End of training

README.md

@@ -3,9 +3,15 @@ library_name: transformers
 license: gemma
 base_model: google/gemma-7b
 tags:
+- alignment-handbook
 - trl
 - orpo
 - generated_from_trainer
+- trl
+- orpo
+- generated_from_trainer
+datasets:
+- silviasapora/low_quality_dpo7k
 model-index:
 - name: gemma-7b-borpo-low-quality-v4
   results: []
@@ -16,7 +22,7 @@ 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: 1.9080
 - Rewards/chosen: -0.6019

all_results.json

@@ -1,5 +1,21 @@
 {
     "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,
+    "eval_samples": 553,
+    "eval_samples_per_second": 1.571,
+    "eval_steps_per_second": 0.395,
     "total_flos": 0.0,
     "train_loss": 1.686337627812536,
     "train_runtime": 31593.308,

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

@@ -0,0 +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,
+    "eval_samples": 553,
+    "eval_samples_per_second": 1.571,
+    "eval_steps_per_second": 0.395
+}

runs/Sep22_17-15-08_zizgpu06.cpu.stats.ox.ac.uk/events.out.tfevents.1727054119.zizgpu06.cpu.stats.ox.ac.uk.3932700.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:457332324a3663880b08e48486fd6d39ad2e5e89fd0073dfc41f5cef9fda98e2
+size 997