End of training

README.md

@@ -3,9 +3,15 @@ library_name: transformers
 license: apache-2.0
 base_model: alignment-handbook/zephyr-7b-sft-full
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- data/zephyr_uf_rlced_conifer_ref
 model-index:
 - name: zephyr-7b-uf-rlced-conifer-group-dpo-2e-alr-0.01
   results: []
@@ -16,7 +22,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-uf-rlced-conifer-group-dpo-2e-alr-0.01
 
-This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on an unknown dataset.
+This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the data/zephyr_uf_rlced_conifer_ref dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.2395
 - Rewards/chosen: -2.8511

all_results.json

@@ -1,5 +1,41 @@
 {
     "epoch": 1.9986120749479528,
+    "eval_alpha_0_uf": 0.5116097927093506,
+    "eval_alpha_1_rlced_conifer": 0.4883902370929718,
+    "eval_excess_loss": 0.03742467654742917,
+    "eval_logits/chosen": 1.9375910758972168,
+    "eval_logits/chosen_0_uf": 3.0999701023101807,
+    "eval_logits/chosen_1_rlced_conifer": 1.572981357574463,
+    "eval_logits/rejected": 3.877798080444336,
+    "eval_logits/rejected_0_uf": 5.175272464752197,
+    "eval_logits/rejected_1_rlced_conifer": 3.567551374435425,
+    "eval_logps/chosen": -677.5836791992188,
+    "eval_logps/chosen_0_uf": -497.45623779296875,
+    "eval_logps/chosen_1_rlced_conifer": -728.9337158203125,
+    "eval_logps/rejected": -1262.6171875,
+    "eval_logps/rejected_0_uf": -584.9677124023438,
+    "eval_logps/rejected_1_rlced_conifer": -1451.01318359375,
+    "eval_loss": 0.23950526118278503,
+    "eval_rewards/accuracies": 0.8778195381164551,
+    "eval_rewards/accuracies_0_uf": 0.7632744312286377,
+    "eval_rewards/accuracies_1_rlced_conifer": 0.9097297191619873,
+    "eval_rewards/chosen": -2.8511428833007812,
+    "eval_rewards/chosen_0_uf": -2.081967353820801,
+    "eval_rewards/chosen_1_rlced_conifer": -3.0535335540771484,
+    "eval_rewards/margins": 5.737672805786133,
+    "eval_rewards/margins_0_uf": 1.3516454696655273,
+    "eval_rewards/margins_1_rlced_conifer": 6.98124361038208,
+    "eval_rewards/rejected": -8.588815689086914,
+    "eval_rewards/rejected_0_uf": -3.433612823486328,
+    "eval_rewards/rejected_1_rlced_conifer": -10.03477668762207,
+    "eval_runtime": 389.2096,
+    "eval_samples": 8491,
+    "eval_samples_per_second": 21.816,
+    "eval_steps_per_second": 0.342,
+    "eval_task_excess_loss_0_uf": 0.0723980620206721,
+    "eval_task_excess_loss_1_rlced_conifer": 0.04272522438225138,
+    "eval_task_loss_0_uf": 0.5184707641601562,
+    "eval_task_loss_1_rlced_conifer": 0.17867246270179749,
     "total_flos": 0.0,
     "train_loss": 0.14047848768532276,
     "train_runtime": 41218.966,

config.json

@@ -22,6 +22,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,39 @@
+{
+    "epoch": 1.9986120749479528,
+    "eval_alpha_0_uf": 0.5116097927093506,
+    "eval_alpha_1_rlced_conifer": 0.4883902370929718,
+    "eval_excess_loss": 0.03742467654742917,
+    "eval_logits/chosen": 1.9375910758972168,
+    "eval_logits/chosen_0_uf": 3.0999701023101807,
+    "eval_logits/chosen_1_rlced_conifer": 1.572981357574463,
+    "eval_logits/rejected": 3.877798080444336,
+    "eval_logits/rejected_0_uf": 5.175272464752197,
+    "eval_logits/rejected_1_rlced_conifer": 3.567551374435425,
+    "eval_logps/chosen": -677.5836791992188,
+    "eval_logps/chosen_0_uf": -497.45623779296875,
+    "eval_logps/chosen_1_rlced_conifer": -728.9337158203125,
+    "eval_logps/rejected": -1262.6171875,
+    "eval_logps/rejected_0_uf": -584.9677124023438,
+    "eval_logps/rejected_1_rlced_conifer": -1451.01318359375,
+    "eval_loss": 0.23950526118278503,
+    "eval_rewards/accuracies": 0.8778195381164551,
+    "eval_rewards/accuracies_0_uf": 0.7632744312286377,
+    "eval_rewards/accuracies_1_rlced_conifer": 0.9097297191619873,
+    "eval_rewards/chosen": -2.8511428833007812,
+    "eval_rewards/chosen_0_uf": -2.081967353820801,
+    "eval_rewards/chosen_1_rlced_conifer": -3.0535335540771484,
+    "eval_rewards/margins": 5.737672805786133,
+    "eval_rewards/margins_0_uf": 1.3516454696655273,
+    "eval_rewards/margins_1_rlced_conifer": 6.98124361038208,
+    "eval_rewards/rejected": -8.588815689086914,
+    "eval_rewards/rejected_0_uf": -3.433612823486328,
+    "eval_rewards/rejected_1_rlced_conifer": -10.03477668762207,
+    "eval_runtime": 389.2096,
+    "eval_samples": 8491,
+    "eval_samples_per_second": 21.816,
+    "eval_steps_per_second": 0.342,
+    "eval_task_excess_loss_0_uf": 0.0723980620206721,
+    "eval_task_excess_loss_1_rlced_conifer": 0.04272522438225138,
+    "eval_task_loss_0_uf": 0.5184707641601562,
+    "eval_task_loss_1_rlced_conifer": 0.17867246270179749
+}