End of training

README.md

@@ -18,12 +18,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 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 the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8406
-- Eval/rewards/chosen: -0.5872
-- Eval/logps/chosen: -198.6973
-- Eval/rewards/rejected: -0.6580
-- Eval/logps/rejected: -236.4490
-- Eval/rewards/margins: 0.0708
+- Loss: 0.8136
+- Eval/rewards/chosen: -15.7423
+- Eval/logps/chosen: -350.2482
+- Eval/rewards/rejected: -21.6619
+- Eval/logps/rejected: -446.4881
+- Eval/rewards/margins: 5.9196
 - Eval/kl: 0.0
 
 ## Model description
@@ -43,7 +43,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
+- learning_rate: 0.0001
 - train_batch_size: 1
 - eval_batch_size: 2
 - seed: 1234
@@ -56,14 +56,14 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss |        |
-|:-------------:|:-----:|:----:|:---------------:|:------:|
-| 1.7241        | 0.96  | 12   | 0.8748          | 0.0178 |
-| 1.1636        | 2.0   | 25   | 0.8626          | 0.0    |
-| 1.367         | 2.96  | 37   | 0.8516          | 0.0    |
-| 1.0187        | 4.0   | 50   | 0.8411          | 0.0    |
-| 1.1771        | 4.96  | 62   | 0.8434          | 0.0    |
-| 0.9371        | 5.76  | 72   | 0.8406          | 0.0    |
+| Training Loss | Epoch | Step | Validation Loss |     |
+|:-------------:|:-----:|:----:|:---------------:|:---:|
+| 0.4474        | 0.96  | 12   | 0.8052          | 0.0 |
+| 0.625         | 2.0   | 25   | 0.8136          | 0.0 |
+| 0.375         | 2.96  | 37   | 0.8136          | 0.0 |
+| 0.625         | 4.0   | 50   | 0.8136          | 0.0 |
+| 0.375         | 4.96  | 62   | 0.8136          | 0.0 |
+| 0.625         | 5.76  | 72   | 0.8136          | 0.0 |
 
 
 ### Framework versions

adapter_config.json

@@ -20,9 +20,9 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_proj",
-    "v_proj",
     "q_proj",
+    "v_proj",
+    "k_proj",
     "o_proj"
   ],
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:046bb2b7b57c5714cedb4b3c986808f2255591f70a96d9ebd415d6921b6339a9
+oid sha256:6762a0a26fc0e7ab6290206c8f7f28688d71b8bbe0bebfd84cf7b3dde3572737
 size 27297544

metrics.jsonl

@@ -46,3 +46,9 @@
 {"epoch": 4.0, "precision": 0.5454545438016529, "recall": 0.9999999944444444, "fold": 0}
 {"epoch": 4.96, "precision": 0.49999999852941174, "recall": 0.9444444391975308, "fold": 0}
 {"epoch": 5.76, "precision": 0.47058823391003457, "recall": 0.8888888839506173, "fold": 0}
+{"epoch": 0.96, "precision": 0.7999999946666667, "recall": 0.666666662962963, "fold": 0}
+{"epoch": 2.0, "precision": 0.9999999500000026, "recall": 0.11111111049382716, "fold": 0}
+{"epoch": 2.96, "precision": 0.0, "recall": 0.0, "fold": 0}
+{"epoch": 4.0, "precision": 0.0, "recall": 0.0, "fold": 0}
+{"epoch": 4.96, "precision": 0.0, "recall": 0.0, "fold": 0}
+{"epoch": 5.76, "precision": 0.0, "recall": 0.0, "fold": 0}

metrics_epoch_0.96_fold_0_lr_0.0001_seed_1234_weight_10.0.json

@@ -0,0 +1 @@
+{"epoch": 0.96, "precision": 0.7999999946666667, "recall": 0.666666662962963, "fold": 0}

metrics_epoch_2.0_fold_0_lr_0.0001_seed_1234_weight_10.0.json

@@ -0,0 +1 @@
+{"epoch": 2.0, "precision": 0.9999999500000026, "recall": 0.11111111049382716, "fold": 0}

metrics_epoch_2.96_fold_0_lr_0.0001_seed_1234_weight_10.0.json

@@ -0,0 +1 @@
+{"epoch": 2.96, "precision": 0.0, "recall": 0.0, "fold": 0}

metrics_epoch_4.0_fold_0_lr_0.0001_seed_1234_weight_10.0.json

@@ -0,0 +1 @@
+{"epoch": 4.0, "precision": 0.0, "recall": 0.0, "fold": 0}

metrics_epoch_4.96_fold_0_lr_0.0001_seed_1234_weight_10.0.json

@@ -0,0 +1 @@
+{"epoch": 4.96, "precision": 0.0, "recall": 0.0, "fold": 0}

metrics_epoch_5.76_fold_0_lr_0.0001_seed_1234_weight_10.0.json

@@ -0,0 +1 @@
+{"epoch": 5.76, "precision": 0.0, "recall": 0.0, "fold": 0}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb64b17200bf3c0a8265b814db333d9ae5539134e2f0eacfd343e63ae8debe78
+oid sha256:50b340257b7c876c7bc2e3b43481f52c813b86aef192525a86f64b55c23483da
 size 5688