Training in progress, step 8000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc111bfdca8ed66f8f79b1bd7bfe63b080b3cd0c7813c6a04d18b038095b2076
+oid sha256:b6c29be95fa54dd4f69e3ce5004267e101dda94af373aaf70d214f9638f89adb
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:012e342ea6ebe917c7ae83cdc62c745f971ca211463bb0b9f52f2a90259f7532
+oid sha256:160ce5599809f8d46aece51b073c165f9acf6e1fe58b27e320879b3a36e071e8
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b7086e9014350d6181c560b7f34a0e20e6a473f5b7d4ab3f99a9989189826cf1
+oid sha256:0b010e746bb61fb9d87a1e1dd8c3322ee496b9630835db02b085aa96f590f941
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2eb03be752bb38e04dcd3624dd94e7f08bf42e1daf7a9f0b1188fbafdad08914
+oid sha256:1fbdeae10eb93c5869f1e3bdcd9e48de4b232e3e091208a64c7200c54dba28a3
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.6897113629162455,
+  "epoch": 1.9311224104893099,
   "eval_steps": 500,
-  "global_step": 7000,
+  "global_step": 8000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1534,11 +1534,229 @@
       "eval_steps_per_second": 20.529,
       "num_input_tokens_seen": 3382174368,
       "step": 7000
+    },
+    {
+      "epoch": 1.7017819152948985,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.590614154217595e-05,
+      "loss": 2.0955,
+      "mean_token_accuracy": 0.55618498865515,
+      "num_input_tokens_seen": 3406371872,
+      "num_tokens": 1435550428.0,
+      "step": 7050
+    },
+    {
+      "epoch": 1.713852467673552,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.571751923947488e-05,
+      "loss": 2.0919,
+      "mean_token_accuracy": 0.5553148340806365,
+      "num_input_tokens_seen": 3430521536,
+      "num_tokens": 1445655664.0,
+      "step": 7100
+    },
+    {
+      "epoch": 1.725923020052205,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.5528896936773805e-05,
+      "loss": 2.0906,
+      "mean_token_accuracy": 0.5547146466746926,
+      "num_input_tokens_seen": 3454573728,
+      "num_tokens": 1455866347.0,
+      "step": 7150
+    },
+    {
+      "epoch": 1.7379935724308584,
+      "grad_norm": 0.265625,
+      "learning_rate": 3.534027463407274e-05,
+      "loss": 2.098,
+      "mean_token_accuracy": 0.5544127273187042,
+      "num_input_tokens_seen": 3478641648,
+      "num_tokens": 1465996080.0,
+      "step": 7200
+    },
+    {
+      "epoch": 1.7500641248095117,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.515165233137166e-05,
+      "loss": 2.1039,
+      "mean_token_accuracy": 0.5536015385761857,
+      "num_input_tokens_seen": 3502632176,
+      "num_tokens": 1476133852.0,
+      "step": 7250
+    },
+    {
+      "epoch": 1.7621346771881647,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 3.496303002867059e-05,
+      "loss": 2.0974,
+      "mean_token_accuracy": 0.5540728243440389,
+      "num_input_tokens_seen": 3526775024,
+      "num_tokens": 1486346988.0,
+      "step": 7300
+    },
+    {
+      "epoch": 1.774205229566818,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.477440772596952e-05,
+      "loss": 2.0986,
+      "mean_token_accuracy": 0.5539619905874134,
+      "num_input_tokens_seen": 3551057232,
+      "num_tokens": 1496540907.0,
+      "step": 7350
+    },
+    {
+      "epoch": 1.7862757819454713,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.4585785423268445e-05,
+      "loss": 2.097,
+      "mean_token_accuracy": 0.5543709811195732,
+      "num_input_tokens_seen": 3575145120,
+      "num_tokens": 1506733776.0,
+      "step": 7400
+    },
+    {
+      "epoch": 1.7983463343241244,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.4397163120567377e-05,
+      "loss": 2.1054,
+      "mean_token_accuracy": 0.5529748990386725,
+      "num_input_tokens_seen": 3599324048,
+      "num_tokens": 1516881364.0,
+      "step": 7450
+    },
+    {
+      "epoch": 1.810416886702778,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.420854081786631e-05,
+      "loss": 2.107,
+      "num_input_tokens_seen": 3623572960,
+      "step": 7500
+    },
+    {
+      "epoch": 1.810416886702778,
+      "eval_loss": 1.9704335927963257,
+      "eval_mean_token_accuracy": 0.5782234972207915,
+      "eval_num_tokens": 1527213100.0,
+      "eval_runtime": 130.0921,
+      "eval_samples_per_second": 82.342,
+      "eval_steps_per_second": 20.585,
+      "num_input_tokens_seen": 3623572960,
+      "step": 7500
+    },
+    {
+      "epoch": 1.822487439081431,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.4019918515165234e-05,
+      "loss": 2.0859,
+      "mean_token_accuracy": 0.5543704128451645,
+      "num_input_tokens_seen": 3647842624,
+      "num_tokens": 1537348409.0,
+      "step": 7550
+    },
+    {
+      "epoch": 1.8345579914600842,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.3831296212464166e-05,
+      "loss": 2.0964,
+      "mean_token_accuracy": 0.5545977150648832,
+      "num_input_tokens_seen": 3671838176,
+      "num_tokens": 1547405691.0,
+      "step": 7600
+    },
+    {
+      "epoch": 1.8466285438387375,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.364267390976309e-05,
+      "loss": 2.1049,
+      "mean_token_accuracy": 0.5531406961008907,
+      "num_input_tokens_seen": 3696051376,
+      "num_tokens": 1557614207.0,
+      "step": 7650
+    },
+    {
+      "epoch": 1.8586990962173906,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.345405160706202e-05,
+      "loss": 2.0976,
+      "mean_token_accuracy": 0.5547824421525002,
+      "num_input_tokens_seen": 3720242912,
+      "num_tokens": 1567813391.0,
+      "step": 7700
+    },
+    {
+      "epoch": 1.8707696485960439,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.326542930436095e-05,
+      "loss": 2.0941,
+      "mean_token_accuracy": 0.5552040388435125,
+      "num_input_tokens_seen": 3744303744,
+      "num_tokens": 1577952759.0,
+      "step": 7750
+    },
+    {
+      "epoch": 1.8828402009746972,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.3076807001659874e-05,
+      "loss": 2.107,
+      "mean_token_accuracy": 0.5526882111281156,
+      "num_input_tokens_seen": 3768596640,
+      "num_tokens": 1588260331.0,
+      "step": 7800
+    },
+    {
+      "epoch": 1.8949107533533502,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.2888184698958806e-05,
+      "loss": 2.0974,
+      "mean_token_accuracy": 0.5548465251550079,
+      "num_input_tokens_seen": 3792634928,
+      "num_tokens": 1598444722.0,
+      "step": 7850
+    },
+    {
+      "epoch": 1.9069813057320035,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.269956239625773e-05,
+      "loss": 2.0957,
+      "mean_token_accuracy": 0.554438531845808,
+      "num_input_tokens_seen": 3816849872,
+      "num_tokens": 1608557410.0,
+      "step": 7900
+    },
+    {
+      "epoch": 1.9190518581106568,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 3.251094009355666e-05,
+      "loss": 2.108,
+      "mean_token_accuracy": 0.5528679783269763,
+      "num_input_tokens_seen": 3840903600,
+      "num_tokens": 1618721583.0,
+      "step": 7950
+    },
+    {
+      "epoch": 1.9311224104893099,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.2322317790855595e-05,
+      "loss": 2.0908,
+      "num_input_tokens_seen": 3864935104,
+      "step": 8000
+    },
+    {
+      "epoch": 1.9311224104893099,
+      "eval_loss": 1.9698705673217773,
+      "eval_mean_token_accuracy": 0.5783014823866923,
+      "eval_num_tokens": 1628824365.0,
+      "eval_runtime": 130.0192,
+      "eval_samples_per_second": 82.388,
+      "eval_steps_per_second": 20.597,
+      "num_input_tokens_seen": 3864935104,
+      "step": 8000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 3382174368,
+  "num_input_tokens_seen": 3864935104,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -1553,7 +1771,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 9.047646534618317e+17,
+  "total_flos": 1.033907862086615e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null