End of training

README.md

@@ -17,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # train_siqa_42_1760637600
 
-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 siqa dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3844
+- Loss: 0.3242
 - Num Input Tokens Seen: 60302568
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 0.32417818903923035,
+    "eval_runtime": 69.9327,
+    "eval_samples_per_second": 47.775,
+    "eval_steps_per_second": 11.954,
+    "num_input_tokens_seen": 60302568,
+    "total_flos": 2.715397382208946e+18,
+    "train_loss": 0.15006305239016296,
+    "train_runtime": 29717.5338,
+    "train_samples_per_second": 20.237,
+    "train_steps_per_second": 5.06
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "eval_loss": 0.32417818903923035,
+    "eval_runtime": 69.9327,
+    "eval_samples_per_second": 47.775,
+    "eval_steps_per_second": 11.954,
+    "num_input_tokens_seen": 60302568
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "num_input_tokens_seen": 60302568,
+    "total_flos": 2.715397382208946e+18,
+    "train_loss": 0.15006305239016296,
+    "train_runtime": 29717.5338,
+    "train_samples_per_second": 20.237,
+    "train_steps_per_second": 5.06
+}