End of training

README.md

@@ -3,10 +3,16 @@ library_name: transformers
 license: gemma
 base_model: google/gemma-7b
 tags:
+- alignment-handbook
+- trl
+- orpo
+- generated_from_trainer
 - trl
 - orpo
 - alignment-handbook
 - generated_from_trainer
+datasets:
+- silviasapora/low_quality_dpo7k
 model-index:
 - name: gemma-7b-borpo-low-quality-v4
   results: []
@@ -17,7 +23,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gemma-7b-borpo-low-quality-v4
 
-This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on an unknown dataset.
+This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the silviasapora/low_quality_dpo7k dataset.
 It achieves the following results on the evaluation set:
 - Loss: 1.8577
 - Rewards/chosen: -0.5993

all_results.json

@@ -1,21 +1,21 @@
 {
     "epoch": 3.0,
-    "eval_log_odds_chosen": 0.42297032475471497,
-    "eval_log_odds_ratio": -0.6477048993110657,
-    "eval_logits/chosen": 282.00848388671875,
-    "eval_logits/rejected": 247.6262664794922,
-    "eval_logps/chosen": -1.2038370370864868,
-    "eval_logps/rejected": -1.5014722347259521,
-    "eval_loss": 1.9079890251159668,
-    "eval_nll_loss": 1.5545235872268677,
-    "eval_rewards/accuracies": 0.6258992552757263,
-    "eval_rewards/chosen": -0.6019185185432434,
-    "eval_rewards/margins": 0.14881758391857147,
-    "eval_rewards/rejected": -0.7507361173629761,
-    "eval_runtime": 351.9295,
+    "eval_log_odds_chosen": 0.4396049678325653,
+    "eval_log_odds_ratio": -0.6420542597770691,
+    "eval_logits/chosen": 301.1214599609375,
+    "eval_logits/rejected": 240.39068603515625,
+    "eval_logps/chosen": -1.19857656955719,
+    "eval_logps/rejected": -1.5204962491989136,
+    "eval_loss": 1.8577181100845337,
+    "eval_nll_loss": 1.5532194375991821,
+    "eval_rewards/accuracies": 0.6142857074737549,
+    "eval_rewards/chosen": -0.599288284778595,
+    "eval_rewards/margins": 0.16095994412899017,
+    "eval_rewards/rejected": -0.7602481245994568,
+    "eval_runtime": 201.8893,
     "eval_samples": 553,
-    "eval_samples_per_second": 1.571,
-    "eval_steps_per_second": 0.395,
+    "eval_samples_per_second": 2.739,
+    "eval_steps_per_second": 0.347,
     "total_flos": 0.0,
     "train_loss": 2.2429193542117165,
     "train_runtime": 13126.4564,

config.json

@@ -24,6 +24,6 @@
   "rope_theta": 10000.0,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 256000
 }

eval_results.json

@@ -1,19 +1,19 @@
 {
-    "epoch": 2.9865871833084947,
-    "eval_log_odds_chosen": 0.42297032475471497,
-    "eval_log_odds_ratio": -0.6477048993110657,
-    "eval_logits/chosen": 282.00848388671875,
-    "eval_logits/rejected": 247.6262664794922,
-    "eval_logps/chosen": -1.2038370370864868,
-    "eval_logps/rejected": -1.5014722347259521,
-    "eval_loss": 1.9079890251159668,
-    "eval_nll_loss": 1.5545235872268677,
-    "eval_rewards/accuracies": 0.6258992552757263,
-    "eval_rewards/chosen": -0.6019185185432434,
-    "eval_rewards/margins": 0.14881758391857147,
-    "eval_rewards/rejected": -0.7507361173629761,
-    "eval_runtime": 351.9295,
+    "epoch": 3.0,
+    "eval_log_odds_chosen": 0.4396049678325653,
+    "eval_log_odds_ratio": -0.6420542597770691,
+    "eval_logits/chosen": 301.1214599609375,
+    "eval_logits/rejected": 240.39068603515625,
+    "eval_logps/chosen": -1.19857656955719,
+    "eval_logps/rejected": -1.5204962491989136,
+    "eval_loss": 1.8577181100845337,
+    "eval_nll_loss": 1.5532194375991821,
+    "eval_rewards/accuracies": 0.6142857074737549,
+    "eval_rewards/chosen": -0.599288284778595,
+    "eval_rewards/margins": 0.16095994412899017,
+    "eval_rewards/rejected": -0.7602481245994568,
+    "eval_runtime": 201.8893,
     "eval_samples": 553,
-    "eval_samples_per_second": 1.571,
-    "eval_steps_per_second": 0.395
+    "eval_samples_per_second": 2.739,
+    "eval_steps_per_second": 0.347
 }

runs/Sep23_17-22-00_zizgpu06.cpu.stats.ox.ac.uk/events.out.tfevents.1727122310.zizgpu06.cpu.stats.ox.ac.uk.54426.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93e86b0da25626dc246ca7e96384edf6422f7dd0d6cf46c9149a602054c80823
+size 997