Model save

README.md

@@ -1,17 +1,11 @@
 ---
 library_name: transformers
-license: apache-2.0
-base_model: mistralai/Mistral-7B-Instruct-v0.2
+license: llama3
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
 tags:
-- alignment-handbook
 - trl
 - simpo
 - generated_from_trainer
-- trl
-- simpo
-- generated_from_trainer
-datasets:
-- princeton-nlp/mistral-instruct-ultrafeedback
 model-index:
 - name: simpo-exps
   results: []
@@ -22,17 +16,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # simpo-exps
 
-This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the princeton-nlp/mistral-instruct-ultrafeedback dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7905
-- Rewards/chosen: -1.7087
-- Rewards/rejected: -1.8663
-- Rewards/accuracies: 0.6051
-- Rewards/margins: 0.1576
-- Logps/rejected: -0.7465
-- Logps/chosen: -0.6835
-- Logits/rejected: -2.7543
-- Logits/chosen: -2.7563
+- Loss: 1.1795
+- Rewards/chosen: -17.3452
+- Rewards/rejected: -23.3517
+- Rewards/accuracies: 0.8435
+- Rewards/margins: 6.0065
+- Logps/rejected: -2.3352
+- Logps/chosen: -1.7345
+- Logits/rejected: -1.4942
+- Logits/chosen: -1.4832
 
 ## Model description
 
@@ -51,7 +45,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 5e-07
+- learning_rate: 1e-06
 - train_batch_size: 2
 - eval_batch_size: 4
 - seed: 42
@@ -69,7 +63,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
 |:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
-| 0.7494        | 0.8573 | 400  | 0.7909          | -1.6908        | -1.8447          | 0.6024             | 0.1538          | -0.7379        | -0.6763      | -2.7573         | -2.7593       |
+| 1.1882        | 0.8550 | 400  | 1.1795          | -17.3452       | -23.3517         | 0.8435             | 6.0065          | -2.3352        | -1.7345      | -1.4942         | -1.4832       |
 
 
 ### Framework versions

all_results.json

@@ -1,22 +1,9 @@
 {
-    "epoch": 0.9987943737441393,
-    "eval_logits/chosen": -2.756258010864258,
-    "eval_logits/rejected": -2.754289388656616,
-    "eval_logps/chosen": -0.6834676861763,
-    "eval_logps/rejected": -0.7465164661407471,
-    "eval_loss": 0.7904670238494873,
-    "eval_rewards/accuracies": 0.605053186416626,
-    "eval_rewards/chosen": -1.7086691856384277,
-    "eval_rewards/margins": 0.15762200951576233,
-    "eval_rewards/rejected": -1.8662911653518677,
-    "eval_runtime": 102.313,
-    "eval_samples": 2994,
-    "eval_samples_per_second": 29.263,
-    "eval_steps_per_second": 1.837,
+    "epoch": 0.9982631930527722,
     "total_flos": 0.0,
-    "train_loss": 0.7914171927476645,
-    "train_runtime": 8402.7408,
-    "train_samples": 59720,
-    "train_samples_per_second": 7.107,
-    "train_steps_per_second": 0.055
+    "train_loss": 1.7882541670789045,
+    "train_runtime": 8154.3088,
+    "train_samples": 59876,
+    "train_samples_per_second": 7.343,
+    "train_steps_per_second": 0.057
 }

generation_config.json

@@ -1,6 +1,12 @@
 {
-  "_from_model_config": true,
-  "bos_token_id": 1,
-  "eos_token_id": 2,
+  "bos_token_id": 128000,
+  "do_sample": true,
+  "eos_token_id": [
+    128001,
+    128009
+  ],
+  "max_length": 4096,
+  "temperature": 0.6,
+  "top_p": 0.9,
   "transformers_version": "4.44.2"
 }

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 0.9987943737441393,
+    "epoch": 0.9982631930527722,
     "total_flos": 0.0,
-    "train_loss": 0.7914171927476645,
-    "train_runtime": 8402.7408,
-    "train_samples": 59720,
-    "train_samples_per_second": 7.107,
-    "train_steps_per_second": 0.055
+    "train_loss": 1.7882541670789045,
+    "train_runtime": 8154.3088,
+    "train_samples": 59876,
+    "train_samples_per_second": 7.343,
+    "train_steps_per_second": 0.057
 }