Training in progress, step 4500, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:935c0c9e6c9bad61c25203243a2dced5424fa2c750c3b63d12e6a9555bd1d414
+oid sha256:19f937f8605011a6423fdcd193fb32164c380c1076f0eed9d33dcc3767b4df87
 size 891558696

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5b6424579b540963ca5dff78ef0b1bc07be8ee502bd08a007f4baf7bfc0962f2
+oid sha256:5ffcdfd14da9f025b2c66271aa5f608d4bbfd5d88721668100d16d60c271d53f
 size 1783272762

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:994ecdd24463592bb7a469b126032a8d12a6fc0731753b561060d5f56a11ebf4
+oid sha256:831ec110a79680c26e00dbf10946e10090e2d0fde59d1cc953456efe8ddb0df7
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86f6fa01b61cfbdf0a99f96e2c8520500066cb2276b7fc4c693828106716e38a
+oid sha256:c460047761054bd92ad62049610941db058ccd6f71ff714487704282f9a72984
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.08466340601444244,
-  "best_model_checkpoint": "./fine-tuned/checkpoint-4000",
-  "epoch": 0.64,
+  "best_metric": 0.0842796117067337,
+  "best_model_checkpoint": "./fine-tuned/checkpoint-4500",
+  "epoch": 0.72,
   "eval_steps": 500,
-  "global_step": 4000,
+  "global_step": 4500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -81,9 +81,9 @@
     {
       "epoch": 0.08,
       "eval_loss": 0.09235642850399017,
-      "eval_runtime": 116.7651,
-      "eval_samples_per_second": 17.128,
-      "eval_steps_per_second": 2.141,
+      "eval_runtime": 109.274,
+      "eval_samples_per_second": 18.303,
+      "eval_steps_per_second": 2.288,
       "step": 500
     },
     {
@@ -159,9 +159,9 @@
     {
       "epoch": 0.16,
       "eval_loss": 0.08949962258338928,
-      "eval_runtime": 116.7407,
-      "eval_samples_per_second": 17.132,
-      "eval_steps_per_second": 2.141,
+      "eval_runtime": 109.2536,
+      "eval_samples_per_second": 18.306,
+      "eval_steps_per_second": 2.288,
       "step": 1000
     },
     {
@@ -237,9 +237,9 @@
     {
       "epoch": 0.24,
       "eval_loss": 0.08808805048465729,
-      "eval_runtime": 116.8722,
-      "eval_samples_per_second": 17.113,
-      "eval_steps_per_second": 2.139,
+      "eval_runtime": 109.2355,
+      "eval_samples_per_second": 18.309,
+      "eval_steps_per_second": 2.289,
       "step": 1500
     },
     {
@@ -315,9 +315,9 @@
     {
       "epoch": 0.32,
       "eval_loss": 0.08704760670661926,
-      "eval_runtime": 116.8362,
-      "eval_samples_per_second": 17.118,
-      "eval_steps_per_second": 2.14,
+      "eval_runtime": 109.4348,
+      "eval_samples_per_second": 18.276,
+      "eval_steps_per_second": 2.284,
       "step": 2000
     },
     {
@@ -393,9 +393,9 @@
     {
       "epoch": 0.4,
       "eval_loss": 0.08615937829017639,
-      "eval_runtime": 116.9591,
-      "eval_samples_per_second": 17.1,
-      "eval_steps_per_second": 2.137,
+      "eval_runtime": 109.2621,
+      "eval_samples_per_second": 18.305,
+      "eval_steps_per_second": 2.288,
       "step": 2500
     },
     {
@@ -471,9 +471,9 @@
     {
       "epoch": 0.48,
       "eval_loss": 0.08551913499832153,
-      "eval_runtime": 116.545,
-      "eval_samples_per_second": 17.161,
-      "eval_steps_per_second": 2.145,
+      "eval_runtime": 109.2626,
+      "eval_samples_per_second": 18.305,
+      "eval_steps_per_second": 2.288,
       "step": 3000
     },
     {
@@ -549,9 +549,9 @@
     {
       "epoch": 0.56,
       "eval_loss": 0.08540560305118561,
-      "eval_runtime": 116.9131,
-      "eval_samples_per_second": 17.107,
-      "eval_steps_per_second": 2.138,
+      "eval_runtime": 109.3641,
+      "eval_samples_per_second": 18.288,
+      "eval_steps_per_second": 2.286,
       "step": 3500
     },
     {
@@ -627,10 +627,88 @@
     {
       "epoch": 0.64,
       "eval_loss": 0.08466340601444244,
-      "eval_runtime": 116.6411,
-      "eval_samples_per_second": 17.147,
-      "eval_steps_per_second": 2.143,
+      "eval_runtime": 109.2066,
+      "eval_samples_per_second": 18.314,
+      "eval_steps_per_second": 2.289,
       "step": 4000
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 7243.01611328125,
+      "learning_rate": 2.3520000000000002e-05,
+      "loss": 0.0622,
+      "step": 4050
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 7986.32568359375,
+      "learning_rate": 2.344e-05,
+      "loss": 0.0678,
+      "step": 4100
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 9114.8974609375,
+      "learning_rate": 2.336e-05,
+      "loss": 0.0671,
+      "step": 4150
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 8830.62109375,
+      "learning_rate": 2.328e-05,
+      "loss": 0.0679,
+      "step": 4200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9311.2412109375,
+      "learning_rate": 2.32e-05,
+      "loss": 0.063,
+      "step": 4250
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 31307.103515625,
+      "learning_rate": 2.3120000000000002e-05,
+      "loss": 0.0649,
+      "step": 4300
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 9040.0126953125,
+      "learning_rate": 2.304e-05,
+      "loss": 0.0633,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 7183.91650390625,
+      "learning_rate": 2.296e-05,
+      "loss": 0.0582,
+      "step": 4400
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 6460.2998046875,
+      "learning_rate": 2.288e-05,
+      "loss": 0.0672,
+      "step": 4450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6104.8671875,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.0597,
+      "step": 4500
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.0842796117067337,
+      "eval_runtime": 109.48,
+      "eval_samples_per_second": 18.268,
+      "eval_steps_per_second": 2.284,
+      "step": 4500
     }
   ],
   "logging_steps": 50,
@@ -650,7 +728,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.948665249792e+16,
+  "total_flos": 2.192248406016e+16,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null