Training in progress, epoch 3, checkpoint

last-checkpoint/global_step189/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13f2d28b244603c8211889c9afcd695067df8233359e5bf233ac00738cc0dfe1
+size 18483417132

last-checkpoint/global_step189/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e8a803f2b04a8276b1eb66810ddbc61190564b23303a92512f9ed7afbb4fcb7
+size 18483417132

last-checkpoint/global_step189/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2adf3fe11d4864749006592786bac2389b529c2fb573de7fc3dde322436ed508
+size 18483417132

last-checkpoint/global_step189/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34a51614e95c914d61cf1aa8c0b9f4aae4a0a9d65868f482622ed0db85587ed7
+size 18483417132

last-checkpoint/global_step189/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d9e3111d3d26ecb49527b8be396833f19c85a4a9035479056db556ad4a57810
+size 239998

last-checkpoint/global_step189/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:057ecdfa0fe058bb849b670b85f4e7901d31dfc556c46fecc4be81caba38acf2
+size 239934

last-checkpoint/global_step189/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:922227d611126b51889f9e50b1f09afa5d2366bd6caaf2fdd49d6c48fd8172eb
+size 239934

last-checkpoint/global_step189/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b46671cd1d4110dbb9dd43fddab33325dddfa3bef14669e729c74cc1d4c3674e
+size 239934

last-checkpoint/latest

@@ -1 +1 @@
-global_step63
+global_step189

last-checkpoint/model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a3faa501ab33a80380f2180b1f8fb7f16d78a7f89b516c1fb239de055d43bed
+oid sha256:ea1b6d7fce5305bfacd4bb0f50d40e3bbe0238a3bb00f40af276051bf7fc5df1
 size 4903351912

last-checkpoint/model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3fbbc0ff5deb82216230034c2e2eedcb7e66b8b01c863e39977722cca24e6bc
+oid sha256:51ea4315a74315b512266770598430ee290cc9c17281a3c02e07e4d7f3345713
 size 4947570872

last-checkpoint/model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89355447e0788040c35f2751b524c4c98dd37271ed58c929b2c3c8ca9fa681fe
+oid sha256:9d4602515e8507cbdaa8a21892d38fe5c4bb7d7d205e4484fee3bc16458d146f
 size 4962221464

last-checkpoint/model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65397289baba5c364fecf387a6c902463375bf07a37fda7b7ea59587c5bfbd1c
+oid sha256:03baccf59188e4af4695f188e99336c9705d69575ac58cf26555a9a706c08e3a
 size 3670322200

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f26ed720ddc0139bd34500d6a56d90b1325ef50284b8f9e9bcefa8c01d87e400
+oid sha256:084292ad91359496cd16eff7072d86abd759541271e83fa50d494de3dbdf77ee
 size 14960

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2bcae6573442578b752be3e988d4ccb38056b45c31b5c02eb579ec6cebcfa62e
+oid sha256:f7b183324e8227a51a9556d86b2ad893a8c4c52205ed4a737356c6611dac7353
 size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:398ff45072fee2975e88b9e078f915103d75d94bc08753303cf855b915973623
+oid sha256:ac69e994090f4818cb1fa6f6cefa363178552c3c731c6507ff195bcb07fd5bef
 size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f4b210bc832d4c648e3bee8c21dca26e5a8b365d6ec90c638062005a052e57b
+oid sha256:f68e61b63402f8afb1f69c960f7944965655dac11e3ccf29919c282f23931f86
 size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e2d72f8efaf9223a5f3a90e86eb233b5729e521882d016722304afbd046b461c
+oid sha256:129a41200f0c6d6e44aca4acb3d831836fddce423bc461718262ff68762bbd3e
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.0,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 63,
+  "global_step": 189,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -128,6 +128,218 @@
       "eval_samples_per_second": 10.597,
       "eval_steps_per_second": 0.371,
       "step": 63
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 20.770261653452945,
+      "learning_rate": 3.5e-07,
+      "logits/chosen": -1.6171875,
+      "logits/rejected": -0.87109375,
+      "logps/chosen": -62.25,
+      "logps/rejected": -85.0,
+      "loss": 0.3381,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.578125,
+      "rewards/margins": 1.953125,
+      "rewards/rejected": -1.375,
+      "step": 70
+    },
+    {
+      "epoch": 1.2698412698412698,
+      "grad_norm": 18.43753108370381,
+      "learning_rate": 3.205882352941177e-07,
+      "logits/chosen": -1.34375,
+      "logits/rejected": -0.84375,
+      "logps/chosen": -49.5,
+      "logps/rejected": -99.0,
+      "loss": 0.3254,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.3671875,
+      "rewards/margins": 2.390625,
+      "rewards/rejected": -2.015625,
+      "step": 80
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 16.839062370433073,
+      "learning_rate": 2.911764705882353e-07,
+      "logits/chosen": -1.5078125,
+      "logits/rejected": -0.78515625,
+      "logps/chosen": -57.5,
+      "logps/rejected": -72.5,
+      "loss": 0.2715,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.84765625,
+      "rewards/margins": 2.03125,
+      "rewards/rejected": -1.1875,
+      "step": 90
+    },
+    {
+      "epoch": 1.5873015873015874,
+      "grad_norm": 16.580044861649466,
+      "learning_rate": 2.6176470588235295e-07,
+      "logits/chosen": -2.09375,
+      "logits/rejected": -1.3359375,
+      "logps/chosen": -51.0,
+      "logps/rejected": -99.5,
+      "loss": 0.2519,
+      "rewards/accuracies": 0.8374999761581421,
+      "rewards/chosen": 0.345703125,
+      "rewards/margins": 1.9375,
+      "rewards/rejected": -1.59375,
+      "step": 100
+    },
+    {
+      "epoch": 1.746031746031746,
+      "grad_norm": 16.88214421094986,
+      "learning_rate": 2.323529411764706e-07,
+      "logits/chosen": -2.3125,
+      "logits/rejected": -1.171875,
+      "logps/chosen": -55.75,
+      "logps/rejected": -90.5,
+      "loss": 0.2389,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.197265625,
+      "rewards/margins": 1.9296875,
+      "rewards/rejected": -1.734375,
+      "step": 110
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 12.829485946331816,
+      "learning_rate": 2.0294117647058823e-07,
+      "logits/chosen": -1.8828125,
+      "logits/rejected": -0.9453125,
+      "logps/chosen": -50.5,
+      "logps/rejected": -102.5,
+      "loss": 0.2831,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 0.10888671875,
+      "rewards/margins": 2.578125,
+      "rewards/rejected": -2.46875,
+      "step": 120
+    },
+    {
+      "epoch": 2.0,
+      "eval_logits/chosen": -1.84375,
+      "eval_logits/rejected": -1.171875,
+      "eval_logps/chosen": -66.0,
+      "eval_logps/rejected": -97.5,
+      "eval_loss": 0.6284375190734863,
+      "eval_rewards/accuracies": 0.6071428656578064,
+      "eval_rewards/chosen": -0.453125,
+      "eval_rewards/margins": 0.75,
+      "eval_rewards/rejected": -1.203125,
+      "eval_runtime": 23.6296,
+      "eval_samples_per_second": 8.464,
+      "eval_steps_per_second": 0.296,
+      "step": 126
+    },
+    {
+      "epoch": 2.0634920634920633,
+      "grad_norm": 7.418629403738186,
+      "learning_rate": 1.7352941176470587e-07,
+      "logits/chosen": -2.109375,
+      "logits/rejected": -0.73046875,
+      "logps/chosen": -61.25,
+      "logps/rejected": -106.0,
+      "loss": 0.1749,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.4296875,
+      "rewards/margins": 2.625,
+      "rewards/rejected": -2.203125,
+      "step": 130
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 5.917946365968759,
+      "learning_rate": 1.441176470588235e-07,
+      "logits/chosen": -2.390625,
+      "logits/rejected": -1.2890625,
+      "logps/chosen": -72.0,
+      "logps/rejected": -100.0,
+      "loss": 0.1132,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.6015625,
+      "rewards/margins": 3.65625,
+      "rewards/rejected": -3.0625,
+      "step": 140
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 7.62179839987552,
+      "learning_rate": 1.1470588235294116e-07,
+      "logits/chosen": -2.125,
+      "logits/rejected": -1.296875,
+      "logps/chosen": -54.0,
+      "logps/rejected": -92.5,
+      "loss": 0.125,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.66796875,
+      "rewards/margins": 3.171875,
+      "rewards/rejected": -2.5,
+      "step": 150
+    },
+    {
+      "epoch": 2.5396825396825395,
+      "grad_norm": 7.689817277750616,
+      "learning_rate": 8.529411764705883e-08,
+      "logits/chosen": -1.8671875,
+      "logits/rejected": -2.3125,
+      "logps/chosen": -50.25,
+      "logps/rejected": -92.5,
+      "loss": 0.118,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.3203125,
+      "rewards/margins": 3.109375,
+      "rewards/rejected": -2.78125,
+      "step": 160
+    },
+    {
+      "epoch": 2.6984126984126986,
+      "grad_norm": 5.583079792162674,
+      "learning_rate": 5.5882352941176474e-08,
+      "logits/chosen": -2.21875,
+      "logits/rejected": -1.1328125,
+      "logps/chosen": -66.0,
+      "logps/rejected": -114.5,
+      "loss": 0.145,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.388671875,
+      "rewards/margins": 3.671875,
+      "rewards/rejected": -3.28125,
+      "step": 170
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 10.79399873370834,
+      "learning_rate": 2.6470588235294116e-08,
+      "logits/chosen": -1.6640625,
+      "logits/rejected": -1.5234375,
+      "logps/chosen": -59.0,
+      "logps/rejected": -87.5,
+      "loss": 0.1329,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.310546875,
+      "rewards/margins": 2.75,
+      "rewards/rejected": -2.4375,
+      "step": 180
+    },
+    {
+      "epoch": 3.0,
+      "eval_logits/chosen": -2.21875,
+      "eval_logits/rejected": -1.4921875,
+      "eval_logps/chosen": -69.5,
+      "eval_logps/rejected": -102.5,
+      "eval_loss": 0.6340625286102295,
+      "eval_rewards/accuracies": 0.6607142686843872,
+      "eval_rewards/chosen": -0.76953125,
+      "eval_rewards/margins": 0.921875,
+      "eval_rewards/rejected": -1.6953125,
+      "eval_runtime": 21.8558,
+      "eval_samples_per_second": 9.151,
+      "eval_steps_per_second": 0.32,
+      "step": 189
     }
   ],
   "logging_steps": 10,
@@ -142,7 +354,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }