Replicate is joining Cloudflare
  1. Dashboard
  2. Predictions
  3. b1987qje55rme0cp1zjs3zy3ec
Model
andreasjansson/create-dynamic-system:132839d0
ID
b1987qje55rme0cp1zjs3zy3ec
Status
Succeeded
Source
Web
Hardware
CPU
Total duration
Created
Webhook
–

Input

target1
Output
https://replicate.delivery/xezq/Oc2TFqk1rLq2ExmRjdm9CH6HOHX3py6Cb8lAGm0BdK7Gi2HF/out-0.webp
target2
Output
https://replicate.delivery/pbxt/MnLZrd7mMDrttoNJvY5vfJmx7rTrQljig10cmOUpeQTynJ7U/image.png
max_attempts
20
width
128
height
128
training_steps
1000
max_variables
200
cycle_length
20
loss_cycles
4
total_cycles
6
learning_rate
0.005
return_animation
true
train_timeout
600

Output

Object output with 10 properties

code
class WaveInterferenceOscillator(DynamicSystem):
    """A dynamic system that generates oscillating patterns through wave interactions and
    non-linear feedback, with very few parameters."""
    
    def __init__(self):
        super().__init__()
        
        # Use a small grid to reduce parameter count
        self.grid_size = 16
        
        # Define only 3 channels (RGB)
        self.channels = 3
        
        # Wave parameters (wavelength, direction, phase)
        self.wave_params = nn.Parameter(torch.rand(3, 4))
        
        # Interaction parameters between channels
        self.interaction = nn.Parameter(0.2 * torch.randn(3, 3))
        
        # Feedback strength and bias
        self.feedback = nn.Parameter(0.1 * torch.randn(3))
        self.bias = nn.Parameter(0.1 * torch.randn(3))
        
        # Non-linearity control
        self.activation = nn.Parameter(torch.ones(1) * 2.0)
        
        # Memory decay factor
        self.memory = nn.Parameter(torch.tensor([0.7]))
        
        # Core pattern - small 4x4 pattern that we'll tile for efficiency
        self.pattern_seed = nn.Parameter(0.1 * torch.randn(3, 4, 4))
        
        # Reset state variables
        self.reset_state()
    
    def reset_state(self):
        # Initialize state by tiling the pattern seed
        self.state = self.pattern_seed.repeat(1, 
                                            self.grid_size // 4, 
                                            self.grid_size // 4)
        self.prev_state = self.state.clone()
        self.phase = torch.zeros(self.channels, device=self.pattern_seed.device)
    
    def spatial_filter(self, x):
        # Create simple spatial filtering with minimal parameters
        # This is much more parameter-efficient than a full Conv2d
        padded = torch.nn.functional.pad(x, (1, 1, 1, 1), mode='circular')
        
        # Apply blurring/diffusion using a 3x3 kernel
        neighbors = 0
        neighbors += padded[:, :-2, 1:-1]  # left
        neighbors += padded[:, 2:, 1:-1]   # right
        neighbors += padded[:, 1:-1, :-2]  # top
        neighbors += padded[:, 1:-1, 2:]   # bottom
        
        return 0.1 * neighbors + 0.6 * x
    
    def step(self):
        # Update phase for wave effects
        self.phase = (self.phase + 0.1) % (2 * torch.pi)
        
        # Store previous state with memory decay
        self.prev_state = self.memory * self.prev_state + (1 - self.memory) * self.state
        
        # Generate wave influence at each point in the grid
        y_grid, x_grid = torch.meshgrid(
            torch.linspace(0, 1, self.grid_size, device=self.phase.device),
            torch.linspace(0, 1, self.grid_size, device=self.phase.device),
            indexing='ij'
        )
        
        wave_effect = torch.zeros_like(self.state)
        for c in range(self.channels):
            # Different wave patterns for each channel
            wx, wy, freq, phase_offset = self.wave_params[c]
            wave = torch.sin(
                2 * torch.pi * freq * (wx * x_grid + wy * y_grid) + 
                self.phase[c] + phase_offset
            )
            wave_effect[c] = 0.1 * wave
            
        # Calculate interactions between channels
        channel_interaction = torch.einsum('ij,jkl->ikl', self.interaction, self.state)
        
        # Add feedback from previous state
        feedback_term = self.feedback.view(-1, 1, 1) * self.prev_state
        
        # Update state with diffusion, interactions and waves
        diffused = self.spatial_filter(self.state)
        state_update = (
            diffused + 
            channel_interaction + 
            wave_effect +
            feedback_term + 
            self.bias.view(-1, 1, 1)
        )
        
        # Apply non-linear activation with learnable steepness
        self.state = torch.tanh(self.activation * state_update)
        
    def get_image(self, width, height):
        # Ensure values are in [0,1] range
        img = (self.state + 1) * 0.5
        img = torch.clamp(img, 0, 1)
        
        # Resize to requested dimensions
        if self.grid_size != height or self.grid_size != width:
            img = torch.nn.functional.interpolate(img.unsqueeze(0), size=(height, width), 
                                                 mode='bilinear', align_corners=False).squeeze(0)
        
        # Return in [H, W, C] format
        return img.permute(1, 2, 0)
critique
NO, this dynamic system needs significant improvement. Let me provide a detailed critique:

1. Regarding vividness and color:
   - The generated images do have colorful elements, but they lack crispness and definition
   - The colors appear blurred and smeared together without clear boundaries
   - The vibrance is there, but without structure to make it visually meaningful

2. Regarding similarity to targets:
   - Target 1 shows a unicorn with rainbow elements and distinct features, but the odd cycles (1, 3, 5) show only vague pink/purple/blue color blobs without any recognizable unicorn shape
   - Target 2 shows a rainbow/prismatic landscape with green base, but even cycles (2, 4, 6) only capture the general blue/green/yellow color palette without the structural elements
   - The "Distance to Targets Over Steps" graph confirms this - there's cyclical behavior but the loss never gets particularly low for either target

3. Overall assessment:
   - The system does successfully oscillate between two different states
   - The loss graph shows the training converged, which is positive
   - However, the quality of reproduction is extremely poor - these are at best abstract color fields loosely inspired by the targets
   - The system captures only the broadest color distributions but none of the structural elements or details

This appears to be a very simple dynamic system that can capture basic color patterns but lacks the complexity needed to generate the detailed features in the target images. You would need a more sophisticated model with higher dimensional state representation to capture the detailed structures in the targets.
animation(tmp6znu0y58.mp4)
Download output
model_state(tmp14uty2nk.pt)
Download output
Preview of file unavailable
Download output
cycle_images
[
  "https://replicate.delivery/xezq/VueK5b8KfBsYNUxllHGlSJZ4fE0v2WcMKjjSbisB0QxOVsApA/tmpk5lcy2hy.png",
  "https://replicate.delivery/xezq/ReYhPotn5evmuk3qe7QStv7i8AycqeGtvtsGafECnhkfpiFIF/tmpd70wcw6r.png",
  "https://replicate.delivery/xezq/dEYfy3KfFRqqaEUXzPQN3LrGovGAnzV5xJlaI3FPm1LnKWgUA/tmp45wears8.png",
  "https://replicate.delivery/xezq/QUs8C4eSSEVER63ZGWy0xMTWcqadqdSnhmODNBDPxeXnKWgUA/tmp8v8da_ef.png",
  "https://replicate.delivery/xezq/J5TrtDV9ooZXLFJpQgV8sAmIYbWYWnh8BHLJ1BWK2A7piFIF/tmpd65ssnef.png",
  "https://replicate.delivery/xezq/BFRZGhImPJpOBN5FNZLCalyeaeVKBuvkyvwKAiaOkXLnKWgUA/tmp7i0a6vun.png"
]
visualization(tmprr5abrng.png)
Download output
optimizer_state(tmpgkm07hl4.pt)
Download output
Preview of file unavailable
Download output
historical_losses
[
  0.12666499614715576,
  0.1411130577325821,
  0.123764768242836,
  0.12161019444465637,
  0.1199849545955658,
  0.11803129315376282,
  0.12170319259166718,
  0.08971667289733887,
  0.07579156756401062,
  0.07163667678833008,
  0.06596732139587402,
  0.06747419387102127,
  0.07704654335975647,
  0.08119787275791168,
  0.07865089178085327,
  0.0716547742486,
  0.06249312311410904,
  0.05517628788948059,
  0.050741538405418396,
  0.05181753635406494,
  0.0541224330663681,
  0.053816720843315125,
  0.050614431500434875,
  0.046315476298332214,
  0.044002875685691833,
  0.043613284826278687,
  0.043761759996414185,
  0.04432697594165802,
  0.043627262115478516,
  0.04150910675525665,
  0.03962543606758118,
  0.034852758049964905,
  0.037527114152908325,
  0.048876985907554626,
  0.043850257992744446,
  0.030366376042366028,
  0.031566694378852844,
  0.034226834774017334,
  0.03305734694004059,
  0.02728457748889923,
  0.028389722108840942,
  0.03817082941532135,
  0.032919347286224365,
  0.022231489419937134,
  0.028281718492507935,
  0.032507359981536865,
  0.030147269368171692,
  0.022934123873710632,
  0.023630082607269287,
  0.030751585960388184,
  0.027689337730407715,
  0.02019841969013214,
  0.0202159583568573,
  0.02298635244369507,
  0.02090558409690857,
  0.01669439673423767,
  0.0171116441488266,
  0.018655985593795776,
  0.016467034816741943,
  0.015396833419799805,
  0.01675310730934143,
  0.015454709529876709,
  0.013570040464401245,
  0.01494482159614563,
  0.014364928007125854,
  0.012387663125991821,
  0.013045325875282288,
  0.012881994247436523,
  0.01234610378742218,
  0.012619942426681519,
  0.011992484331130981,
  0.012237876653671265,
  0.011895895004272461,
  0.012207940220832825,
  0.011885866522789001,
  0.012137085199356079,
  0.011699125170707703,
  0.011741980910301208,
  0.01131071150302887,
  0.011270135641098022,
  0.010994642972946167,
  0.01097753643989563,
  0.010660424828529358,
  0.010557636618614197,
  0.010688155889511108,
  0.012033790349960327,
  0.011458486318588257,
  0.010353267192840576,
  0.010810792446136475,
  0.00992538034915924,
  0.010432153940200806,
  0.009609267115592957,
  0.010019659996032715,
  0.009469091892242432,
  0.009812414646148682,
  0.009259253740310669,
  0.00939837098121643,
  0.00896066427230835,
  0.009160116314888,
  0.00875827670097351,
  0.008872762322425842,
  0.008558511734008789,
  0.008589208126068115,
  0.008397072553634644,
  0.008298933506011963,
  0.008295997977256775,
  0.008082389831542969,
  0.00809238851070404,
  0.007878541946411133,
  0.007892206311225891,
  0.007729098200798035,
  0.007715895771980286,
  0.007571294903755188,
  0.007514134049415588,
  0.007399439811706543,
  0.007322832942008972,
  0.007252246141433716,
  0.007148802280426025,
  0.007082164287567139,
  0.006956368684768677,
  0.006906464695930481,
  0.006797507405281067,
  0.006712168455123901,
  0.006641000509262085,
  0.006526842713356018,
  0.006456315517425537,
  0.006358355283737183,
  0.0062650591135025024,
  0.00619150698184967,
  0.006087824702262878,
  0.0060006678104400635,
  0.005909234285354614,
  0.005804404616355896,
  0.005718886852264404,
  0.00561976432800293,
  0.005516871809959412,
  0.005425944924354553,
  0.005322560667991638,
  0.005221635103225708,
  0.0051302313804626465,
  0.00503009557723999,
  0.004930362105369568,
  0.004839092493057251,
  0.004743173718452454,
  0.004644930362701416,
  0.004553809762001038,
  0.0044628530740737915,
  0.004367649555206299,
  0.0042746663093566895,
  0.004185006022453308,
  0.004093140363693237,
  0.003998681902885437,
  0.0039060860872268677,
  0.003815189003944397,
  0.003722548484802246,
  0.0036278367042541504,
  0.003532230854034424,
  0.0034371018409729004,
  0.003342047333717346,
  0.0032455772161483765,
  0.0031473934650421143,
  0.0030469000339508057,
  0.0029447972774505615,
  0.002840667963027954,
  0.0027348995208740234,
  0.002627149224281311,
  0.0025181472301483154,
  0.002408832311630249,
  0.0023041367530822754,
  0.0022199153900146484,
  0.0021957606077194214,
  0.0023967623710632324,
  0.0026453733444213867,
  0.001843094825744629,
  0.0015596747398376465,
  0.0016973167657852173,
  0.002234905958175659,
  0.0012674927711486816,
  0.0010945051908493042,
  0.0014987140893936157,
  0.0014584064483642578,
  0.0006192624568939209,
  0.0005590915679931641,
  0.0009069442749023438,
  0.0004767179489135742,
  -0.00007209181785583496,
  -0.00021922588348388672,
  -0.00006866455078125,
  -0.000021919608116149902,
  -0.0006359219551086426,
  -0.0010325908660888672,
  -0.0011619031429290771,
  -0.0010530799627304077,
  -0.0006805956363677979,
  -0.0017021149396896362,
  -0.0018919557332992554,
  -0.001388639211654663,
  -0.0016346275806427002,
  -0.002360612154006958,
  -0.0026371628046035767,
  -0.002772510051727295,
  -0.0026842057704925537,
  -0.0017146021127700806,
  -0.003272324800491333,
  -0.0012204945087432861,
  -0.0035430043935775757,
  -0.0012421905994415283,
  -0.0033625811338424683,
  0.00381644070148468,
  0.006192848086357117,
  -0.002988100051879883,
  0.004841849207878113,
  0.01262025535106659,
  0.0050936490297317505,
  -0.0034552067518234253,
  0.0033809691667556763,
  0.004214435815811157,
  -0.0029556751251220703,
  0.0035092532634735107,
  0.011759087443351746,
  0.006071910262107849,
  -0.0038707107305526733,
  0.006714120507240295,
  0.017359107732772827,
  0.012010455131530762,
  -0.0016583800315856934,
  0.0017048418521881104,
  0.012416794896125793,
  0.009996339678764343,
  -0.0009480118751525879,
  -0.00016611814498901367,
  0.007127478718757629,
  0.00455041229724884,
  -0.0031211823225021362,
  0.0024453401565551758,
  0.010130614042282104,
  0.006137490272521973,
  -0.0031899958848953247,
  0.0027441680431365967,
  0.012931540608406067,
  0.011016562581062317,
  -0.00014188885688781738,
  -0.002969026565551758,
  0.002528354525566101,
  0.0003887265920639038,
  -0.0047904253005981445,
  0.00007268786430358887,
  0.002664223313331604,
  -0.0024736225605010986,
  -0.003890708088874817,
  -0.00155600905418396,
  -0.004323214292526245,
  -0.003085613250732422,
  -0.000897139310836792,
  -0.00467275083065033,
  -0.002632319927215576,
  0.0005559176206588745,
  -0.003398701548576355,
  -0.0048342496156692505,
  -0.0030458271503448486,
  -0.005803778767585754,
  -0.002846077084541321,
  -0.0013127624988555908,
  -0.0054382383823394775,
  -0.0019684582948684692,
  0.002314448356628418,
  -0.002019941806793213,
  -0.006306305527687073,
  -0.004303842782974243,
  -0.006114736199378967,
  -0.0037367790937423706,
  -0.00202295184135437,
  -0.0061646997928619385,
  -0.0028862804174423218,
  0.0012230873107910156,
  -0.0031915605068206787,
  -0.0068527162075042725,
  -0.005191907286643982,
  -0.007178634405136108,
  -0.004971221089363098,
  -0.004834085702896118,
  -0.007623776793479919,
  -0.005311340093612671,
  -0.006711125373840332,
  -0.006489425897598267,
  -0.006048768758773804,
  -0.008025318384170532,
  -0.007048606872558594,
  -0.00828513503074646,
  -0.007799357175827026,
  -0.008228018879890442,
  -0.00861753523349762,
  -0.008546322584152222,
  -0.008371293544769287,
  -0.008825257420539856,
  -0.00902402400970459,
  -0.008927956223487854,
  -0.008821472525596619,
  -0.009214043617248535,
  -0.00945940613746643,
  -0.009475484490394592,
  -0.00942029058933258,
  -0.0094451904296875,
  -0.009738564491271973,
  -0.009973600506782532,
  -0.010048463940620422,
  -0.010042861104011536,
  -0.01008424162864685,
  -0.010246559977531433,
  -0.010386347770690918,
  -0.010462895035743713,
  -0.01048983633518219,
  -0.01050366461277008,
  -0.010550811886787415,
  -0.010652169585227966,
  -0.01077619194984436,
  -0.010885089635848999,
  -0.010949462652206421,
  -0.010981649160385132,
  -0.011017769575119019,
  -0.011089920997619629,
  -0.011187136173248291,
  -0.01128588616847992,
  -0.011363238096237183,
  -0.011418238282203674,
  -0.011469200253486633,
  -0.01153506338596344,
  -0.011611953377723694,
  -0.01168641448020935,
  -0.01174783706665039,
  -0.01180008053779602,
  -0.011854887008666992,
  -0.011916771531105042,
  -0.011979162693023682,
  -0.012033313512802124,
  -0.012080579996109009,
  -0.012129604816436768,
  -0.012183517217636108,
  -0.012237399816513062,
  -0.012285321950912476,
  -0.012328490614891052,
  -0.012372612953186035,
  -0.012419462203979492,
  -0.012465506792068481,
  -0.012507617473602295,
  -0.01254718005657196,
  -0.012587472796440125,
  -0.012629121541976929,
  -0.012670323252677917,
  -0.012709394097328186,
  -0.012746855616569519,
  -0.012784287333488464,
  -0.012822270393371582,
  -0.012860357761383057,
  -0.012897521257400513,
  -0.012933552265167236,
  -0.012968957424163818,
  -0.01300443708896637,
  -0.013040199875831604,
  -0.013075590133666992,
  -0.013110250234603882,
  -0.01314420998096466,
  -0.013178035616874695,
  -0.013212054967880249,
  -0.013245999813079834,
  -0.013279572129249573,
  -0.013312652707099915,
  -0.013345479965209961,
  -0.013378322124481201,
  -0.01341111958026886,
  -0.01344384253025055,
  -0.013476371765136719,
  -0.013508722186088562,
  -0.013541042804718018,
  -0.013573378324508667,
  -0.013605877757072449,
  -0.013638496398925781,
  -0.013671264052391052,
  -0.013704255223274231,
  -0.013737544417381287,
  -0.01377132534980774,
  -0.013805866241455078,
  -0.013841480016708374,
  -0.013878494501113892,
  -0.013917699456214905,
  -0.013959884643554688,
  -0.014005854725837708,
  -0.01405365765094757,
  -0.014096587896347046,
  -0.014140188694000244,
  -0.014189034700393677,
  -0.014228060841560364,
  -0.014270201325416565,
  -0.014310359954833984,
  -0.014342248439788818,
  -0.014381304383277893,
  -0.014413982629776001,
  -0.014453187584877014,
  -0.014485940337181091,
  -0.01452028751373291,
  -0.014551296830177307,
  -0.01458171010017395,
  -0.014609009027481079,
  -0.014629870653152466,
  -0.01464080810546875,
  -0.014629587531089783,
  -0.014576882123947144,
  -0.014427900314331055,
  -0.014118403196334839,
  -0.014763325452804565,
  -0.014794126152992249,
  -0.014363855123519897,
  -0.014292091131210327,
  -0.014889925718307495,
  -0.014867156744003296,
  -0.01433078944683075,
  -0.014686569571495056,
  -0.014805465936660767,
  -0.014807149767875671,
  -0.014692112803459167,
  -0.01450762152671814,
  -0.014973700046539307,
  -0.01511932909488678,
  -0.015067577362060547,
  -0.014854997396469116,
  -0.014254599809646606,
  -0.015178695321083069,
  -0.01456749439239502,
  -0.014624089002609253,
  -0.015229254961013794,
  -0.01494838297367096,
  -0.013834431767463684,
  -0.015033602714538574,
  -0.01138395071029663,
  -0.011148825287818909,
  -0.015302523970603943,
  -0.011931180953979492,
  -0.013609558343887329,
  -0.013536378741264343,
  -0.0129280686378479,
  -0.015237092971801758,
  -0.01455804705619812,
  -0.01489187777042389,
  -0.015415161848068237,
  -0.015069633722305298,
  -0.01415964961051941,
  -0.01519089937210083,
  -0.011509358882904053,
  -0.011136531829833984,
  -0.015411317348480225,
  -0.011795058846473694,
  -0.012388095259666443,
  -0.015193164348602295,
  -0.014537841081619263,
  -0.014971181750297546,
  -0.015528909862041473,
  -0.014806598424911499,
  -0.015398219227790833,
  -0.015560649335384369,
  -0.015249788761138916,
  -0.014632850885391235,
  -0.015651345252990723,
  -0.014385715126991272,
  -0.015468984842300415,
  -0.014251559972763062,
  -0.015489235520362854,
  -0.012283608317375183,
  -0.012951046228408813,
  -0.01513950526714325,
  -0.015271447598934174,
  -0.013753458857536316,
  -0.014161497354507446,
  -0.01447083055973053,
  -0.014966323971748352,
  -0.014148682355880737,
  -0.014384984970092773,
  -0.014352619647979736,
  -0.014551311731338501,
  -0.014604926109313965,
  -0.014980942010879517,
  -0.014214150607585907,
  -0.014281019568443298,
  -0.014982402324676514,
  -0.015268728137016296,
  -0.014052286744117737,
  -0.014261908829212189,
  -0.015039458870887756,
  -0.015284575521945953,
  -0.014053449034690857,
  -0.014301635324954987,
  -0.015241049230098724,
  -0.015412017703056335,
  -0.014167338609695435,
  -0.01445028930902481,
  -0.015212014317512512,
  -0.015403464436531067,
  -0.014209926128387451,
  -0.01444978266954422,
  -0.015416733920574188,
  -0.01558908075094223,
  -0.014100484549999237,
  -0.014400139451026917,
  -0.015470169484615326,
  -0.015635676681995392,
  -0.014166168868541718,
  -0.01443561166524887,
  -0.01557132601737976,
  -0.015720263123512268,
  -0.014211080968379974,
  -0.014510184526443481,
  -0.015585683286190033,
  -0.01574702560901642,
  -0.014293424785137177,
  -0.014549992978572845,
  -0.015676066279411316,
  -0.01583634316921234,
  -0.014287076890468597,
  -0.014567829668521881,
  -0.015730656683444977,
  -0.015894271433353424,
  -0.014297790825366974,
  -0.014540955424308777,
  -0.015846967697143555,
  -0.016001909971237183,
  -0.014319188892841339,
  -0.014646559953689575,
  -0.01582089066505432,
  -0.01596367359161377,
  -0.014517650008201599,
  -0.014792487025260925,
  -0.015836596488952637,
  -0.015987485647201538,
  -0.014618992805480957,
  -0.014885962009429932,
  -0.015868425369262695,
  -0.016033872961997986,
  -0.01464584469795227,
  -0.014886140823364258,
  -0.01596340537071228,
  -0.01612432301044464,
  -0.014614909887313843,
  -0.01486082375049591,
  -0.016059644520282745,
  -0.016218475997447968,
  -0.014553174376487732,
  -0.014790073037147522,
  -0.016180187463760376,
  -0.01632539927959442,
  -0.01485595852136612,
  -0.015382640063762665,
  -0.015652671456336975,
  -0.015680328011512756,
  -0.015787072479724884,
  -0.01609373837709427,
  -0.014978393912315369,
  -0.015097714960575104,
  -0.016234874725341797,
  -0.016413912177085876,
  -0.014866113662719727,
  -0.015313588082790375,
  -0.01600005477666855,
  -0.01610010862350464,
  -0.015499062836170197,
  -0.015809401869773865,
  -0.01569370925426483,
  -0.015831217169761658,
  -0.015906453132629395,
  -0.016143441200256348,
  -0.015419259667396545,
  -0.015604279935359955,
  -0.016166329383850098,
  -0.016361676156520844,
  -0.01521187275648117,
  -0.015401087701320648,
  -0.016395434737205505,
  -0.016560159623622894,
  -0.01503589004278183,
  -0.015289075672626495,
  -0.016493260860443115,
  -0.01663697510957718,
  -0.015096180140972137,
  -0.015378221869468689,
  -0.016504570841789246,
  -0.016631893813610077,
  -0.015260115265846252,
  -0.015517055988311768,
  -0.016519926488399506,
  -0.01666855812072754,
  -0.015317708253860474,
  -0.015560105443000793,
  -0.016583524644374847,
  -0.01672406494617462,
  -0.015365928411483765,
  -0.015608735382556915,
  -0.016647078096866608,
  -0.016789935529232025,
  -0.015376247465610504,
  -0.015618987381458282,
  -0.016722865402698517,
  -0.016862064599990845,
  -0.015380099415779114,
  -0.015615686774253845,
  -0.016810975968837738,
  -0.0169467031955719,
  -0.015605337917804718,
  -0.01602780818939209,
  -0.016502253711223602,
  -0.016573287546634674,
  -0.016269870102405548,
  -0.01655447483062744,
  -0.016135118901729584,
  -0.016269981861114502,
  -0.01662600040435791,
  -0.01682616025209427,
  -0.01593036949634552,
  -0.016124770045280457,
  -0.016794241964817047,
  -0.016963444650173187,
  -0.015852555632591248,
  -0.01607143133878708,
  -0.016897916793823242,
  -0.0170498788356781,
  -0.015865668654441833,
  -0.01609736680984497,
  -0.016943320631980896,
  -0.01710212230682373,
  -0.01591651141643524,
  -0.016162142157554626,
  -0.0169379860162735,
  -0.017110571265220642,
  -0.016073942184448242,
  -0.016263194382190704,
  -0.016912013292312622,
  -0.017092660069465637,
  -0.016224876046180725,
  -0.016460329294204712,
  -0.016728855669498444,
  -0.016963757574558258,
  -0.016467556357383728,
  -0.016619570553302765,
  -0.016730666160583496,
  -0.01697412133216858,
  -0.016711093485355377,
  -0.01691928505897522,
  -0.016426339745521545,
  -0.016645029187202454,
  -0.016945019364356995,
  -0.017138153314590454,
  -0.016537241637706757,
  -0.016671881079673767,
  -0.01714952290058136,
  -0.017364032566547394,
  -0.016231834888458252,
  -0.016385674476623535,
  -0.01738917827606201,
  -0.017553314566612244,
  -0.016301758587360382,
  -0.01656973361968994,
  -0.01735268533229828,
  -0.017492808401584625,
  -0.016493983566761017,
  -0.01672758162021637,
  -0.017361141741275787,
  -0.017504438757896423,
  -0.0166509747505188,
  -0.01686175912618637,
  -0.017413541674613953,
  -0.017571136355400085,
  -0.016660600900650024,
  -0.016879647970199585,
  -0.01749803125858307,
  -0.017638131976127625,
  -0.01671682298183441,
  -0.016933031380176544,
  -0.017571859061717987,
  -0.017714783549308777,
  -0.016753852367401123,
  -0.01698918640613556,
  -0.017619222402572632,
  -0.017753615975379944,
  -0.016844742000102997,
  -0.017062753438949585,
  -0.017671436071395874,
  -0.017811298370361328,
  -0.016925789415836334,
  -0.01715337485074997,
  -0.017704740166664124,
  -0.017847910523414612,
  -0.017012298107147217,
  -0.0172262042760849,
  -0.017756976187229156,
  -0.01790090650320053,
  -0.01708783209323883,
  -0.017302416265010834,
  -0.01780712604522705,
  -0.017954491078853607,
  -0.017146185040473938,
  -0.017357245087623596,
  -0.017870426177978516,
  -0.018015146255493164,
  -0.017201587557792664,
  -0.017409242689609528,
  -0.01793856918811798,
  -0.018083877861499786,
  -0.017246589064598083,
  -0.017458036541938782,
  -0.01800394058227539,
  -0.018146805465221405,
  -0.017297059297561646,
  -0.01750582456588745,
  -0.018072500824928284,
  -0.018213145434856415,
  -0.017350375652313232,
  -0.017563097178936005,
  -0.018133297562599182,
  -0.018272846937179565,
  -0.01740957796573639,
  -0.017623253166675568,
  -0.018191471695899963,
  -0.018328547477722168,
  -0.017478853464126587,
  -0.017691947519779205,
  -0.018245652318000793,
  -0.018382824957370758,
  -0.017550349235534668,
  -0.01776409149169922,
  -0.01829613745212555,
  -0.01843317598104477,
  -0.017625555396080017,
  -0.01783653348684311,
  -0.01834747940301895,
  -0.01848457008600235,
  -0.01770162582397461,
  -0.017911873757839203,
  -0.01839607208967209,
  -0.018533840775489807,
  -0.017778724431991577,
  -0.017987720668315887,
  -0.018443040549755096,
  -0.018580615520477295,
  -0.017860040068626404,
  -0.018067091703414917,
  -0.01848718523979187,
  -0.018625423312187195,
  -0.017943978309631348,
  -0.018149837851524353,
  -0.01852729171514511,
  -0.018666185438632965,
  -0.018031612038612366,
  -0.018234580755233765,
  -0.018565088510513306,
  -0.018704988062381744,
  -0.0181213840842247,
  -0.018321730196475983,
  -0.01859964430332184,
  -0.018741048872470856,
  -0.01821228861808777,
  -0.01840939372777939,
  -0.018631957471370697,
  -0.01877473294734955,
  -0.018304601311683655,
  -0.01849822700023651,
  -0.018661513924598694,
  -0.018806010484695435,
  -0.018398001790046692,
  -0.018588267266750336,
  -0.018687725067138672,
  -0.018833965063095093,
  -0.018492944538593292,
  -0.018679179251194,
  -0.01871095597743988,
  -0.018859148025512695,
  -0.018588826060295105,
  -0.018770985305309296,
  -0.01873055100440979,
  -0.018881037831306458,
  -0.018685534596443176,
  -0.018863119184970856,
  -0.018747225403785706,
  -0.018900111317634583,
  -0.018782474100589752,
  -0.01895540952682495,
  -0.01876060664653778,
  -0.018916018307209015,
  -0.018879666924476624,
  -0.019047684967517853,
  -0.018770530819892883,
  -0.018928632140159607,
  -0.018977120518684387,
  -0.019140079617500305,
  -0.018776744604110718,
  -0.01893763244152069,
  -0.019074946641921997,
  -0.019232667982578278,
  -0.01877860724925995,
  -0.01894250512123108,
  -0.019173167645931244,
  -0.01932547241449356,
  -0.018775999546051025,
  -0.018942996859550476,
  -0.019271790981292725,
  -0.01941843330860138,
  -0.01876823604106903,
  -0.01893850415945053,
  -0.01937093585729599,
  -0.019511617720127106,
  -0.01875504106283188,
  -0.018928758800029755,
  -0.019470572471618652,
  -0.019605189561843872,
  -0.018772363662719727,
  -0.01897817850112915,
  -0.01950802654027939,
  -0.019631095230579376,
  -0.01885797828435898,
  -0.01904742419719696,
  -0.019557975232601166,
  -0.01968790590763092,
  -0.018893122673034668,
  -0.019074514508247375,
  -0.019634008407592773,
  -0.019764333963394165,
  -0.01896882802248001,
  -0.01920311152935028,
  -0.019593417644500732,
  -0.019706428050994873,
  -0.019169464707374573,
  -0.01936512440443039,
  -0.019569575786590576,
  -0.019703835248947144,
  -0.019277654588222504,
  -0.019453346729278564,
  -0.01959618180990219,
  -0.01974046230316162,
  -0.019336402416229248,
  -0.019503779709339142,
  -0.019650466740131378,
  -0.019797690212726593,
  -0.019370771944522858,
  -0.019535623490810394,
  -0.019717589020729065,
  -0.01986459642648697,
  -0.019391663372516632,
  -0.019555553793907166,
  -0.019794225692749023,
  -0.019940055906772614,
  -0.019397884607315063,
  -0.01956161856651306,
  -0.019879981875419617,
  -0.02002359926700592,
  -0.019388258457183838,
  -0.019551873207092285,
  -0.019977062940597534,
  -0.020116202533245087,
  -0.019359998404979706,
  -0.01952652633190155,
  -0.020078107714653015,
  -0.020212166011333466,
  -0.01948940008878708,
  -0.019781380891799927,
  -0.01985519379377365,
  -0.01993531733751297,
  -0.019903093576431274,
  -0.020098567008972168,
  -0.01962520182132721,
  -0.019761309027671814,
  -0.020112834870815277,
  -0.020261377096176147,
  -0.01952601969242096,
  -0.01969442516565323,
  -0.020224519073963165,
  -0.02035363018512726,
  -0.019651077687740326,
  -0.019939199090003967,
  -0.020023778080940247,
  -0.020097829401493073,
  -0.02007291465997696,
  -0.020288556814193726,
  -0.019737333059310913,
  -0.01983511447906494,
  -0.02037312090396881,
  -0.020525604486465454,
  -0.019880928099155426,
  -0.020284436643123627,
  -0.01971590518951416,
  -0.019680708646774292,
  -0.020623445510864258,
  -0.020507335662841797,
  -0.02011185884475708,
  -0.020503051578998566,
  -0.019516289234161377,
  -0.01940511167049408,
  -0.02084040641784668,
  -0.020135536789894104,
  -0.02093443274497986,
  -0.02047751098871231,
  -0.020824439823627472,
  -0.020898684859275818,
  -0.020628303289413452,
  -0.0210067480802536,
  -0.020862415432929993,
  -0.020781755447387695,
  -0.02106255292892456,
  -0.02094881236553192,
  -0.020890414714813232,
  -0.021102190017700195,
  -0.021052181720733643,
  -0.020981892943382263,
  -0.021133534610271454,
  -0.02114035189151764,
  -0.021063849329948425,
  -0.021157443523406982,
  -0.021210499107837677,
  -0.02114710956811905,
  -0.02118605375289917,
  -0.021260373294353485,
  -0.02122553437948227,
  -0.021223604679107666,
  -0.021294526755809784,
  -0.021295711398124695,
  -0.021272800862789154,
  -0.021318882703781128,
  -0.021352067589759827,
  -0.021335557103157043,
  -0.021347373723983765,
  -0.021387532353401184,
  -0.02139744907617569,
  -0.02139192819595337,
  -0.021413832902908325,
  -0.021441981196403503,
  -0.021445035934448242,
  -0.021449655294418335,
  -0.021474294364452362,
  -0.02149173617362976,
  -0.021495066583156586,
  -0.021507911384105682,
  -0.021529018878936768,
  -0.02154006063938141,
  -0.02154705673456192,
  -0.021563082933425903,
  -0.021579772233963013,
  -0.021588660776615143,
  -0.021599173545837402,
  -0.021615520119667053,
  -0.021628670394420624,
  -0.02163798362016678,
  -0.021650828421115875,
  -0.021665707230567932,
  -0.021677076816558838,
  -0.021687708795070648,
  -0.021701306104660034,
  -0.021714642643928528,
  -0.021725594997406006,
  -0.021737322211265564,
  -0.02175077795982361,
  -0.02176307886838913,
  -0.021774277091026306,
  -0.021786659955978394,
  -0.02179960161447525,
  -0.02181144803762436,
  -0.021823078393936157,
  -0.021835647523403168,
  -0.021848157048225403,
  -0.021859891712665558,
  -0.02187187969684601,
  -0.021884411573410034,
  -0.02189663052558899,
  -0.021908484399318695,
  -0.02192065119743347,
  -0.021933116018772125,
  -0.021945230662822723,
  -0.021957233548164368,
  -0.021969527006149292,
  -0.021981894969940186,
  -0.021994054317474365,
  -0.022006206214427948,
  -0.022018566727638245,
  -0.022030934691429138,
  -0.02204316109418869,
  -0.02205546945333481,
  -0.02206788957118988,
  -0.022080302238464355,
  -0.02209267020225525,
  -0.022105097770690918,
  -0.022117622196674347,
  -0.022130146622657776,
  -0.022142626345157623,
  -0.022155210375785828,
  -0.022167861461639404,
  -0.02218054234981537,
  -0.022193193435668945,
  -0.02220594882965088,
  -0.022218763828277588,
  -0.02223159372806549,
  -0.02224445343017578,
  -0.02225738763809204,
  -0.022270411252975464,
  -0.022283457219600677,
  -0.022296547889709473,
  -0.022309720516204834
]
cycle_losses_target1
[
  0.1358814537525177,
  0.23682372272014618,
  0.11118382215499878,
  0.23416918516159058,
  0.22229433059692383,
  0.1737680286169052
]
cycle_losses_target2
[
  0.3061038851737976,
  0.11514925956726074,
  0.3049167990684509,
  0.08771948516368866,
  0.2762921452522278,
  0.20045599341392517
]
Generated in
Tweak itReport