End of training

README.md

@@ -17,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # train_math_qa_123_1760637725
 
-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: 1.0669
+- Loss: 1.0601
 - Num Input Tokens Seen: 77961608
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 1.0600605010986328,
+    "eval_runtime": 43.6406,
+    "eval_samples_per_second": 68.377,
+    "eval_steps_per_second": 17.094,
+    "num_input_tokens_seen": 77961608,
+    "total_flos": 3.510700511079039e+18,
+    "train_loss": 1.0956058679626284,
+    "train_runtime": 26091.7688,
+    "train_samples_per_second": 20.584,
+    "train_steps_per_second": 5.146
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 1.0600605010986328,
+    "eval_runtime": 43.6406,
+    "eval_samples_per_second": 68.377,
+    "eval_steps_per_second": 17.094,
+    "num_input_tokens_seen": 77961608
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "num_input_tokens_seen": 77961608,
+    "total_flos": 3.510700511079039e+18,
+    "train_loss": 1.0956058679626284,
+    "train_runtime": 26091.7688,
+    "train_samples_per_second": 20.584,
+    "train_steps_per_second": 5.146
+}