
     "count": 1,
-    "self": 0.3221229049813701,
+    "self": 0.270132729987381,
     "children": {
         "run_training.setup": {
-            "total": 0.017381374011165462,
+            "total": 0.018506341002648696,
             "count": 1,
-            "self": 0.017381374011165462
+            "self": 0.018506341002648696
         },
         "TrainerController.start_learning": {
-            "total": 2782.7422186839976,
+            "total": 9738.966224290009,
             "count": 1,
-            "self": 1.4529281045834068,
+            "self": 4.652608350283117,
             "children": {
                 "TrainerController._reset_env": {
-                    "total": 4.399289562992635,
-                    "count": 3,
-                    "self": 4.399289562992635
+                    "total": 4.244304390012985,
+                    "count": 7,
+                    "self": 4.244304390012985
                 },
                 "TrainerController.advance": {
-                    "total": 2776.7682698464196,
-                    "count": 71154,
-                    "self": 1.2270896326372167,
+                    "total": 9729.947026144713,
+                    "count": 239073,
+                    "self": 4.148107524248189,
                     "children": {
                         "env_step": {
-                            "total": 1269.4150066098955,
-                            "count": 71154,
-                            "self": 1077.1618315594678,
+                            "total": 4354.3673745146225,
+                            "count": 239073,
+                            "self": 3721.0493522615725,
                             "children": {
                                 "SubprocessEnvManager._take_step": {
-                                    "total": 191.40557155759598,
-                                    "count": 71154,
-                                    "self": 5.71247391810175,
+                                    "total": 630.4299217515072,
+                                    "count": 239073,
+                                    "self": 19.02124206413282,
                                     "children": {
                                         "TorchPolicy.evaluate": {
-                                            "total": 185.69309763949423,
-                                            "count": 125058,
-                                            "self": 185.69309763949423
+                                            "total": 611.4086796873744,
+                                            "count": 418182,
+                                            "self": 611.4086796873744
                                         }
                                     }
                                 },
                                 "workers": {
-                                    "total": 0.8476034928316949,
-                                    "count": 71154,
+                                    "total": 2.8881005015427945,
+                                    "count": 239073,
                                     "self": 0.0,
                                     "children": {
                                         "worker_root": {
-                                            "total": 2780.0714288110757,
-                                            "count": 71154,
+                                            "total": 9731.739510110943,
+                                            "count": 239073,
                                             "is_parallel": true,
-                                            "self": 1856.7726158039295,
+                                            "self": 6533.360267171403,
                                             "children": {
                                                 "run_training.setup": {
                                                     "total": 0.0,
@@ -265,48 +265,48 @@
                                                     "self": 0.0,
                                                     "children": {
                                                         "steps_from_proto": {
-                                                            "total": 0.06613134100916795,
+                                                            "total": 0.06389564499841072,
                                                             "count": 2,
                                                             "is_parallel": true,
-                                                            "self": 0.00047555101627949625,
+                                                            "self": 0.000593470991589129,
                                                             "children": {
                                                                 "_process_rank_one_or_two_observation": {
-                                                                    "total": 0.06565578999288846,
+                                                                    "total": 0.06330217400682159,
                                                                     "count": 8,
                                                                     "is_parallel": true,
-                                                                    "self": 0.06565578999288846
+                                                                    "self": 0.06330217400682159
                                                                 }
                                                             }
                                                         },
                                                         "UnityEnvironment.step": {
-                                                            "total": 0.031780133009306155,
+                                                            "total": 0.02612613000383135,
                                                             "count": 1,
                                                             "is_parallel": true,
-                                                            "self": 0.0006475910049630329,
+                                                            "self": 0.000600070008658804,
                                                             "children": {
                                                                 "UnityEnvironment._generate_step_input": {
-                                                                    "total": 0.0005751600110670552,
+                                                                    "total": 0.00040655999328009784,
                                                                     "count": 1,
                                                                     "is_parallel": true,
-                                                                    "self": 0.0005751600110670552
+                                                                    "self": 0.00040655999328009784
                                                                 },
                                                                 "communicator.exchange": {
-                                                                    "total": 0.02851476900104899,
+                                                                    "total": 0.02343449799809605,
                                                                     "count": 1,
                                                                     "is_parallel": true,
-                                                                    "self": 0.02851476900104899
+                                                                    "self": 0.02343449799809605
                                                                 },
                                                                 "steps_from_proto": {
-                                                                    "total": 0.0020426129922270775,
+                                                                    "total": 0.0016850020037963986,
                                                                     "count": 2,
                                                                     "is_parallel": true,
-                                                                    "self": 0.0004071999865118414,
+                                                                    "self": 0.0003057300054933876,
                                                                     "children": {
                                                                         "_process_rank_one_or_two_observation": {
-                                                                            "total": 0.001635413005715236,
+                                                                            "total": 0.001379271998303011,
                                                                             "count": 8,
                                                                             "is_parallel": true,
-                                                                            "self": 0.001635413005715236
+                                                                            "self": 0.001379271998303011
                                                                         }
                                                                     }
                                                                 }
@@ -315,48 +315,48 @@
                                                     }
                                                 },
                                                 "steps_from_proto": {
-                                                    "total": 0.003298083000117913,
-                                                    "count": 4,
+                                                    "total": 0.012151043978519738,
+                                                    "count": 12,
                                                     "is_parallel": true,
-                                                    "self": 0.0005663799965986982,
+                                                    "self": 0.002059131977148354,
                                                     "children": {
                                                         "_process_rank_one_or_two_observation": {
-                                                            "total": 0.0027317030035192147,
-                                                            "count": 16,
+                                                            "total": 0.010091912001371384,
+                                                            "count": 48,
                                                             "is_parallel": true,
-                                                            "self": 0.0027317030035192147
+                                                            "self": 0.010091912001371384
                                                         }
                                                     }
                                                 },
                                                 "UnityEnvironment.step": {
-                                                    "total": 923.2955149241461,
-                                                    "count": 71153,
+                                                    "total": 3198.3670918955613,
+                                                    "count": 239072,
                                                     "is_parallel": true,
-                                                    "self": 50.94944198477606,
+                                                    "self": 175.16570653802773,
                                                     "children": {
                                                         "UnityEnvironment._generate_step_input": {
-                                                            "total": 31.86183314715163,
-                                                            "count": 71153,
+                                                            "total": 110.43290661888022,
+                                                            "count": 239072,
                                                             "is_parallel": true,
-                                                            "self": 31.86183314715163
+                                                            "self": 110.43290661888022
                                                         },
                                                         "communicator.exchange": {
-                                                            "total": 713.6419240085379,
-                                                            "count": 71153,
+                                                            "total": 2470.8126502879313,
+                                                            "count": 239072,
                                                             "is_parallel": true,
-                                                            "self": 713.6419240085379
+                                                            "self": 2470.8126502879313
                                                         },
                                                         "steps_from_proto": {
-                                                            "total": 126.84231578368053,
-                                                            "count": 142306,
+                                                            "total": 441.95582845072204,
+                                                            "count": 478144,
                                                             "is_parallel": true,
-                                                            "self": 22.54734104256204,
+                                                            "self": 78.3687699053844,
                                                             "children": {
                                                                 "_process_rank_one_or_two_observation": {
-                                                                    "total": 104.2949747411185,
-                                                                    "count": 569224,
+                                                                    "total": 363.58705854533764,
+                                                                    "count": 1912576,
                                                                     "is_parallel": true,
-                                                                    "self": 104.2949747411185
+                                                                    "self": 363.58705854533764
                                                                 }
                                                             }
                                                         }
@@ -369,31 +369,31 @@
                             }
                         },
                         "trainer_advance": {
-                            "total": 1506.1261736038869,
-                            "count": 71154,
-                            "self": 8.044354920290061,
+                            "total": 5371.4315441058425,
+                            "count": 239073,
+                            "self": 26.83088273082103,
                             "children": {
                                 "process_trajectory": {
-                                    "total": 323.5915293196158,
-                                    "count": 71154,
-                                    "self": 323.3412643446063,
+                                    "total": 1094.2377518599678,
+                                    "count": 239073,
+                                    "self": 1093.3974272879568,
                                     "children": {
                                         "RLTrainer._checkpoint": {
-                                            "total": 0.2502649750094861,
-                                            "count": 2,
-                                            "self": 0.2502649750094861
+                                            "total": 0.8403245720110135,
+                                            "count": 7,
+                                            "self": 0.8403245720110135
                                         }
                                     }
                                 },
                                 "_update_policy": {
-                                    "total": 1174.490289363981,
-                                    "count": 15,
-                                    "self": 185.54603042497183,
+                                    "total": 4250.362909515054,
+                                    "count": 51,
+                                    "self": 652.032682771809,
                                     "children": {
                                         "TorchPOCAOptimizer.update": {
-                                            "total": 988.9442589390092,
-                                            "count": 1200,
-                                            "self": 988.9442589390092
+                                            "total": 3598.3302267432446,
+                                            "count": 4080,
+                                            "self": 3598.3302267432446
                                         }
                                     }
                                 }
@@ -402,19 +402,19 @@
                     }
                 },
                 "trainer_threads": {
-                    "total": 8.600036380812526e-07,
+                    "total": 9.200011845678091e-07,
                     "count": 1,
-                    "self": 8.600036380812526e-07
+                    "self": 9.200011845678091e-07
                 },
                 "TrainerController._save_models": {
-                    "total": 0.12173030999838375,
+                    "total": 0.12228448499809019,
                     "count": 1,
-                    "self": 0.002498182002454996,
+                    "self": 0.0026952329935738817,
                     "children": {
                         "RLTrainer._checkpoint": {
-                            "total": 0.11923212799592875,
+                            "total": 0.11958925200451631,
                             "count": 1,
-                            "self": 0.11923212799592875
+                            "self": 0.11958925200451631
                         }
                     }
                 }

run_logs/training_status.json

@@ -2,59 +2,59 @@
     "SoccerTwos": {
         "checkpoints": [
             {
-                "steps": 4999994,
-                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-4999994.onnx",
+                "steps": 8499978,
+                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-8499978.onnx",
                 "reward": 0.0,
-                "creation_time": 1710305448.3632262,
+                "creation_time": 1710349255.752344,
                 "auxillary_file_paths": [
-                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-4999994.pt"
+                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-8499978.pt"
                 ]
             },
             {
-                "steps": 5000026,
-                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-5000026.onnx",
+                "steps": 8999988,
+                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-8999988.onnx",
                 "reward": 0.0,
-                "creation_time": 1710305448.48951,
+                "creation_time": 1710350638.0742688,
                 "auxillary_file_paths": [
-                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-5000026.pt"
+                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-8999988.pt"
                 ]
             },
             {
-                "steps": 5499990,
-                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-5499990.onnx",
+                "steps": 9499993,
+                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-9499993.onnx",
                 "reward": 0.0,
-                "creation_time": 1710334432.864333,
+                "creation_time": 1710352100.153916,
                 "auxillary_file_paths": [
-                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-5499990.pt"
+                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-9499993.pt"
                 ]
             },
             {
-                "steps": 5999976,
-                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-5999976.onnx",
-                "reward": 0.0,
-                "creation_time": 1710335864.5550687,
+                "steps": 9999982,
+                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-9999982.onnx",
+                "reward": null,
+                "creation_time": 1710353568.109763,
                 "auxillary_file_paths": [
-                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-5999976.pt"
+                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-9999982.pt"
                 ]
             },
             {
-                "steps": 6000104,
-                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-6000104.onnx",
-                "reward": 0.0,
-                "creation_time": 1710335864.7100072,
+                "steps": 10000046,
+                "file_path": "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-10000046.onnx",
+                "reward": null,
+                "creation_time": 1710353568.2486105,
                 "auxillary_file_paths": [
-                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-6000104.pt"
+                    "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-10000046.pt"
                 ]
             }
         ],
-        "elo": 1561.4600643759927,
+        "elo": 1607.9046472284606,
         "final_checkpoint": {
-            "steps": 6000104,
+            "steps": 10000046,
             "file_path": "results/SoccerTwos-read-deal/SoccerTwos.onnx",
-            "reward": 0.0,
-            "creation_time": 1710335864.7100072,
+            "reward": null,
+            "creation_time": 1710353568.2486105,
             "auxillary_file_paths": [
-                "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-6000104.pt"
+                "results/SoccerTwos-read-deal/SoccerTwos/SoccerTwos-10000046.pt"
             ]
         }
     },