Training in progress, step 2000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4ceb1eb66e59b2be3b7b56c92ceeee0dff3d3afb66f4e798a3c7f7a973769da
+oid sha256:48c8ccaea45630ddf9ea1b88f0156cd1fedd61905cf18898819db2276984fcb7
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:300510ba8259c34ff91864564df369f7ef5da88e12e952b5fc750f0733f75299
+oid sha256:d7b50e0d348776f6bfde587ad7885adcff251381a86d03c221fe493591258125
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dec81976f45ae46b24f8606fae652d67e12c656a1fb7001a6621b0275456c831
+oid sha256:ddd99e2663b559b91167c1d3714d334dd46c05cbb133b69d7b52f7b883548de2
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ebbd97a5f85f91a146bd613ec7c9e8b4ff3314b259cf17fa7c45467f21162760
+oid sha256:7286e34301379ab62676eaae8fceab28248cb123c9029e5c4370161a2e81133d
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.24141104757306456,
+  "epoch": 0.4828220951461291,
   "eval_steps": 500,
-  "global_step": 1000,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -226,11 +226,229 @@
       "eval_steps_per_second": 21.33,
       "num_input_tokens_seen": 482701824,
       "step": 1000
+    },
+    {
+      "epoch": 0.2534815999517178,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.584188292094146e-05,
+      "loss": 2.557,
+      "mean_token_accuracy": 0.5126943441852927,
+      "num_input_tokens_seen": 506798000,
+      "num_tokens": 213604239.0,
+      "step": 1050
+    },
+    {
+      "epoch": 0.265552152330371,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.659625829812915e-05,
+      "loss": 2.5533,
+      "mean_token_accuracy": 0.5126326360553503,
+      "num_input_tokens_seen": 530786496,
+      "num_tokens": 223786306.0,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2776227047090242,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7350633675316838e-05,
+      "loss": 2.5528,
+      "mean_token_accuracy": 0.5130741761997342,
+      "num_input_tokens_seen": 554798768,
+      "num_tokens": 233857737.0,
+      "step": 1150
+    },
+    {
+      "epoch": 0.28969325708767746,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8105009052504525e-05,
+      "loss": 2.5433,
+      "mean_token_accuracy": 0.5140750538557768,
+      "num_input_tokens_seen": 578877696,
+      "num_tokens": 243994162.0,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3017638094663307,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8859384429692215e-05,
+      "loss": 2.5414,
+      "mean_token_accuracy": 0.5151798555627465,
+      "num_input_tokens_seen": 603141184,
+      "num_tokens": 254196491.0,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3138343618449839,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9613759806879906e-05,
+      "loss": 2.5383,
+      "mean_token_accuracy": 0.5148227337375283,
+      "num_input_tokens_seen": 627221936,
+      "num_tokens": 264309463.0,
+      "step": 1300
+    },
+    {
+      "epoch": 0.32590491422363715,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.0368135184067593e-05,
+      "loss": 2.5292,
+      "mean_token_accuracy": 0.5156710411980748,
+      "num_input_tokens_seen": 651386848,
+      "num_tokens": 274462595.0,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3379754666022904,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.1122510561255283e-05,
+      "loss": 2.523,
+      "mean_token_accuracy": 0.5161588852107525,
+      "num_input_tokens_seen": 675579984,
+      "num_tokens": 284636254.0,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3500460189809436,
+      "grad_norm": 6.375,
+      "learning_rate": 2.187688593844297e-05,
+      "loss": 2.528,
+      "mean_token_accuracy": 0.5156370849534869,
+      "num_input_tokens_seen": 699751424,
+      "num_tokens": 294907617.0,
+      "step": 1450
+    },
+    {
+      "epoch": 0.36211657135959685,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.2631261315630658e-05,
+      "loss": 2.5232,
+      "num_input_tokens_seen": 723883888,
+      "step": 1500
+    },
+    {
+      "epoch": 0.36211657135959685,
+      "eval_loss": 2.42679500579834,
+      "eval_mean_token_accuracy": 0.5353444569074368,
+      "eval_num_tokens": 305087605.0,
+      "eval_runtime": 125.481,
+      "eval_samples_per_second": 85.367,
+      "eval_steps_per_second": 21.342,
+      "num_input_tokens_seen": 723883888,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3741871237382501,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.3385636692818348e-05,
+      "loss": 2.5178,
+      "mean_token_accuracy": 0.5161234551295638,
+      "num_input_tokens_seen": 748106064,
+      "num_tokens": 315240015.0,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3862576761169033,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.4140012070006035e-05,
+      "loss": 2.5157,
+      "mean_token_accuracy": 0.5167792574688792,
+      "num_input_tokens_seen": 772242320,
+      "num_tokens": 325393577.0,
+      "step": 1600
+    },
+    {
+      "epoch": 0.39832822849555655,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4894387447193726e-05,
+      "loss": 2.505,
+      "mean_token_accuracy": 0.5181788290664554,
+      "num_input_tokens_seen": 796536976,
+      "num_tokens": 335636146.0,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4103987808742098,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5648762824381413e-05,
+      "loss": 2.5034,
+      "mean_token_accuracy": 0.5191374982148409,
+      "num_input_tokens_seen": 820727792,
+      "num_tokens": 345721073.0,
+      "step": 1700
+    },
+    {
+      "epoch": 0.422469333252863,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.64031382015691e-05,
+      "loss": 2.5049,
+      "mean_token_accuracy": 0.5170740441232919,
+      "num_input_tokens_seen": 844900704,
+      "num_tokens": 355941787.0,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4345398856315162,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.7157513578756787e-05,
+      "loss": 2.4944,
+      "mean_token_accuracy": 0.5191365649551153,
+      "num_input_tokens_seen": 868882304,
+      "num_tokens": 366140276.0,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4466104380101694,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.791188895594448e-05,
+      "loss": 2.4852,
+      "mean_token_accuracy": 0.519771606773138,
+      "num_input_tokens_seen": 893201120,
+      "num_tokens": 376397263.0,
+      "step": 1850
+    },
+    {
+      "epoch": 0.45868099038882265,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.866626433313217e-05,
+      "loss": 2.4242,
+      "mean_token_accuracy": 0.5218987537547946,
+      "num_input_tokens_seen": 917474880,
+      "num_tokens": 386613084.0,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4707515427674759,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.942063971031986e-05,
+      "loss": 2.4118,
+      "mean_token_accuracy": 0.5205260647833347,
+      "num_input_tokens_seen": 941694304,
+      "num_tokens": 396813412.0,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4828220951461291,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.0175015087507546e-05,
+      "loss": 2.3856,
+      "num_input_tokens_seen": 965894272,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4828220951461291,
+      "eval_loss": 2.281926155090332,
+      "eval_mean_token_accuracy": 0.5452906315098778,
+      "eval_num_tokens": 407043305.0,
+      "eval_runtime": 126.3646,
+      "eval_samples_per_second": 84.771,
+      "eval_steps_per_second": 21.193,
+      "num_input_tokens_seen": 965894272,
+      "step": 2000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 482701824,
+  "num_input_tokens_seen": 965894272,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -245,7 +463,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.2912744908980224e+17,
+  "total_flos": 2.5838614488809472e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null