Training in progress, step 9000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6c29be95fa54dd4f69e3ce5004267e101dda94af373aaf70d214f9638f89adb
+oid sha256:af73c4ec485cd7fa414342b390a5c634c47a31d116a73e322cc418d51a330596
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:160ce5599809f8d46aece51b073c165f9acf6e1fe58b27e320879b3a36e071e8
+oid sha256:f8057071e9c871132fbac5baaaef5c6aca4c49e2663c7a32995eef4dffca1eb9
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0b010e746bb61fb9d87a1e1dd8c3322ee496b9630835db02b085aa96f590f941
+oid sha256:1674983d22ea028f37f625821f1fca77be67adb3636d14701707c06c0fbac379
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1fbdeae10eb93c5869f1e3bdcd9e48de4b232e3e091208a64c7200c54dba28a3
+oid sha256:fb36ce4646595e0955071e0d49fcfefa2b2d576fde195d65fd821c4d2bebc721
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.9311224104893099,
+  "epoch": 2.172367487967168,
   "eval_steps": 500,
-  "global_step": 8000,
+  "global_step": 9000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1752,11 +1752,229 @@
       "eval_steps_per_second": 20.597,
       "num_input_tokens_seen": 3864935104,
       "step": 8000
+    },
+    {
+      "epoch": 1.9431929628679634,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.213369548815452e-05,
+      "loss": 2.096,
+      "mean_token_accuracy": 0.5548218312300741,
+      "num_input_tokens_seen": 3889193776,
+      "num_tokens": 1638986755.0,
+      "step": 8050
+    },
+    {
+      "epoch": 1.9552635152466165,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.194507318545345e-05,
+      "loss": 2.0821,
+      "mean_token_accuracy": 0.5565256755426526,
+      "num_input_tokens_seen": 3913273664,
+      "num_tokens": 1649209062.0,
+      "step": 8100
+    },
+    {
+      "epoch": 1.9673340676252697,
+      "grad_norm": 0.267578125,
+      "learning_rate": 3.175645088275238e-05,
+      "loss": 2.1026,
+      "mean_token_accuracy": 0.5532021636888385,
+      "num_input_tokens_seen": 3937597616,
+      "num_tokens": 1659435713.0,
+      "step": 8150
+    },
+    {
+      "epoch": 1.979404620003923,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.156782858005131e-05,
+      "loss": 2.0984,
+      "mean_token_accuracy": 0.554893646761775,
+      "num_input_tokens_seen": 3961803792,
+      "num_tokens": 1669582457.0,
+      "step": 8200
+    },
+    {
+      "epoch": 1.991475172382576,
+      "grad_norm": 0.25,
+      "learning_rate": 3.1379206277350235e-05,
+      "loss": 2.0818,
+      "mean_token_accuracy": 0.5563116483017803,
+      "num_input_tokens_seen": 3985940144,
+      "num_tokens": 1679783022.0,
+      "step": 8250
+    },
+    {
+      "epoch": 2.003379754666023,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.119058397464916e-05,
+      "loss": 2.0986,
+      "mean_token_accuracy": 0.5548089014411124,
+      "num_input_tokens_seen": 4009775217,
+      "num_tokens": 1689934301.0,
+      "step": 8300
+    },
+    {
+      "epoch": 2.0154503070446763,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.100196167194809e-05,
+      "loss": 2.101,
+      "mean_token_accuracy": 0.554502502605319,
+      "num_input_tokens_seen": 4033881233,
+      "num_tokens": 1700106136.0,
+      "step": 8350
+    },
+    {
+      "epoch": 2.0275208594233294,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.081333936924702e-05,
+      "loss": 2.0952,
+      "mean_token_accuracy": 0.5547482476383447,
+      "num_input_tokens_seen": 4058061905,
+      "num_tokens": 1710288179.0,
+      "step": 8400
+    },
+    {
+      "epoch": 2.0395914118019824,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.062471706654595e-05,
+      "loss": 2.0997,
+      "mean_token_accuracy": 0.5538438270241022,
+      "num_input_tokens_seen": 4082118001,
+      "num_tokens": 1720402394.0,
+      "step": 8450
+    },
+    {
+      "epoch": 2.051661964180636,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.043609476384488e-05,
+      "loss": 2.0904,
+      "num_input_tokens_seen": 4106175169,
+      "step": 8500
+    },
+    {
+      "epoch": 2.051661964180636,
+      "eval_loss": 1.969247817993164,
+      "eval_mean_token_accuracy": 0.5783566591497497,
+      "eval_num_tokens": 1730551722.0,
+      "eval_runtime": 130.525,
+      "eval_samples_per_second": 82.069,
+      "eval_steps_per_second": 20.517,
+      "num_input_tokens_seen": 4106175169,
+      "step": 8500
+    },
+    {
+      "epoch": 2.063732516559289,
+      "grad_norm": 0.244140625,
+      "learning_rate": 3.0247472461143806e-05,
+      "loss": 2.1045,
+      "mean_token_accuracy": 0.5544156692735851,
+      "num_input_tokens_seen": 4130212065,
+      "num_tokens": 1740781983.0,
+      "step": 8550
+    },
+    {
+      "epoch": 2.075803068937942,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.0058850158442735e-05,
+      "loss": 2.0968,
+      "mean_token_accuracy": 0.5541778185963631,
+      "num_input_tokens_seen": 4154342289,
+      "num_tokens": 1750886868.0,
+      "step": 8600
+    },
+    {
+      "epoch": 2.0878736213165956,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.9870227855741667e-05,
+      "loss": 2.0825,
+      "mean_token_accuracy": 0.5571553486213088,
+      "num_input_tokens_seen": 4178622977,
+      "num_tokens": 1761024406.0,
+      "step": 8650
+    },
+    {
+      "epoch": 2.0999441736952487,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.9681605553040592e-05,
+      "loss": 2.0936,
+      "mean_token_accuracy": 0.554418184608221,
+      "num_input_tokens_seen": 4202833969,
+      "num_tokens": 1771241652.0,
+      "step": 8700
+    },
+    {
+      "epoch": 2.112014726073902,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.9492983250339524e-05,
+      "loss": 2.0959,
+      "mean_token_accuracy": 0.5550215977802873,
+      "num_input_tokens_seen": 4226915233,
+      "num_tokens": 1781349215.0,
+      "step": 8750
+    },
+    {
+      "epoch": 2.1240852784525552,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.930436094763845e-05,
+      "loss": 2.1008,
+      "mean_token_accuracy": 0.553907332457602,
+      "num_input_tokens_seen": 4251137105,
+      "num_tokens": 1791502343.0,
+      "step": 8800
+    },
+    {
+      "epoch": 2.1361558308312083,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.9115738644937378e-05,
+      "loss": 2.0903,
+      "mean_token_accuracy": 0.5557647632434964,
+      "num_input_tokens_seen": 4275224433,
+      "num_tokens": 1801620139.0,
+      "step": 8850
+    },
+    {
+      "epoch": 2.148226383209862,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.892711634223631e-05,
+      "loss": 2.105,
+      "mean_token_accuracy": 0.5533201249688864,
+      "num_input_tokens_seen": 4299490545,
+      "num_tokens": 1811829692.0,
+      "step": 8900
+    },
+    {
+      "epoch": 2.160296935588515,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.8738494039535235e-05,
+      "loss": 2.0771,
+      "mean_token_accuracy": 0.5572306806966663,
+      "num_input_tokens_seen": 4323603905,
+      "num_tokens": 1822011165.0,
+      "step": 8950
+    },
+    {
+      "epoch": 2.172367487967168,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.8549871736834167e-05,
+      "loss": 2.0766,
+      "num_input_tokens_seen": 4347894913,
+      "step": 9000
+    },
+    {
+      "epoch": 2.172367487967168,
+      "eval_loss": 1.968759536743164,
+      "eval_mean_token_accuracy": 0.5784085558767724,
+      "eval_num_tokens": 1832187809.0,
+      "eval_runtime": 130.1161,
+      "eval_samples_per_second": 82.326,
+      "eval_steps_per_second": 20.582,
+      "num_input_tokens_seen": 4347894913,
+      "step": 9000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 3864935104,
+  "num_input_tokens_seen": 4347894913,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -1771,7 +1989,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.033907862086615e+18,
+  "total_flos": 1.163104324681851e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null