Training in progress, epoch 22, checkpoint

last-checkpoint/global_step4336/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:122969b01761ed6b674a240a3433e44537a8fbb7f7b680d19211aedd554d48d2
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:627b2f0779529a60749eae5ad349a15086d1c58779dfc872a5451bc5ab500226
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f2af4720f840507596e015e20fc7cc1cb591bb20e79388b45f5f39bd33f8234
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70c67a73bb7e9537f7ba83cf013f913fe26adc83693252b080cb5bdcff894596
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fdda6bcc90558c85abff6693a008dfa23077f7f5f0b1545c046874c5591bf6d
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e29a77fdca370f5b14ff4efa6cbac041a0d1d7b330bda45b7bcadde108aedca6
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5c261d04d6bb797c064dc8574c8ee038f8e920fa3c4d3a75fa5467be7b934b4
+size 2506176112

last-checkpoint/global_step4336/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9edfbb323f7c40c47fdbd11271a00f0b39bae5f5d797e5bcbba152b3f4210465
+size 2506176112

last-checkpoint/global_step4336/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64755bc06cd33e1c88ff74f3c159857529029311c67288f5471d98eff348459c
+size 85570

last-checkpoint/global_step4336/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b58f9c8779ad8bfe590df7057bcde445a98a20b5776d73ff2987618cc374d8e
+size 85506

last-checkpoint/global_step4336/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f25315d367d54554dc801dfb165ddcab611fdade28d481ed0298099738d8a449
+size 85506

last-checkpoint/global_step4336/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f55554e29d3f9ae44cff8cc312e475f3b4120a2984728dcb04ba7c0b67af0a75
+size 85506

last-checkpoint/global_step4336/zero_pp_rank_4_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1936da3305b03057e3add1edec9f7ebab06e9ed9639b5abe4dee9db744d77f2f
+size 85506

last-checkpoint/global_step4336/zero_pp_rank_5_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1ff12d44664d8f6a572c90dddd61e29fe111388f59f740fd1d1294673fe3c4e
+size 85506

last-checkpoint/global_step4336/zero_pp_rank_6_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b476024c0562a709cc9cb0ff78f5aff769960f36401a6220235d64f50a8059b
+size 85506

last-checkpoint/global_step4336/zero_pp_rank_7_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2956087d52aaa66c7dc7bbbb48e07266a3e736f1446b04bd959db45d1579eb6d
+size 85506

last-checkpoint/latest

@@ -1 +1 @@
-global_step4138
+global_step4336

last-checkpoint/model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c3b64130a2b1328ad48739587bb6da6724b5b1efb9925e63b321bad9bfc0fc3
+oid sha256:da6a4c69beed061f5c9128d0fc377ffac0dd574b2fd31065d9fac99fd04c06f7
 size 4945242264

last-checkpoint/model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0679ad497e055b66159ad85f0138459c80b2a0410c5f8ecf9be82beb3ab863a5
+oid sha256:a92fd02d10813ce3cef6dbcbb14b9a87a8544f0439ddb53aa479c171a1393b3a
 size 67121608

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e130d1863d00ae1dd3fcda6965ee955535fc41fccbf0e8e5180feeae82230859
+oid sha256:6ffbd38f2626212a6ee67166039a4e916fc424a3ef78a112b81ecba34404f3d1
 size 15984

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:38a0bd73f24f92e9f35076e28562a135316e508c049328542ef38455ec5d9e27
+oid sha256:761cc4f38afe955241d2348fca72dd9656d9ebe889e2a62241d58bd4bc402ab3
 size 15984

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d61eb034a8cd4e97cff715cad1115d6f82a8d574822a796ff36bc37add8eb7c5
+oid sha256:df49e54f3e60ca7b97616f5b3ec776cb88dd34df15356bfe0565e073195eabb5
 size 15984

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9514ccff01a64d3814b885efa3d70c91036573ae2800e7f6e079eb57bb2c3a4
+oid sha256:3a34a02b568d14f0c9040613f420faf5a53edb4c287e30c4c7c5812d2c661932
 size 15984

last-checkpoint/rng_state_4.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0e7809a9b06e7f3cd7bdaa78797c848d3852c6fa406da741fe8f6ca848b06788
+oid sha256:dbb893a42ae156daa2dddcfcf2ef773c576ec3c274aaf9e5a216bd1d659a190b
 size 15984

last-checkpoint/rng_state_5.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f279045fb5076c6bded286c635cef1a4ed4a9eb0b04fdbdf9d7f0d8d58002fe
+oid sha256:b966d4dca42c479a7d82dd6ce6a32209ee7e19048adffff6ca6d5cbfafe89c52
 size 15984

last-checkpoint/rng_state_6.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:25937bc8ad7c01f12ab5bbb4608d4476d51c9a1009b84fc9c92f223235e135ad
+oid sha256:dfdec22eeccd05416d2a00acd09a775df394351b2aa4e3b1d32df8f7f4a8b07f
 size 15984

last-checkpoint/rng_state_7.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:683f3da1be97f6622bb41c0c819a35cc250ff0f8e5d0d0048c42366f7fc913e3
+oid sha256:89e852c764f530a6e44842564809b78387dfa34f7ec6fcece34742866c5d452e
 size 15984

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f86c9ec22746a9970fcd3444f420512e5e9006e0fba26ae30af3ab1b930074d4
+oid sha256:3a4a381db01b95441bf8a0f5bad46ab9935674f44c9e22dc758217b09cacfca8
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.99746835443038,
+  "epoch": 22.0,
   "eval_steps": 100,
-  "global_step": 4138,
+  "global_step": 4336,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -6217,6 +6217,306 @@
       "rewards/margins": 0.5419777631759644,
       "rewards/rejected": -0.34173792600631714,
       "step": 4130
+    },
+    {
+      "epoch": 21.00759493670886,
+      "grad_norm": 610764.2416650191,
+      "learning_rate": 2.3127546223754308e-07,
+      "logits/chosen": -0.7396095395088196,
+      "logits/rejected": -0.635381281375885,
+      "logps/chosen": -43.1786994934082,
+      "logps/rejected": -567.4010620117188,
+      "loss": 13604.9312,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19861166179180145,
+      "rewards/margins": 0.5279361605644226,
+      "rewards/rejected": -0.32932454347610474,
+      "step": 4140
+    },
+    {
+      "epoch": 21.058227848101264,
+      "grad_norm": 475061.8763801183,
+      "learning_rate": 2.3049200877467878e-07,
+      "logits/chosen": -0.8914744257926941,
+      "logits/rejected": -0.421735942363739,
+      "logps/chosen": -35.262672424316406,
+      "logps/rejected": -577.42919921875,
+      "loss": 14494.0125,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.1971307098865509,
+      "rewards/margins": 0.5426384210586548,
+      "rewards/rejected": -0.3455076515674591,
+      "step": 4150
+    },
+    {
+      "epoch": 21.10886075949367,
+      "grad_norm": 645277.3437759997,
+      "learning_rate": 2.2970855531181446e-07,
+      "logits/chosen": -0.8478671908378601,
+      "logits/rejected": -0.030678223818540573,
+      "logps/chosen": -41.63772964477539,
+      "logps/rejected": -578.0343017578125,
+      "loss": 14695.0688,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.2051258385181427,
+      "rewards/margins": 0.5436097979545593,
+      "rewards/rejected": -0.338483989238739,
+      "step": 4160
+    },
+    {
+      "epoch": 21.159493670886075,
+      "grad_norm": 585994.7230623597,
+      "learning_rate": 2.2892510184895017e-07,
+      "logits/chosen": -0.41463834047317505,
+      "logits/rejected": 0.02628953382372856,
+      "logps/chosen": -42.1683464050293,
+      "logps/rejected": -583.2857055664062,
+      "loss": 14024.5219,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19314590096473694,
+      "rewards/margins": 0.542751669883728,
+      "rewards/rejected": -0.3496057987213135,
+      "step": 4170
+    },
+    {
+      "epoch": 21.21012658227848,
+      "grad_norm": 484376.19543389586,
+      "learning_rate": 2.2814164838608585e-07,
+      "logits/chosen": -1.1306734085083008,
+      "logits/rejected": -1.8836634159088135,
+      "logps/chosen": -37.82279968261719,
+      "logps/rejected": -602.1395874023438,
+      "loss": 14403.9312,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20070350170135498,
+      "rewards/margins": 0.561983048915863,
+      "rewards/rejected": -0.36127954721450806,
+      "step": 4180
+    },
+    {
+      "epoch": 21.260759493670886,
+      "grad_norm": 1094104.6320160846,
+      "learning_rate": 2.2735819492322155e-07,
+      "logits/chosen": -1.173640251159668,
+      "logits/rejected": -1.2174631357192993,
+      "logps/chosen": -41.484169006347656,
+      "logps/rejected": -603.4227294921875,
+      "loss": 13658.0031,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20206312835216522,
+      "rewards/margins": 0.5591098070144653,
+      "rewards/rejected": -0.3570466637611389,
+      "step": 4190
+    },
+    {
+      "epoch": 21.31139240506329,
+      "grad_norm": 479054.6529765657,
+      "learning_rate": 2.2657474146035723e-07,
+      "logits/chosen": 0.22174246609210968,
+      "logits/rejected": 0.7805773615837097,
+      "logps/chosen": -36.59177780151367,
+      "logps/rejected": -585.4400634765625,
+      "loss": 13599.2641,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20017173886299133,
+      "rewards/margins": 0.5513723492622375,
+      "rewards/rejected": -0.3512006402015686,
+      "step": 4200
+    },
+    {
+      "epoch": 21.362025316455696,
+      "grad_norm": 397662.76305916836,
+      "learning_rate": 2.2579128799749294e-07,
+      "logits/chosen": -0.08221355825662613,
+      "logits/rejected": 0.6834055185317993,
+      "logps/chosen": -34.845909118652344,
+      "logps/rejected": -582.8208618164062,
+      "loss": 13710.1328,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19616791605949402,
+      "rewards/margins": 0.548448920249939,
+      "rewards/rejected": -0.35228094458580017,
+      "step": 4210
+    },
+    {
+      "epoch": 21.4126582278481,
+      "grad_norm": 558368.8389175668,
+      "learning_rate": 2.2500783453462862e-07,
+      "logits/chosen": -1.6244313716888428,
+      "logits/rejected": -1.1929272413253784,
+      "logps/chosen": -40.19694900512695,
+      "logps/rejected": -582.0238647460938,
+      "loss": 13864.4125,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.20411738753318787,
+      "rewards/margins": 0.5427777767181396,
+      "rewards/rejected": -0.3386603891849518,
+      "step": 4220
+    },
+    {
+      "epoch": 21.463291139240507,
+      "grad_norm": 677734.2581549523,
+      "learning_rate": 2.2422438107176433e-07,
+      "logits/chosen": -1.662096381187439,
+      "logits/rejected": -1.0047109127044678,
+      "logps/chosen": -35.72515106201172,
+      "logps/rejected": -602.2333374023438,
+      "loss": 14032.5453,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.2048303186893463,
+      "rewards/margins": 0.5686389207839966,
+      "rewards/rejected": -0.36380860209465027,
+      "step": 4230
+    },
+    {
+      "epoch": 21.513924050632912,
+      "grad_norm": 1168191.91590756,
+      "learning_rate": 2.234409276089e-07,
+      "logits/chosen": -1.4097559452056885,
+      "logits/rejected": -1.5273138284683228,
+      "logps/chosen": -41.850120544433594,
+      "logps/rejected": -581.159912109375,
+      "loss": 14159.3531,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.19850656390190125,
+      "rewards/margins": 0.5419961810112,
+      "rewards/rejected": -0.3434896469116211,
+      "step": 4240
+    },
+    {
+      "epoch": 21.564556962025318,
+      "grad_norm": 458950.46774975123,
+      "learning_rate": 2.226574741460357e-07,
+      "logits/chosen": 0.21833649277687073,
+      "logits/rejected": 0.5494757294654846,
+      "logps/chosen": -34.96293258666992,
+      "logps/rejected": -556.4089965820312,
+      "loss": 14176.0922,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.189616397023201,
+      "rewards/margins": 0.5252028107643127,
+      "rewards/rejected": -0.33558645844459534,
+      "step": 4250
+    },
+    {
+      "epoch": 21.615189873417723,
+      "grad_norm": 421239.1758950125,
+      "learning_rate": 2.218740206831714e-07,
+      "logits/chosen": -1.4874672889709473,
+      "logits/rejected": -0.608989417552948,
+      "logps/chosen": -41.988407135009766,
+      "logps/rejected": -571.3652954101562,
+      "loss": 13889.2,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.19781097769737244,
+      "rewards/margins": 0.534324586391449,
+      "rewards/rejected": -0.33651357889175415,
+      "step": 4260
+    },
+    {
+      "epoch": 21.665822784810125,
+      "grad_norm": 454253.5294425473,
+      "learning_rate": 2.2109056722030712e-07,
+      "logits/chosen": -2.9821505546569824,
+      "logits/rejected": -2.786928415298462,
+      "logps/chosen": -34.64727020263672,
+      "logps/rejected": -584.9675903320312,
+      "loss": 13704.1344,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20551709830760956,
+      "rewards/margins": 0.5548884868621826,
+      "rewards/rejected": -0.3493713140487671,
+      "step": 4270
+    },
+    {
+      "epoch": 21.71645569620253,
+      "grad_norm": 452082.6783455646,
+      "learning_rate": 2.203071137574428e-07,
+      "logits/chosen": 0.3382144868373871,
+      "logits/rejected": 0.4318400025367737,
+      "logps/chosen": -45.613182067871094,
+      "logps/rejected": -586.7680053710938,
+      "loss": 13852.6578,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.20163491368293762,
+      "rewards/margins": 0.5407856702804565,
+      "rewards/rejected": -0.3391507565975189,
+      "step": 4280
+    },
+    {
+      "epoch": 21.767088607594935,
+      "grad_norm": 514094.8538962621,
+      "learning_rate": 2.195236602945785e-07,
+      "logits/chosen": -0.4016874432563782,
+      "logits/rejected": 0.6381920576095581,
+      "logps/chosen": -43.911903381347656,
+      "logps/rejected": -575.604736328125,
+      "loss": 13636.45,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.19898727536201477,
+      "rewards/margins": 0.5374493598937988,
+      "rewards/rejected": -0.33846214413642883,
+      "step": 4290
+    },
+    {
+      "epoch": 21.81772151898734,
+      "grad_norm": 566656.255024603,
+      "learning_rate": 2.187402068317142e-07,
+      "logits/chosen": 1.3056986331939697,
+      "logits/rejected": 1.5420660972595215,
+      "logps/chosen": -26.3211669921875,
+      "logps/rejected": -560.0467529296875,
+      "loss": 13776.3141,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19464771449565887,
+      "rewards/margins": 0.5358431935310364,
+      "rewards/rejected": -0.3411955237388611,
+      "step": 4300
+    },
+    {
+      "epoch": 21.868354430379746,
+      "grad_norm": 615988.2584437894,
+      "learning_rate": 2.179567533688499e-07,
+      "logits/chosen": -1.382976770401001,
+      "logits/rejected": -1.09381902217865,
+      "logps/chosen": -39.40836715698242,
+      "logps/rejected": -589.1639404296875,
+      "loss": 14209.4484,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20153430104255676,
+      "rewards/margins": 0.5491803884506226,
+      "rewards/rejected": -0.3476461172103882,
+      "step": 4310
+    },
+    {
+      "epoch": 21.91898734177215,
+      "grad_norm": 537804.0354341647,
+      "learning_rate": 2.1717329990598557e-07,
+      "logits/chosen": -0.1827346831560135,
+      "logits/rejected": 0.2996447682380676,
+      "logps/chosen": -45.03169250488281,
+      "logps/rejected": -602.3880615234375,
+      "loss": 13524.8656,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.20245595276355743,
+      "rewards/margins": 0.5563360452651978,
+      "rewards/rejected": -0.3538800776004791,
+      "step": 4320
+    },
+    {
+      "epoch": 21.969620253164557,
+      "grad_norm": 1296539.8068217032,
+      "learning_rate": 2.1638984644312128e-07,
+      "logits/chosen": -1.9990384578704834,
+      "logits/rejected": -2.4059531688690186,
+      "logps/chosen": -43.4345588684082,
+      "logps/rejected": -604.287841796875,
+      "loss": 13463.9719,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.21004056930541992,
+      "rewards/margins": 0.5605840086936951,
+      "rewards/rejected": -0.35054340958595276,
+      "step": 4330
     }
   ],
   "logging_steps": 10,