dmcooller/neural-matia-phi-ft-3

README.md

@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2938
+- Loss: 0.3371
 
 ## Model description
 
@@ -44,22 +44,17 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 2
-- num_epochs: 10
+- num_epochs: 5
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.9354        | 1.0   | 9    | 0.8417          |
-| 0.6176        | 2.0   | 18   | 0.3875          |
-| 0.382         | 3.0   | 27   | 0.3327          |
-| 0.3369        | 4.0   | 36   | 0.3130          |
-| 0.3179        | 5.0   | 45   | 0.3053          |
-| 0.3108        | 6.0   | 54   | 0.3008          |
-| 0.3043        | 7.0   | 63   | 0.2974          |
-| 0.3015        | 8.0   | 72   | 0.2952          |
-| 0.2982        | 9.0   | 81   | 0.2941          |
-| 0.2966        | 10.0  | 90   | 0.2938          |
+| 2.7573        | 1.0   | 9    | 1.8932          |
+| 1.3209        | 2.0   | 18   | 0.6054          |
+| 0.5213        | 3.0   | 27   | 0.3828          |
+| 0.3903        | 4.0   | 36   | 0.3464          |
+| 0.3588        | 5.0   | 45   | 0.3371          |
 
 
 ### Framework versions

adapter_config.json

@@ -2,7 +2,7 @@
   "alpha_pattern": {},
   "auto_mapping": null,
   "base_model_name_or_path": "microsoft/phi-2",
-  "bias": "none",
+  "bias": "all",
   "fan_in_fan_out": false,
   "inference_mode": true,
   "init_lora_weights": true,
@@ -20,10 +20,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "dense",
     "k_proj",
-    "v_proj"
+    "q_proj",
+    "v_proj",
+    "dense"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a5d0e0fe3ad8d57b4c0785987f95e2c08d75b4a780975a9e2e6878afe4f71e76
-size 83920464
+oid sha256:88dd44df2c6fa668f634807a17ff6f6553c84daa1a8d911d87c613bde513327c
+size 87440736

runs/Apr08_09-27-31_ba950c75e1e3/events.out.tfevents.1712568458.ba950c75e1e3.75.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c560845d4341022688c0846d354fddf93b638b1333fc2ca5c27204075bebaa22
+size 5277

runs/Apr08_09-27-31_ba950c75e1e3/events.out.tfevents.1712568510.ba950c75e1e3.75.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4d2442767f50773aaf6038b6efeb3da3abbf1a225907c95ec4c827da139ba4e
+size 7990

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9eabb7cb8f3f9d859feb44979cc23d6d80f7e79f68197cbe706cb2f1c02d4ece
+oid sha256:2223e83f9336243cc114fdf3e44ef4335fdff8b0d59b1f406f227096db3d23ae
 size 4920