Training in progress, epoch 4, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a6835e2ea0abd83fb5c0d5321f0bc3130e0335aae2b5383db169e2267aeb11a9
+oid sha256:353197bbe3f951952d2d6809feecb26d1a448bd0eac3711bb6db781ffb726eac
 size 1761071472

last-checkpoint/global_step1150/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54e062749ee2b3b23ac27231a97244fafd1b366a48babbd7c47f40f9c32ae2e8
+size 1327659941

last-checkpoint/global_step1150/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba6537dd75e372215aaf0c676577aaef014c9ba96df3d150209e05edd5c8edee
+size 1327660005

last-checkpoint/global_step1150/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bb4c78af53604cf1c2ab7bd6fe9cebc960b533b059b87af968f0b3ca1405a6d
+size 1327660005

last-checkpoint/global_step1150/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47da212a07ba477418d7733fe9b3b0c26679fdc177f4fd6adb3462f01f3edcef
+size 1327660005

last-checkpoint/global_step1150/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48aa696e1c33f11e21886743fe6f7639b5c4bc801ecfaeeab3f3246829ce102c
+size 1761341721

last-checkpoint/latest

@@ -1 +1 @@
-global_step1000
+global_step1150

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9424aa4ca4f5d5d7ed4b28ad103f73e5ec01fd92ffd10f2484182599afcc96b2
+oid sha256:0ff16acfda6bbc5cf50b99aa265031ad0b4a77a4ea996fe62e798c2b245c7d95
 size 15429

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:954f7886cbcde89a6b6186033056d5c3a9c094aa23ee419948f79eebeb5cd45a
+oid sha256:8a794a90df9c5b0631ebe2e7987dab57982f6a583f0010d55836d1074ee1a40d
 size 15429

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:329e7d395b6d8d6c6b83b6a3ac4c77be31bb3e8987830afc155d97e75103bb01
+oid sha256:bd6668393a5bc1b0de5d288b1691ef425f75f4c9702525195c76ff72f801ebb1
 size 15429

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:952f52236c0f7e07ba2124032aecfeb075608fba1d6e24c0a1939670b1b3d0ac
+oid sha256:f24d02dcb6989a54f9b8b9339108aafa5df8c99759987bdb33d7a9ef54169400
 size 15429

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:582b877be82ff9c13c06000e75cdcdb0df53e4eba3f609826f42d5175bddacf6
+oid sha256:ff8a318056a4dd572db7424ca40c1d5e6272063bff10a8e6b8a68fd386c676db
 size 1401

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
   "best_global_step": null,
-  "best_metric": 2.3136022718972526e-05,
+  "best_metric": 1.926410914165899e-05,
   "best_model_checkpoint": null,
-  "epoch": 3.5842293906810037,
+  "epoch": 4.121863799283154,
   "eval_steps": 50,
-  "global_step": 1000,
+  "global_step": 1150,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3328,6 +3328,504 @@
       "eval_samples_per_second": 55.234,
       "eval_steps_per_second": 1.862,
       "step": 1000
+    },
+    {
+      "epoch": 3.6021505376344085,
+      "grad_norm": 0.0003056343994103372,
+      "learning_rate": 6.333639026527406e-06,
+      "logits/chosen": -4.217955589294434,
+      "logits/rejected": -4.45806884765625,
+      "logps/chosen": -311.0265808105469,
+      "logps/rejected": -737.9935302734375,
+      "loss": 0.0003,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.399871826171875,
+      "rewards/margins": 47.924896240234375,
+      "rewards/rejected": -77.32476806640625,
+      "step": 1005
+    },
+    {
+      "epoch": 3.6200716845878134,
+      "grad_norm": 0.20004621148109436,
+      "learning_rate": 6.332183735178689e-06,
+      "logits/chosen": -4.212627410888672,
+      "logits/rejected": -4.426999092102051,
+      "logps/chosen": -272.0925598144531,
+      "logps/rejected": -717.3326416015625,
+      "loss": 0.0009,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -24.56581687927246,
+      "rewards/margins": 50.211341857910156,
+      "rewards/rejected": -74.77716064453125,
+      "step": 1010
+    },
+    {
+      "epoch": 3.6379928315412187,
+      "grad_norm": 0.008566437289118767,
+      "learning_rate": 6.330720641326227e-06,
+      "logits/chosen": -4.1919050216674805,
+      "logits/rejected": -4.385049343109131,
+      "logps/chosen": -232.0782012939453,
+      "logps/rejected": -636.6551513671875,
+      "loss": 0.0003,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -20.43191909790039,
+      "rewards/margins": 45.23147201538086,
+      "rewards/rejected": -65.66339111328125,
+      "step": 1015
+    },
+    {
+      "epoch": 3.6559139784946235,
+      "grad_norm": 0.012521971017122269,
+      "learning_rate": 6.3292497486533546e-06,
+      "logits/chosen": -4.2089738845825195,
+      "logits/rejected": -4.356890678405762,
+      "logps/chosen": -250.1702880859375,
+      "logps/rejected": -661.1627197265625,
+      "loss": 0.0001,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -22.228679656982422,
+      "rewards/margins": 46.03176498413086,
+      "rewards/rejected": -68.26045227050781,
+      "step": 1020
+    },
+    {
+      "epoch": 3.673835125448029,
+      "grad_norm": 0.010390724055469036,
+      "learning_rate": 6.327771060863043e-06,
+      "logits/chosen": -4.249519348144531,
+      "logits/rejected": -4.361356258392334,
+      "logps/chosen": -309.2774353027344,
+      "logps/rejected": -672.4700317382812,
+      "loss": 0.0002,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.068859100341797,
+      "rewards/margins": 40.909908294677734,
+      "rewards/rejected": -69.978759765625,
+      "step": 1025
+    },
+    {
+      "epoch": 3.6917562724014337,
+      "grad_norm": 0.006015332881361246,
+      "learning_rate": 6.326284581677888e-06,
+      "logits/chosen": -4.25397253036499,
+      "logits/rejected": -4.409300804138184,
+      "logps/chosen": -315.4617004394531,
+      "logps/rejected": -725.4534912109375,
+      "loss": 0.0001,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.897472381591797,
+      "rewards/margins": 46.05516815185547,
+      "rewards/rejected": -75.95263671875,
+      "step": 1030
+    },
+    {
+      "epoch": 3.709677419354839,
+      "grad_norm": 0.46787169575691223,
+      "learning_rate": 6.324790314840098e-06,
+      "logits/chosen": -4.260027885437012,
+      "logits/rejected": -4.418359279632568,
+      "logps/chosen": -335.4932861328125,
+      "logps/rejected": -802.9205322265625,
+      "loss": 0.0002,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -31.905109405517578,
+      "rewards/margins": 52.64836883544922,
+      "rewards/rejected": -84.55348205566406,
+      "step": 1035
+    },
+    {
+      "epoch": 3.727598566308244,
+      "grad_norm": 0.011817414313554764,
+      "learning_rate": 6.323288264111487e-06,
+      "logits/chosen": -4.29123067855835,
+      "logits/rejected": -4.481266021728516,
+      "logps/chosen": -315.4663391113281,
+      "logps/rejected": -852.9619140625,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.964025497436523,
+      "rewards/margins": 60.6983528137207,
+      "rewards/rejected": -90.6623764038086,
+      "step": 1040
+    },
+    {
+      "epoch": 3.7455197132616487,
+      "grad_norm": 0.00018397330131847411,
+      "learning_rate": 6.321778433273468e-06,
+      "logits/chosen": -4.287988662719727,
+      "logits/rejected": -4.4838433265686035,
+      "logps/chosen": -336.77215576171875,
+      "logps/rejected": -855.5325927734375,
+      "loss": 0.0004,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -32.178916931152344,
+      "rewards/margins": 58.75712966918945,
+      "rewards/rejected": -90.93604278564453,
+      "step": 1045
+    },
+    {
+      "epoch": 3.763440860215054,
+      "grad_norm": 0.007025348953902721,
+      "learning_rate": 6.3202608261270374e-06,
+      "logits/chosen": -4.282968997955322,
+      "logits/rejected": -4.504518985748291,
+      "logps/chosen": -317.99346923828125,
+      "logps/rejected": -874.0140380859375,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -30.056011199951172,
+      "rewards/margins": 62.95256805419922,
+      "rewards/rejected": -93.00858306884766,
+      "step": 1050
+    },
+    {
+      "epoch": 3.763440860215054,
+      "eval_logits/chosen": -4.260556697845459,
+      "eval_logits/rejected": -4.4869794845581055,
+      "eval_logps/chosen": -332.7008361816406,
+      "eval_logps/rejected": -877.4202880859375,
+      "eval_loss": 0.00016538219642825425,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/chosen": -31.80805778503418,
+      "eval_rewards/margins": 61.57281494140625,
+      "eval_rewards/rejected": -93.38087463378906,
+      "eval_runtime": 5.2799,
+      "eval_samples_per_second": 50.57,
+      "eval_steps_per_second": 1.705,
+      "step": 1050
+    },
+    {
+      "epoch": 3.781362007168459,
+      "grad_norm": 8.211416570702568e-05,
+      "learning_rate": 6.318735446492769e-06,
+      "logits/chosen": -4.2552900314331055,
+      "logits/rejected": -4.489185333251953,
+      "logps/chosen": -314.22955322265625,
+      "logps/rejected": -890.9273681640625,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.863494873046875,
+      "rewards/margins": 65.31141662597656,
+      "rewards/rejected": -95.1749038696289,
+      "step": 1055
+    },
+    {
+      "epoch": 3.7992831541218637,
+      "grad_norm": 1.5417823306052014e-05,
+      "learning_rate": 6.317202298210805e-06,
+      "logits/chosen": -4.282965660095215,
+      "logits/rejected": -4.5105462074279785,
+      "logps/chosen": -316.37835693359375,
+      "logps/rejected": -842.4830322265625,
+      "loss": 0.0038,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -30.062692642211914,
+      "rewards/margins": 59.639610290527344,
+      "rewards/rejected": -89.70230865478516,
+      "step": 1060
+    },
+    {
+      "epoch": 3.817204301075269,
+      "grad_norm": 0.0011977432295680046,
+      "learning_rate": 6.315661385140844e-06,
+      "logits/chosen": -4.225736141204834,
+      "logits/rejected": -4.4905266761779785,
+      "logps/chosen": -281.7426452636719,
+      "logps/rejected": -859.9666137695312,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -26.312814712524414,
+      "rewards/margins": 65.23667907714844,
+      "rewards/rejected": -91.54948425292969,
+      "step": 1065
+    },
+    {
+      "epoch": 3.835125448028674,
+      "grad_norm": 2.4491560907335952e-05,
+      "learning_rate": 6.314112711162133e-06,
+      "logits/chosen": -4.2156500816345215,
+      "logits/rejected": -4.521008014678955,
+      "logps/chosen": -271.5885314941406,
+      "logps/rejected": -847.3681640625,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -24.896705627441406,
+      "rewards/margins": 65.2315902709961,
+      "rewards/rejected": -90.1282958984375,
+      "step": 1070
+    },
+    {
+      "epoch": 3.8530465949820787,
+      "grad_norm": 3.7467976653715596e-05,
+      "learning_rate": 6.312556280173458e-06,
+      "logits/chosen": -4.194138526916504,
+      "logits/rejected": -4.490732669830322,
+      "logps/chosen": -262.0039978027344,
+      "logps/rejected": -836.509765625,
+      "loss": 0.002,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -23.526700973510742,
+      "rewards/margins": 65.3176040649414,
+      "rewards/rejected": -88.84429931640625,
+      "step": 1075
+    },
+    {
+      "epoch": 3.870967741935484,
+      "grad_norm": 0.00016050839622039348,
+      "learning_rate": 6.310992096093129e-06,
+      "logits/chosen": -4.209008693695068,
+      "logits/rejected": -4.46913480758667,
+      "logps/chosen": -303.6595153808594,
+      "logps/rejected": -873.8640747070312,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -28.550806045532227,
+      "rewards/margins": 64.56785583496094,
+      "rewards/rejected": -93.11866760253906,
+      "step": 1080
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 32.48484802246094,
+      "learning_rate": 6.309420162858981e-06,
+      "logits/chosen": -4.154443264007568,
+      "logits/rejected": -4.440248966217041,
+      "logps/chosen": -312.61505126953125,
+      "logps/rejected": -977.4017333984375,
+      "loss": 0.0051,
+      "rewards/accuracies": 0.9937499761581421,
+      "rewards/chosen": -29.2774658203125,
+      "rewards/margins": 75.44593811035156,
+      "rewards/rejected": -104.72340393066406,
+      "step": 1085
+    },
+    {
+      "epoch": 3.9068100358422937,
+      "grad_norm": 0.001882254146039486,
+      "learning_rate": 6.307840484428351e-06,
+      "logits/chosen": -4.128855228424072,
+      "logits/rejected": -4.383482933044434,
+      "logps/chosen": -347.81744384765625,
+      "logps/rejected": -889.56787109375,
+      "loss": 0.1115,
+      "rewards/accuracies": 0.9937499761581421,
+      "rewards/chosen": -33.82958984375,
+      "rewards/margins": 61.577186584472656,
+      "rewards/rejected": -95.40676879882812,
+      "step": 1090
+    },
+    {
+      "epoch": 3.924731182795699,
+      "grad_norm": 0.051850225776433945,
+      "learning_rate": 6.30625306477808e-06,
+      "logits/chosen": -4.084771156311035,
+      "logits/rejected": -4.34722375869751,
+      "logps/chosen": -317.49847412109375,
+      "logps/rejected": -800.1726684570312,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -30.48468589782715,
+      "rewards/margins": 54.538116455078125,
+      "rewards/rejected": -85.02281188964844,
+      "step": 1095
+    },
+    {
+      "epoch": 3.942652329749104,
+      "grad_norm": 0.0009036629926413298,
+      "learning_rate": 6.304657907904495e-06,
+      "logits/chosen": -4.059517860412598,
+      "logits/rejected": -4.292267799377441,
+      "logps/chosen": -307.61102294921875,
+      "logps/rejected": -788.194091796875,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.0938720703125,
+      "rewards/margins": 53.90114212036133,
+      "rewards/rejected": -82.9950180053711,
+      "step": 1100
+    },
+    {
+      "epoch": 3.942652329749104,
+      "eval_logits/chosen": -4.063472270965576,
+      "eval_logits/rejected": -4.276122093200684,
+      "eval_logps/chosen": -321.2622375488281,
+      "eval_logps/rejected": -787.795654296875,
+      "eval_loss": 0.00043061067117378116,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/chosen": -30.482948303222656,
+      "eval_rewards/margins": 52.51531982421875,
+      "eval_rewards/rejected": -82.99827575683594,
+      "eval_runtime": 4.8153,
+      "eval_samples_per_second": 55.448,
+      "eval_steps_per_second": 1.869,
+      "step": 1100
+    },
+    {
+      "epoch": 3.9605734767025087,
+      "grad_norm": 0.0006791537161916494,
+      "learning_rate": 6.303055017823399e-06,
+      "logits/chosen": -4.0789361000061035,
+      "logits/rejected": -4.320033073425293,
+      "logps/chosen": -300.1755065917969,
+      "logps/rejected": -787.130126953125,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -28.024511337280273,
+      "rewards/margins": 55.07674026489258,
+      "rewards/rejected": -83.10125732421875,
+      "step": 1105
+    },
+    {
+      "epoch": 3.978494623655914,
+      "grad_norm": 0.1369067132472992,
+      "learning_rate": 6.301444398570068e-06,
+      "logits/chosen": -4.058089256286621,
+      "logits/rejected": -4.238375186920166,
+      "logps/chosen": -315.3334655761719,
+      "logps/rejected": -788.9442138671875,
+      "loss": 0.0001,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -29.589107513427734,
+      "rewards/margins": 53.730003356933594,
+      "rewards/rejected": -83.3191146850586,
+      "step": 1110
+    },
+    {
+      "epoch": 3.996415770609319,
+      "grad_norm": 0.0027310196310281754,
+      "learning_rate": 6.299826054199234e-06,
+      "logits/chosen": -4.069396018981934,
+      "logits/rejected": -4.289578437805176,
+      "logps/chosen": -296.4201354980469,
+      "logps/rejected": -808.119384765625,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -27.70052719116211,
+      "rewards/margins": 57.1780891418457,
+      "rewards/rejected": -84.87861633300781,
+      "step": 1115
+    },
+    {
+      "epoch": 4.014336917562724,
+      "grad_norm": 0.11653987318277359,
+      "learning_rate": 6.298199988785076e-06,
+      "logits/chosen": -4.069671154022217,
+      "logits/rejected": -4.300382137298584,
+      "logps/chosen": -292.45477294921875,
+      "logps/rejected": -830.7077026367188,
+      "loss": 0.0006,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -27.467620849609375,
+      "rewards/margins": 60.38385009765625,
+      "rewards/rejected": -87.85147094726562,
+      "step": 1120
+    },
+    {
+      "epoch": 4.032258064516129,
+      "grad_norm": 2.3456966857793304e-07,
+      "learning_rate": 6.296566206421215e-06,
+      "logits/chosen": -4.103507995605469,
+      "logits/rejected": -4.336733341217041,
+      "logps/chosen": -338.1702880859375,
+      "logps/rejected": -885.9920654296875,
+      "loss": 0.0004,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -32.37504577636719,
+      "rewards/margins": 61.923828125,
+      "rewards/rejected": -94.29887390136719,
+      "step": 1125
+    },
+    {
+      "epoch": 4.050179211469534,
+      "grad_norm": 0.0002948185720015317,
+      "learning_rate": 6.294924711220691e-06,
+      "logits/chosen": -4.136487007141113,
+      "logits/rejected": -4.365674018859863,
+      "logps/chosen": -343.9933776855469,
+      "logps/rejected": -931.3427734375,
+      "loss": 0.0001,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -33.29169464111328,
+      "rewards/margins": 66.65242004394531,
+      "rewards/rejected": -99.9441146850586,
+      "step": 1130
+    },
+    {
+      "epoch": 4.068100358422939,
+      "grad_norm": 6.268872311920859e-06,
+      "learning_rate": 6.293275507315969e-06,
+      "logits/chosen": -4.170263290405273,
+      "logits/rejected": -4.390472412109375,
+      "logps/chosen": -342.8783264160156,
+      "logps/rejected": -962.1148681640625,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -33.151588439941406,
+      "rewards/margins": 70.10566711425781,
+      "rewards/rejected": -103.25726318359375,
+      "step": 1135
+    },
+    {
+      "epoch": 4.086021505376344,
+      "grad_norm": 6.805603334214538e-06,
+      "learning_rate": 6.2916185988589185e-06,
+      "logits/chosen": -4.181471824645996,
+      "logits/rejected": -4.400230884552002,
+      "logps/chosen": -403.2137756347656,
+      "logps/rejected": -1049.491455078125,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -39.57488250732422,
+      "rewards/margins": 73.46653747558594,
+      "rewards/rejected": -113.04142761230469,
+      "step": 1140
+    },
+    {
+      "epoch": 4.103942652329749,
+      "grad_norm": 0.6307092308998108,
+      "learning_rate": 6.289953990020803e-06,
+      "logits/chosen": -4.245871543884277,
+      "logits/rejected": -4.452054977416992,
+      "logps/chosen": -416.23065185546875,
+      "logps/rejected": -996.2029418945312,
+      "loss": 0.0003,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -41.286231994628906,
+      "rewards/margins": 65.87135314941406,
+      "rewards/rejected": -107.1575927734375,
+      "step": 1145
+    },
+    {
+      "epoch": 4.121863799283154,
+      "grad_norm": 0.16407367587089539,
+      "learning_rate": 6.288281684992274e-06,
+      "logits/chosen": -4.201587677001953,
+      "logits/rejected": -4.478522300720215,
+      "logps/chosen": -342.016845703125,
+      "logps/rejected": -968.5989379882812,
+      "loss": 0.0,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -32.899147033691406,
+      "rewards/margins": 70.87158966064453,
+      "rewards/rejected": -103.7707290649414,
+      "step": 1150
+    },
+    {
+      "epoch": 4.121863799283154,
+      "eval_logits/chosen": -4.203995227813721,
+      "eval_logits/rejected": -4.490315914154053,
+      "eval_logps/chosen": -316.3565368652344,
+      "eval_logps/rejected": -893.77880859375,
+      "eval_loss": 1.926410914165899e-05,
+      "eval_rewards/accuracies": 1.0,
+      "eval_rewards/chosen": -29.91464614868164,
+      "eval_rewards/margins": 65.36128997802734,
+      "eval_rewards/rejected": -95.27592468261719,
+      "eval_runtime": 4.7988,
+      "eval_samples_per_second": 55.638,
+      "eval_steps_per_second": 1.875,
+      "step": 1150
     }
   ],
   "logging_steps": 5,
@@ -3342,7 +3840,7 @@
         "early_stopping_threshold": 0.0001
       },
       "attributes": {
-        "early_stopping_patience_counter": 2
+        "early_stopping_patience_counter": 5
       }
     },
     "TrainerControl": {
@@ -3351,7 +3849,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }