Files
explore-tis-minp-40-8B/training_logs/20260612_111844_metrics_explore-tis-minp_736735.csv
ModelHub XC cab2418769 初始化项目,由ModelHub XC社区提供模型
Model: laion/explore-tis-minp-40-8B
Source: Original Platform
2026-08-09 12:53:53 +08:00

2.5 KiB

1async/discard_rateasync/discarded_countasync/effective_batch_groupsasync/effective_batch_samplesasync/staleness_maxasync/staleness_meanasync/staleness_minasync/staleness_ratiogenerate/avg_num_tokensgenerate/avg_tokens_non_zero_rewardsgenerate/avg_tokens_zero_rewardsgenerate/max_num_tokensgenerate/min_num_tokensgenerate/std_num_tokensgenerate/tis/aligned_tokensgenerate/tis/alignment_fail_countgenerate/tis/exact_match_fractiongenerate/tis/lcs_fallback_fractiongenerate/tis/lcs_fallback_messagesgenerate/tis/unaligned_fractionloss/avg_final_rewardsloss/avg_raw_advantagesloss/avg_raw_advantages_abspolicy/final_losspolicy/log_ratio_abs_maxpolicy/log_ratio_abs_meanpolicy/log_ratio_abs_p99policy/log_ratio_abs_pos00policy/log_ratio_abs_pos10policy/log_ratio_abs_pos20policy/log_ratio_abs_pos30policy/log_ratio_abs_pos40policy/log_ratio_abs_pos50policy/log_ratio_abs_pos60policy/log_ratio_abs_pos70policy/log_ratio_abs_pos80policy/log_ratio_abs_pos90policy/n_tokens_dp_gt_10pctpolicy/n_tokens_dp_gt_1pctpolicy/n_tokens_dp_gt_50pctpolicy/policy_entropypolicy/policy_losspolicy/policy_lrpolicy/policy_update_stepspolicy/ppo_clip_ratiopolicy/raw_grad_normpolicy/rollout_train_prob_diff_meanpolicy/rollout_train_prob_diff_stdpolicy/tis/imp_ratio_capped_fractionpolicy/tis/imp_ratio_meanpolicy/tis/log_ratio_abs_meanreward/avg_pass_at_8reward/avg_raw_rewardsystem/process_rss_gbsystem/process_vms_gbsystem/ram_available_gbsystem/ram_percentsystem/ram_total_gbsystem/ram_used_gbtiming/cleanup_old_checkpointstiming/compute_advantages_and_returnstiming/convert_to_training_inputtiming/fwd_logprobs_values_rewardtiming/policy_traintiming/run_trainingtiming/save_checkpointstiming/steptiming/sync_weightstiming/train_critic_and_policytiming/wait_for_generation_buffertis/batch_skipped_no_logprobstis/skipped_fractiontrainer/epochtrainer/global_stepbatch_errors/total_batchesbatch_errors/total_instancesbatch_errors/total_successfulbatch_errors/total_failedbatch_errors/total_maskedbatch_errors/avg_VerifierTimeoutErrorbatch_errors/total_VerifierTimeoutError
20.006451200.000.05063.84574657.29016406.5042258868383897.34461699520.057.00.96380.00.00.03620.7676-0.01180.10470.00.00.00.00.00.00.00.00.00.00.00.00.00.00.00.00.00.220.00.01.00.00.0150736.9062166815696.00.00.99630.01430.82810.767611.883650.8716576.68432.8857.9675281.283612.14850.10785.03440.6516198.9476239.913319.74731825.934724.537199.15341556.45010.00.018420160157100.153