Training in progress, step 2000, checkpoint

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1568527f327df07256f5a6c4e9bbd9b77a59a0494cbe6ba8a95a0e7c9329df65
-size 1976267
+oid sha256:f8bf065dce895a71030a8256ac77e86f07171b7a40886a69f80a202e40a5e3c4
+size 2252747

last-checkpoint/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4c813450e0c95fd384d2ff2898b39d648674b117ab52e8fd7daff5eed04513e6
-size 741336195
+oid sha256:dc94ca058a6bc953fb721f096ed27d747b42b54f2e3663b43f7186796afba942
+size 847648899

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa51b6645b4eccecb101cadb683ef0e972138fa3a9aba5da3ee559d38423cf09
+oid sha256:a7d35dbfbf266d337501eed62097b3a3a9e66befb826a069c58e0d8414eae6cb
 size 14455

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d04230f28149ed35951faa6c0efeb666df4e8906da2d72b8b274de5d8730258
+oid sha256:b716fb1dcbc77ddc6a0ae7c4c98d559e36f1986fff006ef6982bd75247833235
 size 1465

last-checkpoint/trainer_state.json

@@ -2,328 +2,170 @@
   "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": 12.660503387451172,
-      "learning_rate": 7.920000000000001e-07,
-      "loss": 10.891383056640626,
+      "grad_norm": 3.6465723514556885,
+      "learning_rate": 7.920000000000001e-05,
+      "loss": 9.041119384765626,
       "step": 100
     },
     {
       "epoch": 0.004,
-      "grad_norm": 7.186215400695801,
-      "learning_rate": 1.5920000000000002e-06,
-      "loss": 10.36431640625,
+      "grad_norm": 3.7281060218811035,
+      "learning_rate": 0.00015920000000000002,
+      "loss": 7.32277099609375,
       "step": 200
     },
     {
       "epoch": 0.006,
-      "grad_norm": 3.448742389678955,
-      "learning_rate": 2.392e-06,
-      "loss": 9.655166625976562,
+      "grad_norm": 5.1717400550842285,
+      "learning_rate": 0.00023920000000000001,
+      "loss": 7.0370635986328125,
       "step": 300
     },
     {
       "epoch": 0.008,
-      "grad_norm": 3.1286885738372803,
-      "learning_rate": 3.192e-06,
-      "loss": 9.2309130859375,
+      "grad_norm": 277.63214111328125,
+      "learning_rate": 0.0003192,
+      "loss": 7.320344848632812,
       "step": 400
     },
     {
       "epoch": 0.01,
-      "grad_norm": 2.5423293113708496,
-      "learning_rate": 3.992e-06,
-      "loss": 9.036063232421874,
+      "grad_norm": 3302.212646484375,
+      "learning_rate": 0.0003992,
+      "loss": 7.335491943359375,
       "step": 500
     },
     {
       "epoch": 0.012,
-      "grad_norm": 3.8436737060546875,
-      "learning_rate": 4.792000000000001e-06,
-      "loss": 8.7992333984375,
+      "grad_norm": 5685008.0,
+      "learning_rate": 0.00047920000000000005,
+      "loss": 8.384228515625,
       "step": 600
     },
     {
       "epoch": 0.014,
-      "grad_norm": 3.467728614807129,
-      "learning_rate": 5.592000000000001e-06,
-      "loss": 8.584224853515625,
+      "grad_norm": 82818592.0,
+      "learning_rate": 0.0005592,
+      "loss": 9.99034912109375,
       "step": 700
     },
     {
       "epoch": 0.016,
-      "grad_norm": 2.465226411819458,
-      "learning_rate": 6.392000000000001e-06,
-      "loss": 8.331553955078125,
+      "grad_norm": NaN,
+      "learning_rate": 0.0006392,
+      "loss": NaN,
       "step": 800
     },
     {
       "epoch": 0.018,
-      "grad_norm": 8.884796142578125,
-      "learning_rate": 7.192e-06,
-      "loss": 8.120615234375,
+      "grad_norm": NaN,
+      "learning_rate": 0.0007191999999999999,
+      "loss": NaN,
       "step": 900
     },
     {
       "epoch": 0.02,
-      "grad_norm": 3.531571388244629,
-      "learning_rate": 7.992e-06,
-      "loss": 7.909391479492188,
+      "grad_norm": NaN,
+      "learning_rate": 0.0007992,
+      "loss": NaN,
       "step": 1000
     },
     {
       "epoch": 0.02,
-      "eval_accuracy": 0.10772407045009785,
-      "eval_loss": 7.779632091522217,
-      "eval_runtime": 20.2829,
-      "eval_samples_per_second": 49.303,
-      "eval_steps_per_second": 6.163,
+      "eval_accuracy": 0.0,
+      "eval_loss": NaN,
+      "eval_runtime": 21.1749,
+      "eval_samples_per_second": 47.226,
+      "eval_steps_per_second": 5.903,
       "step": 1000
     },
     {
       "epoch": 0.022,
-      "grad_norm": 3.3351476192474365,
-      "learning_rate": 8.792e-06,
-      "loss": 7.69487060546875,
+      "grad_norm": NaN,
+      "learning_rate": 0.0008792,
+      "loss": NaN,
       "step": 1100
     },
     {
       "epoch": 0.024,
-      "grad_norm": 6.731584072113037,
-      "learning_rate": 9.592e-06,
-      "loss": 7.469768676757813,
+      "grad_norm": NaN,
+      "learning_rate": 0.0009592000000000001,
+      "loss": NaN,
       "step": 1200
     },
     {
       "epoch": 0.026,
-      "grad_norm": 2.172372817993164,
-      "learning_rate": 1.0392e-05,
-      "loss": 7.39029052734375,
+      "grad_norm": NaN,
+      "learning_rate": 0.0010391999999999999,
+      "loss": NaN,
       "step": 1300
     },
     {
       "epoch": 0.028,
-      "grad_norm": 2.362459421157837,
-      "learning_rate": 1.1192e-05,
-      "loss": 7.290626220703125,
+      "grad_norm": NaN,
+      "learning_rate": 0.0011192,
+      "loss": NaN,
       "step": 1400
     },
     {
       "epoch": 0.03,
-      "grad_norm": 2.927532196044922,
-      "learning_rate": 1.1992000000000001e-05,
-      "loss": 7.201221313476562,
+      "grad_norm": NaN,
+      "learning_rate": 0.0011992,
+      "loss": NaN,
       "step": 1500
     },
     {
       "epoch": 0.032,
-      "grad_norm": 3.7425637245178223,
-      "learning_rate": 1.2792e-05,
-      "loss": 7.103947143554688,
+      "grad_norm": NaN,
+      "learning_rate": 0.0012791999999999999,
+      "loss": NaN,
       "step": 1600
     },
     {
       "epoch": 0.034,
-      "grad_norm": 3.8855597972869873,
-      "learning_rate": 1.3592000000000001e-05,
-      "loss": 7.126583862304687,
+      "grad_norm": NaN,
+      "learning_rate": 0.0013592,
+      "loss": NaN,
       "step": 1700
     },
     {
       "epoch": 0.036,
-      "grad_norm": 3.833674669265747,
-      "learning_rate": 1.4392000000000002e-05,
-      "loss": 7.025943603515625,
+      "grad_norm": NaN,
+      "learning_rate": 0.0014392,
+      "loss": NaN,
       "step": 1800
     },
     {
       "epoch": 0.038,
-      "grad_norm": 3.688350200653076,
-      "learning_rate": 1.5192000000000003e-05,
-      "loss": 6.976273803710938,
+      "grad_norm": NaN,
+      "learning_rate": 0.0015192,
+      "loss": NaN,
       "step": 1900
     },
     {
       "epoch": 0.04,
-      "grad_norm": 3.1060683727264404,
-      "learning_rate": 1.5992000000000002e-05,
-      "loss": 6.903162841796875,
+      "grad_norm": NaN,
+      "learning_rate": 0.0015992,
+      "loss": NaN,
       "step": 2000
     },
     {
       "epoch": 0.04,
-      "eval_accuracy": 0.14267514677103718,
-      "eval_loss": 6.938798904418945,
-      "eval_runtime": 8.7308,
-      "eval_samples_per_second": 114.538,
-      "eval_steps_per_second": 14.317,
+      "eval_accuracy": 0.0,
+      "eval_loss": NaN,
+      "eval_runtime": 9.5954,
+      "eval_samples_per_second": 104.217,
+      "eval_steps_per_second": 13.027,
       "step": 2000
-    },
-    {
-      "epoch": 0.042,
-      "grad_norm": 5.402246475219727,
-      "learning_rate": 1.6792e-05,
-      "loss": 6.878953857421875,
-      "step": 2100
-    },
-    {
-      "epoch": 0.044,
-      "grad_norm": 5.448638916015625,
-      "learning_rate": 1.7592000000000004e-05,
-      "loss": 6.885161743164063,
-      "step": 2200
-    },
-    {
-      "epoch": 0.046,
-      "grad_norm": 5.395037651062012,
-      "learning_rate": 1.8392e-05,
-      "loss": 6.817017822265625,
-      "step": 2300
-    },
-    {
-      "epoch": 0.048,
-      "grad_norm": 7.231200695037842,
-      "learning_rate": 1.9192000000000002e-05,
-      "loss": 6.765516357421875,
-      "step": 2400
-    },
-    {
-      "epoch": 0.05,
-      "grad_norm": 4.857005596160889,
-      "learning_rate": 1.9992e-05,
-      "loss": 6.7409033203125,
-      "step": 2500
-    },
-    {
-      "epoch": 0.052,
-      "grad_norm": 4.149160861968994,
-      "learning_rate": 1.9999785636239034e-05,
-      "loss": 6.67716064453125,
-      "step": 2600
-    },
-    {
-      "epoch": 0.054,
-      "grad_norm": 3.6910853385925293,
-      "learning_rate": 1.9999133871331222e-05,
-      "loss": 6.673251342773438,
-      "step": 2700
-    },
-    {
-      "epoch": 0.056,
-      "grad_norm": 5.89624547958374,
-      "learning_rate": 1.9998044711915177e-05,
-      "loss": 6.649310302734375,
-      "step": 2800
-    },
-    {
-      "epoch": 0.058,
-      "grad_norm": 5.415604114532471,
-      "learning_rate": 1.999651820563426e-05,
-      "loss": 6.534013061523438,
-      "step": 2900
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 7.183228969573975,
-      "learning_rate": 1.99945544192628e-05,
-      "loss": 6.565379028320312,
-      "step": 3000
-    },
-    {
-      "epoch": 0.06,
-      "eval_accuracy": 0.1556751467710372,
-      "eval_loss": 6.592763900756836,
-      "eval_runtime": 8.7378,
-      "eval_samples_per_second": 114.445,
-      "eval_steps_per_second": 14.306,
-      "step": 3000
-    },
-    {
-      "epoch": 0.062,
-      "grad_norm": 8.597551345825195,
-      "learning_rate": 1.9992153438703174e-05,
-      "loss": 6.53469970703125,
-      "step": 3100
-    },
-    {
-      "epoch": 0.064,
-      "grad_norm": 3.4981071949005127,
-      "learning_rate": 1.9989315368982053e-05,
-      "loss": 6.47679931640625,
-      "step": 3200
-    },
-    {
-      "epoch": 0.066,
-      "grad_norm": 6.5537614822387695,
-      "learning_rate": 1.99860403342458e-05,
-      "loss": 6.47298828125,
-      "step": 3300
-    },
-    {
-      "epoch": 0.068,
-      "grad_norm": 5.503767967224121,
-      "learning_rate": 1.9982328477755036e-05,
-      "loss": 6.516907348632812,
-      "step": 3400
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 11.562784194946289,
-      "learning_rate": 1.9978179961878405e-05,
-      "loss": 6.511883544921875,
-      "step": 3500
-    },
-    {
-      "epoch": 0.072,
-      "grad_norm": 36.503265380859375,
-      "learning_rate": 1.9973594968085412e-05,
-      "loss": 6.553932495117188,
-      "step": 3600
-    },
-    {
-      "epoch": 0.074,
-      "grad_norm": 199.27720642089844,
-      "learning_rate": 1.996857369693855e-05,
-      "loss": 7.200830688476563,
-      "step": 3700
-    },
-    {
-      "epoch": 0.076,
-      "grad_norm": 1094.3033447265625,
-      "learning_rate": 1.996311636808449e-05,
-      "loss": 7.753666381835938,
-      "step": 3800
-    },
-    {
-      "epoch": 0.078,
-      "grad_norm": 472.0617370605469,
-      "learning_rate": 1.9957223220244462e-05,
-      "loss": 7.77503173828125,
-      "step": 3900
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 270.61187744140625,
-      "learning_rate": 1.9950894511203855e-05,
-      "loss": 7.76434814453125,
-      "step": 4000
-    },
-    {
-      "epoch": 0.08,
-      "eval_accuracy": 0.04172015655577299,
-      "eval_loss": 7.808525085449219,
-      "eval_runtime": 8.5336,
-      "eval_samples_per_second": 117.184,
-      "eval_steps_per_second": 14.648,
-      "step": 4000
     }
   ],
   "logging_steps": 100,

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0551fcc809086177b55c2e14e6be9041103a4dee9033ded1a7be27d9690e908d
+oid sha256:f3f1a617986bbcc3791eb0bf2535cb60187eec748b89d00f28d3cc6070a2d716
 size 5329