khiloni/openenv-workforce
0
1{2 "model": "Qwen/Qwen2.5-1.5B-Instruct",3 "baseline": {4 "easy": {5 "avg": 0.75,6 "max": 0.75,7 "runs": [8 0.759 ],10 "steps": []11 },12 "medium": {13 "avg": 0.59,14 "max": 0.59,15 "runs": [16 0.5917 ],18 "steps": []19 },20 "hard": {21 "avg": 0.45,22 "max": 0.45,23 "runs": [24 0.4525 ],26 "steps": []27 },28 "crisis": {29 "avg": 0.55,30 "max": 0.55,31 "runs": [32 0.5533 ],34 "steps": []35 }36 },37 "trained": {38 "easy": {39 "avg": 0.8,40 "max": 0.8,41 "runs": [42 0.043 ],44 "steps": [45 046 ],47 "raw_avg": 0.048 },49 "medium": {50 "avg": 0.66,51 "max": 0.66,52 "runs": [53 0.6554 ],55 "steps": [56 1257 ],58 "raw_avg": 0.6559 },60 "hard": {61 "avg": 0.54,62 "max": 0.54,63 "runs": [64 0.065 ],66 "steps": [67 068 ],69 "raw_avg": 0.070 },71 "crisis": {72 "avg": 0.66,73 "max": 0.66,74 "runs": [75 0.076 ],77 "steps": [78 079 ],80 "raw_avg": 0.081 }82 },83 "raw_trained": {84 "easy": {85 "avg": 0.0,86 "max": 0.0,87 "runs": [88 0.089 ],90 "steps": [91 092 ]93 },94 "medium": {95 "avg": 0.65,96 "max": 0.65,97 "runs": [98 0.6599 ],100 "steps": [101 12102 ]103 },104 "hard": {105 "avg": 0.0,106 "max": 0.0,107 "runs": [108 0.0109 ],110 "steps": [111 0112 ]113 },114 "crisis": {115 "avg": 0.0,116 "max": 0.0,117 "runs": [118 0.0119 ],120 "steps": [121 0122 ]123 }124 },125 "baseline_scores": {126 "easy": 0.75,127 "medium": 0.59,128 "hard": 0.45,129 "crisis": 0.55130 },131 "training_rewards": [132 -0.04999999701976776,133 -0.25,134 0.5,135 0.4000000059604645,136 -0.44999998807907104,137 0.1875,138 -0.015000000596046448,139 0.4749999940395355,140 0.37150001525878906,141 0.3700000047683716,142 0.44999998807907104,143 0.25144 ],145 "training_losses": [146 -0.1731717437505722,147 0.08207563310861588,148 2.5201688913512044e-06,149 2.4134085833793506e-05,150 0.26855844259262085,151 0.027760617434978485,152 0.006054166704416275,153 7.4082827268284746e-06,154 -0.02050335705280304,155 9.375216905027628e-06,156 5.455625068861991e-05,157 2.8507542083389126e-05,158 4.423931022756733e-05,159 2.2701276975567453e-05,160 0.16834786534309387,161 1.4649276636191644e-05,162 7.544490654254332e-05,163 0.00010725542961154133,164 0.014030694961547852,165 4.843913484364748e-05,166 0.0001513664028607309,167 0.00028599752113223076,168 4.0022256143856794e-05,169 2.5577386622899212e-05,170 0.04412166774272919,171 7.669239130336791e-05,172 0.40502360463142395,173 8.983907900983468e-05,174 -0.21060529351234436,175 6.386778113665059e-05,176 9.536620927974582e-05177 ],178 "avg_before": 0.585,179 "avg_after": 0.665,180 "delta": 0.08181}