End of training

README.md

@@ -4,6 +4,7 @@ license: llama3
 base_model: meta-llama/Meta-Llama-3-8B-Instruct
 tags:
 - llama-factory
+- ia3
 - generated_from_trainer
 datasets:
 - super_glue
@@ -17,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # train_cb_1745950312
 
-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 super_glue 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 cb dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7426
+- Loss: 0.1586
 - Num Input Tokens Seen: 22164464
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 701.7610619469026,
+    "eval_loss": 0.15857546031475067,
+    "eval_runtime": 1.0555,
+    "eval_samples_per_second": 23.684,
+    "eval_steps_per_second": 12.316,
+    "num_input_tokens_seen": 22164464,
+    "total_flos": 9.980819425571635e+17,
+    "train_loss": 0.006749326340724289,
+    "train_runtime": 12825.8192,
+    "train_samples_per_second": 12.475,
+    "train_steps_per_second": 3.119
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 701.7610619469026,
+    "eval_loss": 0.15857546031475067,
+    "eval_runtime": 1.0555,
+    "eval_samples_per_second": 23.684,
+    "eval_steps_per_second": 12.316,
+    "num_input_tokens_seen": 22164464
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 701.7610619469026,
+    "num_input_tokens_seen": 22164464,
+    "total_flos": 9.980819425571635e+17,
+    "train_loss": 0.006749326340724289,
+    "train_runtime": 12825.8192,
+    "train_samples_per_second": 12.475,
+    "train_steps_per_second": 3.119
+}