End of training

README.md

@@ -2,6 +2,8 @@
 base_model: mock_training_run/llama_configs/config.json
 tags:
 - generated_from_trainer
+datasets:
+- wikitext
 model-index:
 - name: llama-wikitext
   results: []
@@ -12,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama-wikitext
 
-This model is a fine-tuned version of [mock_training_run/llama_configs/config.json](https://huggingface.co/mock_training_run/llama_configs/config.json) on the None dataset.
+This model is a fine-tuned version of [mock_training_run/llama_configs/config.json](https://huggingface.co/mock_training_run/llama_configs/config.json) on the wikitext wikitext-103-v1 dataset.
 
 ## Model description
 

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "train_loss": 5.974732427886038,
-    "train_runtime": 1543.433,
+    "epoch": 3.0,
+    "train_loss": 4.541526129751494,
+    "train_runtime": 4638.0405,
     "train_samples": 67643,
-    "train_samples_per_second": 43.826,
-    "train_steps_per_second": 0.086
+    "train_samples_per_second": 43.753,
+    "train_steps_per_second": 0.085
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "train_loss": 5.974732427886038,
-    "train_runtime": 1543.433,
+    "epoch": 3.0,
+    "train_loss": 4.541526129751494,
+    "train_runtime": 4638.0405,
     "train_samples": 67643,
-    "train_samples_per_second": 43.826,
-    "train_steps_per_second": 0.086
+    "train_samples_per_second": 43.753,
+    "train_steps_per_second": 0.085
 }

trainer_state.json

@@ -1,106 +1,262 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9990539262062441,
+  "epoch": 2.9971617786187323,
   "eval_steps": 100.0,
-  "global_step": 132,
+  "global_step": 396,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.08,
-      "learning_rate": 0.00029279999999999996,
-      "loss": 9.0737,
+      "learning_rate": 0.00015,
+      "loss": 9.4494,
       "step": 10
     },
     {
       "epoch": 0.15,
-      "learning_rate": 0.0002688,
-      "loss": 7.1389,
+      "learning_rate": 0.0003,
+      "loss": 7.3931,
       "step": 20
     },
     {
       "epoch": 0.23,
-      "learning_rate": 0.0002448,
-      "loss": 6.6649,
+      "learning_rate": 0.0002920212765957447,
+      "loss": 6.6623,
       "step": 30
     },
     {
       "epoch": 0.3,
-      "learning_rate": 0.00022079999999999997,
-      "loss": 6.3274,
+      "learning_rate": 0.00028404255319148934,
+      "loss": 6.2849,
       "step": 40
     },
     {
       "epoch": 0.38,
-      "learning_rate": 0.00019679999999999999,
-      "loss": 6.0455,
+      "learning_rate": 0.00027606382978723404,
+      "loss": 5.968,
       "step": 50
     },
     {
       "epoch": 0.45,
-      "learning_rate": 0.00017279999999999997,
-      "loss": 5.7722,
+      "learning_rate": 0.0002680851063829787,
+      "loss": 5.6831,
       "step": 60
     },
     {
       "epoch": 0.53,
-      "learning_rate": 0.00014879999999999998,
-      "loss": 5.6097,
+      "learning_rate": 0.0002601063829787234,
+      "loss": 5.4843,
       "step": 70
     },
     {
       "epoch": 0.61,
-      "learning_rate": 0.00012479999999999997,
-      "loss": 5.4476,
+      "learning_rate": 0.00025212765957446806,
+      "loss": 5.2955,
       "step": 80
     },
     {
       "epoch": 0.68,
-      "learning_rate": 0.0001008,
-      "loss": 5.3298,
+      "learning_rate": 0.0002441489361702127,
+      "loss": 5.1497,
       "step": 90
     },
     {
       "epoch": 0.76,
-      "learning_rate": 7.68e-05,
-      "loss": 5.2202,
+      "learning_rate": 0.00023617021276595742,
+      "loss": 5.0148,
       "step": 100
     },
     {
       "epoch": 0.83,
-      "learning_rate": 5.279999999999999e-05,
-      "loss": 5.1345,
+      "learning_rate": 0.0002281914893617021,
+      "loss": 4.8996,
       "step": 110
     },
     {
       "epoch": 0.91,
-      "learning_rate": 2.88e-05,
-      "loss": 5.067,
+      "learning_rate": 0.00022021276595744679,
+      "loss": 4.7704,
       "step": 120
     },
     {
       "epoch": 0.98,
-      "learning_rate": 4.8e-06,
-      "loss": 5.0283,
+      "learning_rate": 0.0002122340425531915,
+      "loss": 4.663,
       "step": 130
     },
     {
-      "epoch": 1.0,
-      "step": 132,
-      "total_flos": 3.244642363842232e+17,
-      "train_loss": 5.974732427886038,
-      "train_runtime": 1543.433,
-      "train_samples_per_second": 43.826,
-      "train_steps_per_second": 0.086
+      "epoch": 1.06,
+      "learning_rate": 0.00020425531914893615,
+      "loss": 4.5895,
+      "step": 140
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00019627659574468083,
+      "loss": 4.488,
+      "step": 150
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001882978723404255,
+      "loss": 4.3955,
+      "step": 160
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001803191489361702,
+      "loss": 4.3052,
+      "step": 170
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001723404255319149,
+      "loss": 4.2514,
+      "step": 180
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016436170212765956,
+      "loss": 4.1705,
+      "step": 190
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00015638297872340426,
+      "loss": 4.0962,
+      "step": 200
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00014840425531914892,
+      "loss": 4.0549,
+      "step": 210
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0001404255319148936,
+      "loss": 4.0031,
+      "step": 220
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00013244680851063828,
+      "loss": 3.9261,
+      "step": 230
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00012446808510638296,
+      "loss": 3.8936,
+      "step": 240
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00011648936170212764,
+      "loss": 3.8438,
+      "step": 250
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010851063829787234,
+      "loss": 3.8208,
+      "step": 260
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00010053191489361702,
+      "loss": 3.7774,
+      "step": 270
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.25531914893617e-05,
+      "loss": 3.7364,
+      "step": 280
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.457446808510637e-05,
+      "loss": 3.709,
+      "step": 290
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 7.659574468085105e-05,
+      "loss": 3.6827,
+      "step": 300
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 6.861702127659574e-05,
+      "loss": 3.6612,
+      "step": 310
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 6.063829787234042e-05,
+      "loss": 3.632,
+      "step": 320
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5.26595744680851e-05,
+      "loss": 3.6141,
+      "step": 330
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 4.468085106382978e-05,
+      "loss": 3.5908,
+      "step": 340
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.670212765957446e-05,
+      "loss": 3.5763,
+      "step": 350
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.8723404255319147e-05,
+      "loss": 3.5524,
+      "step": 360
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.0744680851063828e-05,
+      "loss": 3.5515,
+      "step": 370
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.276595744680851e-05,
+      "loss": 3.5409,
+      "step": 380
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.7872340425531906e-06,
+      "loss": 3.5372,
+      "step": 390
+    },
+    {
+      "epoch": 3.0,
+      "step": 396,
+      "total_flos": 9.733927091526697e+17,
+      "train_loss": 4.541526129751494,
+      "train_runtime": 4638.0405,
+      "train_samples_per_second": 43.753,
+      "train_steps_per_second": 0.085
     }
   ],
   "logging_steps": 10,
-  "max_steps": 132,
-  "num_train_epochs": 1,
+  "max_steps": 396,
+  "num_train_epochs": 3,
   "save_steps": 100,
-  "total_flos": 3.244642363842232e+17,
+  "total_flos": 9.733927091526697e+17,
   "trial_name": null,
   "trial_params": null
 }