End of training

README.md

@@ -17,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # train_gsm8k_123_1760637708
 
-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 gsm8k dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.5359
 - Num Input Tokens Seen: 30837792

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 0.5358567833900452,
+    "eval_runtime": 37.4984,
+    "eval_samples_per_second": 39.868,
+    "eval_steps_per_second": 9.974,
+    "num_input_tokens_seen": 30837792,
+    "total_flos": 1.3886118360648253e+18,
+    "train_loss": 0.49428450607934526,
+    "train_runtime": 7026.5426,
+    "train_samples_per_second": 17.015,
+    "train_steps_per_second": 4.255
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 0.5358567833900452,
+    "eval_runtime": 37.4984,
+    "eval_samples_per_second": 39.868,
+    "eval_steps_per_second": 9.974,
+    "num_input_tokens_seen": 30837792
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "num_input_tokens_seen": 30837792,
+    "total_flos": 1.3886118360648253e+18,
+    "train_loss": 0.49428450607934526,
+    "train_runtime": 7026.5426,
+    "train_samples_per_second": 17.015,
+    "train_steps_per_second": 4.255
+}