Upload exp_phase8_t3_rl_edit_tight_gspo_FIXED_s1_20260506_123025/log.jsonl with huggingface_hub
Browse files
exp_phase8_t3_rl_edit_tight_gspo_FIXED_s1_20260506_123025/log.jsonl
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"rl_step": 1, "loss": 4.470348358154297e-08, "mean_return": 43.53395080566406, "max_return": 60.37014389038086, "mean_motif_chained": 43.533950402632215, "gnorm": 0.0002727508544921875, "elapsed_s": 7.875335931777954, "cell_type": "Ex", "edit_type": "cell_type_transfer", "target_tf": "OLIG2", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -4.470348358154297e-08, "adv_std": 0.9999999403953552, "adv_min": -1.2214254140853882, "adv_max": 0.9708089232444763, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 43.533950402632215, "comp_kd_mean": 0.0, "comp_total_mean": 43.533950402632215, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 2 |
+
{"rl_step": 5, "loss": -2.9802322387695312e-08, "mean_return": 53.38246154785156, "max_return": 92.57649230957031, "mean_motif_chained": 53.38246198921896, "gnorm": 0.0004177093505859375, "elapsed_s": 35.48500442504883, "cell_type": "Ex", "edit_type": "cell_type_transfer", "target_tf": "SOX6", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 2.9802322387695312e-08, "adv_std": 0.9999999403953552, "adv_min": -0.9416700005531311, "adv_max": 1.371871829032898, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 53.38246198921896, "comp_kd_mean": 0.0, "comp_total_mean": 53.38246198921896, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 3 |
+
{"rl_step": 10, "loss": 1.7881393432617188e-07, "mean_return": 57.358150482177734, "max_return": 74.49144744873047, "mean_motif_chained": 57.358148742058674, "gnorm": 0.000286102294921875, "elapsed_s": 69.94463539123535, "cell_type": "Mic", "edit_type": "promoter_retarget", "target_tf": "RUNX1", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -1.7881393432617188e-07, "adv_std": 0.9999998807907104, "adv_min": -0.7957872748374939, "adv_max": 1.441671371459961, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 57.358148742058674, "comp_kd_mean": 0.0, "comp_total_mean": 57.358148742058674, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 4 |
+
{"rl_step": 15, "loss": -1.341104507446289e-07, "mean_return": 41.18315124511719, "max_return": 64.73605346679688, "mean_motif_chained": 41.1831540279867, "gnorm": 0.0003681182861328125, "elapsed_s": 107.19862771034241, "cell_type": "Mic", "edit_type": "activity_boost", "target_tf": "IRF8", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 1.341104507446289e-07, "adv_std": 1.0, "adv_min": -1.0109411478042603, "adv_max": 1.3203370571136475, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 41.1831540279867, "comp_kd_mean": 0.0, "comp_total_mean": 41.1831540279867, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 5 |
+
{"rl_step": 20, "loss": -1.4901161193847656e-08, "mean_return": 56.54112243652344, "max_return": 72.63003540039062, "mean_motif_chained": 56.54112256143473, "gnorm": 0.00023365020751953125, "elapsed_s": 144.1853425502777, "cell_type": "Ex", "edit_type": "activity_boost", "target_tf": "TBR1", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 1.4901161193847656e-08, "adv_std": 0.9999998807907104, "adv_min": -0.9874871373176575, "adv_max": 1.0929124355316162, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 56.54112256143473, "comp_kd_mean": 0.0, "comp_total_mean": 56.54112256143473, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 6 |
+
{"rl_step": 25, "loss": -2.9802322387695312e-08, "mean_return": 42.61048889160156, "max_return": 58.691070556640625, "mean_motif_chained": 42.610488937425934, "gnorm": 0.0002498626708984375, "elapsed_s": 180.23224759101868, "cell_type": "End", "edit_type": "cell_type_transfer", "target_tf": "RORB", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 2.9802322387695312e-08, "adv_std": 0.9999998807907104, "adv_min": -1.2856539487838745, "adv_max": 1.139735460281372, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 42.610488937425934, "comp_kd_mean": 0.0, "comp_total_mean": 42.610488937425934, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 7 |
+
{"rl_step": 30, "loss": -2.980232238769531e-07, "mean_return": 40.367958068847656, "max_return": 48.97187805175781, "mean_motif_chained": 40.367960541127715, "gnorm": 0.0003757476806640625, "elapsed_s": 216.33601307868958, "cell_type": "Mic", "edit_type": "promoter_retarget", "target_tf": "SPI1", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 2.980232238769531e-07, "adv_std": 0.9999998211860657, "adv_min": -1.2129565477371216, "adv_max": 1.1524946689605713, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 40.367960541127715, "comp_kd_mean": 0.0, "comp_total_mean": 40.367960541127715, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 8 |
+
{"rl_step": 35, "loss": 0.0, "mean_return": 41.823974609375, "max_return": 54.5073127746582, "mean_motif_chained": 41.82397488110995, "gnorm": 0.000255584716796875, "elapsed_s": 252.38608050346375, "cell_type": "Oli", "edit_type": "activity_boost", "target_tf": "OLIG2", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 0.0, "adv_std": 0.9999998211860657, "adv_min": -1.3842597007751465, "adv_max": 1.005548119544983, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 41.82397488110995, "comp_kd_mean": 0.0, "comp_total_mean": 41.82397488110995, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 9 |
+
{"rl_step": 40, "loss": -7.450580596923828e-08, "mean_return": 55.48974609375, "max_return": 83.69243621826172, "mean_motif_chained": 55.489748534383146, "gnorm": 0.0001983642578125, "elapsed_s": 288.6922183036804, "cell_type": "Ex", "edit_type": "promoter_retarget", "target_tf": "NEUROD2", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 7.450580596923828e-08, "adv_std": 1.0, "adv_min": -1.253061056137085, "adv_max": 1.0642616748809814, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 55.489748534383146, "comp_kd_mean": 0.0, "comp_total_mean": 55.489748534383146, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 10 |
+
{"rl_step": 45, "loss": -1.043081283569336e-07, "mean_return": 34.461036682128906, "max_return": 60.45305252075195, "mean_motif_chained": 34.46103859663317, "gnorm": 0.0003948211669921875, "elapsed_s": 325.1855092048645, "cell_type": "Oli", "edit_type": "cell_type_transfer", "target_tf": "MEF2C", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 1.043081283569336e-07, "adv_std": 0.9999998807907104, "adv_min": -0.8410735726356506, "adv_max": 1.34917414188385, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 34.46103859663317, "comp_kd_mean": 0.0, "comp_total_mean": 34.46103859663317, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 11 |
+
{"rl_step": 50, "loss": -5.960464477539063e-08, "mean_return": 59.62983322143555, "max_return": 79.74868774414062, "mean_motif_chained": 59.6298348352339, "gnorm": 0.00043487548828125, "elapsed_s": 361.577166557312, "cell_type": "Ex", "edit_type": "activity_boost", "target_tf": "MEF2C", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 5.960464477539063e-08, "adv_std": 1.0, "adv_min": -1.046620488166809, "adv_max": 1.2909044027328491, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 59.6298348352339, "comp_kd_mean": 0.0, "comp_total_mean": 59.6298348352339, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 12 |
+
{"rl_step": 55, "loss": 5.960464477539063e-08, "mean_return": 39.854896545410156, "max_return": 80.7538070678711, "mean_motif_chained": 39.85489595435867, "gnorm": 0.0003376007080078125, "elapsed_s": 398.0036880970001, "cell_type": "Oli", "edit_type": "activity_boost", "target_tf": "MYRF", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -5.960464477539063e-08, "adv_std": 0.9999998807907104, "adv_min": -0.6806336045265198, "adv_max": 1.485324740409851, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 39.85489595435867, "comp_kd_mean": 0.0, "comp_total_mean": 39.85489595435867, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 13 |
+
{"rl_step": 60, "loss": -2.9802322387695312e-08, "mean_return": 40.19462203979492, "max_return": 58.39018630981445, "mean_motif_chained": 40.19462262343721, "gnorm": 0.00029754638671875, "elapsed_s": 434.71662521362305, "cell_type": "Ex", "edit_type": "promoter_retarget", "target_tf": "MEF2C", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 2.9802322387695312e-08, "adv_std": 1.0, "adv_min": -1.1277062892913818, "adv_max": 1.2748939990997314, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 40.19462262343721, "comp_kd_mean": 0.0, "comp_total_mean": 40.19462262343721, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 14 |
+
{"rl_step": 65, "loss": -2.9802322387695312e-08, "mean_return": 40.81182098388672, "max_return": 64.96154022216797, "mean_motif_chained": 40.81182135568146, "gnorm": 0.0003261566162109375, "elapsed_s": 471.0648469924927, "cell_type": "Ast", "edit_type": "promoter_retarget", "target_tf": "AP1", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 2.9802322387695312e-08, "adv_std": 0.9999998211860657, "adv_min": -0.8274224400520325, "adv_max": 1.2032634019851685, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 40.81182135568146, "comp_kd_mean": 0.0, "comp_total_mean": 40.81182135568146, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 15 |
+
{"rl_step": 70, "loss": 1.4156103134155273e-07, "mean_return": 63.125396728515625, "max_return": 78.79481506347656, "mean_motif_chained": 63.125395065352905, "gnorm": 0.000270843505859375, "elapsed_s": 507.3769814968109, "cell_type": "In", "edit_type": "promoter_retarget", "target_tf": "DLX2", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -1.4156103134155273e-07, "adv_std": 0.9999999403953552, "adv_min": -1.2886886596679688, "adv_max": 1.0804476737976074, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 63.125395065352905, "comp_kd_mean": 0.0, "comp_total_mean": 63.125395065352905, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 16 |
+
{"rl_step": 75, "loss": 2.9802322387695312e-08, "mean_return": 53.28898620605469, "max_return": 91.96192932128906, "mean_motif_chained": 53.28898398998666, "gnorm": 0.0002536773681640625, "elapsed_s": 543.9116055965424, "cell_type": "OPC", "edit_type": "cell_type_transfer", "target_tf": "LHX6", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -2.9802322387695312e-08, "adv_std": 0.9999998211860657, "adv_min": -0.6980793476104736, "adv_max": 1.460349440574646, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 53.28898398998666, "comp_kd_mean": 0.0, "comp_total_mean": 53.28898398998666, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 17 |
+
{"rl_step": 80, "loss": -1.1920928955078125e-07, "mean_return": 47.64537048339844, "max_return": 58.64678955078125, "mean_motif_chained": 47.64537230561942, "gnorm": 0.0004558563232421875, "elapsed_s": 578.088757276535, "cell_type": "Mic", "edit_type": "promoter_retarget", "target_tf": "IRF8", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 1.1920928955078125e-07, "adv_std": 0.9999998807907104, "adv_min": -1.2026324272155762, "adv_max": 1.1631641387939453, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 47.64537230561942, "comp_kd_mean": 0.0, "comp_total_mean": 47.64537230561942, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 18 |
+
{"rl_step": 85, "loss": 1.4901161193847656e-07, "mean_return": 53.682594299316406, "max_return": 75.31249237060547, "mean_motif_chained": 53.682593190576426, "gnorm": 0.00046539306640625, "elapsed_s": 612.592129945755, "cell_type": "Ex", "edit_type": "promoter_retarget", "target_tf": "TBR1", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -1.4901161193847656e-07, "adv_std": 1.0, "adv_min": -0.8405500650405884, "adv_max": 1.4466736316680908, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 53.682593190576426, "comp_kd_mean": 0.0, "comp_total_mean": 53.682593190576426, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 19 |
+
{"rl_step": 90, "loss": -8.940696716308594e-08, "mean_return": 37.36669921875, "max_return": 47.160789489746094, "mean_motif_chained": 37.36670026318524, "gnorm": 0.0003566741943359375, "elapsed_s": 649.3024468421936, "cell_type": "Mic", "edit_type": "activity_boost", "target_tf": "SPI1", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 8.940696716308594e-08, "adv_std": 0.9999998807907104, "adv_min": -1.413169503211975, "adv_max": 0.7373439073562622, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 37.36670026318524, "comp_kd_mean": 0.0, "comp_total_mean": 37.36670026318524, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 20 |
+
{"rl_step": 95, "loss": 6.705522537231445e-08, "mean_return": 39.73131561279297, "max_return": 55.03954315185547, "mean_motif_chained": 39.73131465290243, "gnorm": 0.0002899169921875, "elapsed_s": 686.887937784195, "cell_type": "OPC", "edit_type": "cell_type_transfer", "target_tf": "DLX2", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": -6.705522537231445e-08, "adv_std": 0.9999998807907104, "adv_min": -1.3257139921188354, "adv_max": 1.1026084423065186, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 39.73131465290243, "comp_kd_mean": 0.0, "comp_total_mean": 39.73131465290243, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|
| 21 |
+
{"rl_step": 100, "loss": -5.960464477539063e-08, "mean_return": 64.7710952758789, "max_return": 97.39196014404297, "mean_motif_chained": 64.77109626684509, "gnorm": 0.000354766845703125, "elapsed_s": 723.3714847564697, "cell_type": "Oli", "edit_type": "promoter_retarget", "target_tf": "NKX2-2", "kd_active": false, "kd_tf": "", "kd_score_mean": 0.0, "ratio_mean": 1.0, "ratio_min": 1.0, "ratio_max": 1.0, "log_ratio_mean": 0.0, "adv_mean": 5.960464477539063e-08, "adv_std": 0.9999998807907104, "adv_min": -0.7337556481361389, "adv_max": 1.469343662261963, "comp_oracle_mean": 0.0, "comp_moods_chained_mean": 64.77109626684509, "comp_kd_mean": 0.0, "comp_total_mean": 64.77109626684509, "_algo": "gspo", "_lora_r": 16, "_peak_lr": 1e-05, "_temperature": 1.0, "_seq_len_mean": 360.0}
|