Training in progress, step 4500, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:09f46755ffeadc3d999dd23fd7022daf35bf2fbdd0db7cbfb92677a532f0a38f
+oid sha256:4b4f912370d1bedd71274be25ed29670ef08c3ac85ef3bed57515bfb8a68ade1
 size 577789320

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4766a7bbe36c3cafb785ca7ab63a31dbb580385e53a5efb78b1082ff7c77f718
+oid sha256:d66b9b6cd4ef7c4691763fae34990a87238cf81f4159ed58131692b71c37b63e
 size 1155772233

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:56cc8e2d77d54e9e58cdbed4bf8357f729e20efe110f0bdfcd24e4976aebb3b0
+oid sha256:3ab3c7bbc7d4e63c4ce7251d758f5b983c3c682908320ebf240583aab9c890e9
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f3323bc562f08c84a55f57ffe42cad4e899b9bed240a410d83f006b05ef914d
+oid sha256:abd54cd94767f93e5d208ede71f2b5ed0172a8c421b769cc48d09cb32a6db474
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.33282962441444397,
-  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_s8000/checkpoint-4000",
-  "epoch": 6.701570680628272,
+  "best_metric": 0.33003953099250793,
+  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_s8000/checkpoint-4500",
+  "epoch": 7.539267015706806,
   "eval_steps": 500,
-  "global_step": 4000,
+  "global_step": 4500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -631,6 +631,84 @@
       "eval_samples_per_second": 31.254,
       "eval_steps_per_second": 3.91,
       "step": 4000
+    },
+    {
+      "epoch": 6.785340314136126,
+      "grad_norm": 0.9167115092277527,
+      "learning_rate": 6.586666666666666e-05,
+      "loss": 0.3618,
+      "step": 4050
+    },
+    {
+      "epoch": 6.869109947643979,
+      "grad_norm": 1.0496320724487305,
+      "learning_rate": 6.503333333333333e-05,
+      "loss": 0.3624,
+      "step": 4100
+    },
+    {
+      "epoch": 6.952879581151833,
+      "grad_norm": 1.7927188873291016,
+      "learning_rate": 6.42e-05,
+      "loss": 0.3621,
+      "step": 4150
+    },
+    {
+      "epoch": 7.036649214659686,
+      "grad_norm": 1.3362383842468262,
+      "learning_rate": 6.336666666666667e-05,
+      "loss": 0.3631,
+      "step": 4200
+    },
+    {
+      "epoch": 7.12041884816754,
+      "grad_norm": 1.5686593055725098,
+      "learning_rate": 6.253333333333333e-05,
+      "loss": 0.3672,
+      "step": 4250
+    },
+    {
+      "epoch": 7.204188481675392,
+      "grad_norm": 2.042232036590576,
+      "learning_rate": 6.170000000000001e-05,
+      "loss": 0.3598,
+      "step": 4300
+    },
+    {
+      "epoch": 7.287958115183246,
+      "grad_norm": 1.2040599584579468,
+      "learning_rate": 6.086666666666667e-05,
+      "loss": 0.3628,
+      "step": 4350
+    },
+    {
+      "epoch": 7.371727748691099,
+      "grad_norm": 1.3659878969192505,
+      "learning_rate": 6.003333333333334e-05,
+      "loss": 0.3626,
+      "step": 4400
+    },
+    {
+      "epoch": 7.455497382198953,
+      "grad_norm": 1.9217562675476074,
+      "learning_rate": 5.92e-05,
+      "loss": 0.3636,
+      "step": 4450
+    },
+    {
+      "epoch": 7.539267015706806,
+      "grad_norm": 1.5920127630233765,
+      "learning_rate": 5.836666666666667e-05,
+      "loss": 0.3575,
+      "step": 4500
+    },
+    {
+      "epoch": 7.539267015706806,
+      "eval_loss": 0.33003953099250793,
+      "eval_runtime": 274.282,
+      "eval_samples_per_second": 30.95,
+      "eval_steps_per_second": 3.872,
+      "step": 4500
     }
   ],
   "logging_steps": 50,
@@ -650,7 +728,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 7.164514367145274e+16,
+  "total_flos": 8.06136922967207e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null