Training in progress, step 3000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3a6f08fab5c9cd3e3824141f661d04fab419e16f0f27aa8be427a55a045a5024
+oid sha256:b39bfb04c37041840837102213b584e6ae1204eb3c1f47b35a8ac4f25c8411ea
 size 577789320

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59d2fcaa714da200edbccb39a9f2728d1eac302590a189b7a1ad33e51ddded74
+oid sha256:55c130523c1814ac02576b7bcd8ddb8224d21012fa3dc0b1b8581ea711f9eefd
 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:b382f91d6a62c99969921e0d8014a3b1f89a198a6a81ab888e1194bc144c13d5
+oid sha256:390813e12fc409286b76f8ab397b98e6ba085504e2b094a343edebab876251b0
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.3528364896774292,
-  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_s8000/checkpoint-2500",
-  "epoch": 4.18848167539267,
+  "best_metric": 0.3480445444583893,
+  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_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": 31.122,
       "eval_steps_per_second": 3.893,
       "step": 2500
+    },
+    {
+      "epoch": 4.272251308900524,
+      "grad_norm": 1.0409568548202515,
+      "learning_rate": 9.086666666666666e-05,
+      "loss": 0.3894,
+      "step": 2550
+    },
+    {
+      "epoch": 4.356020942408377,
+      "grad_norm": 2.047908306121826,
+      "learning_rate": 9.003333333333333e-05,
+      "loss": 0.3862,
+      "step": 2600
+    },
+    {
+      "epoch": 4.439790575916231,
+      "grad_norm": 1.767168641090393,
+      "learning_rate": 8.92e-05,
+      "loss": 0.3832,
+      "step": 2650
+    },
+    {
+      "epoch": 4.523560209424084,
+      "grad_norm": 1.1918233633041382,
+      "learning_rate": 8.836666666666667e-05,
+      "loss": 0.3839,
+      "step": 2700
+    },
+    {
+      "epoch": 4.607329842931938,
+      "grad_norm": 0.9553185105323792,
+      "learning_rate": 8.753333333333334e-05,
+      "loss": 0.3807,
+      "step": 2750
+    },
+    {
+      "epoch": 4.69109947643979,
+      "grad_norm": 1.1633609533309937,
+      "learning_rate": 8.67e-05,
+      "loss": 0.3836,
+      "step": 2800
+    },
+    {
+      "epoch": 4.774869109947644,
+      "grad_norm": 1.1633822917938232,
+      "learning_rate": 8.586666666666668e-05,
+      "loss": 0.3815,
+      "step": 2850
+    },
+    {
+      "epoch": 4.858638743455497,
+      "grad_norm": 1.7152239084243774,
+      "learning_rate": 8.503333333333334e-05,
+      "loss": 0.3809,
+      "step": 2900
+    },
+    {
+      "epoch": 4.942408376963351,
+      "grad_norm": 1.2379584312438965,
+      "learning_rate": 8.42e-05,
+      "loss": 0.3801,
+      "step": 2950
+    },
+    {
+      "epoch": 5.026178010471204,
+      "grad_norm": 1.1686451435089111,
+      "learning_rate": 8.336666666666667e-05,
+      "loss": 0.3807,
+      "step": 3000
+    },
+    {
+      "epoch": 5.026178010471204,
+      "eval_loss": 0.3480445444583893,
+      "eval_runtime": 272.7064,
+      "eval_samples_per_second": 31.129,
+      "eval_steps_per_second": 3.894,
+      "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