amburger66's picture
LoRA fine-tune on RobotSmith task03 tool 12_long_2
f160d1d verified
{
"step": 1000,
"metrics": {
"eval_rew_align/success_auprc_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6195081082466317,
"eval_rew_align/positive_success_acc_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_rew_align/negative_success_acc_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.952161913523459,
"eval_rew_align/loss_amburger66_robotsmith_rbm_12_long_2_robotsmith": 3.8970178365707397,
"eval_rew_align/pearson_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.980936971524903,
"eval_p_rank/kendall_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/kendall_rewind_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/avg_succ_subopt_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/min_succ_subopt_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/max_succ_subopt_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/avg_subopt_fail_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/min_subopt_fail_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/max_subopt_fail_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/avg_succ_fail_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/min_succ_fail_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/max_succ_fail_diff_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/ranking_acc_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_all_pairs_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_failure_vs_suboptimal_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_suboptimal_vs_successful_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_failure_vs_successful_last_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/kendall_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/kendall_rewind_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/avg_succ_subopt_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/min_succ_subopt_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/max_succ_subopt_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/avg_subopt_fail_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/min_subopt_fail_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/max_subopt_fail_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/avg_succ_fail_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/min_succ_fail_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/max_succ_fail_diff_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/ranking_acc_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_all_pairs_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_failure_vs_suboptimal_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_suboptimal_vs_successful_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_failure_vs_successful_avg_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/kendall_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/kendall_rewind_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/avg_succ_subopt_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/min_succ_subopt_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/max_succ_subopt_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.22012983560562138,
"eval_p_rank/avg_subopt_fail_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/min_subopt_fail_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/max_subopt_fail_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.6151127217337489,
"eval_p_rank/avg_succ_fail_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/min_succ_fail_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/max_succ_fail_diff_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 0.8352425573393703,
"eval_p_rank/ranking_acc_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_all_pairs_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_failure_vs_suboptimal_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_suboptimal_vs_successful_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"eval_p_rank/ranking_acc_failure_vs_successful_sum_amburger66_robotsmith_rbm_12_long_2_robotsmith": 1.0,
"time/custom_evaluations": 42.60612651426345
}
}