End of training

README.md

@@ -2,12 +2,13 @@
 license: gemma
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - sft
 - generated_from_trainer
 base_model: google/gemma-7b
 datasets:
-- generator
+- chansung/no_robots_only_coding
 model-index:
 - name: gemma-7b-sft-qlora-1
   results: []
@@ -18,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gemma-7b-sft-qlora-1
 
-This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the generator dataset.
+This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the chansung/no_robots_only_coding dataset.
 It achieves the following results on the evaluation set:
 - Loss: 2.1615
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 22.73,
+    "eval_loss": 2.161546468734741,
+    "eval_runtime": 0.6065,
+    "eval_samples": 16,
+    "eval_samples_per_second": 3.297,
+    "eval_steps_per_second": 1.649,
     "train_loss": 3.26698664855957,
     "train_runtime": 470.6237,
     "train_samples": 926,

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 22.73,
-    "eval_loss": 2.209489345550537,
-    "eval_runtime": 0.6001,
+    "eval_loss": 2.161546468734741,
+    "eval_runtime": 0.6065,
     "eval_samples": 16,
-    "eval_samples_per_second": 3.333,
-    "eval_steps_per_second": 1.667
+    "eval_samples_per_second": 3.297,
+    "eval_steps_per_second": 1.649
 }

runs/Apr11_22-49-38_deep-diver-main-rare-husky-1-0-0/events.out.tfevents.1712890767.deep-diver-main-rare-husky-1-0-0.520.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13449748ea5d8a81a0584aa238af8f17ead7ea046647303b4a679b46a4a0ff8b
+size 354