Training in progress, step 16000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:15f637ff72e852c00df336464cba31267a78c2fec942618a4cf3dbc081150cb8
+oid sha256:aeeda4b1ee34efcb8cc9870b2c843e6649254ccabfeca77e86d631e6043a12c7
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:11255a9366d03d2ecf115313602ca401e81860858d4e1ecad341feef41b0e95b
+oid sha256:ad619e41a755fe598f7951c93afa49e976ac527f447e6a0b8b599bd265a02a63
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea828a56e17bf773dc8e4fa2c22d13619b805c6b9321028dd494ff57e5daf8e6
+oid sha256:ce8b7e7bf463c8aae6a90e1a089b32f412ca31160cb21f8d8c8b0ee26b9ca28d
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1819c72414dd202fc7a5b387187559436ac1d66f4c4de3f13c18065ffbdf0216
+oid sha256:ab47855ede8e2538e1a83a739503d4b0055ccb1a5bd772cbc798bb06eb73f571
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.620667803310349,
+  "epoch": 3.8620788508834134,
   "eval_steps": 500,
-  "global_step": 15000,
+  "global_step": 16000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3278,11 +3278,229 @@
       "eval_steps_per_second": 20.492,
       "num_input_tokens_seen": 7245951473,
       "step": 15000
+    },
+    {
+      "epoch": 3.632738355689002,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.726573110004527e-06,
+      "loss": 2.0923,
+      "mean_token_accuracy": 0.554327048882842,
+      "num_input_tokens_seen": 7269914417,
+      "num_tokens": 3063722766.0,
+      "step": 15050
+    },
+    {
+      "epoch": 3.6448089080676556,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.5379508073034565e-06,
+      "loss": 2.0861,
+      "mean_token_accuracy": 0.5561711810901762,
+      "num_input_tokens_seen": 7293956449,
+      "num_tokens": 3073837103.0,
+      "step": 15100
+    },
+    {
+      "epoch": 3.6568794604463086,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.349328504602384e-06,
+      "loss": 2.0949,
+      "mean_token_accuracy": 0.5547518468275666,
+      "num_input_tokens_seen": 7318063713,
+      "num_tokens": 3083990662.0,
+      "step": 15150
+    },
+    {
+      "epoch": 3.668950012824962,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.160706201901313e-06,
+      "loss": 2.0874,
+      "mean_token_accuracy": 0.555564073510468,
+      "num_input_tokens_seen": 7342169889,
+      "num_tokens": 3094140237.0,
+      "step": 15200
+    },
+    {
+      "epoch": 3.681020565203615,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.9720838992002415e-06,
+      "loss": 2.1014,
+      "mean_token_accuracy": 0.5540463343262673,
+      "num_input_tokens_seen": 7366510321,
+      "num_tokens": 3104349542.0,
+      "step": 15250
+    },
+    {
+      "epoch": 3.6930911175822683,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.78346159649917e-06,
+      "loss": 2.1128,
+      "mean_token_accuracy": 0.5523605942726135,
+      "num_input_tokens_seen": 7390741489,
+      "num_tokens": 3114615388.0,
+      "step": 15300
+    },
+    {
+      "epoch": 3.7051616699609218,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.594839293798099e-06,
+      "loss": 2.0836,
+      "mean_token_accuracy": 0.5556057692691684,
+      "num_input_tokens_seen": 7414960897,
+      "num_tokens": 3124852831.0,
+      "step": 15350
+    },
+    {
+      "epoch": 3.717232222339575,
+      "grad_norm": 0.234375,
+      "learning_rate": 4.406216991097027e-06,
+      "loss": 2.0832,
+      "mean_token_accuracy": 0.557135313116014,
+      "num_input_tokens_seen": 7439141473,
+      "num_tokens": 3135118289.0,
+      "step": 15400
+    },
+    {
+      "epoch": 3.729302774718228,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.217594688395956e-06,
+      "loss": 2.0993,
+      "mean_token_accuracy": 0.554307484254241,
+      "num_input_tokens_seen": 7463114529,
+      "num_tokens": 3145345547.0,
+      "step": 15450
+    },
+    {
+      "epoch": 3.7413733270968814,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.028972385694885e-06,
+      "loss": 2.0916,
+      "num_input_tokens_seen": 7487502401,
+      "step": 15500
+    },
+    {
+      "epoch": 3.7413733270968814,
+      "eval_loss": 1.9680771827697754,
+      "eval_mean_token_accuracy": 0.5784905481975056,
+      "eval_num_tokens": 3155655029.0,
+      "eval_runtime": 131.7036,
+      "eval_samples_per_second": 81.334,
+      "eval_steps_per_second": 20.334,
+      "num_input_tokens_seen": 7487502401,
+      "step": 15500
+    },
+    {
+      "epoch": 3.7534438794755345,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.840350082993813e-06,
+      "loss": 2.0919,
+      "mean_token_accuracy": 0.5555992320179939,
+      "num_input_tokens_seen": 7511609537,
+      "num_tokens": 3165956280.0,
+      "step": 15550
+    },
+    {
+      "epoch": 3.7655144318541875,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.6517277802927423e-06,
+      "loss": 2.096,
+      "mean_token_accuracy": 0.5547948920354248,
+      "num_input_tokens_seen": 7535873665,
+      "num_tokens": 3176126007.0,
+      "step": 15600
+    },
+    {
+      "epoch": 3.777584984232841,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.463105477591671e-06,
+      "loss": 2.0963,
+      "mean_token_accuracy": 0.5549294283241033,
+      "num_input_tokens_seen": 7560110225,
+      "num_tokens": 3186367919.0,
+      "step": 15650
+    },
+    {
+      "epoch": 3.789655536611494,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.274483174890599e-06,
+      "loss": 2.0986,
+      "mean_token_accuracy": 0.5547161266207695,
+      "num_input_tokens_seen": 7584157249,
+      "num_tokens": 3196441494.0,
+      "step": 15700
+    },
+    {
+      "epoch": 3.801726088990147,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.0858608721895278e-06,
+      "loss": 2.0952,
+      "mean_token_accuracy": 0.554626210257411,
+      "num_input_tokens_seen": 7608282737,
+      "num_tokens": 3206560665.0,
+      "step": 15750
+    },
+    {
+      "epoch": 3.8137966413688007,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.8972385694884564e-06,
+      "loss": 2.0874,
+      "mean_token_accuracy": 0.5568051477894187,
+      "num_input_tokens_seen": 7632420385,
+      "num_tokens": 3216638197.0,
+      "step": 15800
+    },
+    {
+      "epoch": 3.8258671937474538,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.708616266787385e-06,
+      "loss": 2.0935,
+      "mean_token_accuracy": 0.5554833044111729,
+      "num_input_tokens_seen": 7656546049,
+      "num_tokens": 3226721685.0,
+      "step": 15850
+    },
+    {
+      "epoch": 3.837937746126107,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.5199939640863136e-06,
+      "loss": 2.1087,
+      "mean_token_accuracy": 0.5525853624939918,
+      "num_input_tokens_seen": 7680650689,
+      "num_tokens": 3236963270.0,
+      "step": 15900
+    },
+    {
+      "epoch": 3.8500082985047603,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.3313716613852423e-06,
+      "loss": 2.0952,
+      "mean_token_accuracy": 0.5550886183232069,
+      "num_input_tokens_seen": 7704888257,
+      "num_tokens": 3247161827.0,
+      "step": 15950
+    },
+    {
+      "epoch": 3.8620788508834134,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.142749358684171e-06,
+      "loss": 2.0914,
+      "num_input_tokens_seen": 7729128049,
+      "step": 16000
+    },
+    {
+      "epoch": 3.8620788508834134,
+      "eval_loss": 1.9680593013763428,
+      "eval_mean_token_accuracy": 0.5785026788667034,
+      "eval_num_tokens": 3257367290.0,
+      "eval_runtime": 130.8705,
+      "eval_samples_per_second": 81.852,
+      "eval_steps_per_second": 20.463,
+      "num_input_tokens_seen": 7729128049,
+      "step": 16000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 7245951473,
+  "num_input_tokens_seen": 7729128049,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -3297,7 +3515,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.9383627395138765e+18,
+  "total_flos": 2.0676171893972582e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null