End of training

README.md

@@ -18,13 +18,13 @@ 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: 1.0466
-- Eval/rewards/chosen: -0.2063
-- Eval/logps/chosen: -199.5112
-- Eval/rewards/rejected: -0.2128
-- Eval/logps/rejected: -256.5334
-- Eval/rewards/margins: 0.0065
-- Eval/kl: 0.0
+- Loss: 0.8384
+- Eval/rewards/chosen: 4.1202
+- Eval/logps/chosen: -159.3281
+- Eval/rewards/rejected: 5.6445
+- Eval/logps/rejected: -183.8697
+- Eval/rewards/margins: -1.5243
+- Eval/kl: 46.6855
 
 ## Model description
 
@@ -43,10 +43,10 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
+- learning_rate: 0.001
 - train_batch_size: 1
 - eval_batch_size: 2
-- seed: 1234
+- seed: 5678
 - gradient_accumulation_steps: 8
 - total_train_batch_size: 8
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
@@ -56,10 +56,10 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss |        |
-|:-------------:|:-----:|:----:|:---------------:|:------:|
-| 1.7092        | 0.96  | 12   | 1.0871          | 0.0293 |
-| 1.1814        | 2.0   | 25   | 1.0466          | 0.0    |
+| Training Loss | Epoch | Step | Validation Loss |         |
+|:-------------:|:-----:|:----:|:---------------:|:-------:|
+| 0.5473        | 0.96  | 12   | 0.7074          | 21.7741 |
+| 0.571         | 2.0   | 25   | 0.8384          | 46.6855 |
 
 
 ### Framework versions

adapter_config.json

@@ -21,8 +21,8 @@
   "revision": null,
   "target_modules": [
     "q_proj",
-    "k_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 sha256:27903a3b3043c02e2d7dc42d3e7f7c844117961a2c15068be632dd5d9ddfb031
+oid sha256:33f4253903a33fecdab9905931d778edf5ca9afefb0ed414f36196e361c13176
 size 27297544

metrics.jsonl

@@ -10,3 +10,5 @@
 {"epoch": 2.0, "precision": 0.8333333287037037, "recall": 0.4999999983333333, "fold": 0}
 {"epoch": 0.96, "precision": 0.4915254228957196, "recall": 0.9666666634444444, "fold": 0}
 {"epoch": 2.0, "precision": 0.5686274498654363, "recall": 0.9666666634444444, "fold": 0}
+{"epoch": 0.96, "precision": 0.3582089546892404, "recall": 0.7999999973333333, "fold": 0}
+{"epoch": 2.0, "precision": 0.624999998046875, "recall": 0.6666666644444444, "fold": 0}

metrics_epoch_0.96_fold_0_lr_0.001_seed_5678_weight_2.0.json

@@ -0,0 +1 @@
+{"epoch": 0.96, "precision": 0.3582089546892404, "recall": 0.7999999973333333, "fold": 0}

metrics_epoch_2.0_fold_0_lr_0.001_seed_5678_weight_2.0.json

@@ -0,0 +1 @@
+{"epoch": 2.0, "precision": 0.624999998046875, "recall": 0.6666666644444444, "fold": 0}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d7fb04fe53a69c0947a1c162ecc7667487c5218a7fa30cd0338ed7127717252
+oid sha256:aa3015864a02e074df16dca1b82335c7f1f02b788ef694b1ccbec216ecd71ce4
 size 5688