Training in progress, step 12000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd378ab1a42d536af5db20740b7c6ba4c863b9ff3eeb07dfe7d4b811a689ab5f
+oid sha256:20ce803e4fe3b56cd4360f4db9480f38ae31a8a1afff95c81fdf52f10ba0ac27
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d5f198053dbdfaa8c376a0fdaef1cec44750b494d68cc275559bc743db6f9c6
+oid sha256:57fe321ef28d6e68df66c93386db17dd5a5a544cc930d8f0b18e8919d55948ac
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fe27678952c245c0bb175fc5ebd37cf8ebfcd743a407e4957181be4dbbc6146b
+oid sha256:b9ed029202dee104daa8ff402cceceb0954b69804876f1536e7023f304ea17a5
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c68adf80be0bee4802e3498e2b20587f8f5db858b7307e722582d5bdeff1cda7
+oid sha256:26e93f6b8bba63ca3cafe7523f1374f6bb0c88ef8963cfe1774bbd335ba7bcbe
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.6551895831132972,
+  "epoch": 2.8966006306863616,
   "eval_steps": 500,
-  "global_step": 11000,
+  "global_step": 12000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2406,11 +2406,229 @@
       "eval_steps_per_second": 20.413,
       "num_input_tokens_seen": 5314253297,
       "step": 11000
+    },
+    {
+      "epoch": 2.6672601354919503,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.0816357326090238e-05,
+      "loss": 2.0981,
+      "mean_token_accuracy": 0.5546219968609511,
+      "num_input_tokens_seen": 5338466017,
+      "num_tokens": 2249988177.0,
+      "step": 11050
+    },
+    {
+      "epoch": 2.679330687870604,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.0627735023389166e-05,
+      "loss": 2.0921,
+      "mean_token_accuracy": 0.5557056156918406,
+      "num_input_tokens_seen": 5362616753,
+      "num_tokens": 2260143864.0,
+      "step": 11100
+    },
+    {
+      "epoch": 2.691401240249257,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.0439112720688095e-05,
+      "loss": 2.0957,
+      "mean_token_accuracy": 0.555001782849431,
+      "num_input_tokens_seen": 5386751937,
+      "num_tokens": 2270269015.0,
+      "step": 11150
+    },
+    {
+      "epoch": 2.7034717926279104,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.0250490417987024e-05,
+      "loss": 2.0879,
+      "mean_token_accuracy": 0.5558207688108087,
+      "num_input_tokens_seen": 5410821777,
+      "num_tokens": 2280359409.0,
+      "step": 11200
+    },
+    {
+      "epoch": 2.7155423450065634,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.0061868115285952e-05,
+      "loss": 2.0862,
+      "mean_token_accuracy": 0.5561293217167258,
+      "num_input_tokens_seen": 5435109553,
+      "num_tokens": 2290676732.0,
+      "step": 11250
+    },
+    {
+      "epoch": 2.7276128973852165,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.987324581258488e-05,
+      "loss": 2.096,
+      "mean_token_accuracy": 0.5545283930376173,
+      "num_input_tokens_seen": 5459182481,
+      "num_tokens": 2300830078.0,
+      "step": 11300
+    },
+    {
+      "epoch": 2.73968344976387,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.968462350988381e-05,
+      "loss": 2.0944,
+      "mean_token_accuracy": 0.5546667322888971,
+      "num_input_tokens_seen": 5483343233,
+      "num_tokens": 2310996379.0,
+      "step": 11350
+    },
+    {
+      "epoch": 2.751754002142523,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.9496001207182738e-05,
+      "loss": 2.0962,
+      "mean_token_accuracy": 0.5551789667457342,
+      "num_input_tokens_seen": 5507547265,
+      "num_tokens": 2321208987.0,
+      "step": 11400
+    },
+    {
+      "epoch": 2.763824554521176,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9307378904481667e-05,
+      "loss": 2.0824,
+      "mean_token_accuracy": 0.5572770998999477,
+      "num_input_tokens_seen": 5531617505,
+      "num_tokens": 2331304919.0,
+      "step": 11450
+    },
+    {
+      "epoch": 2.7758951068998297,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9118756601780595e-05,
+      "loss": 2.1095,
+      "num_input_tokens_seen": 5555689025,
+      "step": 11500
+    },
+    {
+      "epoch": 2.7758951068998297,
+      "eval_loss": 1.9681649208068848,
+      "eval_mean_token_accuracy": 0.5785138376329798,
+      "eval_num_tokens": 2341489447.0,
+      "eval_runtime": 130.3192,
+      "eval_samples_per_second": 82.198,
+      "eval_steps_per_second": 20.55,
+      "num_input_tokens_seen": 5555689025,
+      "step": 11500
+    },
+    {
+      "epoch": 2.7879656592784827,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8930134299079524e-05,
+      "loss": 2.0913,
+      "mean_token_accuracy": 0.553754635732621,
+      "num_input_tokens_seen": 5579808081,
+      "num_tokens": 2351725815.0,
+      "step": 11550
+    },
+    {
+      "epoch": 2.800036211657136,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8741511996378453e-05,
+      "loss": 2.0961,
+      "mean_token_accuracy": 0.5546263293549418,
+      "num_input_tokens_seen": 5604043089,
+      "num_tokens": 2362015471.0,
+      "step": 11600
+    },
+    {
+      "epoch": 2.8121067640357893,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.855288969367738e-05,
+      "loss": 2.0887,
+      "mean_token_accuracy": 0.555464554913342,
+      "num_input_tokens_seen": 5628213761,
+      "num_tokens": 2372168279.0,
+      "step": 11650
+    },
+    {
+      "epoch": 2.8241773164144424,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.836426739097631e-05,
+      "loss": 2.1049,
+      "mean_token_accuracy": 0.5529934700578452,
+      "num_input_tokens_seen": 5652297233,
+      "num_tokens": 2382280922.0,
+      "step": 11700
+    },
+    {
+      "epoch": 2.8362478687930954,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.817564508827524e-05,
+      "loss": 2.1025,
+      "mean_token_accuracy": 0.5528503654524684,
+      "num_input_tokens_seen": 5676338817,
+      "num_tokens": 2392486600.0,
+      "step": 11750
+    },
+    {
+      "epoch": 2.848318421171749,
+      "grad_norm": 0.25,
+      "learning_rate": 1.7987022785574167e-05,
+      "loss": 2.1018,
+      "mean_token_accuracy": 0.5537810071185231,
+      "num_input_tokens_seen": 5700543889,
+      "num_tokens": 2402663993.0,
+      "step": 11800
+    },
+    {
+      "epoch": 2.860388973550402,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.7798400482873096e-05,
+      "loss": 2.0891,
+      "mean_token_accuracy": 0.5560761171206832,
+      "num_input_tokens_seen": 5724764289,
+      "num_tokens": 2412816351.0,
+      "step": 11850
+    },
+    {
+      "epoch": 2.872459525929055,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.7609778180172024e-05,
+      "loss": 2.0974,
+      "mean_token_accuracy": 0.5541906878352165,
+      "num_input_tokens_seen": 5749009809,
+      "num_tokens": 2423083966.0,
+      "step": 11900
+    },
+    {
+      "epoch": 2.8845300783077086,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.7421155877470953e-05,
+      "loss": 2.0996,
+      "mean_token_accuracy": 0.5539322036504746,
+      "num_input_tokens_seen": 5773128241,
+      "num_tokens": 2433212460.0,
+      "step": 11950
+    },
+    {
+      "epoch": 2.8966006306863616,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.723253357476988e-05,
+      "loss": 2.0986,
+      "num_input_tokens_seen": 5797304337,
+      "step": 12000
+    },
+    {
+      "epoch": 2.8966006306863616,
+      "eval_loss": 1.9681628942489624,
+      "eval_mean_token_accuracy": 0.5785647708146406,
+      "eval_num_tokens": 2443385171.0,
+      "eval_runtime": 130.5866,
+      "eval_samples_per_second": 82.03,
+      "eval_steps_per_second": 20.507,
+      "num_input_tokens_seen": 5797304337,
+      "step": 12000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 5314253297,
+  "num_input_tokens_seen": 5797304337,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -2425,7 +2643,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.4216146240596787e+18,
+  "total_flos": 1.550835491837829e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null