Training in progress, step 13000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:20ce803e4fe3b56cd4360f4db9480f38ae31a8a1afff95c81fdf52f10ba0ac27
+oid sha256:23be2c11c244c72601ea6f47dd507781736231ff1da2289fe5f8ba433277cb99
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:57fe321ef28d6e68df66c93386db17dd5a5a544cc930d8f0b18e8919d55948ac
+oid sha256:04943bdcad0923c88796f61e80a911b94cde9c121a1bb27006e82c8a584a0c44
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9ed029202dee104daa8ff402cceceb0954b69804876f1536e7023f304ea17a5
+oid sha256:ff84b2998c9ce4e6e3eaf03e775fc93a7c11be8195c0bb3abb7a8b9a1cec86e5
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26e93f6b8bba63ca3cafe7523f1374f6bb0c88ef8963cfe1774bbd335ba7bcbe
+oid sha256:71a524f67e79e2b512d6d818f94e2b528e5b7f4447259f3966ae44cdba439db5
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.8966006306863616,
+  "epoch": 3.1378457081642197,
   "eval_steps": 500,
-  "global_step": 12000,
+  "global_step": 13000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2624,11 +2624,229 @@
       "eval_steps_per_second": 20.507,
       "num_input_tokens_seen": 5797304337,
       "step": 12000
+    },
+    {
+      "epoch": 2.908671183065015,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.704391127206881e-05,
+      "loss": 2.0958,
+      "mean_token_accuracy": 0.5546299646422267,
+      "num_input_tokens_seen": 5821442385,
+      "num_tokens": 2453453845.0,
+      "step": 12050
+    },
+    {
+      "epoch": 2.920741735443668,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.685528896936774e-05,
+      "loss": 2.0926,
+      "mean_token_accuracy": 0.5549974143505096,
+      "num_input_tokens_seen": 5845686961,
+      "num_tokens": 2463776050.0,
+      "step": 12100
+    },
+    {
+      "epoch": 2.9328122878223217,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 2.1015,
+      "mean_token_accuracy": 0.5541527543962002,
+      "num_input_tokens_seen": 5869745137,
+      "num_tokens": 2473828195.0,
+      "step": 12150
+    },
+    {
+      "epoch": 2.944882840200975,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.6478044363965596e-05,
+      "loss": 2.1041,
+      "mean_token_accuracy": 0.5541104365140199,
+      "num_input_tokens_seen": 5893803025,
+      "num_tokens": 2483915340.0,
+      "step": 12200
+    },
+    {
+      "epoch": 2.956953392579628,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.6289422061264525e-05,
+      "loss": 2.0922,
+      "mean_token_accuracy": 0.5555301706120371,
+      "num_input_tokens_seen": 5918068641,
+      "num_tokens": 2494208052.0,
+      "step": 12250
+    },
+    {
+      "epoch": 2.9690239449582814,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.6100799758563453e-05,
+      "loss": 2.0938,
+      "mean_token_accuracy": 0.5548020200431347,
+      "num_input_tokens_seen": 5942257041,
+      "num_tokens": 2504393372.0,
+      "step": 12300
+    },
+    {
+      "epoch": 2.9810944973369344,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.5912177455862382e-05,
+      "loss": 2.0843,
+      "mean_token_accuracy": 0.5566597804427147,
+      "num_input_tokens_seen": 5966422081,
+      "num_tokens": 2514627798.0,
+      "step": 12350
+    },
+    {
+      "epoch": 2.9931650497155875,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.572355515316131e-05,
+      "loss": 2.0886,
+      "mean_token_accuracy": 0.5566082544624805,
+      "num_input_tokens_seen": 5990574321,
+      "num_tokens": 2524805007.0,
+      "step": 12400
+    },
+    {
+      "epoch": 3.005069631999034,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.553493285046024e-05,
+      "loss": 2.1001,
+      "mean_token_accuracy": 0.5549402527407246,
+      "num_input_tokens_seen": 6014380145,
+      "num_tokens": 2534738802.0,
+      "step": 12450
+    },
+    {
+      "epoch": 3.0171401843776877,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.5346310547759168e-05,
+      "loss": 2.092,
+      "num_input_tokens_seen": 6038556753,
+      "step": 12500
+    },
+    {
+      "epoch": 3.0171401843776877,
+      "eval_loss": 1.9681233167648315,
+      "eval_mean_token_accuracy": 0.5784891846355349,
+      "eval_num_tokens": 2544886437.0,
+      "eval_runtime": 130.6689,
+      "eval_samples_per_second": 81.978,
+      "eval_steps_per_second": 20.495,
+      "num_input_tokens_seen": 6038556753,
+      "step": 12500
+    },
+    {
+      "epoch": 3.029210736756341,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.5157688245058096e-05,
+      "loss": 2.0925,
+      "mean_token_accuracy": 0.5550393326207995,
+      "num_input_tokens_seen": 6062857617,
+      "num_tokens": 2555112151.0,
+      "step": 12550
+    },
+    {
+      "epoch": 3.041281289134994,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4969065942357025e-05,
+      "loss": 2.0957,
+      "mean_token_accuracy": 0.5551863227039575,
+      "num_input_tokens_seen": 6087077841,
+      "num_tokens": 2565388515.0,
+      "step": 12600
+    },
+    {
+      "epoch": 3.0533518415136474,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.4780443639655952e-05,
+      "loss": 2.0858,
+      "mean_token_accuracy": 0.5563259933143854,
+      "num_input_tokens_seen": 6111161617,
+      "num_tokens": 2575504513.0,
+      "step": 12650
+    },
+    {
+      "epoch": 3.0654223938923004,
+      "grad_norm": 0.25,
+      "learning_rate": 1.4591821336954884e-05,
+      "loss": 2.101,
+      "mean_token_accuracy": 0.5549140437319875,
+      "num_input_tokens_seen": 6135170369,
+      "num_tokens": 2585570498.0,
+      "step": 12700
+    },
+    {
+      "epoch": 3.077492946270954,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.4403199034253811e-05,
+      "loss": 2.0935,
+      "mean_token_accuracy": 0.5543564364686608,
+      "num_input_tokens_seen": 6159397985,
+      "num_tokens": 2595740107.0,
+      "step": 12750
+    },
+    {
+      "epoch": 3.089563498649607,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.421457673155274e-05,
+      "loss": 2.0928,
+      "mean_token_accuracy": 0.5548016136884689,
+      "num_input_tokens_seen": 6183511137,
+      "num_tokens": 2605900153.0,
+      "step": 12800
+    },
+    {
+      "epoch": 3.10163405102826,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.4025954428851668e-05,
+      "loss": 2.0862,
+      "mean_token_accuracy": 0.5555924268066883,
+      "num_input_tokens_seen": 6207630993,
+      "num_tokens": 2616105592.0,
+      "step": 12850
+    },
+    {
+      "epoch": 3.1137046034069136,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.3837332126150595e-05,
+      "loss": 2.0938,
+      "mean_token_accuracy": 0.554584386125207,
+      "num_input_tokens_seen": 6231763217,
+      "num_tokens": 2626268256.0,
+      "step": 12900
+    },
+    {
+      "epoch": 3.1257751557855666,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3648709823449527e-05,
+      "loss": 2.1042,
+      "mean_token_accuracy": 0.553115917481482,
+      "num_input_tokens_seen": 6255995041,
+      "num_tokens": 2636461653.0,
+      "step": 12950
+    },
+    {
+      "epoch": 3.1378457081642197,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.3460087520748454e-05,
+      "loss": 2.0952,
+      "num_input_tokens_seen": 6280158129,
+      "step": 13000
+    },
+    {
+      "epoch": 3.1378457081642197,
+      "eval_loss": 1.9681209325790405,
+      "eval_mean_token_accuracy": 0.5785721040555485,
+      "eval_num_tokens": 2646712354.0,
+      "eval_runtime": 130.3881,
+      "eval_samples_per_second": 82.155,
+      "eval_steps_per_second": 20.539,
+      "num_input_tokens_seen": 6280158129,
+      "step": 13000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 5797304337,
+  "num_input_tokens_seen": 6280158129,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -2643,7 +2861,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.550835491837829e+18,
+  "total_flos": 1.680003593850839e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null