Model save

README.md

@@ -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 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 None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6931
-- Rewards/chosen: 0.0
-- Rewards/rejected: 0.0
-- Rewards/accuracies: 0.0
-- Rewards/margins: 0.0
-- Logps/rejected: -179.8250
-- Logps/chosen: -194.3670
-- Logits/rejected: -3.0430
-- Logits/chosen: -3.0079
+- 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
 
 ## Model description
 
@@ -42,25 +42,27 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.05
+- learning_rate: 5e-07
 - train_batch_size: 8
 - eval_batch_size: 4
 - seed: 42
 - distributed_type: multi-GPU
-- num_devices: 2
+- num_devices: 4
 - gradient_accumulation_steps: 2
-- total_train_batch_size: 32
-- total_eval_batch_size: 8
+- total_train_batch_size: 64
+- total_eval_batch_size: 16
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 1
+- num_epochs: 3
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
 |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
-| 0.3466        | 1.0   | 1    | 0.6931          | 0.0            | 0.0              | 0.0                | 0.0             | -179.8250      | -194.3670    | -3.0430         | -3.0079       |
+| 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       |
 
 
 ### Framework versions

all_results.json

@@ -1,8 +1,21 @@
 {
-    "epoch": 0.86,
-    "train_loss": 1.0374332269032795,
-    "train_runtime": 116.0419,
+    "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,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 4.503,
+    "eval_steps_per_second": 0.281,
+    "train_loss": 0.36701411283100355,
+    "train_runtime": 84636.1866,
     "train_samples": 61966,
-    "train_samples_per_second": 0.862,
-    "train_steps_per_second": 0.026
+    "train_samples_per_second": 2.196,
+    "train_steps_per_second": 0.034
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 1.0,
-    "eval_logits/chosen": -3.0078587532043457,
-    "eval_logits/rejected": -3.042999744415283,
-    "eval_logps/chosen": -194.36697387695312,
-    "eval_logps/rejected": -179.82501220703125,
-    "eval_loss": 0.6931471824645996,
-    "eval_rewards/accuracies": 0.0,
-    "eval_rewards/chosen": 0.0,
-    "eval_rewards/margins": 0.0,
-    "eval_rewards/rejected": 0.0,
-    "eval_runtime": 5.0964,
-    "eval_samples": 10,
-    "eval_samples_per_second": 1.962,
-    "eval_steps_per_second": 0.392
+    "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,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 4.503,
+    "eval_steps_per_second": 0.281
 }

runs/Dec10_14-39-55_uclaml03.cs.ucla.edu/events.out.tfevents.1702333147.uclaml03.cs.ucla.edu.2921932.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98c72e0be945c8e8579e85873da199ac30c8349d4cc8c0f46bc59098496e7072
+size 828

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 0.86,
-    "train_loss": 1.0374332269032795,
-    "train_runtime": 116.0419,
+    "epoch": 3.0,
+    "train_loss": 0.36701411283100355,
+    "train_runtime": 84636.1866,
     "train_samples": 61966,
-    "train_samples_per_second": 0.862,
-    "train_steps_per_second": 0.026
+    "train_samples_per_second": 2.196,
+    "train_steps_per_second": 0.034
 }