Model save

README.md

@@ -2,14 +2,10 @@
 license: apache-2.0
 library_name: peft
 tags:
-- alignment-handbook
-- generated_from_trainer
 - trl
 - dpo
-base_model: mistralai/Mistral-7B-v0.1
-datasets:
-- updated
-- original
+- generated_from_trainer
+base_model: alignment-handbook/zephyr-7b-sft-full
 model-index:
 - name: zephyr-7b-dpo-qlora-min-pi-part-0
   results: []
@@ -20,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-dpo-qlora-min-pi-part-0
 
-This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-qlora](https://huggingface.co/alignment-handbook/zephyr-7b-sft-qlora) on the updated and the original datasets.
+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 None dataset.
 
 ## Model description
 

all_results.json

@@ -1,21 +1,8 @@
 {
-    "epoch": 1.0,
-    "eval_logits/chosen": -2.498661994934082,
-    "eval_logits/rejected": -2.4008209705352783,
-    "eval_logps/chosen": -269.65570068359375,
-    "eval_logps/rejected": -257.6377868652344,
-    "eval_loss": 0.6512373089790344,
-    "eval_rewards/accuracies": 0.6639999747276306,
-    "eval_rewards/chosen": -0.06204665079712868,
-    "eval_rewards/margins": 0.10243026912212372,
-    "eval_rewards/rejected": -0.1644769161939621,
-    "eval_runtime": 542.9483,
-    "eval_samples": 2000,
-    "eval_samples_per_second": 3.684,
-    "eval_steps_per_second": 0.46,
-    "train_loss": 0.0,
-    "train_runtime": 0.0278,
+    "epoch": 0.77,
+    "train_loss": 0.692988375822703,
+    "train_runtime": 257.3561,
     "train_samples": 10000,
-    "train_samples_per_second": 359461.447,
-    "train_steps_per_second": 1401.9
+    "train_samples_per_second": 3.886,
+    "train_steps_per_second": 0.012
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "train_loss": 0.0,
-    "train_runtime": 0.0278,
+    "epoch": 0.77,
+    "train_loss": 0.692988375822703,
+    "train_runtime": 257.3561,
     "train_samples": 10000,
-    "train_samples_per_second": 359461.447,
-    "train_steps_per_second": 1401.9
+    "train_samples_per_second": 3.886,
+    "train_steps_per_second": 0.012
 }

trainer_state.json

@@ -1,20 +1,20 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9984,
+  "epoch": 0.768,
   "eval_steps": 100,
-  "global_step": 39,
+  "global_step": 3,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.03,
-      "learning_rate": 1.25e-06,
-      "logits/chosen": -2.3930084705352783,
-      "logits/rejected": -2.4061639308929443,
-      "logps/chosen": -245.26866149902344,
-      "logps/rejected": -261.33184814453125,
+      "epoch": 0.26,
+      "learning_rate": 5e-06,
+      "logits/chosen": -2.7575411796569824,
+      "logits/rejected": -2.648534059524536,
+      "logps/chosen": -262.0839538574219,
+      "logps/rejected": -262.1346435546875,
       "loss": 0.6931,
       "rewards/accuracies": 0.0,
       "rewards/chosen": 0.0,
@@ -22,60 +22,18 @@
       "rewards/rejected": 0.0,
       "step": 1
     },
-    {
-      "epoch": 0.26,
-      "learning_rate": 4.646121984004666e-06,
-      "logits/chosen": -2.525505542755127,
-      "logits/rejected": -2.416348695755005,
-      "logps/chosen": -264.3957214355469,
-      "logps/rejected": -243.22918701171875,
-      "loss": 0.6859,
-      "rewards/accuracies": 0.5711805820465088,
-      "rewards/chosen": 0.037347737699747086,
-      "rewards/margins": 0.01542865764349699,
-      "rewards/rejected": 0.02191907912492752,
-      "step": 10
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 2.835583164544139e-06,
-      "logits/chosen": -2.4794507026672363,
-      "logits/rejected": -2.397329568862915,
-      "logps/chosen": -257.61175537109375,
-      "logps/rejected": -238.9801025390625,
-      "loss": 0.6674,
-      "rewards/accuracies": 0.699999988079071,
-      "rewards/chosen": 0.03777569904923439,
-      "rewards/margins": 0.06583560258150101,
-      "rewards/rejected": -0.028059903532266617,
-      "step": 20
-    },
     {
       "epoch": 0.77,
-      "learning_rate": 7.723433775328385e-07,
-      "logits/chosen": -2.5016331672668457,
-      "logits/rejected": -2.392857074737549,
-      "logps/chosen": -282.13555908203125,
-      "logps/rejected": -262.0860595703125,
-      "loss": 0.6557,
-      "rewards/accuracies": 0.671875,
-      "rewards/chosen": -0.02580374851822853,
-      "rewards/margins": 0.10012584924697876,
-      "rewards/rejected": -0.125929594039917,
-      "step": 30
-    },
-    {
-      "epoch": 1.0,
-      "step": 39,
+      "step": 3,
       "total_flos": 0.0,
-      "train_loss": 0.0,
-      "train_runtime": 0.0278,
-      "train_samples_per_second": 359461.447,
-      "train_steps_per_second": 1401.9
+      "train_loss": 0.692988375822703,
+      "train_runtime": 257.3561,
+      "train_samples_per_second": 3.886,
+      "train_steps_per_second": 0.012
     }
   ],
   "logging_steps": 10,
-  "max_steps": 39,
+  "max_steps": 3,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 100,