Training in progress, step 5000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:522262e27734a9378c8b08aa0c633e807d9e84add79bfe63e4bd216a33a75a21
+oid sha256:e89ff2481874bf269cba7ac84ce957b44c8fa49e2c675e9e6b16008fdb7a55c7
 size 577789320

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c7330d372f27ce4d7c5f07492d9ce5dde9db6730eef727936aabfaef3c74195d
+oid sha256:2b314243ca4b21fc0b329b39dc658d58545fdaa26b0dc513de3ccf6b20a972ec
 size 1155772233

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3ab3c7bbc7d4e63c4ce7251d758f5b983c3c682908320ebf240583aab9c890e9
+oid sha256:8f4e204979057d17355f455adeaa53e6157f86fff91322a4565250d75055d836
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d9ea6ac8e84dcaff37314a282ec7aba7995babdb11261372593bc800ef835cf
+oid sha256:51ae9e871201b5140a75803631a3b447a66295ef984c1c9cd1719a8febdf7390
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.37984684109687805,
-  "best_model_checkpoint": "mikhail-panzo/zlm_b128_le5_s8000/checkpoint-4500",
-  "epoch": 7.539267015706806,
+  "best_metric": 0.37838345766067505,
+  "best_model_checkpoint": "mikhail-panzo/zlm_b128_le5_s8000/checkpoint-5000",
+  "epoch": 8.37696335078534,
   "eval_steps": 500,
-  "global_step": 4500,
+  "global_step": 5000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -709,6 +709,84 @@
       "eval_samples_per_second": 33.831,
       "eval_steps_per_second": 4.232,
       "step": 4500
+    },
+    {
+      "epoch": 7.62303664921466,
+      "grad_norm": 1.3911954164505005,
+      "learning_rate": 5.755000000000001e-06,
+      "loss": 0.42,
+      "step": 4550
+    },
+    {
+      "epoch": 7.706806282722513,
+      "grad_norm": 1.3781930208206177,
+      "learning_rate": 5.671666666666668e-06,
+      "loss": 0.4234,
+      "step": 4600
+    },
+    {
+      "epoch": 7.790575916230367,
+      "grad_norm": 1.5225090980529785,
+      "learning_rate": 5.588333333333334e-06,
+      "loss": 0.4242,
+      "step": 4650
+    },
+    {
+      "epoch": 7.87434554973822,
+      "grad_norm": 1.4195870161056519,
+      "learning_rate": 5.505000000000001e-06,
+      "loss": 0.4243,
+      "step": 4700
+    },
+    {
+      "epoch": 7.958115183246074,
+      "grad_norm": 1.9713116884231567,
+      "learning_rate": 5.4216666666666676e-06,
+      "loss": 0.4158,
+      "step": 4750
+    },
+    {
+      "epoch": 8.041884816753926,
+      "grad_norm": 1.3215075731277466,
+      "learning_rate": 5.3383333333333345e-06,
+      "loss": 0.4141,
+      "step": 4800
+    },
+    {
+      "epoch": 8.12565445026178,
+      "grad_norm": 4.483395576477051,
+      "learning_rate": 5.2550000000000005e-06,
+      "loss": 0.4165,
+      "step": 4850
+    },
+    {
+      "epoch": 8.209424083769633,
+      "grad_norm": 1.4759974479675293,
+      "learning_rate": 5.171666666666667e-06,
+      "loss": 0.4291,
+      "step": 4900
+    },
+    {
+      "epoch": 8.293193717277488,
+      "grad_norm": 1.3336604833602905,
+      "learning_rate": 5.088333333333334e-06,
+      "loss": 0.4134,
+      "step": 4950
+    },
+    {
+      "epoch": 8.37696335078534,
+      "grad_norm": 1.5618665218353271,
+      "learning_rate": 5.0049999999999995e-06,
+      "loss": 0.4222,
+      "step": 5000
+    },
+    {
+      "epoch": 8.37696335078534,
+      "eval_loss": 0.37838345766067505,
+      "eval_runtime": 256.1892,
+      "eval_samples_per_second": 33.136,
+      "eval_steps_per_second": 4.145,
+      "step": 5000
     }
   ],
   "logging_steps": 50,
@@ -728,7 +806,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 8.06136922967207e+16,
+  "total_flos": 8.958474220139136e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null