End of training

README.md

@@ -17,6 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 # ft-google-gemma-2b-it-qlora
 
 This model is a fine-tuned version of [google/gemma-2b-it](https://huggingface.co/google/gemma-2b-it) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.7355
 
 ## Model description
 
@@ -43,11 +45,16 @@ The following hyperparameters were used during training:
 - total_train_batch_size: 128
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: constant
-- num_epochs: 2
+- num_epochs: 10
 - mixed_precision_training: Native AMP
 
 ### Training results
 
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 0.1332        | 3.0   | 3    | 2.5378          |
+| 0.0683        | 6.0   | 6    | 2.5858          |
+| 0.0236        | 9.0   | 9    | 2.7355          |
 
 
 ### Framework versions

adapter_config.json

@@ -22,13 +22,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "gate_proj",
-    "v_proj",
     "o_proj",
+    "up_proj",
     "down_proj",
+    "gate_proj",
     "k_proj",
-    "up_proj"
+    "v_proj",
+    "q_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:d5a1e80cdf4d9589cb3464c0ebbefc672fd365fff758efa6a085bd5fb8a89761
+oid sha256:1306fc825bd6b4779454a9c73f3b63a6ec5b4c832cf30bb173fa1849e5b3e3e4
 size 4272785368

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a5e1b95d13f1f742f86006934a11e493977572ccebd7789b6dfba52eb445091
+oid sha256:b2387bc0ec1e091240f40e33e26537876055316d5030ebd4f152ea413ca6d9d0
 size 4920