End of training

README.md

@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # train_hellaswag_42_1760637627
 
-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 an unknown 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 hellaswag dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2200
+- Loss: 0.0582
 - Num Input Tokens Seen: 218263888
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 0.058249082416296005,
+    "eval_runtime": 125.1498,
+    "eval_samples_per_second": 31.89,
+    "eval_steps_per_second": 7.974,
+    "num_input_tokens_seen": 218263888,
+    "total_flos": 9.847548995415441e+18,
+    "train_loss": 0.01583954824622065,
+    "train_runtime": 78549.6897,
+    "train_samples_per_second": 9.144,
+    "train_steps_per_second": 2.286
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 0.058249082416296005,
+    "eval_runtime": 125.1498,
+    "eval_samples_per_second": 31.89,
+    "eval_steps_per_second": 7.974,
+    "num_input_tokens_seen": 218263888
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "num_input_tokens_seen": 218263888,
+    "total_flos": 9.847548995415441e+18,
+    "train_loss": 0.01583954824622065,
+    "train_runtime": 78549.6897,
+    "train_samples_per_second": 9.144,
+    "train_steps_per_second": 2.286
+}