Training in progress, step 30, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e3d252e57d25fd55cc6c135f9d0462441bdc56ecbab5859792ece52ecd4f625
+oid sha256:ca9f8c408985a41e46f49ceef2211c4ed59d9aaac07d6694fb94ba569a6460c0
 size 1572956792

last-checkpoint/global_step30/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:e1950b38f55041c0501446210e01d47350dfc4e666a68c2014e2ac30b586c688
+size 4718632304

last-checkpoint/global_step30/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:a190f9dce639e2356e2a8f1c6565700db5eabec3b6fa1185940f6c066a85a18d
+size 4718632432

last-checkpoint/global_step30/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1696327fc14811140114b5dffdbae6863bb05f043428602c5698598e5b2adb7
+size 1573384300

last-checkpoint/latest

@@ -1 +1 @@
-global_step20
+global_step30

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c2058a6b08b6b8fddd5fb26f48f2c04b4abd7596bbd7fd0f7dbc31810d6e6dd
+oid sha256:b59c5bc1e8c71589354c12a03089e1981f2c08909912d42022b93f822df32096
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.009342520144809062,
+  "epoch": 0.014013780217213593,
   "eval_steps": 500,
-  "global_step": 20,
+  "global_step": 30,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -308,6 +308,156 @@
       "rewards/margins": -0.08396562933921814,
       "rewards/rejected": -2.374962329864502,
       "step": 20
+    },
+    {
+      "epoch": 0.009809646152049515,
+      "grad_norm": 68.49879455566406,
+      "learning_rate": 3e-05,
+      "logits/chosen": -3.1144859790802,
+      "logits/rejected": -2.9034385681152344,
+      "logps/chosen": -216.12496948242188,
+      "logps/rejected": -190.55833435058594,
+      "loss": 1.5832,
+      "rewards/accuracies": 0.46875,
+      "rewards/chosen": -3.5558087825775146,
+      "rewards/margins": -0.48250633478164673,
+      "rewards/rejected": -3.0733022689819336,
+      "step": 21
+    },
+    {
+      "epoch": 0.010276772159289968,
+      "grad_norm": 61.08313751220703,
+      "learning_rate": 2.9999922925895862e-05,
+      "logits/chosen": -3.1215250492095947,
+      "logits/rejected": -2.597733974456787,
+      "logps/chosen": -219.85137939453125,
+      "logps/rejected": -196.4637908935547,
+      "loss": 1.4738,
+      "rewards/accuracies": 0.40625,
+      "rewards/chosen": -3.155580997467041,
+      "rewards/margins": -0.5727983117103577,
+      "rewards/rejected": -2.582782745361328,
+      "step": 22
+    },
+    {
+      "epoch": 0.010743898166530421,
+      "grad_norm": 41.851783752441406,
+      "learning_rate": 2.999969170437549e-05,
+      "logits/chosen": -3.0340187549591064,
+      "logits/rejected": -2.84521484375,
+      "logps/chosen": -177.37728881835938,
+      "logps/rejected": -183.36668395996094,
+      "loss": 1.0908,
+      "rewards/accuracies": 0.46875,
+      "rewards/chosen": -1.8575204610824585,
+      "rewards/margins": 0.3737794756889343,
+      "rewards/rejected": -2.231299877166748,
+      "step": 23
+    },
+    {
+      "epoch": 0.011211024173770874,
+      "grad_norm": 39.802371978759766,
+      "learning_rate": 2.9999306337815055e-05,
+      "logits/chosen": -2.8932125568389893,
+      "logits/rejected": -3.1646995544433594,
+      "logps/chosen": -176.2078399658203,
+      "logps/rejected": -199.5233612060547,
+      "loss": 1.0934,
+      "rewards/accuracies": 0.53125,
+      "rewards/chosen": -1.9304168224334717,
+      "rewards/margins": 0.2455454170703888,
+      "rewards/rejected": -2.175962448120117,
+      "step": 24
+    },
+    {
+      "epoch": 0.011678150181011327,
+      "grad_norm": 29.43556785583496,
+      "learning_rate": 2.999876683017479e-05,
+      "logits/chosen": -2.940061092376709,
+      "logits/rejected": -2.888075590133667,
+      "logps/chosen": -193.4315948486328,
+      "logps/rejected": -197.52081298828125,
+      "loss": 1.1539,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": -1.5905036926269531,
+      "rewards/margins": 0.9262561202049255,
+      "rewards/rejected": -2.5167598724365234,
+      "step": 25
+    },
+    {
+      "epoch": 0.01214527618825178,
+      "grad_norm": 54.44322204589844,
+      "learning_rate": 2.999807318699897e-05,
+      "logits/chosen": -2.6563143730163574,
+      "logits/rejected": -2.6990714073181152,
+      "logps/chosen": -156.08848571777344,
+      "logps/rejected": -182.99139404296875,
+      "loss": 1.2892,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": -1.465782642364502,
+      "rewards/margins": 0.17339667677879333,
+      "rewards/rejected": -1.6391793489456177,
+      "step": 26
+    },
+    {
+      "epoch": 0.012612402195492234,
+      "grad_norm": 33.6212272644043,
+      "learning_rate": 2.999722541541585e-05,
+      "logits/chosen": -2.61735200881958,
+      "logits/rejected": -2.4723963737487793,
+      "logps/chosen": -155.68017578125,
+      "logps/rejected": -158.19589233398438,
+      "loss": 1.0627,
+      "rewards/accuracies": 0.34375,
+      "rewards/chosen": -1.1587157249450684,
+      "rewards/margins": 0.1697230488061905,
+      "rewards/rejected": -1.3284387588500977,
+      "step": 27
+    },
+    {
+      "epoch": 0.013079528202732687,
+      "grad_norm": 19.592470169067383,
+      "learning_rate": 2.99962235241376e-05,
+      "logits/chosen": -2.5871520042419434,
+      "logits/rejected": -2.7644474506378174,
+      "logps/chosen": -184.37257385253906,
+      "logps/rejected": -169.0127716064453,
+      "loss": 0.8552,
+      "rewards/accuracies": 0.53125,
+      "rewards/chosen": -0.5363420248031616,
+      "rewards/margins": 0.4760582447052002,
+      "rewards/rejected": -1.0124002695083618,
+      "step": 28
+    },
+    {
+      "epoch": 0.01354665420997314,
+      "grad_norm": 20.270267486572266,
+      "learning_rate": 2.9995067523460198e-05,
+      "logits/chosen": -2.8748791217803955,
+      "logits/rejected": -3.1334304809570312,
+      "logps/chosen": -157.93455505371094,
+      "logps/rejected": -176.13912963867188,
+      "loss": 0.9041,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": -0.36636897921562195,
+      "rewards/margins": 0.27877098321914673,
+      "rewards/rejected": -0.6451399922370911,
+      "step": 29
+    },
+    {
+      "epoch": 0.014013780217213593,
+      "grad_norm": 43.62104415893555,
+      "learning_rate": 2.9993757425263343e-05,
+      "logits/chosen": -2.7570629119873047,
+      "logits/rejected": -2.771437644958496,
+      "logps/chosen": -157.29891967773438,
+      "logps/rejected": -166.71913146972656,
+      "loss": 1.5401,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": -0.970974862575531,
+      "rewards/margins": -0.46308204531669617,
+      "rewards/rejected": -0.5078927874565125,
+      "step": 30
     }
   ],
   "logging_steps": 1,