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_bsz4096_lr5e-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_bsz4096_lr5e-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: 0.0595
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 0.059506773948669434,
+    "eval_runtime": 88.8314,
+    "eval_samples_per_second": 203.791,
+    "eval_steps_per_second": 0.405,
+    "total_flos": 3375200049561600.0,
+    "train_loss": 0.5100103264290189,
+    "train_runtime": 14999.08,
+    "train_samples_per_second": 68.795,
+    "train_steps_per_second": 0.017
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 0.059506773948669434,
+    "eval_runtime": 88.8314,
+    "eval_samples_per_second": 203.791,
+    "eval_steps_per_second": 0.405
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 3375200049561600.0,
+    "train_loss": 0.5100103264290189,
+    "train_runtime": 14999.08,
+    "train_samples_per_second": 68.795,
+    "train_steps_per_second": 0.017
+}

trainer_state.json

@@ -0,0 +1,241 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 252,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 12.12427822748705,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.8467,
+      "step": 10
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 3.002942598305113,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.7188,
+      "step": 20
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 2.104507046576198,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.6355,
+      "step": 30
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 2.701380557766081,
+      "learning_rate": 4.999030264010747e-06,
+      "loss": 0.592,
+      "step": 40
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 3.43564096357487,
+      "learning_rate": 4.965177193256699e-06,
+      "loss": 0.5666,
+      "step": 50
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 4.482652550424151,
+      "learning_rate": 4.88366999036662e-06,
+      "loss": 0.548,
+      "step": 60
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 2.3119532755575016,
+      "learning_rate": 4.756262083927795e-06,
+      "loss": 0.5365,
+      "step": 70
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 2.185140305123278,
+      "learning_rate": 4.585694344163653e-06,
+      "loss": 0.5257,
+      "step": 80
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.06480678915977478,
+      "eval_runtime": 89.7553,
+      "eval_samples_per_second": 201.693,
+      "eval_steps_per_second": 0.401,
+      "step": 84
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 2.0216909780523364,
+      "learning_rate": 4.375636119798806e-06,
+      "loss": 0.5048,
+      "step": 90
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 2.7025913785100326,
+      "learning_rate": 4.130606300974687e-06,
+      "loss": 0.4898,
+      "step": 100
+    },
+    {
+      "epoch": 1.3095238095238095,
+      "grad_norm": 4.766452560030004,
+      "learning_rate": 3.855876106354553e-06,
+      "loss": 0.4875,
+      "step": 110
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 3.0324776515869174,
+      "learning_rate": 3.5573556857160563e-06,
+      "loss": 0.4833,
+      "step": 120
+    },
+    {
+      "epoch": 1.5476190476190477,
+      "grad_norm": 2.8196920573226,
+      "learning_rate": 3.241466977499929e-06,
+      "loss": 0.4769,
+      "step": 130
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 2.1850077742233243,
+      "learning_rate": 2.915005556474384e-06,
+      "loss": 0.4745,
+      "step": 140
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 1.5856136275210844,
+      "learning_rate": 2.584994443525617e-06,
+      "loss": 0.4716,
+      "step": 150
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 1.4105991640255675,
+      "learning_rate": 2.258533022500071e-06,
+      "loss": 0.4675,
+      "step": 160
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.06080044060945511,
+      "eval_runtime": 89.1509,
+      "eval_samples_per_second": 203.06,
+      "eval_steps_per_second": 0.404,
+      "step": 168
+    },
+    {
+      "epoch": 2.0238095238095237,
+      "grad_norm": 1.4038560901075425,
+      "learning_rate": 1.9426443142839448e-06,
+      "loss": 0.4609,
+      "step": 170
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 1.2144245134712555,
+      "learning_rate": 1.6441238936454482e-06,
+      "loss": 0.4413,
+      "step": 180
+    },
+    {
+      "epoch": 2.261904761904762,
+      "grad_norm": 0.9045048286143268,
+      "learning_rate": 1.369393699025314e-06,
+      "loss": 0.4393,
+      "step": 190
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.6798693516985902,
+      "learning_rate": 1.1243638802011956e-06,
+      "loss": 0.4354,
+      "step": 200
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6407504960086512,
+      "learning_rate": 9.143056558363464e-07,
+      "loss": 0.4341,
+      "step": 210
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.5684219469046241,
+      "learning_rate": 7.437379160722046e-07,
+      "loss": 0.4356,
+      "step": 220
+    },
+    {
+      "epoch": 2.738095238095238,
+      "grad_norm": 0.5252372924853815,
+      "learning_rate": 6.163300096333806e-07,
+      "loss": 0.4342,
+      "step": 230
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.532076614069117,
+      "learning_rate": 5.348228067433014e-07,
+      "loss": 0.4331,
+      "step": 240
+    },
+    {
+      "epoch": 2.9761904761904763,
+      "grad_norm": 0.4782685958794092,
+      "learning_rate": 5.009697359892537e-07,
+      "loss": 0.4307,
+      "step": 250
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.059506773948669434,
+      "eval_runtime": 88.4493,
+      "eval_samples_per_second": 204.671,
+      "eval_steps_per_second": 0.407,
+      "step": 252
+    },
+    {
+      "epoch": 3.0,
+      "step": 252,
+      "total_flos": 3375200049561600.0,
+      "train_loss": 0.5100103264290189,
+      "train_runtime": 14999.08,
+      "train_samples_per_second": 68.795,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 252,
+  "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": 3375200049561600.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}