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: 0.0599
 

all_results.json

@@ -0,0 +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,
+    "total_flos": 3376037568184320.0,
+    "train_loss": 0.5144212938963421,
+    "train_runtime": 17119.0234,
+    "train_samples_per_second": 60.276,
+    "train_steps_per_second": 0.029
+}

eval_results.json

@@ -0,0 +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
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 3376037568184320.0,
+    "train_loss": 0.5144212938963421,
+    "train_runtime": 17119.0234,
+    "train_samples_per_second": 60.276,
+    "train_steps_per_second": 0.029
+}

trainer_state.json

@@ -0,0 +1,416 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 504,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05952380952380952,
+      "grad_norm": 11.642287552521111,
+      "learning_rate": 2.631578947368421e-07,
+      "loss": 0.8829,
+      "step": 10
+    },
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 7.106858407842631,
+      "learning_rate": 5.263157894736842e-07,
+      "loss": 0.8087,
+      "step": 20
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 2.093269967612724,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 0.7296,
+      "step": 30
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 1.2477023318233176,
+      "learning_rate": 1.0526315789473683e-06,
+      "loss": 0.6712,
+      "step": 40
+    },
+    {
+      "epoch": 0.2976190476190476,
+      "grad_norm": 1.0512343429003204,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.6321,
+      "step": 50
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 4.474475922392763,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.6026,
+      "step": 60
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 1.6529279657405915,
+      "learning_rate": 1.8421052631578946e-06,
+      "loss": 0.5884,
+      "step": 70
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 2.7949986781945366,
+      "learning_rate": 1.9996767546702485e-06,
+      "loss": 0.5783,
+      "step": 80
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 2.118745503355743,
+      "learning_rate": 1.996043443883064e-06,
+      "loss": 0.5706,
+      "step": 90
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 2.0357375426196147,
+      "learning_rate": 1.988392397752233e-06,
+      "loss": 0.5598,
+      "step": 100
+    },
+    {
+      "epoch": 0.6547619047619048,
+      "grad_norm": 2.2701188124758,
+      "learning_rate": 1.9767648201496052e-06,
+      "loss": 0.5569,
+      "step": 110
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 2.865482452040676,
+      "learning_rate": 1.961223330122206e-06,
+      "loss": 0.5477,
+      "step": 120
+    },
+    {
+      "epoch": 0.7738095238095238,
+      "grad_norm": 2.7451064629160493,
+      "learning_rate": 1.941851624664209e-06,
+      "loss": 0.5459,
+      "step": 130
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 2.0988347378452,
+      "learning_rate": 1.9187540279759314e-06,
+      "loss": 0.5383,
+      "step": 140
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 1.7836154166022118,
+      "learning_rate": 1.8920549296372686e-06,
+      "loss": 0.5337,
+      "step": 150
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.0480422517342471,
+      "learning_rate": 1.861898114721218e-06,
+      "loss": 0.528,
+      "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,
+      "step": 168
+    },
+    {
+      "epoch": 1.0119047619047619,
+      "grad_norm": 1.8121694809796869,
+      "learning_rate": 1.8284459894551025e-06,
+      "loss": 0.5223,
+      "step": 170
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 1.420187622758248,
+      "learning_rate": 1.7918787065996015e-06,
+      "loss": 0.5032,
+      "step": 180
+    },
+    {
+      "epoch": 1.130952380952381,
+      "grad_norm": 1.5424317018915787,
+      "learning_rate": 1.7523931952557666e-06,
+      "loss": 0.5004,
+      "step": 190
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 1.850588186757041,
+      "learning_rate": 1.7102021003248955e-06,
+      "loss": 0.498,
+      "step": 200
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4232612836192142,
+      "learning_rate": 1.6655326373326793e-06,
+      "loss": 0.4987,
+      "step": 210
+    },
+    {
+      "epoch": 1.3095238095238095,
+      "grad_norm": 1.230835146194867,
+      "learning_rate": 1.6186253687848507e-06,
+      "loss": 0.493,
+      "step": 220
+    },
+    {
+      "epoch": 1.369047619047619,
+      "grad_norm": 1.2193965194507899,
+      "learning_rate": 1.569732908644127e-06,
+      "loss": 0.4923,
+      "step": 230
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 1.5201451501554284,
+      "learning_rate": 1.5191185619053519e-06,
+      "loss": 0.4902,
+      "step": 240
+    },
+    {
+      "epoch": 1.4880952380952381,
+      "grad_norm": 1.2106618553239001,
+      "learning_rate": 1.4670549065952552e-06,
+      "loss": 0.4863,
+      "step": 250
+    },
+    {
+      "epoch": 1.5476190476190477,
+      "grad_norm": 1.073369144919772,
+      "learning_rate": 1.4138223258333096e-06,
+      "loss": 0.4845,
+      "step": 260
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 1.3569659774066425,
+      "learning_rate": 1.3597074978591206e-06,
+      "loss": 0.4823,
+      "step": 270
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 1.1172719170259655,
+      "learning_rate": 1.3050018521581279e-06,
+      "loss": 0.4826,
+      "step": 280
+    },
+    {
+      "epoch": 1.7261904761904763,
+      "grad_norm": 1.057235946491605,
+      "learning_rate": 1.2499999999999999e-06,
+      "loss": 0.4817,
+      "step": 290
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 1.1395484188056737,
+      "learning_rate": 1.1949981478418721e-06,
+      "loss": 0.4763,
+      "step": 300
+    },
+    {
+      "epoch": 1.8452380952380953,
+      "grad_norm": 0.9673060331448388,
+      "learning_rate": 1.1402925021408796e-06,
+      "loss": 0.4751,
+      "step": 310
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 1.2716355726282513,
+      "learning_rate": 1.0861776741666901e-06,
+      "loss": 0.4743,
+      "step": 320
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 0.8670777947729602,
+      "learning_rate": 1.032945093404745e-06,
+      "loss": 0.4738,
+      "step": 330
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.061109066009521484,
+      "eval_runtime": 116.1265,
+      "eval_samples_per_second": 155.89,
+      "eval_steps_per_second": 0.611,
+      "step": 336
+    },
+    {
+      "epoch": 2.0238095238095237,
+      "grad_norm": 1.1781799129689645,
+      "learning_rate": 9.80881438094648e-07,
+      "loss": 0.4626,
+      "step": 340
+    },
+    {
+      "epoch": 2.0833333333333335,
+      "grad_norm": 1.196256029896588,
+      "learning_rate": 9.302670913558731e-07,
+      "loss": 0.4514,
+      "step": 350
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 1.240120670934915,
+      "learning_rate": 8.813746312151494e-07,
+      "loss": 0.4466,
+      "step": 360
+    },
+    {
+      "epoch": 2.2023809523809526,
+      "grad_norm": 1.4515250964073299,
+      "learning_rate": 8.344673626673205e-07,
+      "loss": 0.4461,
+      "step": 370
+    },
+    {
+      "epoch": 2.261904761904762,
+      "grad_norm": 0.8373786848430113,
+      "learning_rate": 7.897978996751046e-07,
+      "loss": 0.4491,
+      "step": 380
+    },
+    {
+      "epoch": 2.3214285714285716,
+      "grad_norm": 1.0346083625080364,
+      "learning_rate": 7.476068047442332e-07,
+      "loss": 0.4443,
+      "step": 390
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.9459213157979981,
+      "learning_rate": 7.081212934003984e-07,
+      "loss": 0.4435,
+      "step": 400
+    },
+    {
+      "epoch": 2.4404761904761907,
+      "grad_norm": 0.9958744135548997,
+      "learning_rate": 6.715540105448972e-07,
+      "loss": 0.4428,
+      "step": 410
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.9392635640773219,
+      "learning_rate": 6.381018852787821e-07,
+      "loss": 0.4427,
+      "step": 420
+    },
+    {
+      "epoch": 2.5595238095238093,
+      "grad_norm": 0.7294108780936411,
+      "learning_rate": 6.079450703627314e-07,
+      "loss": 0.4443,
+      "step": 430
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.8142282440596496,
+      "learning_rate": 5.812459720240681e-07,
+      "loss": 0.4448,
+      "step": 440
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 0.8207822906148434,
+      "learning_rate": 5.581483753357905e-07,
+      "loss": 0.4425,
+      "step": 450
+    },
+    {
+      "epoch": 2.738095238095238,
+      "grad_norm": 0.9245896216419172,
+      "learning_rate": 5.387766698777935e-07,
+      "loss": 0.4438,
+      "step": 460
+    },
+    {
+      "epoch": 2.7976190476190474,
+      "grad_norm": 0.7453831990866285,
+      "learning_rate": 5.232351798503945e-07,
+      "loss": 0.4423,
+      "step": 470
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.8498775475082037,
+      "learning_rate": 5.116076022477671e-07,
+      "loss": 0.4416,
+      "step": 480
+    },
+    {
+      "epoch": 2.9166666666666665,
+      "grad_norm": 0.8873720999371016,
+      "learning_rate": 5.039565561169362e-07,
+      "loss": 0.4385,
+      "step": 490
+    },
+    {
+      "epoch": 2.9761904761904763,
+      "grad_norm": 0.8671697549026851,
+      "learning_rate": 5.003232453297512e-07,
+      "loss": 0.4405,
+      "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,
+      "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_steps_per_second": 0.029
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 504,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3376037568184320.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}