1 line
11 KiB
JSON
1 line
11 KiB
JSON
|
|
{"val-aux/sciknoweval/incorrect_format/best@8/std":0,"val-aux/num_turns/max":2,"self_distillation/feedback_used_fraction":0,"val-aux/sciknoweval/score/best@4/std":0.027759320424956824,"val-aux/sciknoweval/reward/best@2/mean":0.7809255319148936,"rollout_corr/rollout_log_ppl":0.15177409329044167,"val-aux/sciknoweval/incorrect_format/worst@4/std":0,"val-aux/sciknoweval/score/best@16/mean":0.813808510638298,"val-aux/sciknoweval/score/maj@4/std":0.031638291528035434,"val-aux/sciknoweval/score/maj@16/std":0.013016076917955125,"val-aux/sciknoweval/acc/worst@2/mean":0.7494468085106384,"val-aux/sciknoweval/score/maj@4/mean":0.764436170212766,"val-aux/sciknoweval/incorrect_format/worst@4/mean":1,"val-aux/sciknoweval/score/best@4/mean":0.7944468085106383,"response_length/clip_ratio":0,"self_distillation/token_reweight_eps_w":0.20000000000000004,"rollout_corr/log_ppl_diff_max":0.008212178945541382,"self_distillation/token_reweight_w_clipped_mean":0.9991688379086554,"val-aux/sciknoweval/score/maj@16/mean":0.7633617021276595,"timing_s/agent_loop/tool_calls/min":0,"critic/score/min":0,"critic/rewards/mean":0.67578125,"response_length_non_aborted/clip_ratio":0,"timing_s/save_checkpoint":12.537891793996096,"val-aux/sciknoweval/reward/maj@8/std":0.020384144360752216,"timing_s/agent_loop/slowest/generate_sequences":6.074218215420842,"global_seqlen/max":28053,"critic/advantages/min":-0.875,"val-aux/sciknoweval/acc/best@4/mean":0.7944468085106383,"self_distillation/reprompt_sample_fraction":0.6875,"val-core/sciknoweval/acc/best@16/mean":0.813808510638298,"val-aux/sciknoweval/acc/best@2/std":0.03462769102759328,"val-core/sciknoweval/acc/mean@16":0.7646276595744681,"rollout_corr/rollout_is_std":0.029670871794223785,"val-aux/sciknoweval/score/worst@2/mean":0.7494468085106384,"rollout_corr/log_ppl_diff":0.000486337645270396,"val-aux/sciknoweval/acc/worst@8/std":0.018341796156463988,"rollout_corr/log_ppl_abs_diff":0.0015185297597781755,"training/rollout_probs_diff_mean":0.002963797887787223,"critic/returns/mean":-0.0012260939693078399,"val-aux/sciknoweval/score/best@16/std":0.011791035148083585,"prompt_length/min":190,"val-aux/sciknoweval/reward/best@16/std":0.011791035148083585,"val-aux/sciknoweval/score/worst@16/mean":0.7182659574468084,"rollout_corr/log_ppl_diff_min":-0.004179961979389191,"val-aux/sciknoweval/reward/best@4/std":0.027759320424956824,"global_seqlen/balanced_max":23681,"timing_s/update_actor":23.16192770190537,"training/global_step":100,"self_distillation/token_reweight_w_mean":1.0763714532367885,"val-aux/sciknoweval/incorrect_format/maj@4/std":0,"self_distillation/empty_target_batch":0.3125,"val-core/sciknoweval/acc/best@16/std":0.011791035148083585,"critic/score/mean":0.67578125,"val-aux/sciknoweval/incorrect_format/best@16/mean":1,"_timestamp":1.7830844827750568e+09,"perf/max_memory_reserved_gb":140.86328125,"val-aux/sciknoweval/reward/maj@4/std":0.031638291528035434,"val-aux/sciknoweval/incorrect_format/std@16":0,"critic/rewards/max":1,"val-aux/sciknoweval/score/maj@2/mean":0.7651063829787235,"self_distillation/token_reweight_context_sample_fraction":0.6875,"val-aux/num_turns/min":2,"val-core/sciknoweval/acc/maj@16/std":0.013016076917955125,"global_seqlen/minmax_diff":7733,"response_length_non_aborted/mean":348.859375,"val-aux/sciknoweval/incorrect_format/maj@16/mean":1,"val-aux/sciknoweval/score/mean@16":0.7646276595744681,"rollout_corr/rollout_is_seq_max_deviation":0.005880475044250488,"global_seqlen/balanced_min":23678,"timing_s/agent_loop/generate_sequences/max":6.074218215420842,"rollout_corr/training_log_ppl":0.15226043093571207,"rollout_corr/rollout_is_ratio_fraction_low":0,"val-aux/sciknoweval/incorrect_format/best@2/mean":1,"timing_s/testing":40.37994212470949,"training/epoch":3,"self_distillation/success_group_fraction":0.6875,"actor/grad_norm":0.08581123873591423,"val-aux/sciknoweval/acc/std@16":0.0401737640202366,"val-aux/sciknoweval/incorrect_format/worst@16/std":0,"val-aux/sciknoweval/reward/maj@16/std":0.013016076917955125,"self_distillation/feedback_available_fraction":0,"perf/time_per_step":40.37044422328472
|