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_1e2e
 model-index:
 - name: zephyr-7b-uf-rlced-conifer-1e2e-group-dpo-2e
   results: []
@@ -16,7 +22,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-uf-rlced-conifer-1e2e-group-dpo-2e
 
-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_1e2e dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.2626
 - Rewards/chosen: -2.1843

all_results.json

@@ -1,5 +1,41 @@
 {
     "epoch": 1.9986120749479528,
+    "eval_alpha_0_uf": 0.8469963073730469,
+    "eval_alpha_1_rlced_conifer": 0.15300358831882477,
+    "eval_excess_loss": 0.03744262155964691,
+    "eval_logits/chosen": -0.7806382775306702,
+    "eval_logits/chosen_0_uf": -0.1003958135843277,
+    "eval_logits/chosen_1_rlced_conifer": -0.9982388019561768,
+    "eval_logits/rejected": 1.2317651510238647,
+    "eval_logits/rejected_0_uf": 1.824297308921814,
+    "eval_logits/rejected_1_rlced_conifer": 1.1158097982406616,
+    "eval_logps/chosen": -610.9031982421875,
+    "eval_logps/chosen_0_uf": -489.0419006347656,
+    "eval_logps/chosen_1_rlced_conifer": -646.3860473632812,
+    "eval_logps/rejected": -946.61572265625,
+    "eval_logps/rejected_0_uf": -572.521240234375,
+    "eval_logps/rejected_1_rlced_conifer": -1049.993896484375,
+    "eval_loss": 0.2625831067562103,
+    "eval_rewards/accuracies": 0.8684210777282715,
+    "eval_rewards/accuracies_0_uf": 0.7603244781494141,
+    "eval_rewards/accuracies_1_rlced_conifer": 0.8986932039260864,
+    "eval_rewards/chosen": -2.184337615966797,
+    "eval_rewards/chosen_0_uf": -1.9978234767913818,
+    "eval_rewards/chosen_1_rlced_conifer": -2.228055715560913,
+    "eval_rewards/margins": 3.244464159011841,
+    "eval_rewards/margins_0_uf": 1.3113244771957397,
+    "eval_rewards/margins_1_rlced_conifer": 3.79652738571167,
+    "eval_rewards/rejected": -5.428801536560059,
+    "eval_rewards/rejected_0_uf": -3.3091485500335693,
+    "eval_rewards/rejected_1_rlced_conifer": -6.024582862854004,
+    "eval_runtime": 387.2635,
+    "eval_samples": 8491,
+    "eval_samples_per_second": 21.926,
+    "eval_steps_per_second": 0.343,
+    "eval_task_excess_loss_0_uf": 0.04689166452158737,
+    "eval_task_excess_loss_1_rlced_conifer": 0.0474511875690677,
+    "eval_task_loss_0_uf": 0.49440711736679077,
+    "eval_task_loss_1_rlced_conifer": 0.2101784497499466,
     "total_flos": 0.0,
     "train_loss": 0.15370953861210082,
     "train_runtime": 41916.6173,

config.json

@@ -22,6 +22,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,39 @@
+{
+    "epoch": 1.9986120749479528,
+    "eval_alpha_0_uf": 0.8469963073730469,
+    "eval_alpha_1_rlced_conifer": 0.15300358831882477,
+    "eval_excess_loss": 0.03744262155964691,
+    "eval_logits/chosen": -0.7806382775306702,
+    "eval_logits/chosen_0_uf": -0.1003958135843277,
+    "eval_logits/chosen_1_rlced_conifer": -0.9982388019561768,
+    "eval_logits/rejected": 1.2317651510238647,
+    "eval_logits/rejected_0_uf": 1.824297308921814,
+    "eval_logits/rejected_1_rlced_conifer": 1.1158097982406616,
+    "eval_logps/chosen": -610.9031982421875,
+    "eval_logps/chosen_0_uf": -489.0419006347656,
+    "eval_logps/chosen_1_rlced_conifer": -646.3860473632812,
+    "eval_logps/rejected": -946.61572265625,
+    "eval_logps/rejected_0_uf": -572.521240234375,
+    "eval_logps/rejected_1_rlced_conifer": -1049.993896484375,
+    "eval_loss": 0.2625831067562103,
+    "eval_rewards/accuracies": 0.8684210777282715,
+    "eval_rewards/accuracies_0_uf": 0.7603244781494141,
+    "eval_rewards/accuracies_1_rlced_conifer": 0.8986932039260864,
+    "eval_rewards/chosen": -2.184337615966797,
+    "eval_rewards/chosen_0_uf": -1.9978234767913818,
+    "eval_rewards/chosen_1_rlced_conifer": -2.228055715560913,
+    "eval_rewards/margins": 3.244464159011841,
+    "eval_rewards/margins_0_uf": 1.3113244771957397,
+    "eval_rewards/margins_1_rlced_conifer": 3.79652738571167,
+    "eval_rewards/rejected": -5.428801536560059,
+    "eval_rewards/rejected_0_uf": -3.3091485500335693,
+    "eval_rewards/rejected_1_rlced_conifer": -6.024582862854004,
+    "eval_runtime": 387.2635,
+    "eval_samples": 8491,
+    "eval_samples_per_second": 21.926,
+    "eval_steps_per_second": 0.343,
+    "eval_task_excess_loss_0_uf": 0.04689166452158737,
+    "eval_task_excess_loss_1_rlced_conifer": 0.0474511875690677,
+    "eval_task_loss_0_uf": 0.49440711736679077,
+    "eval_task_loss_1_rlced_conifer": 0.2101784497499466
+}