End of training

README.md

@@ -4,6 +4,7 @@ license: llama3
 base_model: meta-llama/Meta-Llama-3-8B-Instruct
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: filtered_cooking_train_data
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # filtered_cooking_train_data
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the None dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the identity and the filtered_cooking_train_data datasets.
 It achieves the following results on the evaluation set:
-- Loss: 0.3808
+- Loss: 0.3951
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 0.9959839357429718,
-    "eval_loss": 0.37736228108406067,
-    "eval_runtime": 2.8915,
-    "eval_samples_per_second": 28.359,
-    "eval_steps_per_second": 2.075,
-    "total_flos": 15726579548160.0,
-    "train_loss": 0.5279561415795357,
-    "train_runtime": 522.1498,
-    "train_samples_per_second": 7.63,
-    "train_steps_per_second": 0.059
+    "epoch": 2.931726907630522,
+    "eval_loss": 0.3951052725315094,
+    "eval_runtime": 2.8855,
+    "eval_samples_per_second": 28.418,
+    "eval_steps_per_second": 2.079,
+    "total_flos": 46305721843712.0,
+    "train_loss": 0.3068104995194302,
+    "train_runtime": 1422.7961,
+    "train_samples_per_second": 8.4,
+    "train_steps_per_second": 0.065
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 0.9959839357429718,
-    "eval_loss": 0.37736228108406067,
-    "eval_runtime": 2.8915,
-    "eval_samples_per_second": 28.359,
-    "eval_steps_per_second": 2.075
+    "epoch": 2.931726907630522,
+    "eval_loss": 0.3951052725315094,
+    "eval_runtime": 2.8855,
+    "eval_samples_per_second": 28.418,
+    "eval_steps_per_second": 2.079
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 0.9959839357429718,
-    "total_flos": 15726579548160.0,
-    "train_loss": 0.5279561415795357,
-    "train_runtime": 522.1498,
-    "train_samples_per_second": 7.63,
-    "train_steps_per_second": 0.059
+    "epoch": 2.931726907630522,
+    "total_flos": 46305721843712.0,
+    "train_loss": 0.3068104995194302,
+    "train_runtime": 1422.7961,
+    "train_samples_per_second": 8.4,
+    "train_steps_per_second": 0.065
 }

trainer_state.json

@@ -1,48 +1,98 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9959839357429718,
+  "epoch": 2.931726907630522,
   "eval_steps": 50,
-  "global_step": 31,
+  "global_step": 93,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.321285140562249,
-      "grad_norm": 5.4018345438242354,
-      "learning_rate": 8.83022221559489e-06,
-      "loss": 0.7064,
+      "grad_norm": 5.656576420158855,
+      "learning_rate": 1e-05,
+      "loss": 0.7327,
       "step": 10
     },
     {
       "epoch": 0.642570281124498,
-      "grad_norm": 4.66985543573099,
-      "learning_rate": 3.5659838364445505e-06,
-      "loss": 0.4659,
+      "grad_norm": 4.889362082789352,
+      "learning_rate": 9.646090361587828e-06,
+      "loss": 0.4739,
       "step": 20
     },
     {
       "epoch": 0.963855421686747,
-      "grad_norm": 3.4130652958692216,
-      "learning_rate": 3.3808211290284886e-08,
-      "loss": 0.4287,
+      "grad_norm": 3.5915730218435247,
+      "learning_rate": 8.634462259215719e-06,
+      "loss": 0.4578,
       "step": 30
     },
     {
-      "epoch": 0.9959839357429718,
-      "step": 31,
-      "total_flos": 15726579548160.0,
-      "train_loss": 0.5279561415795357,
-      "train_runtime": 522.1498,
-      "train_samples_per_second": 7.63,
-      "train_steps_per_second": 0.059
+      "epoch": 1.2570281124497993,
+      "grad_norm": 3.1515081038371155,
+      "learning_rate": 7.10832566725092e-06,
+      "loss": 0.3057,
+      "step": 40
+    },
+    {
+      "epoch": 1.5783132530120483,
+      "grad_norm": 2.931428647353244,
+      "learning_rate": 5.2837263654653715e-06,
+      "loss": 0.2544,
+      "step": 50
+    },
+    {
+      "epoch": 1.5783132530120483,
+      "eval_loss": 0.3807889521121979,
+      "eval_runtime": 2.914,
+      "eval_samples_per_second": 28.14,
+      "eval_steps_per_second": 2.059,
+      "step": 50
+    },
+    {
+      "epoch": 1.8995983935742973,
+      "grad_norm": 3.061268584858154,
+      "learning_rate": 3.4189616655158803e-06,
+      "loss": 0.2408,
+      "step": 60
+    },
+    {
+      "epoch": 2.1927710843373496,
+      "grad_norm": 2.0914870710223967,
+      "learning_rate": 1.7780148476756148e-06,
+      "loss": 0.1671,
+      "step": 70
+    },
+    {
+      "epoch": 2.5140562248995986,
+      "grad_norm": 2.4370061841731814,
+      "learning_rate": 5.931846699267558e-07,
+      "loss": 0.1015,
+      "step": 80
+    },
+    {
+      "epoch": 2.835341365461847,
+      "grad_norm": 2.1110646362602354,
+      "learning_rate": 3.220026018407541e-08,
+      "loss": 0.0912,
+      "step": 90
+    },
+    {
+      "epoch": 2.931726907630522,
+      "step": 93,
+      "total_flos": 46305721843712.0,
+      "train_loss": 0.3068104995194302,
+      "train_runtime": 1422.7961,
+      "train_samples_per_second": 8.4,
+      "train_steps_per_second": 0.065
     }
   ],
   "logging_steps": 10,
-  "max_steps": 31,
+  "max_steps": 93,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -56,7 +106,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 15726579548160.0,
+  "total_flos": 46305721843712.0,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null