End of training

README.md

@@ -3,6 +3,9 @@ library_name: transformers
 license: apache-2.0
 base_model: c14kevincardenas/beit-large-patch16-384-limb
 tags:
+- image-regression
+- human-movement
+- vision
 - generated_from_trainer
 model-index:
 - name: limbxy_pose_4heads_1layers_8embeddim
@@ -14,10 +17,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # limbxy_pose_4heads_1layers_8embeddim
 
-This model is a fine-tuned version of [c14kevincardenas/beit-large-patch16-384-limb](https://huggingface.co/c14kevincardenas/beit-large-patch16-384-limb) on an unknown dataset.
+This model is a fine-tuned version of [c14kevincardenas/beit-large-patch16-384-limb](https://huggingface.co/c14kevincardenas/beit-large-patch16-384-limb) on the c14kevincardenas/beta_caller_284_limbxy_pose dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1451
-- Rmse: 0.3810
+- Loss: 0.1449
+- Rmse: 0.3806
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 20.0,
-    "eval_loss": 0.1417153775691986,
-    "eval_rmse": 0.3764510452747345,
-    "eval_runtime": 9.7727,
-    "eval_samples_per_second": 102.326,
-    "eval_steps_per_second": 1.637,
+    "eval_loss": 0.14487937092781067,
+    "eval_rmse": 0.380630224943161,
+    "eval_runtime": 9.5453,
+    "eval_samples_per_second": 104.764,
+    "eval_steps_per_second": 1.676,
     "total_flos": 0.0,
-    "train_loss": 0.17676029650682815,
-    "train_runtime": 3336.7226,
-    "train_samples_per_second": 33.937,
-    "train_steps_per_second": 0.533
+    "train_loss": 0.16350786438149012,
+    "train_runtime": 3270.5878,
+    "train_samples_per_second": 34.624,
+    "train_steps_per_second": 0.544
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "eval_loss": 0.1417153775691986,
-    "eval_rmse": 0.3764510452747345,
-    "eval_runtime": 9.7727,
-    "eval_samples_per_second": 102.326,
-    "eval_steps_per_second": 1.637
+    "eval_loss": 0.14487937092781067,
+    "eval_rmse": 0.380630224943161,
+    "eval_runtime": 9.5453,
+    "eval_samples_per_second": 104.764,
+    "eval_steps_per_second": 1.676
 }

runs/Feb19_07-59-38_galactica.ad.cirange.net/events.out.tfevents.1739955603.galactica.ad.cirange.net.2838155.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:820971256c9c39f81b23da63a32e102499bc0271a24128744cfb6ab5469e632a
+size 407

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
     "total_flos": 0.0,
-    "train_loss": 0.17676029650682815,
-    "train_runtime": 3336.7226,
-    "train_samples_per_second": 33.937,
-    "train_steps_per_second": 0.533
+    "train_loss": 0.16350786438149012,
+    "train_runtime": 3270.5878,
+    "train_samples_per_second": 34.624,
+    "train_steps_per_second": 0.544
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.1417153775691986,
-  "best_model_checkpoint": "limbxy_pose/checkpoint-1780",
+  "best_metric": 0.14487937092781067,
+  "best_model_checkpoint": "limbxy_pose/checkpoint-1602",
   "epoch": 20.0,
   "eval_steps": 500,
   "global_step": 1780,
@@ -10,689 +10,689 @@
   "log_history": [
     {
       "epoch": 0.2808988764044944,
-      "grad_norm": 5.3874141476076276e-14,
+      "grad_norm": 1.6329148591013448e-13,
       "learning_rate": 5e-06,
       "loss": 0.3282,
       "step": 25
     },
     {
       "epoch": 0.5617977528089888,
-      "grad_norm": 2.9326387238765017e-12,
+      "grad_norm": 609844.4375,
       "learning_rate": 1e-05,
-      "loss": 0.3304,
+      "loss": 0.3197,
       "step": 50
     },
     {
       "epoch": 0.8426966292134831,
-      "grad_norm": 7.455580207024748e-13,
+      "grad_norm": 51261.34765625,
       "learning_rate": 1.5e-05,
-      "loss": 0.3414,
+      "loss": 0.2128,
       "step": 75
     },
     {
       "epoch": 1.0,
-      "eval_loss": 0.33112141489982605,
-      "eval_rmse": 0.5754314661026001,
-      "eval_runtime": 9.1133,
-      "eval_samples_per_second": 109.73,
-      "eval_steps_per_second": 1.756,
+      "eval_loss": 0.17451409995555878,
+      "eval_rmse": 0.4177488386631012,
+      "eval_runtime": 9.5044,
+      "eval_samples_per_second": 105.214,
+      "eval_steps_per_second": 1.683,
       "step": 89
     },
     {
       "epoch": 1.1235955056179776,
-      "grad_norm": 2.835559207142041e-13,
+      "grad_norm": 2084511.375,
       "learning_rate": 2e-05,
-      "loss": 0.3408,
+      "loss": 0.198,
       "step": 100
     },
     {
       "epoch": 1.404494382022472,
-      "grad_norm": 1.3095855780007049e-12,
+      "grad_norm": 721293.5625,
       "learning_rate": 2.5e-05,
-      "loss": 0.3358,
+      "loss": 0.1779,
       "step": 125
     },
     {
       "epoch": 1.6853932584269664,
-      "grad_norm": 1.1560928603326914e-10,
+      "grad_norm": 1280031.75,
       "learning_rate": 3e-05,
-      "loss": 0.3348,
+      "loss": 0.1604,
       "step": 150
     },
     {
       "epoch": 1.9662921348314608,
-      "grad_norm": 6.433731585681057e-12,
+      "grad_norm": 1277506.125,
       "learning_rate": 3.5e-05,
-      "loss": 0.3313,
+      "loss": 0.1574,
       "step": 175
     },
     {
       "epoch": 2.0,
-      "eval_loss": 0.33112141489982605,
-      "eval_rmse": 0.5754314661026001,
-      "eval_runtime": 9.2294,
-      "eval_samples_per_second": 108.35,
-      "eval_steps_per_second": 1.734,
+      "eval_loss": 0.148615300655365,
+      "eval_rmse": 0.3855065405368805,
+      "eval_runtime": 9.2155,
+      "eval_samples_per_second": 108.513,
+      "eval_steps_per_second": 1.736,
       "step": 178
     },
     {
       "epoch": 2.247191011235955,
-      "grad_norm": 654211.5,
+      "grad_norm": 1975641.0,
       "learning_rate": 4e-05,
-      "loss": 0.3285,
+      "loss": 0.1589,
       "step": 200
     },
     {
       "epoch": 2.5280898876404496,
-      "grad_norm": 342525.875,
+      "grad_norm": 2331497.75,
       "learning_rate": 4.5e-05,
-      "loss": 0.2248,
+      "loss": 0.1757,
       "step": 225
     },
     {
       "epoch": 2.808988764044944,
-      "grad_norm": 1131143.125,
+      "grad_norm": 4412882.0,
       "learning_rate": 5e-05,
-      "loss": 0.2277,
+      "loss": 0.2045,
       "step": 250
     },
     {
       "epoch": 3.0,
-      "eval_loss": 0.1668396145105362,
-      "eval_rmse": 0.40846002101898193,
-      "eval_runtime": 9.5775,
-      "eval_samples_per_second": 104.411,
-      "eval_steps_per_second": 1.671,
+      "eval_loss": 0.15186643600463867,
+      "eval_rmse": 0.3897004723548889,
+      "eval_runtime": 9.3829,
+      "eval_samples_per_second": 106.576,
+      "eval_steps_per_second": 1.705,
       "step": 267
     },
     {
       "epoch": 3.0898876404494384,
-      "grad_norm": 3920230.0,
+      "grad_norm": 1961631.5,
       "learning_rate": 4.918300653594771e-05,
-      "loss": 0.2028,
+      "loss": 0.1809,
       "step": 275
     },
     {
       "epoch": 3.370786516853933,
-      "grad_norm": 1259230.75,
+      "grad_norm": 5652000.5,
       "learning_rate": 4.8366013071895424e-05,
-      "loss": 0.2022,
+      "loss": 0.1778,
       "step": 300
     },
     {
       "epoch": 3.6516853932584272,
-      "grad_norm": 380599.59375,
+      "grad_norm": 1794376.5,
       "learning_rate": 4.7549019607843135e-05,
-      "loss": 0.1957,
+      "loss": 0.1861,
       "step": 325
     },
     {
       "epoch": 3.932584269662921,
-      "grad_norm": 928917.8125,
+      "grad_norm": 1615815.75,
       "learning_rate": 4.673202614379085e-05,
-      "loss": 0.1756,
+      "loss": 0.1697,
       "step": 350
     },
     {
       "epoch": 4.0,
-      "eval_loss": 0.15156960487365723,
-      "eval_rmse": 0.38931941986083984,
-      "eval_runtime": 9.6534,
-      "eval_samples_per_second": 103.59,
-      "eval_steps_per_second": 1.657,
+      "eval_loss": 0.16323314607143402,
+      "eval_rmse": 0.4040212333202362,
+      "eval_runtime": 9.3078,
+      "eval_samples_per_second": 107.437,
+      "eval_steps_per_second": 1.719,
       "step": 356
     },
     {
       "epoch": 4.213483146067416,
-      "grad_norm": 4586462.0,
+      "grad_norm": 2846568.25,
       "learning_rate": 4.5915032679738564e-05,
-      "loss": 0.1686,
+      "loss": 0.1743,
       "step": 375
     },
     {
       "epoch": 4.49438202247191,
-      "grad_norm": 1776099.0,
+      "grad_norm": 2384729.0,
       "learning_rate": 4.5098039215686275e-05,
-      "loss": 0.1619,
+      "loss": 0.1723,
       "step": 400
     },
     {
       "epoch": 4.775280898876405,
-      "grad_norm": 3756013.5,
+      "grad_norm": 1569872.375,
       "learning_rate": 4.4281045751633986e-05,
-      "loss": 0.1761,
+      "loss": 0.1818,
       "step": 425
     },
     {
       "epoch": 5.0,
-      "eval_loss": 0.1556388884782791,
-      "eval_rmse": 0.394510954618454,
-      "eval_runtime": 9.5792,
-      "eval_samples_per_second": 104.393,
-      "eval_steps_per_second": 1.67,
+      "eval_loss": 0.19486868381500244,
+      "eval_rmse": 0.44143933057785034,
+      "eval_runtime": 9.2813,
+      "eval_samples_per_second": 107.744,
+      "eval_steps_per_second": 1.724,
       "step": 445
     },
     {
       "epoch": 5.056179775280899,
-      "grad_norm": 2636789.0,
+      "grad_norm": 343663.96875,
       "learning_rate": 4.3464052287581704e-05,
-      "loss": 0.1791,
+      "loss": 0.1845,
       "step": 450
     },
     {
       "epoch": 5.337078651685394,
-      "grad_norm": 2208706.25,
+      "grad_norm": 1288543.625,
       "learning_rate": 4.2647058823529415e-05,
-      "loss": 0.1736,
+      "loss": 0.1941,
       "step": 475
     },
     {
       "epoch": 5.617977528089888,
-      "grad_norm": 499775.21875,
+      "grad_norm": 4547344.0,
       "learning_rate": 4.1830065359477126e-05,
-      "loss": 0.1565,
+      "loss": 0.1685,
       "step": 500
     },
     {
       "epoch": 5.898876404494382,
-      "grad_norm": 1655149.625,
+      "grad_norm": 185522.03125,
       "learning_rate": 4.101307189542484e-05,
-      "loss": 0.1632,
+      "loss": 0.1624,
       "step": 525
     },
     {
       "epoch": 6.0,
-      "eval_loss": 0.14596471190452576,
-      "eval_rmse": 0.38205331563949585,
-      "eval_runtime": 9.775,
-      "eval_samples_per_second": 102.302,
-      "eval_steps_per_second": 1.637,
+      "eval_loss": 0.14749938249588013,
+      "eval_rmse": 0.3840564787387848,
+      "eval_runtime": 9.2186,
+      "eval_samples_per_second": 108.477,
+      "eval_steps_per_second": 1.736,
       "step": 534
     },
     {
       "epoch": 6.179775280898877,
-      "grad_norm": 69511.53125,
+      "grad_norm": 2298808.25,
       "learning_rate": 4.0196078431372555e-05,
-      "loss": 0.1537,
+      "loss": 0.1578,
       "step": 550
     },
     {
       "epoch": 6.460674157303371,
-      "grad_norm": 1434219.875,
+      "grad_norm": 1625852.375,
       "learning_rate": 3.9379084967320266e-05,
-      "loss": 0.1578,
+      "loss": 0.162,
       "step": 575
     },
     {
       "epoch": 6.741573033707866,
-      "grad_norm": 959990.375,
+      "grad_norm": 2281359.0,
       "learning_rate": 3.8562091503267977e-05,
-      "loss": 0.1488,
+      "loss": 0.1645,
       "step": 600
     },
     {
       "epoch": 7.0,
-      "eval_loss": 0.14534823596477509,
-      "eval_rmse": 0.3812456429004669,
-      "eval_runtime": 9.5565,
-      "eval_samples_per_second": 104.641,
-      "eval_steps_per_second": 1.674,
+      "eval_loss": 0.14835500717163086,
+      "eval_rmse": 0.38516879081726074,
+      "eval_runtime": 9.4807,
+      "eval_samples_per_second": 105.477,
+      "eval_steps_per_second": 1.688,
       "step": 623
     },
     {
       "epoch": 7.022471910112359,
-      "grad_norm": 887092.875,
+      "grad_norm": 673713.0625,
       "learning_rate": 3.774509803921569e-05,
-      "loss": 0.1508,
+      "loss": 0.1554,
       "step": 625
     },
     {
       "epoch": 7.303370786516854,
-      "grad_norm": 1917969.375,
+      "grad_norm": 260786.171875,
       "learning_rate": 3.6928104575163405e-05,
-      "loss": 0.1532,
+      "loss": 0.1656,
       "step": 650
     },
     {
       "epoch": 7.584269662921348,
-      "grad_norm": 142502.96875,
+      "grad_norm": 2860977.5,
       "learning_rate": 3.611111111111111e-05,
-      "loss": 0.1501,
+      "loss": 0.1642,
       "step": 675
     },
     {
       "epoch": 7.865168539325842,
-      "grad_norm": 308292.90625,
+      "grad_norm": 1183203.125,
       "learning_rate": 3.529411764705883e-05,
-      "loss": 0.1502,
+      "loss": 0.1655,
       "step": 700
     },
     {
       "epoch": 8.0,
-      "eval_loss": 0.1449001282453537,
-      "eval_rmse": 0.38065749406814575,
-      "eval_runtime": 9.5176,
-      "eval_samples_per_second": 105.069,
-      "eval_steps_per_second": 1.681,
+      "eval_loss": 0.14708983898162842,
+      "eval_rmse": 0.3835229277610779,
+      "eval_runtime": 9.2282,
+      "eval_samples_per_second": 108.363,
+      "eval_steps_per_second": 1.734,
       "step": 712
     },
     {
       "epoch": 8.146067415730338,
-      "grad_norm": 2082431.625,
+      "grad_norm": 2235303.5,
       "learning_rate": 3.447712418300654e-05,
-      "loss": 0.1591,
+      "loss": 0.1543,
       "step": 725
     },
     {
       "epoch": 8.426966292134832,
-      "grad_norm": 2036905.875,
+      "grad_norm": 1311017.875,
       "learning_rate": 3.366013071895425e-05,
-      "loss": 0.1563,
+      "loss": 0.1507,
       "step": 750
     },
     {
       "epoch": 8.707865168539326,
-      "grad_norm": 77962.4765625,
+      "grad_norm": 1330708.25,
       "learning_rate": 3.284313725490196e-05,
-      "loss": 0.1551,
+      "loss": 0.1625,
       "step": 775
     },
     {
       "epoch": 8.98876404494382,
-      "grad_norm": 1343378.0,
+      "grad_norm": 553780.625,
       "learning_rate": 3.202614379084967e-05,
-      "loss": 0.1512,
+      "loss": 0.1594,
       "step": 800
     },
     {
       "epoch": 9.0,
-      "eval_loss": 0.15092459321022034,
-      "eval_rmse": 0.38849014043807983,
-      "eval_runtime": 9.3212,
-      "eval_samples_per_second": 107.283,
-      "eval_steps_per_second": 1.717,
+      "eval_loss": 0.15354213118553162,
+      "eval_rmse": 0.3918445110321045,
+      "eval_runtime": 9.1297,
+      "eval_samples_per_second": 109.533,
+      "eval_steps_per_second": 1.753,
       "step": 801
     },
     {
       "epoch": 9.269662921348315,
-      "grad_norm": 2264512.0,
+      "grad_norm": 208979.359375,
       "learning_rate": 3.120915032679739e-05,
-      "loss": 0.1495,
+      "loss": 0.1518,
       "step": 825
     },
     {
       "epoch": 9.55056179775281,
-      "grad_norm": 1154337.125,
+      "grad_norm": 78840.34375,
       "learning_rate": 3.0392156862745097e-05,
-      "loss": 0.1536,
+      "loss": 0.1552,
       "step": 850
     },
     {
       "epoch": 9.831460674157304,
-      "grad_norm": 1739664.875,
+      "grad_norm": 1873128.625,
       "learning_rate": 2.957516339869281e-05,
-      "loss": 0.1491,
+      "loss": 0.1513,
       "step": 875
     },
     {
       "epoch": 10.0,
-      "eval_loss": 0.1450948417186737,
-      "eval_rmse": 0.3809131681919098,
-      "eval_runtime": 9.5293,
-      "eval_samples_per_second": 104.94,
-      "eval_steps_per_second": 1.679,
+      "eval_loss": 0.1448940634727478,
+      "eval_rmse": 0.38064953684806824,
+      "eval_runtime": 9.4691,
+      "eval_samples_per_second": 105.607,
+      "eval_steps_per_second": 1.69,
       "step": 890
     },
     {
       "epoch": 10.112359550561798,
-      "grad_norm": 1887076.125,
+      "grad_norm": 1098680.625,
       "learning_rate": 2.8758169934640522e-05,
-      "loss": 0.1586,
+      "loss": 0.1551,
       "step": 900
     },
     {
       "epoch": 10.393258426966293,
-      "grad_norm": 2062948.625,
+      "grad_norm": 1874154.875,
       "learning_rate": 2.7941176470588236e-05,
-      "loss": 0.1537,
+      "loss": 0.1516,
       "step": 925
     },
     {
       "epoch": 10.674157303370787,
-      "grad_norm": 622017.4375,
+      "grad_norm": 53160.01953125,
       "learning_rate": 2.7124183006535947e-05,
-      "loss": 0.1586,
+      "loss": 0.1466,
       "step": 950
     },
     {
       "epoch": 10.955056179775282,
-      "grad_norm": 279420.875,
+      "grad_norm": 929561.25,
       "learning_rate": 2.630718954248366e-05,
-      "loss": 0.1615,
+      "loss": 0.1488,
       "step": 975
     },
     {
       "epoch": 11.0,
-      "eval_loss": 0.14514553546905518,
-      "eval_rmse": 0.3809797465801239,
-      "eval_runtime": 9.4448,
-      "eval_samples_per_second": 105.879,
-      "eval_steps_per_second": 1.694,
+      "eval_loss": 0.14547079801559448,
+      "eval_rmse": 0.38140633702278137,
+      "eval_runtime": 9.1414,
+      "eval_samples_per_second": 109.393,
+      "eval_steps_per_second": 1.75,
       "step": 979
     },
     {
       "epoch": 11.235955056179776,
-      "grad_norm": 468609.03125,
+      "grad_norm": 1361347.875,
       "learning_rate": 2.5490196078431373e-05,
-      "loss": 0.1498,
+      "loss": 0.1457,
       "step": 1000
     },
     {
       "epoch": 11.51685393258427,
-      "grad_norm": 420918.21875,
+      "grad_norm": 72723.1953125,
       "learning_rate": 2.4673202614379087e-05,
-      "loss": 0.1511,
+      "loss": 0.1508,
       "step": 1025
     },
     {
       "epoch": 11.797752808988765,
-      "grad_norm": 1311995.125,
+      "grad_norm": 93677.0625,
       "learning_rate": 2.38562091503268e-05,
-      "loss": 0.1549,
+      "loss": 0.1507,
       "step": 1050
     },
     {
       "epoch": 12.0,
-      "eval_loss": 0.14528058469295502,
-      "eval_rmse": 0.38115692138671875,
-      "eval_runtime": 9.5503,
-      "eval_samples_per_second": 104.709,
-      "eval_steps_per_second": 1.675,
+      "eval_loss": 0.1535731852054596,
+      "eval_rmse": 0.3918841481208801,
+      "eval_runtime": 9.5703,
+      "eval_samples_per_second": 104.49,
+      "eval_steps_per_second": 1.672,
       "step": 1068
     },
     {
       "epoch": 12.07865168539326,
-      "grad_norm": 2639625.0,
+      "grad_norm": 1383022.125,
       "learning_rate": 2.303921568627451e-05,
-      "loss": 0.153,
+      "loss": 0.155,
       "step": 1075
     },
     {
       "epoch": 12.359550561797754,
-      "grad_norm": 774080.25,
+      "grad_norm": 1435498.5,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 0.1482,
+      "loss": 0.15,
       "step": 1100
     },
     {
       "epoch": 12.640449438202246,
-      "grad_norm": 55058.171875,
+      "grad_norm": 256395.265625,
       "learning_rate": 2.1405228758169934e-05,
-      "loss": 0.1468,
+      "loss": 0.1465,
       "step": 1125
     },
     {
       "epoch": 12.921348314606742,
-      "grad_norm": 386005.375,
+      "grad_norm": 258689.03125,
       "learning_rate": 2.058823529411765e-05,
-      "loss": 0.1511,
+      "loss": 0.1522,
       "step": 1150
     },
     {
       "epoch": 13.0,
-      "eval_loss": 0.1451251208782196,
-      "eval_rmse": 0.380952924489975,
-      "eval_runtime": 9.5734,
-      "eval_samples_per_second": 104.456,
-      "eval_steps_per_second": 1.671,
+      "eval_loss": 0.14494504034519196,
+      "eval_rmse": 0.380716472864151,
+      "eval_runtime": 9.4294,
+      "eval_samples_per_second": 106.051,
+      "eval_steps_per_second": 1.697,
       "step": 1157
     },
     {
       "epoch": 13.202247191011235,
-      "grad_norm": 1428037.875,
+      "grad_norm": 1009406.375,
       "learning_rate": 1.977124183006536e-05,
-      "loss": 0.1493,
+      "loss": 0.1491,
       "step": 1175
     },
     {
       "epoch": 13.48314606741573,
-      "grad_norm": 1060426.0,
+      "grad_norm": 720892.125,
       "learning_rate": 1.895424836601307e-05,
-      "loss": 0.1475,
+      "loss": 0.1502,
       "step": 1200
     },
     {
       "epoch": 13.764044943820224,
-      "grad_norm": 393123.84375,
+      "grad_norm": 48925.546875,
       "learning_rate": 1.8137254901960785e-05,
-      "loss": 0.146,
+      "loss": 0.1458,
       "step": 1225
     },
     {
       "epoch": 14.0,
-      "eval_loss": 0.14483922719955444,
-      "eval_rmse": 0.3805775046348572,
-      "eval_runtime": 9.5961,
-      "eval_samples_per_second": 104.209,
-      "eval_steps_per_second": 1.667,
+      "eval_loss": 0.14527221024036407,
+      "eval_rmse": 0.3811459243297577,
+      "eval_runtime": 9.2593,
+      "eval_samples_per_second": 108.0,
+      "eval_steps_per_second": 1.728,
       "step": 1246
     },
     {
       "epoch": 14.044943820224718,
-      "grad_norm": 55979.83203125,
+      "grad_norm": 421335.125,
       "learning_rate": 1.7320261437908496e-05,
-      "loss": 0.1474,
+      "loss": 0.1485,
       "step": 1250
     },
     {
       "epoch": 14.325842696629213,
-      "grad_norm": 1408408.875,
+      "grad_norm": 375955.40625,
       "learning_rate": 1.650326797385621e-05,
-      "loss": 0.1483,
+      "loss": 0.1457,
       "step": 1275
     },
     {
       "epoch": 14.606741573033707,
-      "grad_norm": 1193337.625,
+      "grad_norm": 624158.0,
       "learning_rate": 1.568627450980392e-05,
-      "loss": 0.1472,
+      "loss": 0.1498,
       "step": 1300
     },
     {
       "epoch": 14.887640449438202,
-      "grad_norm": 138167.046875,
+      "grad_norm": 115186.8984375,
       "learning_rate": 1.4869281045751634e-05,
-      "loss": 0.1495,
+      "loss": 0.1506,
       "step": 1325
     },
     {
       "epoch": 15.0,
-      "eval_loss": 0.14416830241680145,
-      "eval_rmse": 0.37969499826431274,
-      "eval_runtime": 9.6797,
-      "eval_samples_per_second": 103.309,
-      "eval_steps_per_second": 1.653,
+      "eval_loss": 0.1455306112766266,
+      "eval_rmse": 0.3814847767353058,
+      "eval_runtime": 9.3528,
+      "eval_samples_per_second": 106.92,
+      "eval_steps_per_second": 1.711,
       "step": 1335
     },
     {
       "epoch": 15.168539325842696,
-      "grad_norm": 1532522.0,
+      "grad_norm": 1805258.0,
       "learning_rate": 1.4052287581699347e-05,
-      "loss": 0.1461,
+      "loss": 0.1467,
       "step": 1350
     },
     {
       "epoch": 15.44943820224719,
-      "grad_norm": 562337.5,
+      "grad_norm": 520502.5625,
       "learning_rate": 1.323529411764706e-05,
-      "loss": 0.1492,
+      "loss": 0.1529,
       "step": 1375
     },
     {
       "epoch": 15.730337078651685,
-      "grad_norm": 814187.375,
+      "grad_norm": 2020412.75,
       "learning_rate": 1.2418300653594772e-05,
-      "loss": 0.1463,
+      "loss": 0.1505,
       "step": 1400
     },
     {
       "epoch": 16.0,
-      "eval_loss": 0.1432572454214096,
-      "eval_rmse": 0.37849339842796326,
-      "eval_runtime": 9.6044,
-      "eval_samples_per_second": 104.119,
-      "eval_steps_per_second": 1.666,
+      "eval_loss": 0.1451566517353058,
+      "eval_rmse": 0.3809943199157715,
+      "eval_runtime": 9.215,
+      "eval_samples_per_second": 108.518,
+      "eval_steps_per_second": 1.736,
       "step": 1424
     },
     {
       "epoch": 16.01123595505618,
-      "grad_norm": 670902.375,
+      "grad_norm": 252456.1875,
       "learning_rate": 1.1601307189542485e-05,
-      "loss": 0.148,
+      "loss": 0.1481,
       "step": 1425
     },
     {
       "epoch": 16.292134831460675,
-      "grad_norm": 1786915.75,
+      "grad_norm": 267339.03125,
       "learning_rate": 1.0784313725490197e-05,
       "loss": 0.1464,
       "step": 1450
     },
     {
       "epoch": 16.573033707865168,
-      "grad_norm": 204775.359375,
+      "grad_norm": 482848.15625,
       "learning_rate": 9.96732026143791e-06,
-      "loss": 0.147,
+      "loss": 0.1482,
       "step": 1475
     },
     {
       "epoch": 16.853932584269664,
-      "grad_norm": 434166.53125,
+      "grad_norm": 297641.71875,
       "learning_rate": 9.150326797385621e-06,
-      "loss": 0.1455,
+      "loss": 0.1463,
       "step": 1500
     },
     {
       "epoch": 17.0,
-      "eval_loss": 0.14278554916381836,
-      "eval_rmse": 0.37786978483200073,
-      "eval_runtime": 9.3464,
-      "eval_samples_per_second": 106.993,
-      "eval_steps_per_second": 1.712,
+      "eval_loss": 0.1449102759361267,
+      "eval_rmse": 0.38067084550857544,
+      "eval_runtime": 9.2491,
+      "eval_samples_per_second": 108.119,
+      "eval_steps_per_second": 1.73,
       "step": 1513
     },
     {
       "epoch": 17.134831460674157,
-      "grad_norm": 427259.8125,
+      "grad_norm": 353910.8125,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 0.1476,
+      "loss": 0.1481,
       "step": 1525
     },
     {
       "epoch": 17.415730337078653,
-      "grad_norm": 216315.09375,
+      "grad_norm": 847917.6875,
       "learning_rate": 7.5163398692810456e-06,
-      "loss": 0.1481,
+      "loss": 0.1494,
       "step": 1550
     },
     {
       "epoch": 17.696629213483146,
-      "grad_norm": 842487.625,
+      "grad_norm": 197619.375,
       "learning_rate": 6.699346405228758e-06,
-      "loss": 0.1435,
+      "loss": 0.145,
       "step": 1575
     },
     {
       "epoch": 17.97752808988764,
-      "grad_norm": 376340.8125,
+      "grad_norm": 934886.75,
       "learning_rate": 5.882352941176471e-06,
-      "loss": 0.144,
+      "loss": 0.1463,
       "step": 1600
     },
     {
       "epoch": 18.0,
-      "eval_loss": 0.1422489583492279,
-      "eval_rmse": 0.37715908885002136,
-      "eval_runtime": 9.364,
-      "eval_samples_per_second": 106.792,
-      "eval_steps_per_second": 1.709,
+      "eval_loss": 0.14487937092781067,
+      "eval_rmse": 0.380630224943161,
+      "eval_runtime": 9.3303,
+      "eval_samples_per_second": 107.177,
+      "eval_steps_per_second": 1.715,
       "step": 1602
     },
     {
       "epoch": 18.258426966292134,
-      "grad_norm": 287156.1875,
+      "grad_norm": 178175.765625,
       "learning_rate": 5.065359477124184e-06,
-      "loss": 0.1435,
+      "loss": 0.1456,
       "step": 1625
     },
     {
       "epoch": 18.53932584269663,
-      "grad_norm": 129252.7109375,
+      "grad_norm": 282639.03125,
       "learning_rate": 4.2483660130718954e-06,
-      "loss": 0.142,
+      "loss": 0.145,
       "step": 1650
     },
     {
       "epoch": 18.820224719101123,
-      "grad_norm": 415142.15625,
+      "grad_norm": 338323.84375,
       "learning_rate": 3.431372549019608e-06,
-      "loss": 0.1461,
+      "loss": 0.1494,
       "step": 1675
     },
     {
       "epoch": 19.0,
-      "eval_loss": 0.14274005591869354,
-      "eval_rmse": 0.3778095543384552,
-      "eval_runtime": 9.4422,
-      "eval_samples_per_second": 105.907,
-      "eval_steps_per_second": 1.695,
+      "eval_loss": 0.1456519514322281,
+      "eval_rmse": 0.38164374232292175,
+      "eval_runtime": 9.5481,
+      "eval_samples_per_second": 104.733,
+      "eval_steps_per_second": 1.676,
       "step": 1691
     },
     {
       "epoch": 19.10112359550562,
-      "grad_norm": 202907.5625,
+      "grad_norm": 340422.8125,
       "learning_rate": 2.6143790849673204e-06,
-      "loss": 0.1441,
+      "loss": 0.147,
       "step": 1700
     },
     {
       "epoch": 19.382022471910112,
-      "grad_norm": 269955.5625,
+      "grad_norm": 366319.65625,
       "learning_rate": 1.7973856209150326e-06,
-      "loss": 0.1438,
+      "loss": 0.1471,
       "step": 1725
     },
     {
       "epoch": 19.662921348314608,
-      "grad_norm": 197176.828125,
+      "grad_norm": 642705.4375,
       "learning_rate": 9.80392156862745e-07,
-      "loss": 0.1436,
+      "loss": 0.1467,
       "step": 1750
     },
     {
       "epoch": 19.9438202247191,
-      "grad_norm": 579304.625,
+      "grad_norm": 455884.28125,
       "learning_rate": 1.6339869281045752e-07,
-      "loss": 0.1425,
+      "loss": 0.1454,
       "step": 1775
     },
     {
       "epoch": 20.0,
-      "eval_loss": 0.1417153775691986,
-      "eval_rmse": 0.3764510452747345,
-      "eval_runtime": 9.4376,
-      "eval_samples_per_second": 105.959,
-      "eval_steps_per_second": 1.695,
+      "eval_loss": 0.14513316750526428,
+      "eval_rmse": 0.3809635043144226,
+      "eval_runtime": 9.5295,
+      "eval_samples_per_second": 104.937,
+      "eval_steps_per_second": 1.679,
       "step": 1780
     },
     {
       "epoch": 20.0,
       "step": 1780,
       "total_flos": 0.0,
-      "train_loss": 0.17676029650682815,
-      "train_runtime": 3336.7226,
-      "train_samples_per_second": 33.937,
-      "train_steps_per_second": 0.533
+      "train_loss": 0.16350786438149012,
+      "train_runtime": 3270.5878,
+      "train_samples_per_second": 34.624,
+      "train_steps_per_second": 0.544
     }
   ],
   "logging_steps": 25,