End of training

README.md

@@ -4,6 +4,7 @@ license: apache-2.0
 base_model: mistralai/Mistral-7B-v0.1
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: hp_ablations_grid_mistral_bsz2048_lr2e-6_scheduler-cosine-warmup0.15-minlr5e-7
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hp_ablations_grid_mistral_bsz2048_lr2e-6_scheduler-cosine-warmup0.15-minlr5e-7
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
+This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mlfoundations-dev/oh-dcft-v3-llama3.1-nemotron-70b_shareGPT_format dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0602
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 3.0,
-    "eval_loss": 0.05991922318935394,
-    "eval_runtime": 115.096,
-    "eval_samples_per_second": 157.286,
-    "eval_steps_per_second": 0.617,
+    "eval_loss": 0.06018054857850075,
+    "eval_runtime": 115.6102,
+    "eval_samples_per_second": 156.587,
+    "eval_steps_per_second": 0.614,
     "total_flos": 3376037568184320.0,
-    "train_loss": 0.5144212938963421,
-    "train_runtime": 17119.0234,
-    "train_samples_per_second": 60.276,
+    "train_loss": 0.518157976960379,
+    "train_runtime": 17131.2553,
+    "train_samples_per_second": 60.233,
     "train_steps_per_second": 0.029
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 3.0,
-    "eval_loss": 0.05991922318935394,
-    "eval_runtime": 115.096,
-    "eval_samples_per_second": 157.286,
-    "eval_steps_per_second": 0.617
+    "eval_loss": 0.06018054857850075,
+    "eval_runtime": 115.6102,
+    "eval_samples_per_second": 156.587,
+    "eval_steps_per_second": 0.614
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
     "total_flos": 3376037568184320.0,
-    "train_loss": 0.5144212938963421,
-    "train_runtime": 17119.0234,
-    "train_samples_per_second": 60.276,
+    "train_loss": 0.518157976960379,
+    "train_runtime": 17131.2553,
+    "train_samples_per_second": 60.233,
     "train_steps_per_second": 0.029
 }

trainer_state.json

@@ -10,385 +10,385 @@
   "log_history": [
     {
       "epoch": 0.05952380952380952,
-      "grad_norm": 11.642287552521111,
+      "grad_norm": 11.633337718230726,
       "learning_rate": 2.631578947368421e-07,
-      "loss": 0.8829,
+      "loss": 0.8828,
       "step": 10
     },
     {
       "epoch": 0.11904761904761904,
-      "grad_norm": 7.106858407842631,
+      "grad_norm": 7.106262599818549,
       "learning_rate": 5.263157894736842e-07,
       "loss": 0.8087,
       "step": 20
     },
     {
       "epoch": 0.17857142857142858,
-      "grad_norm": 2.093269967612724,
+      "grad_norm": 2.0015322743055655,
       "learning_rate": 7.894736842105263e-07,
-      "loss": 0.7296,
+      "loss": 0.7295,
       "step": 30
     },
     {
       "epoch": 0.23809523809523808,
-      "grad_norm": 1.2477023318233176,
+      "grad_norm": 1.4753694520182823,
       "learning_rate": 1.0526315789473683e-06,
       "loss": 0.6712,
       "step": 40
     },
     {
       "epoch": 0.2976190476190476,
-      "grad_norm": 1.0512343429003204,
+      "grad_norm": 1.110295332689471,
       "learning_rate": 1.3157894736842106e-06,
-      "loss": 0.6321,
+      "loss": 0.6325,
       "step": 50
     },
     {
       "epoch": 0.35714285714285715,
-      "grad_norm": 4.474475922392763,
+      "grad_norm": 2.62423490064751,
       "learning_rate": 1.5789473684210526e-06,
-      "loss": 0.6026,
+      "loss": 0.6019,
       "step": 60
     },
     {
       "epoch": 0.4166666666666667,
-      "grad_norm": 1.6529279657405915,
+      "grad_norm": 2.320671081832736,
       "learning_rate": 1.8421052631578946e-06,
-      "loss": 0.5884,
+      "loss": 0.5851,
       "step": 70
     },
     {
       "epoch": 0.47619047619047616,
-      "grad_norm": 2.7949986781945366,
+      "grad_norm": 3.572874828764162,
       "learning_rate": 1.9996767546702485e-06,
-      "loss": 0.5783,
+      "loss": 0.5748,
       "step": 80
     },
     {
       "epoch": 0.5357142857142857,
-      "grad_norm": 2.118745503355743,
+      "grad_norm": 2.387950128798658,
       "learning_rate": 1.996043443883064e-06,
-      "loss": 0.5706,
+      "loss": 0.5678,
       "step": 90
     },
     {
       "epoch": 0.5952380952380952,
-      "grad_norm": 2.0357375426196147,
+      "grad_norm": 2.3954023739686083,
       "learning_rate": 1.988392397752233e-06,
-      "loss": 0.5598,
+      "loss": 0.5575,
       "step": 100
     },
     {
       "epoch": 0.6547619047619048,
-      "grad_norm": 2.2701188124758,
+      "grad_norm": 2.9223208257498885,
       "learning_rate": 1.9767648201496052e-06,
-      "loss": 0.5569,
+      "loss": 0.5551,
       "step": 110
     },
     {
       "epoch": 0.7142857142857143,
-      "grad_norm": 2.865482452040676,
+      "grad_norm": 2.3303161266900063,
       "learning_rate": 1.961223330122206e-06,
-      "loss": 0.5477,
+      "loss": 0.5468,
       "step": 120
     },
     {
       "epoch": 0.7738095238095238,
-      "grad_norm": 2.7451064629160493,
+      "grad_norm": 2.5697587476278088,
       "learning_rate": 1.941851624664209e-06,
-      "loss": 0.5459,
+      "loss": 0.5452,
       "step": 130
     },
     {
       "epoch": 0.8333333333333334,
-      "grad_norm": 2.0988347378452,
+      "grad_norm": 2.7124137926277583,
       "learning_rate": 1.9187540279759314e-06,
-      "loss": 0.5383,
+      "loss": 0.5381,
       "step": 140
     },
     {
       "epoch": 0.8928571428571429,
-      "grad_norm": 1.7836154166022118,
+      "grad_norm": 2.6892276446954377,
       "learning_rate": 1.8920549296372686e-06,
-      "loss": 0.5337,
+      "loss": 0.5341,
       "step": 150
     },
     {
       "epoch": 0.9523809523809523,
-      "grad_norm": 1.0480422517342471,
+      "grad_norm": 2.078964942505403,
       "learning_rate": 1.861898114721218e-06,
-      "loss": 0.528,
+      "loss": 0.5294,
       "step": 160
     },
     {
       "epoch": 1.0,
-      "eval_loss": 0.0656253919005394,
-      "eval_runtime": 116.504,
-      "eval_samples_per_second": 155.385,
-      "eval_steps_per_second": 0.609,
+      "eval_loss": 0.06584873795509338,
+      "eval_runtime": 116.4553,
+      "eval_samples_per_second": 155.45,
+      "eval_steps_per_second": 0.61,
       "step": 168
     },
     {
       "epoch": 1.0119047619047619,
-      "grad_norm": 1.8121694809796869,
+      "grad_norm": 1.8862054510010102,
       "learning_rate": 1.8284459894551025e-06,
-      "loss": 0.5223,
+      "loss": 0.5242,
       "step": 170
     },
     {
       "epoch": 1.0714285714285714,
-      "grad_norm": 1.420187622758248,
+      "grad_norm": 2.1399543444499467,
       "learning_rate": 1.7918787065996015e-06,
-      "loss": 0.5032,
+      "loss": 0.506,
       "step": 180
     },
     {
       "epoch": 1.130952380952381,
-      "grad_norm": 1.5424317018915787,
+      "grad_norm": 1.9062388412547935,
       "learning_rate": 1.7523931952557666e-06,
-      "loss": 0.5004,
+      "loss": 0.5037,
       "step": 190
     },
     {
       "epoch": 1.1904761904761905,
-      "grad_norm": 1.850588186757041,
+      "grad_norm": 1.2708808678811239,
       "learning_rate": 1.7102021003248955e-06,
-      "loss": 0.498,
+      "loss": 0.5017,
       "step": 200
     },
     {
       "epoch": 1.25,
-      "grad_norm": 1.4232612836192142,
+      "grad_norm": 1.3542515524166345,
       "learning_rate": 1.6655326373326793e-06,
-      "loss": 0.4987,
+      "loss": 0.5023,
       "step": 210
     },
     {
       "epoch": 1.3095238095238095,
-      "grad_norm": 1.230835146194867,
+      "grad_norm": 2.1610871241580822,
       "learning_rate": 1.6186253687848507e-06,
-      "loss": 0.493,
+      "loss": 0.497,
       "step": 220
     },
     {
       "epoch": 1.369047619047619,
-      "grad_norm": 1.2193965194507899,
+      "grad_norm": 2.0822374960801397,
       "learning_rate": 1.569732908644127e-06,
-      "loss": 0.4923,
+      "loss": 0.4962,
       "step": 230
     },
     {
       "epoch": 1.4285714285714286,
-      "grad_norm": 1.5201451501554284,
+      "grad_norm": 2.0776279888799416,
       "learning_rate": 1.5191185619053519e-06,
-      "loss": 0.4902,
+      "loss": 0.4943,
       "step": 240
     },
     {
       "epoch": 1.4880952380952381,
-      "grad_norm": 1.2106618553239001,
+      "grad_norm": 1.991564712073659,
       "learning_rate": 1.4670549065952552e-06,
-      "loss": 0.4863,
+      "loss": 0.4903,
       "step": 250
     },
     {
       "epoch": 1.5476190476190477,
-      "grad_norm": 1.073369144919772,
+      "grad_norm": 2.037009128388502,
       "learning_rate": 1.4138223258333096e-06,
-      "loss": 0.4845,
+      "loss": 0.4885,
       "step": 260
     },
     {
       "epoch": 1.6071428571428572,
-      "grad_norm": 1.3569659774066425,
+      "grad_norm": 1.820207258608856,
       "learning_rate": 1.3597074978591206e-06,
-      "loss": 0.4823,
+      "loss": 0.4864,
       "step": 270
     },
     {
       "epoch": 1.6666666666666665,
-      "grad_norm": 1.1172719170259655,
+      "grad_norm": 2.1241388881877183,
       "learning_rate": 1.3050018521581279e-06,
-      "loss": 0.4826,
+      "loss": 0.4871,
       "step": 280
     },
     {
       "epoch": 1.7261904761904763,
-      "grad_norm": 1.057235946491605,
+      "grad_norm": 1.78922859395332,
       "learning_rate": 1.2499999999999999e-06,
-      "loss": 0.4817,
+      "loss": 0.4864,
       "step": 290
     },
     {
       "epoch": 1.7857142857142856,
-      "grad_norm": 1.1395484188056737,
+      "grad_norm": 1.7328904839729484,
       "learning_rate": 1.1949981478418721e-06,
-      "loss": 0.4763,
+      "loss": 0.4813,
       "step": 300
     },
     {
       "epoch": 1.8452380952380953,
-      "grad_norm": 0.9673060331448388,
+      "grad_norm": 1.935272666911019,
       "learning_rate": 1.1402925021408796e-06,
-      "loss": 0.4751,
+      "loss": 0.4801,
       "step": 310
     },
     {
       "epoch": 1.9047619047619047,
-      "grad_norm": 1.2716355726282513,
+      "grad_norm": 1.2142950270281057,
       "learning_rate": 1.0861776741666901e-06,
-      "loss": 0.4743,
+      "loss": 0.4795,
       "step": 320
     },
     {
       "epoch": 1.9642857142857144,
-      "grad_norm": 0.8670777947729602,
+      "grad_norm": 1.3120184387970604,
       "learning_rate": 1.032945093404745e-06,
-      "loss": 0.4738,
+      "loss": 0.4791,
       "step": 330
     },
     {
       "epoch": 2.0,
-      "eval_loss": 0.061109066009521484,
-      "eval_runtime": 116.1265,
-      "eval_samples_per_second": 155.89,
+      "eval_loss": 0.061718959361314774,
+      "eval_runtime": 116.2674,
+      "eval_samples_per_second": 155.701,
       "eval_steps_per_second": 0.611,
       "step": 336
     },
     {
       "epoch": 2.0238095238095237,
-      "grad_norm": 1.1781799129689645,
+      "grad_norm": 1.4955203361832747,
       "learning_rate": 9.80881438094648e-07,
-      "loss": 0.4626,
+      "loss": 0.4693,
       "step": 340
     },
     {
       "epoch": 2.0833333333333335,
-      "grad_norm": 1.196256029896588,
+      "grad_norm": 1.1141424267445637,
       "learning_rate": 9.302670913558731e-07,
-      "loss": 0.4514,
+      "loss": 0.4599,
       "step": 350
     },
     {
       "epoch": 2.142857142857143,
-      "grad_norm": 1.240120670934915,
+      "grad_norm": 0.8809215575567357,
       "learning_rate": 8.813746312151494e-07,
-      "loss": 0.4466,
+      "loss": 0.4552,
       "step": 360
     },
     {
       "epoch": 2.2023809523809526,
-      "grad_norm": 1.4515250964073299,
+      "grad_norm": 0.819257933682719,
       "learning_rate": 8.344673626673205e-07,
-      "loss": 0.4461,
+      "loss": 0.4546,
       "step": 370
     },
     {
       "epoch": 2.261904761904762,
-      "grad_norm": 0.8373786848430113,
+      "grad_norm": 0.9199338927704399,
       "learning_rate": 7.897978996751046e-07,
-      "loss": 0.4491,
+      "loss": 0.4574,
       "step": 380
     },
     {
       "epoch": 2.3214285714285716,
-      "grad_norm": 1.0346083625080364,
+      "grad_norm": 0.9400263078511765,
       "learning_rate": 7.476068047442332e-07,
-      "loss": 0.4443,
+      "loss": 0.4527,
       "step": 390
     },
     {
       "epoch": 2.380952380952381,
-      "grad_norm": 0.9459213157979981,
+      "grad_norm": 0.8120316882785873,
       "learning_rate": 7.081212934003984e-07,
-      "loss": 0.4435,
+      "loss": 0.4517,
       "step": 400
     },
     {
       "epoch": 2.4404761904761907,
-      "grad_norm": 0.9958744135548997,
+      "grad_norm": 0.926431884810278,
       "learning_rate": 6.715540105448972e-07,
-      "loss": 0.4428,
+      "loss": 0.4507,
       "step": 410
     },
     {
       "epoch": 2.5,
-      "grad_norm": 0.9392635640773219,
+      "grad_norm": 0.8615015270894472,
       "learning_rate": 6.381018852787821e-07,
-      "loss": 0.4427,
+      "loss": 0.4505,
       "step": 420
     },
     {
       "epoch": 2.5595238095238093,
-      "grad_norm": 0.7294108780936411,
+      "grad_norm": 0.794846867305112,
       "learning_rate": 6.079450703627314e-07,
-      "loss": 0.4443,
+      "loss": 0.4519,
       "step": 430
     },
     {
       "epoch": 2.619047619047619,
-      "grad_norm": 0.8142282440596496,
+      "grad_norm": 0.8659294487925853,
       "learning_rate": 5.812459720240681e-07,
-      "loss": 0.4448,
+      "loss": 0.4523,
       "step": 440
     },
     {
       "epoch": 2.678571428571429,
-      "grad_norm": 0.8207822906148434,
+      "grad_norm": 0.8078214310845254,
       "learning_rate": 5.581483753357905e-07,
-      "loss": 0.4425,
+      "loss": 0.4498,
       "step": 450
     },
     {
       "epoch": 2.738095238095238,
-      "grad_norm": 0.9245896216419172,
+      "grad_norm": 0.9616399301296082,
       "learning_rate": 5.387766698777935e-07,
-      "loss": 0.4438,
+      "loss": 0.451,
       "step": 460
     },
     {
       "epoch": 2.7976190476190474,
-      "grad_norm": 0.7453831990866285,
+      "grad_norm": 0.8093607714101151,
       "learning_rate": 5.232351798503945e-07,
-      "loss": 0.4423,
+      "loss": 0.4495,
       "step": 470
     },
     {
       "epoch": 2.857142857142857,
-      "grad_norm": 0.8498775475082037,
+      "grad_norm": 0.8508798798797214,
       "learning_rate": 5.116076022477671e-07,
-      "loss": 0.4416,
+      "loss": 0.4487,
       "step": 480
     },
     {
       "epoch": 2.9166666666666665,
-      "grad_norm": 0.8873720999371016,
+      "grad_norm": 0.738511448943807,
       "learning_rate": 5.039565561169362e-07,
-      "loss": 0.4385,
+      "loss": 0.4454,
       "step": 490
     },
     {
       "epoch": 2.9761904761904763,
-      "grad_norm": 0.8671697549026851,
+      "grad_norm": 0.7290876762277495,
       "learning_rate": 5.003232453297512e-07,
-      "loss": 0.4405,
+      "loss": 0.4473,
       "step": 500
     },
     {
       "epoch": 3.0,
-      "eval_loss": 0.05991922318935394,
-      "eval_runtime": 116.0997,
-      "eval_samples_per_second": 155.926,
-      "eval_steps_per_second": 0.612,
+      "eval_loss": 0.06018054857850075,
+      "eval_runtime": 118.2486,
+      "eval_samples_per_second": 153.093,
+      "eval_steps_per_second": 0.6,
       "step": 504
     },
     {
       "epoch": 3.0,
       "step": 504,
       "total_flos": 3376037568184320.0,
-      "train_loss": 0.5144212938963421,
-      "train_runtime": 17119.0234,
-      "train_samples_per_second": 60.276,
+      "train_loss": 0.518157976960379,
+      "train_runtime": 17131.2553,
+      "train_samples_per_second": 60.233,
       "train_steps_per_second": 0.029
     }
   ],