Training in progress, epoch 20, checkpoint

last-checkpoint/global_step4138/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f42cf63864dbbb6e7e1a1826651f56af8056eb3a8fb816fb80ef20bc57c79e3
+size 2506176112

last-checkpoint/global_step4138/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2889e786be31e412a63d07934daa5cf75fd040a780d99b56fd6c2e296da81dc3
+size 2506176112

last-checkpoint/global_step4138/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:fa4db83b830290e632e020ab30cfed31b98a308ee23d1107edf89e523229709b
+size 2506176112

last-checkpoint/global_step4138/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b2b12e0220843c43e0d69cdee8ed2ec571d7660594bed3ec7776ed36ec129f3
+size 2506176112

last-checkpoint/global_step4138/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:a96cd0bf067a7c3073eddd7166af9e9a6bde6745137ee5ae012c36af3988d453
+size 2506176112

last-checkpoint/global_step4138/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54ab7742400c2d8245521b9d044e64583f52d0b3e13c3a119afb8dedc28850c0
+size 2506176112

last-checkpoint/global_step4138/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95d3a88dd3805d9d5b7f3d78fafdc0eca2463b386ef1788a2c669e04a85dd3a8
+size 2506176112

last-checkpoint/global_step4138/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:c4e57db8110b0a64ef4ea88bbe2a667a24c1763ec5f52e4a38abc50f7849da46
+size 2506176112

last-checkpoint/global_step4138/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:648f988c5b77d52514ff06912da1cc751ce1aa7f8fa4a6605cc6e10aca79bd6a
+size 85570

last-checkpoint/global_step4138/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c469111cc0bdbba8b56da5bab8e2793d3c6ca36b92da21d77bf1fe0e3c54b44c
+size 85506

last-checkpoint/global_step4138/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c720b81e14bbd833f0b7e4c9fe0c9c93b8e3979e5387c885a5fd9e29516b33c
+size 85506

last-checkpoint/global_step4138/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe59c58af131adc9850352833d7483fdb03834096feca8fff8d4cf7f36ed62bf
+size 85506

last-checkpoint/global_step4138/zero_pp_rank_4_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc003fd4cbd85685134bc73c75fe27fd4eee9abdffa23de1b38788474c8ed7d9
+size 85506

last-checkpoint/global_step4138/zero_pp_rank_5_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce36a7052d23544660d68f11d0c54f17ece7b88b20ab5bda1d89ea91acd1b00c
+size 85506

last-checkpoint/global_step4138/zero_pp_rank_6_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fea18a287f5e8ee9fc2777d10b3097d4b10c293255c278b38921ad5961d823e
+size 85506

last-checkpoint/global_step4138/zero_pp_rank_7_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:221a779a693abfb1ebd2875edb80212e4ed658aa72d0b37665bc9bfe86dbc8c9
+size 85506

last-checkpoint/latest

@@ -1 +1 @@
-global_step3941
+global_step4138

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:acdc933f849726cf9f6626987c11cea6a7678ccfd80dc2d0483bec3e83e0c3a7
+oid sha256:0c3b64130a2b1328ad48739587bb6da6724b5b1efb9925e63b321bad9bfc0fc3
 size 4945242264

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b099839b0f1c5f0d5fb18759ba02999fd787c83dda2d02d2ba36941c44a246b
+oid sha256:0679ad497e055b66159ad85f0138459c80b2a0410c5f8ecf9be82beb3ab863a5
 size 67121608

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f8f81f81a21cdf7d6a9d642f67427b821e87802eb5a4ca4a5c038480e9a673f
+oid sha256:e130d1863d00ae1dd3fcda6965ee955535fc41fccbf0e8e5180feeae82230859
 size 15984

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c967add958d82a4ef6285ec19e0e5a560f82b1ea3488260d2539d200cb5f199b
+oid sha256:38a0bd73f24f92e9f35076e28562a135316e508c049328542ef38455ec5d9e27
 size 15984

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1a5e30f320a5f3e4a4760af275c4a2bb9fdcfc6c661a3757da3b20b4c29a87fc
+oid sha256:d61eb034a8cd4e97cff715cad1115d6f82a8d574822a796ff36bc37add8eb7c5
 size 15984

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3411b849d13fdd9e668c8cc5fafa8402de13fd4a2d5761eaf33f2791780721e4
+oid sha256:b9514ccff01a64d3814b885efa3d70c91036573ae2800e7f6e079eb57bb2c3a4
 size 15984

last-checkpoint/rng_state_4.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1765109deada93bbc4f9e38b5c3c6bbd9afe9a12839bbbd72227ca7ff46467b1
+oid sha256:0e7809a9b06e7f3cd7bdaa78797c848d3852c6fa406da741fe8f6ca848b06788
 size 15984

last-checkpoint/rng_state_5.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6147f79a889f2906a81eea5c06f72acf722f674546fd6be8432d3d70a04392a8
+oid sha256:4f279045fb5076c6bded286c635cef1a4ed4a9eb0b04fdbdf9d7f0d8d58002fe
 size 15984

last-checkpoint/rng_state_6.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a9f5815a19b60dc1bb303c451b515a4ec523fa0faf1c87c69b03ad4032c4baa2
+oid sha256:25937bc8ad7c01f12ab5bbb4608d4476d51c9a1009b84fc9c92f223235e135ad
 size 15984

last-checkpoint/rng_state_7.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:395da160bbc37f8049700e0d82d501cbd1effc732719599c0cd143db7892910e
+oid sha256:683f3da1be97f6622bb41c0c819a35cc250ff0f8e5d0d0048c42366f7fc913e3
 size 15984

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9dd8e459c5f044a7182784611ab9373a71c50f4a912290b4cfdcb531e69b1b46
+oid sha256:f86c9ec22746a9970fcd3444f420512e5e9006e0fba26ae30af3ab1b930074d4
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 20.99746835443038,
   "eval_steps": 100,
-  "global_step": 3941,
+  "global_step": 4138,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -5932,6 +5932,291 @@
       "rewards/margins": 0.5218333005905151,
       "rewards/rejected": -0.3395631015300751,
       "step": 3940
+    },
+    {
+      "epoch": 20.045569620253165,
+      "grad_norm": 585485.6374993185,
+      "learning_rate": 2.461610780319649e-07,
+      "logits/chosen": -0.45847588777542114,
+      "logits/rejected": -0.7163432836532593,
+      "logps/chosen": -52.16509246826172,
+      "logps/rejected": -574.7341918945312,
+      "loss": 14472.725,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.18465857207775116,
+      "rewards/margins": 0.5204340219497681,
+      "rewards/rejected": -0.3357754647731781,
+      "step": 3950
+    },
+    {
+      "epoch": 20.09620253164557,
+      "grad_norm": 732893.4730793714,
+      "learning_rate": 2.453776245691006e-07,
+      "logits/chosen": -0.8532247543334961,
+      "logits/rejected": -1.0177998542785645,
+      "logps/chosen": -34.215248107910156,
+      "logps/rejected": -545.2357788085938,
+      "loss": 14258.1938,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.1821087896823883,
+      "rewards/margins": 0.5106922388076782,
+      "rewards/rejected": -0.3285834789276123,
+      "step": 3960
+    },
+    {
+      "epoch": 20.146835443037975,
+      "grad_norm": 566993.8027250646,
+      "learning_rate": 2.445941711062363e-07,
+      "logits/chosen": 0.08969805389642715,
+      "logits/rejected": 0.7759960889816284,
+      "logps/chosen": -44.99492645263672,
+      "logps/rejected": -555.2744140625,
+      "loss": 13412.7156,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.1912485808134079,
+      "rewards/margins": 0.5124030113220215,
+      "rewards/rejected": -0.3211544454097748,
+      "step": 3970
+    },
+    {
+      "epoch": 20.19746835443038,
+      "grad_norm": 694029.5528637858,
+      "learning_rate": 2.4381071764337194e-07,
+      "logits/chosen": -1.096225380897522,
+      "logits/rejected": -0.07990212738513947,
+      "logps/chosen": -50.788089752197266,
+      "logps/rejected": -593.2210083007812,
+      "loss": 14393.8781,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20465774834156036,
+      "rewards/margins": 0.5500935316085815,
+      "rewards/rejected": -0.34543576836586,
+      "step": 3980
+    },
+    {
+      "epoch": 20.248101265822786,
+      "grad_norm": 732062.5147915592,
+      "learning_rate": 2.430272641805077e-07,
+      "logits/chosen": 2.356168270111084,
+      "logits/rejected": 2.851551055908203,
+      "logps/chosen": -47.21477508544922,
+      "logps/rejected": -550.4632568359375,
+      "loss": 14544.5953,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.1797908991575241,
+      "rewards/margins": 0.5041374564170837,
+      "rewards/rejected": -0.3243466317653656,
+      "step": 3990
+    },
+    {
+      "epoch": 20.29873417721519,
+      "grad_norm": 679068.526799364,
+      "learning_rate": 2.4224381071764335e-07,
+      "logits/chosen": -0.03650767728686333,
+      "logits/rejected": 0.4142078459262848,
+      "logps/chosen": -56.95387649536133,
+      "logps/rejected": -573.4332885742188,
+      "loss": 14252.7656,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19543471932411194,
+      "rewards/margins": 0.5279492139816284,
+      "rewards/rejected": -0.3325144648551941,
+      "step": 4000
+    },
+    {
+      "epoch": 20.349367088607593,
+      "grad_norm": 654544.0953281225,
+      "learning_rate": 2.4146035725477906e-07,
+      "logits/chosen": -1.5134330987930298,
+      "logits/rejected": -1.2428243160247803,
+      "logps/chosen": -38.626304626464844,
+      "logps/rejected": -569.9034423828125,
+      "loss": 14105.2687,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.19262897968292236,
+      "rewards/margins": 0.5324395298957825,
+      "rewards/rejected": -0.3398105204105377,
+      "step": 4010
+    },
+    {
+      "epoch": 20.4,
+      "grad_norm": 497489.50957681873,
+      "learning_rate": 2.4067690379191476e-07,
+      "logits/chosen": -0.28921595215797424,
+      "logits/rejected": 0.4628073573112488,
+      "logps/chosen": -35.01650619506836,
+      "logps/rejected": -566.645263671875,
+      "loss": 14669.4656,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19555968046188354,
+      "rewards/margins": 0.5380457639694214,
+      "rewards/rejected": -0.3424859941005707,
+      "step": 4020
+    },
+    {
+      "epoch": 20.450632911392404,
+      "grad_norm": 813543.9506414626,
+      "learning_rate": 2.3989345032905047e-07,
+      "logits/chosen": 0.39950722455978394,
+      "logits/rejected": 0.7022187113761902,
+      "logps/chosen": -34.654640197753906,
+      "logps/rejected": -572.8896484375,
+      "loss": 14513.1688,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.19765284657478333,
+      "rewards/margins": 0.5385235548019409,
+      "rewards/rejected": -0.34087073802948,
+      "step": 4030
+    },
+    {
+      "epoch": 20.50126582278481,
+      "grad_norm": 492881.9600409883,
+      "learning_rate": 2.391099968661861e-07,
+      "logits/chosen": 0.8350554704666138,
+      "logits/rejected": 1.3902348279953003,
+      "logps/chosen": -55.87943649291992,
+      "logps/rejected": -570.520751953125,
+      "loss": 14474.9922,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.19199691712856293,
+      "rewards/margins": 0.5237919092178345,
+      "rewards/rejected": -0.33179494738578796,
+      "step": 4040
+    },
+    {
+      "epoch": 20.551898734177215,
+      "grad_norm": 550156.3808352741,
+      "learning_rate": 2.3832654340332183e-07,
+      "logits/chosen": 0.10426521301269531,
+      "logits/rejected": -0.0343349352478981,
+      "logps/chosen": -41.53984451293945,
+      "logps/rejected": -587.2073974609375,
+      "loss": 14114.8438,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.20071008801460266,
+      "rewards/margins": 0.5482142567634583,
+      "rewards/rejected": -0.3475041091442108,
+      "step": 4050
+    },
+    {
+      "epoch": 20.60253164556962,
+      "grad_norm": 442438.75618485885,
+      "learning_rate": 2.375430899404575e-07,
+      "logits/chosen": 0.6523551344871521,
+      "logits/rejected": 1.2190606594085693,
+      "logps/chosen": -46.92839431762695,
+      "logps/rejected": -584.9151000976562,
+      "loss": 14072.4,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19940249621868134,
+      "rewards/margins": 0.5458526611328125,
+      "rewards/rejected": -0.3464500606060028,
+      "step": 4060
+    },
+    {
+      "epoch": 20.653164556962025,
+      "grad_norm": 1289102.185567105,
+      "learning_rate": 2.3675963647759321e-07,
+      "logits/chosen": -0.5604708790779114,
+      "logits/rejected": -0.4172240197658539,
+      "logps/chosen": -41.39154815673828,
+      "logps/rejected": -568.0210571289062,
+      "loss": 14125.9813,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.18824639916419983,
+      "rewards/margins": 0.5325849652290344,
+      "rewards/rejected": -0.3443385362625122,
+      "step": 4070
+    },
+    {
+      "epoch": 20.70379746835443,
+      "grad_norm": 796999.4756244586,
+      "learning_rate": 2.3597618301472892e-07,
+      "logits/chosen": -0.1419040858745575,
+      "logits/rejected": 0.26791125535964966,
+      "logps/chosen": -37.67776870727539,
+      "logps/rejected": -578.685302734375,
+      "loss": 14390.0187,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 0.1924736052751541,
+      "rewards/margins": 0.5395643711090088,
+      "rewards/rejected": -0.34709078073501587,
+      "step": 4080
+    },
+    {
+      "epoch": 20.754430379746836,
+      "grad_norm": 730146.0659684967,
+      "learning_rate": 2.3519272955186463e-07,
+      "logits/chosen": -2.1902015209198,
+      "logits/rejected": -1.9143873453140259,
+      "logps/chosen": -49.27891159057617,
+      "logps/rejected": -583.9379272460938,
+      "loss": 14559.9,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.19109120965003967,
+      "rewards/margins": 0.5332453846931458,
+      "rewards/rejected": -0.3421540856361389,
+      "step": 4090
+    },
+    {
+      "epoch": 20.80506329113924,
+      "grad_norm": 722476.022933799,
+      "learning_rate": 2.344092760890003e-07,
+      "logits/chosen": -2.4936270713806152,
+      "logits/rejected": -2.54288649559021,
+      "logps/chosen": -38.10564041137695,
+      "logps/rejected": -595.3966064453125,
+      "loss": 14070.8406,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.20286759734153748,
+      "rewards/margins": 0.5558962821960449,
+      "rewards/rejected": -0.35302871465682983,
+      "step": 4100
+    },
+    {
+      "epoch": 20.855696202531647,
+      "grad_norm": 569046.2654479475,
+      "learning_rate": 2.33625822626136e-07,
+      "logits/chosen": -1.9905935525894165,
+      "logits/rejected": -1.920189619064331,
+      "logps/chosen": -34.30987548828125,
+      "logps/rejected": -561.2472534179688,
+      "loss": 14892.8141,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.18726976215839386,
+      "rewards/margins": 0.5297659039497375,
+      "rewards/rejected": -0.3424961268901825,
+      "step": 4110
+    },
+    {
+      "epoch": 20.906329113924052,
+      "grad_norm": 428009.1713524517,
+      "learning_rate": 2.328423691632717e-07,
+      "logits/chosen": -1.5145037174224854,
+      "logits/rejected": -1.5116671323776245,
+      "logps/chosen": -45.871498107910156,
+      "logps/rejected": -565.3566284179688,
+      "loss": 13932.3125,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.1972968429327011,
+      "rewards/margins": 0.5235085487365723,
+      "rewards/rejected": -0.3262116611003876,
+      "step": 4120
+    },
+    {
+      "epoch": 20.956962025316457,
+      "grad_norm": 648441.2224283866,
+      "learning_rate": 2.320589157004074e-07,
+      "logits/chosen": -0.6003355383872986,
+      "logits/rejected": 0.15587857365608215,
+      "logps/chosen": -56.75775146484375,
+      "logps/rejected": -600.1107177734375,
+      "loss": 14185.1125,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.20023982226848602,
+      "rewards/margins": 0.5419777631759644,
+      "rewards/rejected": -0.34173792600631714,
+      "step": 4130
     }
   ],
   "logging_steps": 10,