Model save

README.md

@@ -1,6 +1,6 @@
 ---
 license: apache-2.0
-base_model: alignment-handbook/zephyr-7b-sft-full
+base_model: mistralai/Mistral-7B-v0.1
 tags:
 - generated_from_trainer
 model-index:
@@ -13,17 +13,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # zephyr-7b-dpo-lora
 
-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.
+This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: -0.2038
-- Rewards/chosen: -1.1628
-- Rewards/rejected: -2.4457
-- Rewards/accuracies: 0.6840
-- Rewards/margins: 1.2829
-- Logps/rejected: -252.9479
-- Logps/chosen: -282.7848
-- Logits/rejected: -2.9400
-- Logits/chosen: -2.9655
+- Loss: 0.6642
+- Rewards/chosen: 0.1042
+- Rewards/rejected: 0.0401
+- Rewards/accuracies: 0.6480
+- Rewards/margins: 0.0641
+- Logps/rejected: -230.4560
+- Logps/chosen: -278.6917
+- Logits/rejected: -2.3987
+- Logits/chosen: -2.4597
 
 ## Model description
 
@@ -54,15 +54,13 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 3
+- num_epochs: 1
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
 |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
-| 0.6137        | 1.0   | 968  | 0.6277          | -0.0287        | -0.4191          | 0.7040             | 0.3905          | -232.6823      | -271.4433    | -2.9989         | -3.0154       |
-| 0.0705        | 2.0   | 1937 | 0.0570          | -0.6708        | -1.6676          | 0.6960             | 0.9968          | -245.1669      | -277.8647    | -2.9609         | -2.9830       |
-| -0.2602       | 3.0   | 2904 | -0.2038         | -1.1628        | -2.4457          | 0.6840             | 1.2829          | -252.9479      | -282.7848    | -2.9400         | -2.9655       |
+| 0.661         | 1.0   | 968  | 0.6642          | 0.1042         | 0.0401           | 0.6480             | 0.0641          | -230.4560      | -278.6917    | -2.3987         | -2.4597       |
 
 
 ### Framework versions

all_results.json

@@ -1,21 +1,21 @@
 {
-    "epoch": 3.0,
-    "eval_logits/chosen": -2.965512990951538,
-    "eval_logits/rejected": -2.9399757385253906,
-    "eval_logps/chosen": -282.7847900390625,
-    "eval_logps/rejected": -252.9479217529297,
-    "eval_loss": -0.203842431306839,
-    "eval_rewards/accuracies": 0.6840000152587891,
-    "eval_rewards/chosen": -1.1628247499465942,
-    "eval_rewards/margins": 1.2828813791275024,
-    "eval_rewards/rejected": -2.4457061290740967,
-    "eval_runtime": 444.1107,
+    "epoch": 1.0,
+    "eval_logits/chosen": -2.4597132205963135,
+    "eval_logits/rejected": -2.398695468902588,
+    "eval_logps/chosen": -278.69171142578125,
+    "eval_logps/rejected": -230.4560089111328,
+    "eval_loss": 0.6642152070999146,
+    "eval_rewards/accuracies": 0.6480000019073486,
+    "eval_rewards/chosen": 0.10415761172771454,
+    "eval_rewards/margins": 0.06405296921730042,
+    "eval_rewards/rejected": 0.04010463133454323,
+    "eval_runtime": 444.8959,
     "eval_samples": 2000,
-    "eval_samples_per_second": 4.503,
+    "eval_samples_per_second": 4.495,
     "eval_steps_per_second": 0.281,
-    "train_loss": 0.36701411283100355,
-    "train_runtime": 84636.1866,
+    "train_loss": 0.6728762634529555,
+    "train_runtime": 27528.1814,
     "train_samples": 61966,
-    "train_samples_per_second": 2.196,
-    "train_steps_per_second": 0.034
+    "train_samples_per_second": 2.251,
+    "train_steps_per_second": 0.035
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 3.0,
-    "eval_logits/chosen": -2.965512990951538,
-    "eval_logits/rejected": -2.9399757385253906,
-    "eval_logps/chosen": -282.7847900390625,
-    "eval_logps/rejected": -252.9479217529297,
-    "eval_loss": -0.203842431306839,
-    "eval_rewards/accuracies": 0.6840000152587891,
-    "eval_rewards/chosen": -1.1628247499465942,
-    "eval_rewards/margins": 1.2828813791275024,
-    "eval_rewards/rejected": -2.4457061290740967,
-    "eval_runtime": 444.1107,
+    "epoch": 1.0,
+    "eval_logits/chosen": -2.4597132205963135,
+    "eval_logits/rejected": -2.398695468902588,
+    "eval_logps/chosen": -278.69171142578125,
+    "eval_logps/rejected": -230.4560089111328,
+    "eval_loss": 0.6642152070999146,
+    "eval_rewards/accuracies": 0.6480000019073486,
+    "eval_rewards/chosen": 0.10415761172771454,
+    "eval_rewards/margins": 0.06405296921730042,
+    "eval_rewards/rejected": 0.04010463133454323,
+    "eval_runtime": 444.8959,
     "eval_samples": 2000,
-    "eval_samples_per_second": 4.503,
+    "eval_samples_per_second": 4.495,
     "eval_steps_per_second": 0.281
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "train_loss": 0.36701411283100355,
-    "train_runtime": 84636.1866,
+    "epoch": 1.0,
+    "train_loss": 0.6728762634529555,
+    "train_runtime": 27528.1814,
     "train_samples": 61966,
-    "train_samples_per_second": 2.196,
-    "train_steps_per_second": 0.034
+    "train_samples_per_second": 2.251,
+    "train_steps_per_second": 0.035
 }