Training in progress, step 5000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b4f912370d1bedd71274be25ed29670ef08c3ac85ef3bed57515bfb8a68ade1
+oid sha256:3fb634d25a31951e441158f54ae0db2c273aaeb251169c41f8467cecc648bf56
 size 577789320

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d66b9b6cd4ef7c4691763fae34990a87238cf81f4159ed58131692b71c37b63e
+oid sha256:4cc68ae6c7c867a9ae1fe921933744282f9631aa37051c6ca0a82c81119ae62e
 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:abd54cd94767f93e5d208ede71f2b5ed0172a8c421b769cc48d09cb32a6db474
+oid sha256:404094f43137b5b2180fadcde3d1168585bcd71ff155f508576088e11be9d3e1
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.33003953099250793,
-  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_s8000/checkpoint-4500",
-  "epoch": 7.539267015706806,
+  "best_metric": 0.3284249007701874,
+  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_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": 30.95,
       "eval_steps_per_second": 3.872,
       "step": 4500
+    },
+    {
+      "epoch": 7.62303664921466,
+      "grad_norm": 1.3059043884277344,
+      "learning_rate": 5.753333333333334e-05,
+      "loss": 0.3603,
+      "step": 4550
+    },
+    {
+      "epoch": 7.706806282722513,
+      "grad_norm": 1.6313170194625854,
+      "learning_rate": 5.6699999999999996e-05,
+      "loss": 0.3622,
+      "step": 4600
+    },
+    {
+      "epoch": 7.790575916230367,
+      "grad_norm": 1.7790875434875488,
+      "learning_rate": 5.5866666666666665e-05,
+      "loss": 0.3635,
+      "step": 4650
+    },
+    {
+      "epoch": 7.87434554973822,
+      "grad_norm": 1.1874761581420898,
+      "learning_rate": 5.5033333333333334e-05,
+      "loss": 0.3635,
+      "step": 4700
+    },
+    {
+      "epoch": 7.958115183246074,
+      "grad_norm": 0.9745219349861145,
+      "learning_rate": 5.420000000000001e-05,
+      "loss": 0.357,
+      "step": 4750
+    },
+    {
+      "epoch": 8.041884816753926,
+      "grad_norm": 1.3074321746826172,
+      "learning_rate": 5.3366666666666665e-05,
+      "loss": 0.3543,
+      "step": 4800
+    },
+    {
+      "epoch": 8.12565445026178,
+      "grad_norm": 1.3384989500045776,
+      "learning_rate": 5.2533333333333334e-05,
+      "loss": 0.3556,
+      "step": 4850
+    },
+    {
+      "epoch": 8.209424083769633,
+      "grad_norm": 1.0133076906204224,
+      "learning_rate": 5.17e-05,
+      "loss": 0.3654,
+      "step": 4900
+    },
+    {
+      "epoch": 8.293193717277488,
+      "grad_norm": 1.194360375404358,
+      "learning_rate": 5.086666666666667e-05,
+      "loss": 0.3547,
+      "step": 4950
+    },
+    {
+      "epoch": 8.37696335078534,
+      "grad_norm": 1.049672245979309,
+      "learning_rate": 5.0033333333333334e-05,
+      "loss": 0.3607,
+      "step": 5000
+    },
+    {
+      "epoch": 8.37696335078534,
+      "eval_loss": 0.3284249007701874,
+      "eval_runtime": 274.7134,
+      "eval_samples_per_second": 30.901,
+      "eval_steps_per_second": 3.866,
+      "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