Training in progress, step 500, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dcd2f3fad4c3ce9d6c36acf3b466e166ed86faf72b1d892078776ed3c63d35e9
+oid sha256:ee3dcc6ee910f9b7091986fac9897234201849f91a6a2d070f0151d31a3f24a4
 size 14483498224

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4982b7dccc5007565a9a90f1e0535e952c2171fac24b29bbe271cd98ad1da1ac
+oid sha256:5590ce31ca67611569f3ee9618b3e4cc8e211008c77245896ee5b02009a5d425
 size 28967187355

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f21a015ae4c84d3f2954fe1a07625effa998cb85df43a679159c0bd2c6d4c0e0
+oid sha256:e10df580efad97bb633668d65429a37f5f53374d3373d28957a85a8b301eb78e
 size 1465

last-checkpoint/trainer_state.json

@@ -2,218 +2,118 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.78125,
+  "epoch": 0.390625,
   "eval_steps": 500,
-  "global_step": 1000,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "entropy": 0.6963505896925927,
+      "entropy": 0.5879646396264434,
       "epoch": 0.0390625,
-      "grad_norm": 2.484375,
-      "learning_rate": 2.5520833333333334e-06,
-      "loss": 0.6638687133789063,
-      "mean_token_accuracy": 0.8503057000041008,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.828125000000001e-06,
+      "loss": 0.5608741760253906,
+      "mean_token_accuracy": 0.8732133361697197,
       "num_tokens": 309908.0,
       "step": 50
     },
     {
-      "entropy": 0.11895681096240879,
+      "entropy": 0.09950160900130868,
       "epoch": 0.078125,
-      "grad_norm": 1.1484375,
-      "learning_rate": 5.156250000000001e-06,
-      "loss": 0.10770008087158203,
-      "mean_token_accuracy": 0.9728262135386467,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.734375e-06,
+      "loss": 0.08675869941711425,
+      "mean_token_accuracy": 0.9764643028378487,
       "num_tokens": 618662.0,
       "step": 100
     },
     {
-      "entropy": 0.07512138553895056,
+      "entropy": 0.06861632030457258,
       "epoch": 0.1171875,
-      "grad_norm": 1.0546875,
-      "learning_rate": 7.760416666666666e-06,
-      "loss": 0.06229952812194824,
-      "mean_token_accuracy": 0.9822412073612213,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.1640625000000002e-05,
+      "loss": 0.059808659553527835,
+      "mean_token_accuracy": 0.9829691171646118,
       "num_tokens": 923343.0,
       "step": 150
     },
     {
-      "entropy": 0.0627696047630161,
+      "entropy": 0.06371763135306537,
       "epoch": 0.15625,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.0364583333333334e-05,
-      "loss": 0.05425459384918213,
-      "mean_token_accuracy": 0.9840827691555023,
+      "grad_norm": 1.125,
+      "learning_rate": 1.5546875e-05,
+      "loss": 0.05767580032348633,
+      "mean_token_accuracy": 0.9834094086289406,
       "num_tokens": 1231860.0,
       "step": 200
     },
     {
-      "entropy": 0.0520358403865248,
+      "entropy": 0.05956988543272018,
       "epoch": 0.1953125,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.2968750000000002e-05,
-      "loss": 0.04450764656066895,
-      "mean_token_accuracy": 0.9865634706616402,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9453125e-05,
+      "loss": 0.052296199798583985,
+      "mean_token_accuracy": 0.9851513150334358,
       "num_tokens": 1539428.0,
       "step": 250
     },
     {
-      "entropy": 0.05185608604922891,
+      "entropy": 0.06213963699527085,
       "epoch": 0.234375,
-      "grad_norm": 0.6171875,
-      "learning_rate": 1.5572916666666668e-05,
-      "loss": 0.04415330410003662,
-      "mean_token_accuracy": 0.9866002556681633,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9626736111111114e-05,
+      "loss": 0.057163243293762205,
+      "mean_token_accuracy": 0.9842589551210403,
       "num_tokens": 1844184.0,
       "step": 300
     },
     {
-      "entropy": 0.056249135248363016,
+      "entropy": 0.051620940826833245,
       "epoch": 0.2734375,
-      "grad_norm": 2.15625,
-      "learning_rate": 1.8177083333333332e-05,
-      "loss": 0.05188354969024658,
-      "mean_token_accuracy": 0.9850953930616378,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9192708333333335e-05,
+      "loss": 0.04214866161346435,
+      "mean_token_accuracy": 0.9875012004375457,
       "num_tokens": 2145404.0,
       "step": 350
     },
     {
-      "entropy": 0.04793156278319657,
+      "entropy": 0.032940625557675955,
       "epoch": 0.3125,
-      "grad_norm": 0.9765625,
-      "learning_rate": 1.9913194444444447e-05,
-      "loss": 0.04377750873565674,
-      "mean_token_accuracy": 0.9867869862914085,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.8758680555555557e-05,
+      "loss": 0.0279494047164917,
+      "mean_token_accuracy": 0.9905313903093338,
       "num_tokens": 2454513.0,
       "step": 400
     },
     {
-      "entropy": 0.04919681121595204,
+      "entropy": 0.03187832637690008,
       "epoch": 0.3515625,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 1.9623842592592593e-05,
-      "loss": 0.04342790603637695,
-      "mean_token_accuracy": 0.9877890127897263,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.8324652777777778e-05,
+      "loss": 0.024850430488586425,
+      "mean_token_accuracy": 0.9914047086238861,
       "num_tokens": 2763494.0,
       "step": 450
     },
     {
-      "entropy": 0.036583751048892735,
+      "entropy": 0.029858413645997645,
       "epoch": 0.390625,
-      "grad_norm": 0.421875,
-      "learning_rate": 1.9334490740740743e-05,
-      "loss": 0.029253978729248047,
-      "mean_token_accuracy": 0.9902601790428162,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.7890625000000003e-05,
+      "loss": 0.022111225128173827,
+      "mean_token_accuracy": 0.9919667416810989,
       "num_tokens": 3070032.0,
       "step": 500
-    },
-    {
-      "entropy": 0.04136734585743398,
-      "epoch": 0.4296875,
-      "grad_norm": 0.185546875,
-      "learning_rate": 1.904513888888889e-05,
-      "loss": 0.031803081035614016,
-      "mean_token_accuracy": 0.9897033905982972,
-      "num_tokens": 3371923.0,
-      "step": 550
-    },
-    {
-      "entropy": 0.03246699841227382,
-      "epoch": 0.46875,
-      "grad_norm": 0.27734375,
-      "learning_rate": 1.875578703703704e-05,
-      "loss": 0.02447201490402222,
-      "mean_token_accuracy": 0.9915765914320945,
-      "num_tokens": 3681052.0,
-      "step": 600
-    },
-    {
-      "entropy": 0.032780851311981676,
-      "epoch": 0.5078125,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 1.8466435185185186e-05,
-      "loss": 0.022467806339263915,
-      "mean_token_accuracy": 0.9919375470280647,
-      "num_tokens": 3985154.0,
-      "step": 650
-    },
-    {
-      "entropy": 0.02399231431540102,
-      "epoch": 0.546875,
-      "grad_norm": 0.103515625,
-      "learning_rate": 1.8177083333333332e-05,
-      "loss": 0.018611854314804076,
-      "mean_token_accuracy": 0.9926936730742455,
-      "num_tokens": 4294960.0,
-      "step": 700
-    },
-    {
-      "entropy": 0.028510651290416716,
-      "epoch": 0.5859375,
-      "grad_norm": 0.0849609375,
-      "learning_rate": 1.7887731481481482e-05,
-      "loss": 0.019694100618362426,
-      "mean_token_accuracy": 0.9926475915312767,
-      "num_tokens": 4596846.0,
-      "step": 750
-    },
-    {
-      "entropy": 0.027509439825080334,
-      "epoch": 0.625,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 1.7598379629629632e-05,
-      "loss": 0.01820646047592163,
-      "mean_token_accuracy": 0.992901928126812,
-      "num_tokens": 4904237.0,
-      "step": 800
-    },
-    {
-      "entropy": 0.028216501204296945,
-      "epoch": 0.6640625,
-      "grad_norm": 0.12109375,
-      "learning_rate": 1.730902777777778e-05,
-      "loss": 0.02013282299041748,
-      "mean_token_accuracy": 0.9926498794555664,
-      "num_tokens": 5205258.0,
-      "step": 850
-    },
-    {
-      "entropy": 0.02984008596278727,
-      "epoch": 0.703125,
-      "grad_norm": 0.10791015625,
-      "learning_rate": 1.701967592592593e-05,
-      "loss": 0.020396327972412108,
-      "mean_token_accuracy": 0.9925442752242088,
-      "num_tokens": 5511048.0,
-      "step": 900
-    },
-    {
-      "entropy": 0.027826988152228296,
-      "epoch": 0.7421875,
-      "grad_norm": 0.1689453125,
-      "learning_rate": 1.6730324074074075e-05,
-      "loss": 0.01839717388153076,
-      "mean_token_accuracy": 0.9930671873688698,
-      "num_tokens": 5824149.0,
-      "step": 950
-    },
-    {
-      "entropy": 0.02724860974121839,
-      "epoch": 0.78125,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 1.6440972222222225e-05,
-      "loss": 0.018133511543273927,
-      "mean_token_accuracy": 0.9931328788399696,
-      "num_tokens": 6133395.0,
-      "step": 1000
     }
   ],
   "logging_steps": 50,
-  "max_steps": 3840,
+  "max_steps": 2560,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -227,7 +127,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.885719471186084e+17,
+  "total_flos": 1.4446972723750502e+17,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f4e19c2cd1784f48b7cb7326b604649ead4dc5195e78095944d8317e4ed5a5c4
+oid sha256:98f86bcfd098aae8e261cc44af568a1b790f321588e8fdcc26b68725adda0314
 size 5713