End of training

README.md

@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # open-o1-sft-original
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the mlfoundations-dev/openo1_sft_original dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.5187
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 2.992412746585736,
-    "eval_loss": 0.5186759829521179,
-    "eval_runtime": 31.1204,
-    "eval_samples_per_second": 71.304,
-    "eval_steps_per_second": 1.125,
+    "eval_loss": 0.5186718702316284,
+    "eval_runtime": 31.6385,
+    "eval_samples_per_second": 70.136,
+    "eval_steps_per_second": 1.106,
     "total_flos": 411849782722560.0,
-    "train_loss": 0.5281505080742565,
-    "train_runtime": 6428.6606,
-    "train_samples_per_second": 19.671,
+    "train_loss": 0.5282489497487138,
+    "train_runtime": 6430.2484,
+    "train_samples_per_second": 19.666,
     "train_steps_per_second": 0.038
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 2.992412746585736,
-    "eval_loss": 0.5186759829521179,
-    "eval_runtime": 31.1204,
-    "eval_samples_per_second": 71.304,
-    "eval_steps_per_second": 1.125
+    "eval_loss": 0.5186718702316284,
+    "eval_runtime": 31.6385,
+    "eval_samples_per_second": 70.136,
+    "eval_steps_per_second": 1.106
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.992412746585736,
     "total_flos": 411849782722560.0,
-    "train_loss": 0.5281505080742565,
-    "train_runtime": 6428.6606,
-    "train_samples_per_second": 19.671,
+    "train_loss": 0.5282489497487138,
+    "train_runtime": 6430.2484,
+    "train_samples_per_second": 19.666,
     "train_steps_per_second": 0.038
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "log_history": [
     {
       "epoch": 0.12139605462822459,
-      "grad_norm": 3.2038055208414558,
+      "grad_norm": 1.9068916730195116,
       "learning_rate": 5e-06,
-      "loss": 0.775,
+      "loss": 0.7721,
       "step": 10
     },
     {
       "epoch": 0.24279210925644917,
-      "grad_norm": 1.113848957777039,
+      "grad_norm": 5.44810384092203,
       "learning_rate": 5e-06,
-      "loss": 0.6633,
+      "loss": 0.6638,
       "step": 20
     },
     {
       "epoch": 0.36418816388467373,
-      "grad_norm": 0.6586351716953581,
+      "grad_norm": 0.8162292861303865,
       "learning_rate": 5e-06,
-      "loss": 0.6218,
+      "loss": 0.6289,
       "step": 30
     },
     {
       "epoch": 0.48558421851289835,
-      "grad_norm": 0.6771410187746066,
+      "grad_norm": 0.724722400146013,
       "learning_rate": 5e-06,
-      "loss": 0.6021,
+      "loss": 0.6062,
       "step": 40
     },
     {
       "epoch": 0.6069802731411229,
-      "grad_norm": 0.7328529987985402,
+      "grad_norm": 0.6325755456980601,
       "learning_rate": 5e-06,
-      "loss": 0.5861,
+      "loss": 0.5886,
       "step": 50
     },
     {
       "epoch": 0.7283763277693475,
-      "grad_norm": 0.7718581101481012,
+      "grad_norm": 0.5102096530669636,
       "learning_rate": 5e-06,
-      "loss": 0.575,
+      "loss": 0.5763,
       "step": 60
     },
     {
       "epoch": 0.849772382397572,
-      "grad_norm": 0.7148733095889417,
+      "grad_norm": 0.6134528530146113,
       "learning_rate": 5e-06,
-      "loss": 0.5628,
+      "loss": 0.5635,
       "step": 70
     },
     {
       "epoch": 0.9711684370257967,
-      "grad_norm": 0.6560270346647304,
+      "grad_norm": 0.6520975040339092,
       "learning_rate": 5e-06,
-      "loss": 0.5574,
+      "loss": 0.5578,
       "step": 80
     },
     {
       "epoch": 0.9954476479514416,
-      "eval_loss": 0.5473940372467041,
-      "eval_runtime": 31.2065,
-      "eval_samples_per_second": 71.107,
+      "eval_loss": 0.5475569367408752,
+      "eval_runtime": 31.1849,
+      "eval_samples_per_second": 71.156,
       "eval_steps_per_second": 1.122,
       "step": 82
     },
     {
       "epoch": 1.095599393019727,
-      "grad_norm": 0.531907664284714,
+      "grad_norm": 0.9030012716394636,
       "learning_rate": 5e-06,
-      "loss": 0.5738,
+      "loss": 0.5739,
       "step": 90
     },
     {
       "epoch": 1.2169954476479514,
-      "grad_norm": 0.7745064014146105,
+      "grad_norm": 0.5546978323548724,
       "learning_rate": 5e-06,
-      "loss": 0.51,
+      "loss": 0.5098,
       "step": 100
     },
     {
       "epoch": 1.338391502276176,
-      "grad_norm": 0.9376249610380036,
+      "grad_norm": 0.7373972665017838,
       "learning_rate": 5e-06,
-      "loss": 0.5103,
+      "loss": 0.5099,
       "step": 110
     },
     {
       "epoch": 1.4597875569044005,
-      "grad_norm": 0.7170111999588846,
+      "grad_norm": 0.8335652060900699,
       "learning_rate": 5e-06,
-      "loss": 0.5091,
+      "loss": 0.5086,
       "step": 120
     },
     {
       "epoch": 1.5811836115326252,
-      "grad_norm": 0.534801962869618,
+      "grad_norm": 0.9482742766383457,
       "learning_rate": 5e-06,
-      "loss": 0.4993,
+      "loss": 0.4988,
       "step": 130
     },
     {
       "epoch": 1.7025796661608497,
-      "grad_norm": 0.7790654226033675,
+      "grad_norm": 0.7107692585969188,
       "learning_rate": 5e-06,
-      "loss": 0.4977,
+      "loss": 0.4973,
       "step": 140
     },
     {
       "epoch": 1.8239757207890743,
-      "grad_norm": 0.730776411072306,
+      "grad_norm": 0.6956293579831972,
       "learning_rate": 5e-06,
-      "loss": 0.5013,
+      "loss": 0.5006,
       "step": 150
     },
     {
       "epoch": 1.945371775417299,
-      "grad_norm": 0.6681683348307889,
+      "grad_norm": 0.5537353905530825,
       "learning_rate": 5e-06,
-      "loss": 0.4967,
+      "loss": 0.4962,
       "step": 160
     },
     {
       "epoch": 1.9939301972685888,
-      "eval_loss": 0.5244549512863159,
-      "eval_runtime": 31.3451,
-      "eval_samples_per_second": 70.793,
-      "eval_steps_per_second": 1.117,
+      "eval_loss": 0.5237926244735718,
+      "eval_runtime": 31.2236,
+      "eval_samples_per_second": 71.068,
+      "eval_steps_per_second": 1.121,
       "step": 164
     },
     {
       "epoch": 2.069802731411229,
-      "grad_norm": 0.9515480070083546,
+      "grad_norm": 0.9676846382246841,
       "learning_rate": 5e-06,
-      "loss": 0.5184,
+      "loss": 0.5181,
       "step": 170
     },
     {
       "epoch": 2.191198786039454,
-      "grad_norm": 0.9367718940586406,
+      "grad_norm": 1.0604432739536909,
       "learning_rate": 5e-06,
-      "loss": 0.4489,
+      "loss": 0.4483,
       "step": 180
     },
     {
       "epoch": 2.3125948406676784,
-      "grad_norm": 0.8897530781776953,
+      "grad_norm": 0.9072874578025836,
       "learning_rate": 5e-06,
-      "loss": 0.4489,
+      "loss": 0.4481,
       "step": 190
     },
     {
       "epoch": 2.433990895295903,
-      "grad_norm": 0.6335454904633909,
+      "grad_norm": 0.93397296136386,
       "learning_rate": 5e-06,
-      "loss": 0.4513,
+      "loss": 0.4503,
       "step": 200
     },
     {
       "epoch": 2.5553869499241273,
-      "grad_norm": 0.7983639127127669,
+      "grad_norm": 0.5608137627707893,
       "learning_rate": 5e-06,
-      "loss": 0.4488,
+      "loss": 0.4475,
       "step": 210
     },
     {
       "epoch": 2.676783004552352,
-      "grad_norm": 0.8751989014248259,
+      "grad_norm": 0.7216376866533744,
       "learning_rate": 5e-06,
-      "loss": 0.4505,
+      "loss": 0.4491,
       "step": 220
     },
     {
       "epoch": 2.7981790591805766,
-      "grad_norm": 0.6443956817031236,
+      "grad_norm": 0.7502162152741092,
       "learning_rate": 5e-06,
-      "loss": 0.4487,
+      "loss": 0.4475,
       "step": 230
     },
     {
       "epoch": 2.919575113808801,
-      "grad_norm": 0.7244641175857807,
+      "grad_norm": 0.6902724753233441,
       "learning_rate": 5e-06,
-      "loss": 0.4516,
+      "loss": 0.4505,
       "step": 240
     },
     {
       "epoch": 2.992412746585736,
-      "eval_loss": 0.5186759829521179,
-      "eval_runtime": 31.5369,
-      "eval_samples_per_second": 70.362,
-      "eval_steps_per_second": 1.11,
+      "eval_loss": 0.5186718702316284,
+      "eval_runtime": 31.6067,
+      "eval_samples_per_second": 70.207,
+      "eval_steps_per_second": 1.107,
       "step": 246
     },
     {
       "epoch": 2.992412746585736,
       "step": 246,
       "total_flos": 411849782722560.0,
-      "train_loss": 0.5281505080742565,
-      "train_runtime": 6428.6606,
-      "train_samples_per_second": 19.671,
+      "train_loss": 0.5282489497487138,
+      "train_runtime": 6430.2484,
+      "train_samples_per_second": 19.666,
       "train_steps_per_second": 0.038
     }
   ],