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 sha256:73736d20e4b5078be3846f488716be6f8e377191d732f9dd074004b19b880cdb
+size 13476836524

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:fff07b006b508703d293b537de80d632bb73464abe548b8db25000cf2ff7472a
+size 13476836524

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 sha256:35f396c83e2e063f7feaadcaa27cc4dfb30ac9e8f652e6ae84b91ae6a15c88c0
+size 13476836524

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 sha256:3c253541de0013ab57f2d4a91feadcdd9fa5668c6eaf3e583eefb06069b45689
+size 13476836524

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:fad6f8e38bd53cc02696fdd324acefc6450b408b94aa0efc4ff10b7c6725a66e
+size 150693

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:f2769a91ee3073096451825a14033196304aa61466a0742d138c51e7daf74827
+size 150693

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 sha256:1a1731958171867a7d85e560aa4daebfb823dbd70dd783e2e66bc88bb7340988
+size 150693

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 sha256:65f5c3867311d6afda2cf705c7b21af911b58685f925c8915171e17d8ab9b14d
+size 150693

last-checkpoint/latest

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

last-checkpoint/model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a18f906e422539d7561fb74fcbe4cda65c8dd8a58f49c0673740a6cbc3312e21
+oid sha256:95525d9c30b0213d4459f12201ac85cc3dcc181006ed5c82f4badf60c04d57c2
 size 4938985352

last-checkpoint/model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0b35678ee867b0ee73246f6cf7e43051df256f2bcc6845be8d2c0606598550c4
+oid sha256:ae02f26857487c0ba98d339de83e4a15d2d435e050d87310ea9215f837474a82
 size 4947390880

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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47ea9884ae8f8df38f905a0c4f62aecc15e57960dcee4e41334eb39453a817ac
+oid sha256:f0fd6210a0d41e7f1f6251abc8cb58672cecda03d6cb318bd0e2a2023af05f6c
 size 3590488816

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1ce770862b76ceecb3453a5d0b9ef65311eab0f6a60f52997726a89d718de08
+oid sha256:ab7cf0d34d60dfda516cf9661904550e2e294e723edd07c25c738f05e8ba92d1
 size 15024

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 sha256:845f8eea760ff90e22dadf5fb7f80c9079f8c2298a8770c5d4c3b3309b38a6ef
+oid sha256:c988df4a306242fbc45c908424c576078313f1b5bb472bee13c3b6371b94095e
 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": 16.341,
       "eval_steps_per_second": 0.572,
       "step": 63
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 13.062522185911641,
+      "learning_rate": 3.5e-07,
+      "logits/chosen": -1.375,
+      "logits/rejected": -1.375,
+      "logps/chosen": -93.0,
+      "logps/rejected": -108.0,
+      "loss": 0.5617,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": 0.1533203125,
+      "rewards/margins": 0.375,
+      "rewards/rejected": -0.220703125,
+      "step": 70
+    },
+    {
+      "epoch": 1.2698412698412698,
+      "grad_norm": 12.793133064546527,
+      "learning_rate": 3.205882352941177e-07,
+      "logits/chosen": -1.2421875,
+      "logits/rejected": -1.359375,
+      "logps/chosen": -82.5,
+      "logps/rejected": -112.5,
+      "loss": 0.5127,
+      "rewards/accuracies": 0.7749999761581421,
+      "rewards/chosen": 0.1455078125,
+      "rewards/margins": 0.515625,
+      "rewards/rejected": -0.37109375,
+      "step": 80
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 13.067935081505256,
+      "learning_rate": 2.911764705882353e-07,
+      "logits/chosen": -1.484375,
+      "logits/rejected": -1.3203125,
+      "logps/chosen": -100.0,
+      "logps/rejected": -98.0,
+      "loss": 0.5069,
+      "rewards/accuracies": 0.7875000238418579,
+      "rewards/chosen": 0.283203125,
+      "rewards/margins": 0.5390625,
+      "rewards/rejected": -0.2578125,
+      "step": 90
+    },
+    {
+      "epoch": 1.5873015873015874,
+      "grad_norm": 15.317610443936815,
+      "learning_rate": 2.6176470588235295e-07,
+      "logits/chosen": -1.3671875,
+      "logits/rejected": -1.375,
+      "logps/chosen": -90.0,
+      "logps/rejected": -111.0,
+      "loss": 0.4966,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/chosen": 0.2353515625,
+      "rewards/margins": 0.5546875,
+      "rewards/rejected": -0.318359375,
+      "step": 100
+    },
+    {
+      "epoch": 1.746031746031746,
+      "grad_norm": 12.447441760307086,
+      "learning_rate": 2.323529411764706e-07,
+      "logits/chosen": -1.4296875,
+      "logits/rejected": -1.296875,
+      "logps/chosen": -97.0,
+      "logps/rejected": -115.0,
+      "loss": 0.4934,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 0.1435546875,
+      "rewards/margins": 0.56640625,
+      "rewards/rejected": -0.423828125,
+      "step": 110
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 13.19848987425151,
+      "learning_rate": 2.0294117647058823e-07,
+      "logits/chosen": -1.34375,
+      "logits/rejected": -1.34375,
+      "logps/chosen": -94.0,
+      "logps/rejected": -117.5,
+      "loss": 0.4805,
+      "rewards/accuracies": 0.8500000238418579,
+      "rewards/chosen": 0.10205078125,
+      "rewards/margins": 0.60546875,
+      "rewards/rejected": -0.50390625,
+      "step": 120
+    },
+    {
+      "epoch": 2.0,
+      "eval_logits/chosen": -1.34375,
+      "eval_logits/rejected": -1.328125,
+      "eval_logps/chosen": -92.0,
+      "eval_logps/rejected": -113.0,
+      "eval_loss": 0.5889843702316284,
+      "eval_rewards/accuracies": 0.6428571343421936,
+      "eval_rewards/chosen": -0.0751953125,
+      "eval_rewards/margins": 0.31640625,
+      "eval_rewards/rejected": -0.392578125,
+      "eval_runtime": 14.7924,
+      "eval_samples_per_second": 13.52,
+      "eval_steps_per_second": 0.473,
+      "step": 126
+    },
+    {
+      "epoch": 2.0634920634920633,
+      "grad_norm": 10.98673276740895,
+      "learning_rate": 1.7352941176470587e-07,
+      "logits/chosen": -1.390625,
+      "logits/rejected": -1.3984375,
+      "logps/chosen": -101.0,
+      "logps/rejected": -116.0,
+      "loss": 0.4562,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 0.193359375,
+      "rewards/margins": 0.67578125,
+      "rewards/rejected": -0.482421875,
+      "step": 130
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 9.416416295176981,
+      "learning_rate": 1.441176470588235e-07,
+      "logits/chosen": -1.4140625,
+      "logits/rejected": -1.3828125,
+      "logps/chosen": -100.5,
+      "logps/rejected": -107.5,
+      "loss": 0.4104,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 0.2099609375,
+      "rewards/margins": 0.76171875,
+      "rewards/rejected": -0.55078125,
+      "step": 140
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 11.858508121730686,
+      "learning_rate": 1.1470588235294116e-07,
+      "logits/chosen": -1.3671875,
+      "logits/rejected": -1.3515625,
+      "logps/chosen": -93.5,
+      "logps/rejected": -113.5,
+      "loss": 0.413,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.2177734375,
+      "rewards/margins": 0.7265625,
+      "rewards/rejected": -0.51171875,
+      "step": 150
+    },
+    {
+      "epoch": 2.5396825396825395,
+      "grad_norm": 9.556697117699082,
+      "learning_rate": 8.529411764705883e-08,
+      "logits/chosen": -1.296875,
+      "logits/rejected": -1.4296875,
+      "logps/chosen": -90.5,
+      "logps/rejected": -108.5,
+      "loss": 0.4203,
+      "rewards/accuracies": 0.800000011920929,
+      "rewards/chosen": 0.1337890625,
+      "rewards/margins": 0.82421875,
+      "rewards/rejected": -0.69140625,
+      "step": 160
+    },
+    {
+      "epoch": 2.6984126984126986,
+      "grad_norm": 10.065550572065934,
+      "learning_rate": 5.5882352941176474e-08,
+      "logits/chosen": -1.3671875,
+      "logits/rejected": -1.3671875,
+      "logps/chosen": -98.5,
+      "logps/rejected": -109.5,
+      "loss": 0.4442,
+      "rewards/accuracies": 0.7749999761581421,
+      "rewards/chosen": 0.09814453125,
+      "rewards/margins": 0.734375,
+      "rewards/rejected": -0.63671875,
+      "step": 170
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 11.316830865980743,
+      "learning_rate": 2.6470588235294116e-08,
+      "logits/chosen": -1.3515625,
+      "logits/rejected": -1.34375,
+      "logps/chosen": -102.0,
+      "logps/rejected": -104.5,
+      "loss": 0.4264,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/chosen": 0.10498046875,
+      "rewards/margins": 0.6328125,
+      "rewards/rejected": -0.52734375,
+      "step": 180
+    },
+    {
+      "epoch": 3.0,
+      "eval_logits/chosen": -1.3515625,
+      "eval_logits/rejected": -1.3359375,
+      "eval_logps/chosen": -92.5,
+      "eval_logps/rejected": -114.0,
+      "eval_loss": 0.5811718702316284,
+      "eval_rewards/accuracies": 0.6428571343421936,
+      "eval_rewards/chosen": -0.1357421875,
+      "eval_rewards/margins": 0.34765625,
+      "eval_rewards/rejected": -0.484375,
+      "eval_runtime": 14.7749,
+      "eval_samples_per_second": 13.537,
+      "eval_steps_per_second": 0.474,
+      "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": {}
     }