End of training

README.md

@@ -17,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # train_math_qa_456_1760637839
 
-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 math_qa dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0866
+- Loss: 1.0762
 - Num Input Tokens Seen: 77891968
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 1.0761665105819702,
+    "eval_runtime": 45.6763,
+    "eval_samples_per_second": 65.329,
+    "eval_steps_per_second": 16.332,
+    "num_input_tokens_seen": 77891968,
+    "total_flos": 3.507564542108369e+18,
+    "train_loss": 1.0954747395889015,
+    "train_runtime": 26429.8341,
+    "train_samples_per_second": 20.32,
+    "train_steps_per_second": 5.081
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 1.0761665105819702,
+    "eval_runtime": 45.6763,
+    "eval_samples_per_second": 65.329,
+    "eval_steps_per_second": 16.332,
+    "num_input_tokens_seen": 77891968
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "num_input_tokens_seen": 77891968,
+    "total_flos": 3.507564542108369e+18,
+    "train_loss": 1.0954747395889015,
+    "train_runtime": 26429.8341,
+    "train_samples_per_second": 20.32,
+    "train_steps_per_second": 5.081
+}