Training in progress, step 2000, checkpoint

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e53425bdc67c0605a853db9bf5bcf93793a8c96cfa5f75f67b2650c95c20081e
+oid sha256:6a97a82677b155799fd7abbfad24bcb6c60ee2e5a7f214b89e844f848122913c
 size 2252747

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d7eac66c44b28b5ff036fd57811c8fd5650cb56fd5b287df7ec209d4660ff3c1
+oid sha256:50812fb37f8b270e85c88b3329307b73066cddfd874f37431b6b883ce08cf517
 size 14455

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76cf98a8cde877f1e49a9e4b884c4dce3d51bb4035429d464e74acc35ddca024
+oid sha256:0a036b9529dfa34fa02854662d61c3e2cd7af540260222cbaee309d1e565d1f4
 size 1465

last-checkpoint/trainer_state.json

@@ -2,159 +2,159 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.08,
+  "epoch": 0.04,
   "eval_steps": 1000,
-  "global_step": 4000,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.002,
-      "grad_norm": 3.6465723514556885,
-      "learning_rate": 7.920000000000001e-05,
-      "loss": 9.041119384765626,
+      "grad_norm": 4.326515197753906,
+      "learning_rate": 7.92e-06,
+      "loss": 10.223335571289063,
       "step": 100
     },
     {
       "epoch": 0.004,
-      "grad_norm": 3.7281060218811035,
-      "learning_rate": 0.00015920000000000002,
-      "loss": 7.32277099609375,
+      "grad_norm": 4.537860870361328,
+      "learning_rate": 1.592e-05,
+      "loss": 8.924765625,
       "step": 200
     },
     {
       "epoch": 0.006,
-      "grad_norm": 5.1717400550842285,
-      "learning_rate": 0.00023920000000000001,
-      "loss": 7.0370635986328125,
+      "grad_norm": 3.0256755352020264,
+      "learning_rate": 2.392e-05,
+      "loss": 8.119463500976563,
       "step": 300
     },
     {
       "epoch": 0.008,
-      "grad_norm": 277.63214111328125,
-      "learning_rate": 0.0003192,
-      "loss": 7.320344848632812,
+      "grad_norm": 2.5825231075286865,
+      "learning_rate": 3.192e-05,
+      "loss": 7.47499755859375,
       "step": 400
     },
     {
       "epoch": 0.01,
-      "grad_norm": 3302.212646484375,
-      "learning_rate": 0.0003992,
-      "loss": 7.335491943359375,
+      "grad_norm": 3.1428987979888916,
+      "learning_rate": 3.9920000000000004e-05,
+      "loss": 7.310105590820313,
       "step": 500
     },
     {
       "epoch": 0.012,
-      "grad_norm": 5685008.0,
-      "learning_rate": 0.00047920000000000005,
-      "loss": 8.384228515625,
+      "grad_norm": 4.996999740600586,
+      "learning_rate": 4.792e-05,
+      "loss": 7.11893798828125,
       "step": 600
     },
     {
       "epoch": 0.014,
-      "grad_norm": 82818592.0,
-      "learning_rate": 0.0005592,
-      "loss": 9.99034912109375,
+      "grad_norm": 3.3079946041107178,
+      "learning_rate": 5.592000000000001e-05,
+      "loss": 7.0108880615234375,
       "step": 700
     },
     {
       "epoch": 0.016,
-      "grad_norm": NaN,
-      "learning_rate": 0.0006392,
-      "loss": NaN,
+      "grad_norm": 4.659358024597168,
+      "learning_rate": 6.392e-05,
+      "loss": 6.869476318359375,
       "step": 800
     },
     {
       "epoch": 0.018,
-      "grad_norm": NaN,
-      "learning_rate": 0.0007191999999999999,
-      "loss": NaN,
+      "grad_norm": 7.537159442901611,
+      "learning_rate": 7.192e-05,
+      "loss": 6.7979345703125,
       "step": 900
     },
     {
       "epoch": 0.02,
-      "grad_norm": NaN,
-      "learning_rate": 0.0007992,
-      "loss": NaN,
+      "grad_norm": 3.9471282958984375,
+      "learning_rate": 7.992000000000001e-05,
+      "loss": 6.715423583984375,
       "step": 1000
     },
     {
       "epoch": 0.02,
-      "eval_accuracy": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 21.1749,
-      "eval_samples_per_second": 47.226,
-      "eval_steps_per_second": 5.903,
+      "eval_accuracy": 0.13923287671232876,
+      "eval_loss": 6.70299768447876,
+      "eval_runtime": 9.4613,
+      "eval_samples_per_second": 105.694,
+      "eval_steps_per_second": 13.212,
       "step": 1000
     },
     {
       "epoch": 0.022,
-      "grad_norm": NaN,
-      "learning_rate": 0.0008792,
-      "loss": NaN,
+      "grad_norm": 32.88319396972656,
+      "learning_rate": 8.792e-05,
+      "loss": 6.7681121826171875,
       "step": 1100
     },
     {
       "epoch": 0.024,
-      "grad_norm": NaN,
-      "learning_rate": 0.0009592000000000001,
-      "loss": NaN,
+      "grad_norm": 1666.3564453125,
+      "learning_rate": 9.592e-05,
+      "loss": 6.879424438476563,
       "step": 1200
     },
     {
       "epoch": 0.026,
-      "grad_norm": NaN,
-      "learning_rate": 0.0010391999999999999,
-      "loss": NaN,
+      "grad_norm": 532019.8125,
+      "learning_rate": 0.00010392,
+      "loss": 9.755108642578126,
       "step": 1300
     },
     {
       "epoch": 0.028,
-      "grad_norm": NaN,
-      "learning_rate": 0.0011192,
-      "loss": NaN,
+      "grad_norm": 7875070.5,
+      "learning_rate": 0.00011192,
+      "loss": 10.504903564453125,
       "step": 1400
     },
     {
       "epoch": 0.03,
-      "grad_norm": NaN,
-      "learning_rate": 0.0011992,
-      "loss": NaN,
+      "grad_norm": 101449456.0,
+      "learning_rate": 0.00011992,
+      "loss": 10.605618896484375,
       "step": 1500
     },
     {
       "epoch": 0.032,
-      "grad_norm": NaN,
-      "learning_rate": 0.0012791999999999999,
-      "loss": NaN,
+      "grad_norm": 1575154432.0,
+      "learning_rate": 0.00012792,
+      "loss": 10.660673828125,
       "step": 1600
     },
     {
       "epoch": 0.034,
-      "grad_norm": NaN,
-      "learning_rate": 0.0013592,
-      "loss": NaN,
+      "grad_norm": 3583042816.0,
+      "learning_rate": 0.00013592,
+      "loss": 10.6936279296875,
       "step": 1700
     },
     {
       "epoch": 0.036,
-      "grad_norm": NaN,
-      "learning_rate": 0.0014392,
-      "loss": NaN,
+      "grad_norm": 5040027136.0,
+      "learning_rate": 0.00014392000000000002,
+      "loss": 10.703485107421875,
       "step": 1800
     },
     {
       "epoch": 0.038,
       "grad_norm": NaN,
-      "learning_rate": 0.0015192,
+      "learning_rate": 0.00015192000000000002,
       "loss": NaN,
       "step": 1900
     },
     {
       "epoch": 0.04,
       "grad_norm": NaN,
-      "learning_rate": 0.0015992,
+      "learning_rate": 0.00015992,
       "loss": NaN,
       "step": 2000
     },
@@ -162,168 +162,10 @@
       "epoch": 0.04,
       "eval_accuracy": 0.0,
       "eval_loss": NaN,
-      "eval_runtime": 9.5954,
-      "eval_samples_per_second": 104.217,
-      "eval_steps_per_second": 13.027,
+      "eval_runtime": 9.3223,
+      "eval_samples_per_second": 107.269,
+      "eval_steps_per_second": 13.409,
       "step": 2000
-    },
-    {
-      "epoch": 0.042,
-      "grad_norm": NaN,
-      "learning_rate": 0.0016792,
-      "loss": NaN,
-      "step": 2100
-    },
-    {
-      "epoch": 0.044,
-      "grad_norm": NaN,
-      "learning_rate": 0.0017592,
-      "loss": NaN,
-      "step": 2200
-    },
-    {
-      "epoch": 0.046,
-      "grad_norm": NaN,
-      "learning_rate": 0.0018392,
-      "loss": NaN,
-      "step": 2300
-    },
-    {
-      "epoch": 0.048,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019192,
-      "loss": NaN,
-      "step": 2400
-    },
-    {
-      "epoch": 0.05,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019992,
-      "loss": NaN,
-      "step": 2500
-    },
-    {
-      "epoch": 0.052,
-      "grad_norm": NaN,
-      "learning_rate": 0.001999978563623903,
-      "loss": NaN,
-      "step": 2600
-    },
-    {
-      "epoch": 0.054,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019999133871331223,
-      "loss": NaN,
-      "step": 2700
-    },
-    {
-      "epoch": 0.056,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019998044711915177,
-      "loss": NaN,
-      "step": 2800
-    },
-    {
-      "epoch": 0.058,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019996518205634257,
-      "loss": NaN,
-      "step": 2900
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019994554419262797,
-      "loss": NaN,
-      "step": 3000
-    },
-    {
-      "epoch": 0.06,
-      "eval_accuracy": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 9.7513,
-      "eval_samples_per_second": 102.551,
-      "eval_steps_per_second": 12.819,
-      "step": 3000
-    },
-    {
-      "epoch": 0.062,
-      "grad_norm": NaN,
-      "learning_rate": 0.001999215343870317,
-      "loss": NaN,
-      "step": 3100
-    },
-    {
-      "epoch": 0.064,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019989315368982054,
-      "loss": NaN,
-      "step": 3200
-    },
-    {
-      "epoch": 0.066,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019986040334245797,
-      "loss": NaN,
-      "step": 3300
-    },
-    {
-      "epoch": 0.068,
-      "grad_norm": NaN,
-      "learning_rate": 0.001998232847775504,
-      "loss": NaN,
-      "step": 3400
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019978179961878404,
-      "loss": NaN,
-      "step": 3500
-    },
-    {
-      "epoch": 0.072,
-      "grad_norm": NaN,
-      "learning_rate": 0.001997359496808541,
-      "loss": NaN,
-      "step": 3600
-    },
-    {
-      "epoch": 0.074,
-      "grad_norm": NaN,
-      "learning_rate": 0.001996857369693855,
-      "loss": NaN,
-      "step": 3700
-    },
-    {
-      "epoch": 0.076,
-      "grad_norm": NaN,
-      "learning_rate": 0.0019963116368084486,
-      "loss": NaN,
-      "step": 3800
-    },
-    {
-      "epoch": 0.078,
-      "grad_norm": NaN,
-      "learning_rate": 0.001995722322024446,
-      "loss": NaN,
-      "step": 3900
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": NaN,
-      "learning_rate": 0.001995089451120385,
-      "loss": NaN,
-      "step": 4000
-    },
-    {
-      "epoch": 0.08,
-      "eval_accuracy": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 9.7912,
-      "eval_samples_per_second": 102.132,
-      "eval_steps_per_second": 12.767,
-      "step": 4000
     }
   ],
   "logging_steps": 100,

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f3f1a617986bbcc3791eb0bf2535cb60187eec748b89d00f28d3cc6070a2d716
+oid sha256:57f2bc41e1c59fea3898f352ecf4a943074eea21936561a82374151ca7bca5f7
 size 5329