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-no-robots15
   results: []
@@ -18,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gemma-7b-sft-qlora-no-robots15
 
-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: 1.2808
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 13.64,
+    "eval_loss": 1.2808171510696411,
+    "eval_runtime": 0.595,
+    "eval_samples": 16,
+    "eval_samples_per_second": 3.361,
+    "eval_steps_per_second": 1.681,
     "train_loss": 4.785474745432536,
     "train_runtime": 283.6494,
     "train_samples": 926,

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 13.64,
-    "eval_loss": 1.2830009460449219,
-    "eval_runtime": 0.6019,
+    "eval_loss": 1.2808171510696411,
+    "eval_runtime": 0.595,
     "eval_samples": 16,
-    "eval_samples_per_second": 3.323,
-    "eval_steps_per_second": 1.661
+    "eval_samples_per_second": 3.361,
+    "eval_steps_per_second": 1.681
 }

runs/Apr11_21-36-25_deep-diver-main-swift-parrot-1-0-0/events.out.tfevents.1712886219.deep-diver-main-swift-parrot-1-0-0.521.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e449f99c3bf9225cde47f9b24e7c3a06e1814f1c4533a866238b1b1885018f4
+size 354