Training in progress, step 4000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e63c3db29c1618638ad881936caf0567d6bf237c2d8358081d7324f323992b88
+oid sha256:205a3991d60a5c28fb1c39f7dbf7a515c4fd4b6685d8240efe51c017acfa36b1
 size 150625560

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c5d06df021b0676b1a33468e46c419e946ec39b40bd40d3bc089bc612bb03c6
+oid sha256:1d50c4bfe654d987803fd6a6960587e83c15bda23187fdd2e49b310d524ae5ac
 size 602335994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:841a17a497b009366c5a93d5dd6be560b5debec016825b3dd528f17c802bd947
+oid sha256:a2fab474267dfdb6f9f735fba3b6956eaa8395da984c318144fab7c0aefa914f
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5361224ea55f23c97a6ce6c66443aa545b9c0f839415b7a0921d0d516508616d
+oid sha256:1982f4530393fe0c872036d1fa81199d7b6bd002acf4619fae87ec9de696f64d
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.7242331427191937,
+  "epoch": 0.9656441902922582,
   "eval_steps": 500,
-  "global_step": 3000,
+  "global_step": 4000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -662,11 +662,229 @@
       "eval_steps_per_second": 21.312,
       "num_input_tokens_seen": 1448897616,
       "step": 3000
+    },
+    {
+      "epoch": 0.7363036950978469,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.6016898008449e-05,
+      "loss": 2.2866,
+      "mean_token_accuracy": 0.535948946569115,
+      "num_input_tokens_seen": 1472853872,
+      "num_tokens": 620644723.0,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7483742474765002,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.6771273385636696e-05,
+      "loss": 2.2713,
+      "mean_token_accuracy": 0.5381996771320701,
+      "num_input_tokens_seen": 1497112768,
+      "num_tokens": 630831881.0,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7604447998551533,
+      "grad_norm": 0.318359375,
+      "learning_rate": 4.752564876282438e-05,
+      "loss": 2.2659,
+      "mean_token_accuracy": 0.5379517000168562,
+      "num_input_tokens_seen": 1521326656,
+      "num_tokens": 640973141.0,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7725153522338066,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.828002414001207e-05,
+      "loss": 2.2534,
+      "mean_token_accuracy": 0.5404072028771043,
+      "num_input_tokens_seen": 1545505712,
+      "num_tokens": 651205945.0,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7845859046124598,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.903439951719976e-05,
+      "loss": 2.2662,
+      "mean_token_accuracy": 0.5395008590817452,
+      "num_input_tokens_seen": 1569597408,
+      "num_tokens": 661422296.0,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7966564569911131,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.978877489438745e-05,
+      "loss": 2.2429,
+      "mean_token_accuracy": 0.5423036898300052,
+      "num_input_tokens_seen": 1593795440,
+      "num_tokens": 671551427.0,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8087270093697663,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.9864191942055236e-05,
+      "loss": 2.234,
+      "mean_token_accuracy": 0.5444167210906744,
+      "num_input_tokens_seen": 1618048688,
+      "num_tokens": 681790219.0,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8207975617484196,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.967556963935416e-05,
+      "loss": 2.2466,
+      "mean_token_accuracy": 0.5417763916775584,
+      "num_input_tokens_seen": 1642159776,
+      "num_tokens": 691978554.0,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8328681141270727,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.9486947336653086e-05,
+      "loss": 2.2358,
+      "mean_token_accuracy": 0.5441526301577687,
+      "num_input_tokens_seen": 1666410544,
+      "num_tokens": 702184762.0,
+      "step": 3450
+    },
+    {
+      "epoch": 0.844938666505726,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.929832503395201e-05,
+      "loss": 2.2274,
+      "num_input_tokens_seen": 1690571888,
+      "step": 3500
+    },
+    {
+      "epoch": 0.844938666505726,
+      "eval_loss": 2.110778331756592,
+      "eval_mean_token_accuracy": 0.5686311067219009,
+      "eval_num_tokens": 712362296.0,
+      "eval_runtime": 127.3714,
+      "eval_samples_per_second": 84.1,
+      "eval_steps_per_second": 21.025,
+      "num_input_tokens_seen": 1690571888,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8570092188843792,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.9109702731250944e-05,
+      "loss": 2.2322,
+      "mean_token_accuracy": 0.5444289642199874,
+      "num_input_tokens_seen": 1714787776,
+      "num_tokens": 722682302.0,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8690797712630324,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.8921080428549876e-05,
+      "loss": 2.2194,
+      "mean_token_accuracy": 0.5459688815101981,
+      "num_input_tokens_seen": 1739009552,
+      "num_tokens": 732823822.0,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8811503236416857,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.87324581258488e-05,
+      "loss": 2.2139,
+      "mean_token_accuracy": 0.5467682545632124,
+      "num_input_tokens_seen": 1763209840,
+      "num_tokens": 743138600.0,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8932208760203388,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.854383582314773e-05,
+      "loss": 2.204,
+      "mean_token_accuracy": 0.5477216844260693,
+      "num_input_tokens_seen": 1787295680,
+      "num_tokens": 753284868.0,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9052914283989921,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.835521352044666e-05,
+      "loss": 2.186,
+      "mean_token_accuracy": 0.5463542007282376,
+      "num_input_tokens_seen": 1811501840,
+      "num_tokens": 763533047.0,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9173619807776453,
+      "grad_norm": 0.294921875,
+      "learning_rate": 4.816659121774559e-05,
+      "loss": 2.1705,
+      "mean_token_accuracy": 0.5472249809652567,
+      "num_input_tokens_seen": 1835579680,
+      "num_tokens": 773772552.0,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9294325331562986,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.797796891504452e-05,
+      "loss": 2.1472,
+      "mean_token_accuracy": 0.5502070318907499,
+      "num_input_tokens_seen": 1859762928,
+      "num_tokens": 783996309.0,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9415030855349518,
+      "grad_norm": 0.30078125,
+      "learning_rate": 4.778934661234345e-05,
+      "loss": 2.1494,
+      "mean_token_accuracy": 0.548779489658773,
+      "num_input_tokens_seen": 1883948656,
+      "num_tokens": 794252324.0,
+      "step": 3900
+    },
+    {
+      "epoch": 0.953573637913605,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.760072430964237e-05,
+      "loss": 2.1484,
+      "mean_token_accuracy": 0.5490481401607394,
+      "num_input_tokens_seen": 1908219840,
+      "num_tokens": 804538128.0,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9656441902922582,
+      "grad_norm": 0.291015625,
+      "learning_rate": 4.7412102006941305e-05,
+      "loss": 2.1447,
+      "num_input_tokens_seen": 1932223680,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9656441902922582,
+      "eval_loss": 2.0152089595794678,
+      "eval_mean_token_accuracy": 0.5734592423989222,
+      "eval_num_tokens": 814660681.0,
+      "eval_runtime": 126.585,
+      "eval_samples_per_second": 84.623,
+      "eval_steps_per_second": 21.156,
+      "num_input_tokens_seen": 1932223680,
+      "step": 4000
     }
   ],
   "logging_steps": 50,
   "max_steps": 16568,
-  "num_input_tokens_seen": 1448897616,
+  "num_input_tokens_seen": 1932223680,
   "num_train_epochs": 4,
   "save_steps": 1000,
   "stateful_callbacks": {
@@ -681,7 +899,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.8759425352073216e+17,
+  "total_flos": 5.168886929031168e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null