You can deploy this model and start earning money today!

siddharthmb/
2026.mechaptcha.full-accuracy-target-90p-1p2m-20260525

Name: 2026.mechaptcha.full-accuracy-target-90p-1p2m-20260525
Availability: InStock

pytorch

captcha

image-classification

jacobcohen/mechaptcha

Views

No views yet

2026.mechaptcha.full-accuracy-target-90p-1p2m-20260525

CAPTCHA CNN trained on the Mechaptcha dataset.

Source Links

Dataset: jacobcohen/mechaptcha
Weights & Biases run: https://wandb.ai/siddharth-stanford/mechaptcha/runs/8jrjqy11

Validation Metrics

Metric	Value
`global_step`	`93760`
`val/final/char_accuracy`	`0.981692`
`val/final/exact_match`	`0.91802`
`val/final/failed_exact_count`	`4099.0`
`val/final/loss`	`0.0611491064453125`

Validation History

`epoch`	`global_step`	`val/char_accuracy`	`val/exact_match`	`val/loss`
`1`	`1000`	`0.562236`	`0.09014`	`1.4095021875`
`1`	`2000`	`0.74062`	`0.30046`	`0.891805078125`
`1`	`3000`	`0.80994`	`0.4285`	`0.673828671875`
`1`	`4000`	`0.853608`	`0.52996`	`0.5310894921875`
`1`	`4688`	`0.87148`	`0.57062`	`0.4664365625`
`2`	`5000`	`0.879512`	`0.5942`	`0.4412981640625`
`2`	`6000`	`0.897796`	`0.64186`	`0.3791025`
`2`	`7000`	`0.905644`	`0.66444`	`0.3449559765625`
`2`	`8000`	`0.916504`	`0.6954`	`0.305776796875`
`2`	`9000`	`0.923976`	`0.71752`	`0.27607462890625`
`2`	`9376`	`0.925404`	`0.7228`	`0.27132375`
`3`	`10000`	`0.929276`	`0.73526`	`0.2524692578125`
`3`	`11000`	`0.932984`	`0.74646`	`0.23880955078125`
`3`	`12000`	`0.937676`	`0.76142`	`0.22484921875`
`3`	`13000`	`0.93892`	`0.76646`	`0.2237569140625`
`3`	`14000`	`0.943236`	`0.77788`	`0.20522333984375`
`3`	`14064`	`0.943532`	`0.78006`	`0.2066888671875`
`4`	`15000`	`0.9468`	`0.7911`	`0.19536048828125`
`4`	`16000`	`0.94812`	`0.79506`	`0.1824358203125`
`4`	`17000`	`0.950128`	`0.80302`	`0.17641548828125`
`4`	`18000`	`0.951592`	`0.80766`	`0.17177212890625`
`4`	`18752`	`0.953336`	`0.8142`	`0.167141640625`
`5`	`19000`	`0.95414`	`0.81624`	`0.1638885546875`
`5`	`20000`	`0.95532`	`0.82046`	`0.158185732421875`
`5`	`21000`	`0.956512`	`0.82384`	`0.1546035546875`
`5`	`22000`	`0.957624`	`0.82768`	`0.15208203125`
`5`	`23000`	`0.959568`	`0.8354`	`0.142680166015625`
`5`	`23440`	`0.96042`	`0.83784`	`0.14071896484375`
`6`	`24000`	`0.959416`	`0.8347`	`0.14087556640625`
`6`	`25000`	`0.961208`	`0.8414`	`0.138833291015625`
`6`	`26000`	`0.961572`	`0.84156`	`0.136232578125`
`6`	`27000`	`0.962636`	`0.84522`	`0.130859365234375`
`6`	`28000`	`0.963896`	`0.85064`	`0.126852666015625`
`6`	`28128`	`0.96402`	`0.84964`	`0.1254619921875`
`7`	`29000`	`0.96406`	`0.85066`	`0.128651318359375`
`7`	`30000`	`0.9649`	`0.85332`	`0.122390654296875`
`7`	`31000`	`0.96522`	`0.85406`	`0.121330751953125`
`7`	`32000`	`0.966556`	`0.8598`	`0.117576416015625`
`7`	`32816`	`0.967288`	`0.86264`	`0.112649169921875`
`8`	`33000`	`0.967312`	`0.86258`	`0.114581064453125`
`8`	`34000`	`0.967588`	`0.86398`	`0.1124984765625`
`8`	`35000`	`0.967756`	`0.86366`	`0.11148232421875`
`8`	`36000`	`0.968692`	`0.86804`	`0.10813322265625`
`8`	`37000`	`0.9694`	`0.86984`	`0.105675234375`
`8`	`37504`	`0.968444`	`0.86602`	`0.109775673828125`
`9`	`38000`	`0.970072`	`0.87342`	`0.10428421875`
`9`	`39000`	`0.97012`	`0.87282`	`0.10288388671875`
`9`	`40000`	`0.970744`	`0.87534`	`0.10092134765625`
`9`	`41000`	`0.97028`	`0.8736`	`0.1011048046875`
`9`	`42000`	`0.970808`	`0.8769`	`0.10053736328125`
`9`	`42192`	`0.971252`	`0.87688`	`0.09793888671875`
`10`	`43000`	`0.971176`	`0.87654`	`0.098364306640625`
`10`	`44000`	`0.972344`	`0.88158`	`0.09464603515625`
`10`	`45000`	`0.97268`	`0.8828`	`0.09272423828125`
`10`	`46000`	`0.973108`	`0.8849`	`0.09314388671875`
`10`	`46880`	`0.973232`	`0.88466`	`0.092936044921875`
`11`	`47000`	`0.9729`	`0.88306`	`0.092447060546875`
`11`	`48000`	`0.9739`	`0.88784`	`0.09053125`
`11`	`49000`	`0.973016`	`0.8836`	`0.091995751953125`
`11`	`50000`	`0.974596`	`0.89014`	`0.086622470703125`
`11`	`51000`	`0.97472`	`0.8901`	`0.08586833984375`
`11`	`51568`	`0.97436`	`0.88892`	`0.085729599609375`
`12`	`52000`	`0.974992`	`0.89138`	`0.08468142578125`
`12`	`53000`	`0.975192`	`0.8931`	`0.08592841796875`
`12`	`54000`	`0.975384`	`0.893`	`0.08231384765625`
`12`	`55000`	`0.9753`	`0.89264`	`0.084165341796875`
`12`	`56000`	`0.9761`	`0.89636`	`0.0813374609375`
`12`	`56256`	`0.976632`	`0.8979`	`0.078527666015625`
`13`	`57000`	`0.976408`	`0.89746`	`0.08015087890625`
`13`	`58000`	`0.976228`	`0.89618`	`0.0808177783203125`
`13`	`59000`	`0.977384`	`0.9011`	`0.077632099609375`
`13`	`60000`	`0.977044`	`0.89968`	`0.0783234228515625`
`13`	`60944`	`0.977636`	`0.90274`	`0.0761007080078125`
`14`	`61000`	`0.97726`	`0.9007`	`0.07717640625`
`14`	`62000`	`0.977028`	`0.8997`	`0.0776199853515625`
`14`	`63000`	`0.976956`	`0.89972`	`0.0777383056640625`
`14`	`64000`	`0.978132`	`0.9039`	`0.07464880859375`
`14`	`65000`	`0.97794`	`0.90346`	`0.073996005859375`
`14`	`65632`	`0.978012`	`0.90422`	`0.073600361328125`
`15`	`66000`	`0.978612`	`0.906`	`0.0727809619140625`
`15`	`67000`	`0.97892`	`0.90736`	`0.0713900927734375`
`15`	`68000`	`0.979016`	`0.90794`	`0.070351611328125`
`15`	`69000`	`0.979048`	`0.90814`	`0.069944736328125`
`15`	`70000`	`0.978864`	`0.907`	`0.0716740771484375`
`15`	`70320`	`0.97868`	`0.90616`	`0.0723883349609375`
`16`	`71000`	`0.978892`	`0.9076`	`0.070965322265625`
`16`	`72000`	`0.979632`	`0.91016`	`0.0687335791015625`
`16`	`73000`	`0.979572`	`0.90942`	`0.068150078125`
`16`	`74000`	`0.979224`	`0.90836`	`0.069789130859375`
`16`	`75000`	`0.979952`	`0.9116`	`0.0666025048828125`
`16`	`75008`	`0.979988`	`0.91194`	`0.0669423974609375`
`17`	`76000`	`0.980156`	`0.9121`	`0.06650501953125`
`17`	`77000`	`0.980188`	`0.91214`	`0.0657359228515625`
`17`	`78000`	`0.980276`	`0.9126`	`0.0655079541015625`
`17`	`79000`	`0.98082`	`0.91512`	`0.064917314453125`
`17`	`79696`	`0.980404`	`0.913`	`0.064996337890625`
`18`	`80000`	`0.980668`	`0.91444`	`0.0643851025390625`
`18`	`81000`	`0.980592`	`0.9144`	`0.064393466796875`
`18`	`82000`	`0.980848`	`0.91506`	`0.063143486328125`
`18`	`83000`	`0.980892`	`0.91534`	`0.06378814453125`
`18`	`84000`	`0.981288`	`0.91632`	`0.0632607861328125`
`18`	`84384`	`0.98126`	`0.91636`	`0.0627262158203125`
`19`	`85000`	`0.980972`	`0.91478`	`0.063327744140625`
`19`	`86000`	`0.980832`	`0.9151`	`0.06304638671875`
`19`	`87000`	`0.980924`	`0.91546`	`0.0628027783203125`
`19`	`88000`	`0.981688`	`0.91856`	`0.06105244140625`
`19`	`89000`	`0.981444`	`0.91732`	`0.0618552978515625`
`19`	`89072`	`0.98164`	`0.91832`	`0.0613931005859375`
`20`	`90000`	`0.981544`	`0.9178`	`0.0606447265625`
`20`	`91000`	`0.981208`	`0.91672`	`0.06265763671875`
`20`	`92000`	`0.9818`	`0.91866`	`0.0608877294921875`
`20`	`93000`	`0.982044`	`0.92016`	`0.0601501513671875`
`20`	`93760`	`0.981692`	`0.91802`	`0.0611491064453125`

Training Summary

Global step: 93760
Best validation exact match: 0.92016

Configuration

The full training and model configuration is included in training_metadata.json.

siddharthmb/2026.mechaptcha.full-accuracy-target-90p-1p2m-20260525