HuggingFaceTB/SmolLM-360M-Instruct

README.md

@@ -20,7 +20,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [HuggingFaceTB/SmolLM-360M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM-360M-Instruct) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.0409
+- Loss: 1.7710
 
 ## Model description
 
@@ -39,7 +39,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.0002
+- learning_rate: 0.001
 - train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
@@ -51,18 +51,32 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: constant
 - lr_scheduler_warmup_ratio: 0.03
-- num_epochs: 6
+- num_epochs: 20
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss |
-|:-------------:|:------:|:----:|:---------------:|
-| 2.3444        | 0.9524 | 10   | 2.3038          |
-| 2.2609        | 2.0    | 21   | 2.1919          |
-| 2.1732        | 2.9524 | 31   | 2.1334          |
-| 2.1255        | 4.0    | 42   | 2.0941          |
-| 2.0831        | 4.9524 | 52   | 2.0632          |
-| 2.0563        | 5.7143 | 60   | 2.0409          |
+| Training Loss | Epoch   | Step | Validation Loss |
+|:-------------:|:-------:|:----:|:---------------:|
+| 2.2354        | 0.9524  | 10   | 2.1179          |
+| 2.0715        | 2.0     | 21   | 2.0075          |
+| 1.9885        | 2.9524  | 31   | 1.9518          |
+| 1.9385        | 4.0     | 42   | 1.9090          |
+| 1.8925        | 4.9524  | 52   | 1.8811          |
+| 1.867         | 6.0     | 63   | 1.8578          |
+| 1.8507        | 6.9524  | 73   | 1.8416          |
+| 1.8314        | 8.0     | 84   | 1.8279          |
+| 1.8086        | 8.9524  | 94   | 1.8177          |
+| 1.7969        | 10.0    | 105  | 1.8085          |
+| 1.7858        | 10.9524 | 115  | 1.8016          |
+| 1.7711        | 12.0    | 126  | 1.7957          |
+| 1.7616        | 12.9524 | 136  | 1.7904          |
+| 1.7504        | 14.0    | 147  | 1.7854          |
+| 1.7399        | 14.9524 | 157  | 1.7821          |
+| 1.732         | 16.0    | 168  | 1.7791          |
+| 1.7218        | 16.9524 | 178  | 1.7755          |
+| 1.7142        | 18.0    | 189  | 1.7735          |
+| 1.6992        | 18.9524 | 199  | 1.7708          |
+| 1.6993        | 19.0476 | 200  | 1.7710          |
 
 
 ### Framework versions

adapter_config.json

@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "gate_proj",
-    "up_proj",
+    "k_proj",
     "o_proj",
     "q_proj",
-    "k_proj",
-    "down_proj",
-    "v_proj"
+    "gate_proj",
+    "v_proj",
+    "up_proj",
+    "down_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:e0216683abaa9c920d78c4c5a90aed2f701cdb4ae4b481dc09618211a026e35d
+oid sha256:31c655e84ff620beca55c5028fc8e5f810800373205f1a8dab2d4379558b1bef
 size 17426248

runs/Aug28_13-15-09_algo-1/events.out.tfevents.1724850924.algo-1.67.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f93a6bb2ccc83a4cd3e3ac02d6438a63a93a6aa05d9ae451b70af47e501e78c
+size 15294

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d57dd6dc0a640f00c6f19e312c5fd1fd07932eb590221acbbcd4acbb9a2c7ba0
+oid sha256:4fef496847ab2924b511e4f67bfdd2b7f90755757b282a49c79c06fd4b763a1a
 size 5240