End of training

README.md

@@ -18,11 +18,11 @@ 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: 1.3382
+- Loss: 0.6324
 - Eval/rewards/chosen: 0.0
-- Eval/logps/chosen: -192.8253
+- Eval/logps/chosen: -203.5358
 - Eval/rewards/rejected: 0.0
-- Eval/logps/rejected: -229.8692
+- Eval/logps/rejected: -215.2117
 - Eval/rewards/margins: 0.0
 - Eval/kl: 0.0
 
@@ -46,7 +46,7 @@ The following hyperparameters were used during training:
 - learning_rate: 0.0
 - 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
@@ -58,7 +58,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |     |
 |:-------------:|:-----:|:----:|:---------------:|:---:|
-| 3.3125        | 0.96  | 12   | 1.3382          | 0.0 |
+| 0.5625        | 0.96  | 12   | 0.6324          | 0.0 |
 
 
 ### Framework versions

adapter_config.json

@@ -20,10 +20,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "o_proj",
+    "q_proj",
     "k_proj",
-    "q_proj"
+    "o_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42d136e0b0e73d38dc7c67bedb6732e527d052b8c2640d6e3f1b100b2c46a2a8
+oid sha256:6c21ab0c04e30634c1e6c3d153a150d25a226a07a986f1724f5da941a4be84dd
 size 27297544

metrics.jsonl

@@ -18,3 +18,5 @@
 {"epoch": 0.96, "precision": 0.41463414533016063, "recall": 0.9444444391975308, "fold": 0}
 {"epoch": -1, "precision": 0.41463414533016063, "recall": 0.9444444391975308, "fold": 0}
 {"epoch": 0.96, "precision": 0.41463414533016063, "recall": 0.9444444391975308, "fold": 0}
+{"epoch": -1, "precision": 0.3529411757785467, "recall": 0.9999999944444444, "fold": 0}
+{"epoch": 0.96, "precision": 0.3529411757785467, "recall": 0.9999999944444444, "fold": 0}

metrics_epoch_-1_fold_0_lr_0_seed_5678_weight_0.json

@@ -0,0 +1 @@
+{"epoch": -1, "precision": 0.3529411757785467, "recall": 0.9999999944444444, "fold": 0}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2167c112c737455967bddb86d7d85056a717aa8b523580a6ee38098b772da94d
+oid sha256:4faa8cc33f1a3342c3fecb3a69ccbadca0ad4e833bf943ad7f9c37ef6ed154a2
 size 5688