End of training

README.md

@@ -4,6 +4,7 @@ library_name: peft
 license: llama3.1
 tags:
 - llama-factory
+- lora
 - generated_from_trainer
 model-index:
 - name: Llama-3.1-8B-Instruct-SFT-500
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-3.1-8B-Instruct-SFT-500
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on the bct_non_cot_sft_500 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0781
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 9.955555555555556,
+    "eval_loss": 0.07805962860584259,
+    "eval_runtime": 1.3061,
+    "eval_samples_per_second": 38.281,
+    "eval_steps_per_second": 19.141,
+    "total_flos": 2.516129863512883e+16,
+    "train_loss": 0.3581957229546138,
+    "train_runtime": 398.237,
+    "train_samples_per_second": 11.3,
+    "train_steps_per_second": 0.703
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 9.955555555555556,
+    "eval_loss": 0.07805962860584259,
+    "eval_runtime": 1.3061,
+    "eval_samples_per_second": 38.281,
+    "eval_steps_per_second": 19.141
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.955555555555556,
+    "total_flos": 2.516129863512883e+16,
+    "train_loss": 0.3581957229546138,
+    "train_runtime": 398.237,
+    "train_samples_per_second": 11.3,
+    "train_steps_per_second": 0.703
+}

trainer_state.json

@@ -0,0 +1,278 @@
+{
+  "best_metric": 0.07805962860584259,
+  "best_model_checkpoint": "saves/Llama-3.1-8B-Instruct/lora/sft-500/checkpoint-250",
+  "epoch": 9.955555555555556,
+  "eval_steps": 50,
+  "global_step": 280,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 7.369814872741699,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 1.7064,
+      "step": 10
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 5.399565696716309,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 1.5556,
+      "step": 20
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 4.430342674255371,
+      "learning_rate": 4.999805731202437e-06,
+      "loss": 1.4196,
+      "step": 30
+    },
+    {
+      "epoch": 1.4222222222222223,
+      "grad_norm": 4.033640384674072,
+      "learning_rate": 4.976529986032632e-06,
+      "loss": 1.1164,
+      "step": 40
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 3.2736451625823975,
+      "learning_rate": 4.914814565722671e-06,
+      "loss": 0.8493,
+      "step": 50
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "eval_loss": 0.818483293056488,
+      "eval_runtime": 1.2942,
+      "eval_samples_per_second": 38.633,
+      "eval_steps_per_second": 19.316,
+      "step": 50
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 2.509134292602539,
+      "learning_rate": 4.815617391525772e-06,
+      "loss": 0.6287,
+      "step": 60
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 2.4819371700286865,
+      "learning_rate": 4.680478160991514e-06,
+      "loss": 0.4584,
+      "step": 70
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 2.9780495166778564,
+      "learning_rate": 4.511494449416671e-06,
+      "loss": 0.3315,
+      "step": 80
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.2511732578277588,
+      "learning_rate": 4.311289152148182e-06,
+      "loss": 0.1892,
+      "step": 90
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 2.088059663772583,
+      "learning_rate": 4.0829697730853505e-06,
+      "loss": 0.1595,
+      "step": 100
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "eval_loss": 0.11228977888822556,
+      "eval_runtime": 1.2903,
+      "eval_samples_per_second": 38.752,
+      "eval_steps_per_second": 19.376,
+      "step": 100
+    },
+    {
+      "epoch": 3.911111111111111,
+      "grad_norm": 1.010196328163147,
+      "learning_rate": 3.830080191288342e-06,
+      "loss": 0.1311,
+      "step": 110
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 0.5358583927154541,
+      "learning_rate": 3.556545654351749e-06,
+      "loss": 0.1064,
+      "step": 120
+    },
+    {
+      "epoch": 4.622222222222222,
+      "grad_norm": 0.7223713994026184,
+      "learning_rate": 3.2666118523333363e-06,
+      "loss": 0.0928,
+      "step": 130
+    },
+    {
+      "epoch": 4.977777777777778,
+      "grad_norm": 0.6626089811325073,
+      "learning_rate": 2.964779017907287e-06,
+      "loss": 0.106,
+      "step": 140
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.5725705027580261,
+      "learning_rate": 2.6557320756121306e-06,
+      "loss": 0.0797,
+      "step": 150
+    },
+    {
+      "epoch": 5.333333333333333,
+      "eval_loss": 0.0811280757188797,
+      "eval_runtime": 1.2941,
+      "eval_samples_per_second": 38.636,
+      "eval_steps_per_second": 19.318,
+      "step": 150
+    },
+    {
+      "epoch": 5.688888888888889,
+      "grad_norm": 1.287990927696228,
+      "learning_rate": 2.3442679243878698e-06,
+      "loss": 0.0918,
+      "step": 160
+    },
+    {
+      "epoch": 6.044444444444444,
+      "grad_norm": 1.3758983612060547,
+      "learning_rate": 2.035220982092714e-06,
+      "loss": 0.1028,
+      "step": 170
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.8700389862060547,
+      "learning_rate": 1.7333881476666648e-06,
+      "loss": 0.0882,
+      "step": 180
+    },
+    {
+      "epoch": 6.7555555555555555,
+      "grad_norm": 0.7997310161590576,
+      "learning_rate": 1.443454345648252e-06,
+      "loss": 0.0792,
+      "step": 190
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 1.6571403741836548,
+      "learning_rate": 1.169919808711659e-06,
+      "loss": 0.0997,
+      "step": 200
+    },
+    {
+      "epoch": 7.111111111111111,
+      "eval_loss": 0.07891152799129486,
+      "eval_runtime": 1.2907,
+      "eval_samples_per_second": 38.739,
+      "eval_steps_per_second": 19.369,
+      "step": 200
+    },
+    {
+      "epoch": 7.466666666666667,
+      "grad_norm": 0.7907149195671082,
+      "learning_rate": 9.170302269146509e-07,
+      "loss": 0.0651,
+      "step": 210
+    },
+    {
+      "epoch": 7.822222222222222,
+      "grad_norm": 0.6304967403411865,
+      "learning_rate": 6.887108478518184e-07,
+      "loss": 0.0977,
+      "step": 220
+    },
+    {
+      "epoch": 8.177777777777777,
+      "grad_norm": 1.6207903623580933,
+      "learning_rate": 4.885055505833291e-07,
+      "loss": 0.0592,
+      "step": 230
+    },
+    {
+      "epoch": 8.533333333333333,
+      "grad_norm": 0.4881339371204376,
+      "learning_rate": 3.1952183900848673e-07,
+      "loss": 0.0816,
+      "step": 240
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 1.008658766746521,
+      "learning_rate": 1.843826084742284e-07,
+      "loss": 0.0896,
+      "step": 250
+    },
+    {
+      "epoch": 8.88888888888889,
+      "eval_loss": 0.07805962860584259,
+      "eval_runtime": 1.2925,
+      "eval_samples_per_second": 38.685,
+      "eval_steps_per_second": 19.342,
+      "step": 250
+    },
+    {
+      "epoch": 9.244444444444444,
+      "grad_norm": 1.0305126905441284,
+      "learning_rate": 8.518543427732951e-08,
+      "loss": 0.0924,
+      "step": 260
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.7199366092681885,
+      "learning_rate": 2.347001396736798e-08,
+      "loss": 0.0848,
+      "step": 270
+    },
+    {
+      "epoch": 9.955555555555556,
+      "grad_norm": 0.5508742332458496,
+      "learning_rate": 1.9426879756284656e-10,
+      "loss": 0.0668,
+      "step": 280
+    },
+    {
+      "epoch": 9.955555555555556,
+      "step": 280,
+      "total_flos": 2.516129863512883e+16,
+      "train_loss": 0.3581957229546138,
+      "train_runtime": 398.237,
+      "train_samples_per_second": 11.3,
+      "train_steps_per_second": 0.703
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 280,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.516129863512883e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}