Training in progress, step 3000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a5a846d47736910ffa390729689c49f8f436012a7ed5267f8423224af8f7153
+oid sha256:9ee760036bd756361470f528d4ec7376115fb68748fa38362e88c40e07ef9522
 size 577789320

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1b11d7e918af896e00bc1d17eef1b86fe0eefd81cc23746df213d340735a9936
+oid sha256:eaf4569976454937eea43c00a16f8f139fec421442a724a7ef012f3b4590cecc
 size 1155772233

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e0a3df38c17b475ea7e47b6c462cb3838a91cdec2a3f47a69fc35d66481cfa21
+oid sha256:b0bd26a66f18a63513b4644dff9bbce4420bc172da7cc36bbc9e2bbc11796d8c
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f21e44438b98faf063216590a5534fa1899dfb57fd798660776c8a525c03fb28
+oid sha256:944131a8cfe7c888a8dee4aec652300305a40382e916372dab3fcd26a65bac19
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.40933898091316223,
-  "best_model_checkpoint": "mikhail-panzo/zlm_b128_le5_s8000/checkpoint-2500",
-  "epoch": 4.18848167539267,
+  "best_metric": 0.39678552746772766,
+  "best_model_checkpoint": "mikhail-panzo/zlm_b128_le5_s8000/checkpoint-3000",
+  "epoch": 5.026178010471204,
   "eval_steps": 500,
-  "global_step": 2500,
+  "global_step": 3000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -397,6 +397,84 @@
       "eval_samples_per_second": 33.578,
       "eval_steps_per_second": 4.201,
       "step": 2500
+    },
+    {
+      "epoch": 4.272251308900524,
+      "grad_norm": 5.594339370727539,
+      "learning_rate": 9.088333333333334e-06,
+      "loss": 0.4592,
+      "step": 2550
+    },
+    {
+      "epoch": 4.356020942408377,
+      "grad_norm": 2.930290699005127,
+      "learning_rate": 9.005000000000001e-06,
+      "loss": 0.4549,
+      "step": 2600
+    },
+    {
+      "epoch": 4.439790575916231,
+      "grad_norm": 2.199352741241455,
+      "learning_rate": 8.921666666666668e-06,
+      "loss": 0.4512,
+      "step": 2650
+    },
+    {
+      "epoch": 4.523560209424084,
+      "grad_norm": 2.435898780822754,
+      "learning_rate": 8.838333333333335e-06,
+      "loss": 0.4557,
+      "step": 2700
+    },
+    {
+      "epoch": 4.607329842931938,
+      "grad_norm": 3.2397773265838623,
+      "learning_rate": 8.755e-06,
+      "loss": 0.449,
+      "step": 2750
+    },
+    {
+      "epoch": 4.69109947643979,
+      "grad_norm": 1.9361001253128052,
+      "learning_rate": 8.671666666666667e-06,
+      "loss": 0.4525,
+      "step": 2800
+    },
+    {
+      "epoch": 4.774869109947644,
+      "grad_norm": 2.1852686405181885,
+      "learning_rate": 8.588333333333334e-06,
+      "loss": 0.4486,
+      "step": 2850
+    },
+    {
+      "epoch": 4.858638743455497,
+      "grad_norm": 1.8136191368103027,
+      "learning_rate": 8.505e-06,
+      "loss": 0.4517,
+      "step": 2900
+    },
+    {
+      "epoch": 4.942408376963351,
+      "grad_norm": 2.694406032562256,
+      "learning_rate": 8.421666666666668e-06,
+      "loss": 0.4505,
+      "step": 2950
+    },
+    {
+      "epoch": 5.026178010471204,
+      "grad_norm": 1.251187801361084,
+      "learning_rate": 8.338333333333335e-06,
+      "loss": 0.4489,
+      "step": 3000
+    },
+    {
+      "epoch": 5.026178010471204,
+      "eval_loss": 0.39678552746772766,
+      "eval_runtime": 256.1618,
+      "eval_samples_per_second": 33.139,
+      "eval_steps_per_second": 4.146,
+      "step": 3000
     }
   ],
   "logging_steps": 50,
@@ -416,7 +494,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.477753664307475e+16,
+  "total_flos": 5.37491824204009e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null