1 line
11 KiB
JSON
1 line
11 KiB
JSON
|
|
{"rollout_corr/log_ppl_diff_min":-0.014117404818534851,"val-aux/sciknoweval/reward/maj@16/mean":0.7674333333333333,"self_distillation/correct_sample_fraction":0.6640625,"val-aux/sciknoweval/incorrect_format/worst@16/std":0.00755187697793512,"timing_s/old_log_prob":2.3089759554713964,"rollout_corr/chi2_token":0.0008595744147896767,"response_length/min":29,"val-aux/sciknoweval/acc/worst@2/mean":0.7324333333333333,"training/epoch":1,"perf/max_memory_reserved_gb":143.0078125,"timing_per_token_ms/adv":0.004661559897271101,"response_length/mean":61.05078125,"self_distillation/empty_target_batch":0.1640625,"val-aux/sciknoweval/score/maj@8/std":0.04659246363692196,"response_length_non_aborted/clip_ratio":0,"val-aux/sciknoweval/reward/best@8/mean":0.8498714285714286,"val-aux/sciknoweval/score/worst@16/std":0.022450981651084227,"val-core/sciknoweval/acc/best@16/std":0.022818983837850748,"val-aux/sciknoweval/score/worst@2/std":0.06710901416436343,"val-aux/sciknoweval/reward/maj@4/std":0.06469916434542414,"prompt_length/min":165,"val-aux/sciknoweval/reward/worst@16/std":0.022450981651084227,"timing_s/agent_loop/slowest/generate_sequences":1.1843877080827951,"timing_s/agent_loop/slowest/prompt_length":359,"val-aux/sciknoweval/incorrect_format/best@8/mean":1,"val-aux/sciknoweval/reward/maj@8/std":0.04659246363692196,"rollout_corr/rollout_is_max":1.6167027950286865,"rollout_corr/rollout_is_mean":1.0001804828643799,"val-aux/sciknoweval/reward/best@16/mean":0.8647904761904761,"val-aux/sciknoweval/score/worst@8/mean":0.6884666666666667,"prompt_length/mean":296.15625,"srpo/correct_sample_fraction":0.6640625,"perf/max_memory_allocated_gb":138.39261770248413,"val-aux/sciknoweval/reward/maj@8/mean":0.7679904761904761,"val-core/sciknoweval/acc/best@16/mean":0.8647904761904761,"val-aux/sciknoweval/reward/best@4/mean":0.828,"val-aux/sciknoweval/acc/best@2/mean":0.7995714285714286,"val-aux/sciknoweval/acc/std@16":0.08180825117524429,"val-aux/sciknoweval/reward/best@8/std":0.03842860858872272,"response_length_non_aborted/mean":61.05078125,"val-aux/sciknoweval/acc/best@8/std":0.03842860858872272,"prompt_length/clip_ratio":0,"timing_s/agent_loop/generate_sequences/min":0.32152901589870453,"val-aux/sciknoweval/incorrect_format/best@2/std":0.0016347712194112146,"srpo/grpo_sample_fraction":0.8203125,"val-aux/sciknoweval/reward/best@2/std":0.07120304368358653,"critic/advantages/min":-0.75,"_step":100,"val-aux/sciknoweval/reward/maj@2/std":0.08209907766512188,"val-aux/sciknoweval/incorrect_format/best@16/std":0,"self_distillation/reprompt_sample_fraction":0.8359375,"val-aux/sciknoweval/reward/worst@8/mean":0.6884666666666667,"val-aux/sciknoweval/score/best@8/mean":0.8498714285714286,"val-aux/sciknoweval/reward/best@16/std":0.022818983837850748,"global_seqlen/min":9718,"critic/score/max":1,"val-aux/sciknoweval/score/maj@2/mean":0.7658571428571429,"actor/entropy":0.07664129883050919,"val-aux/sciknoweval/score/worst@4/mean":0.7069809523809524,"global_seqlen/max":14844,"training/global_step":100,"critic/rewards/min":0,"timing_per_token_ms/gen":0.4833862506069128,"val-aux/num_turns/mean":2,"response_length_non_aborted/max":123,"timing_s/adv":0.4262763448059559,"val-aux/sciknoweval/score/maj@16/mean":0.7674333333333333,"val-aux/sciknoweval/score/worst@4/std":0.04900531157529151,"srpo/routed_token_count":61.05078125,"val-aux/sciknoweval/score/maj@4/mean":0.767504761904762,"val-aux/sciknoweval/acc/worst@4/std":0.04900531157529151,"val-aux/sciknoweval/score/std@16":0.08180825117524429,"actor/pg_loss":-0.056587886293527845,"critic/rewards/max":1,"val-aux/sciknoweval/acc/maj@4/mean":0.767504761904762,"val-aux/sciknoweval/incorrect_format/std@16":0.005316651201494241,"val-aux/sciknoweval/acc/best@2/std":0.07120304368358653,"rollout_corr/rollout_is_seq_std":0.003016919596120715,"srpo/sdpo_loss":0.006881217535010364,"timing_s/agent_loop/slowest/tool_calls":0,"val-aux/sciknoweval/incorrect_format/maj@16/std":0.0002604287412166727,"rollout_corr/chi2_seq":0.05469026346690953,"critic/score/min":0,"rollout_corr/k3_kl":0.00035141411022021965,"global_seqlen/balanced_max":11
|