Rameesha/test_cdp

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/opt-iml-1.3b](https://huggingface.co/facebook/opt-iml-1.3b) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.6184
+- Loss: 0.6240
 
 ## Model description
 
@@ -34,7 +34,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.0003
+- learning_rate: 0.0002
 - train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
@@ -43,37 +43,27 @@ 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: 100
-- training_steps: 400
+- training_steps: 200
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 3.5718        | 2.76  | 20   | 3.6184          |
-| 3.5423        | 5.52  | 40   | 3.6184          |
-| 3.5865        | 8.28  | 60   | 3.6184          |
-| 3.562         | 11.03 | 80   | 3.6184          |
-| 3.5519        | 13.79 | 100  | 3.6184          |
-| 3.5689        | 16.55 | 120  | 3.6184          |
-| 3.5551        | 19.31 | 140  | 3.6184          |
-| 3.5397        | 22.07 | 160  | 3.6184          |
-| 3.5849        | 24.83 | 180  | 3.6184          |
-| 3.5773        | 27.59 | 200  | 3.6184          |
-| 3.5301        | 30.34 | 220  | 3.6184          |
-| 3.5603        | 33.1  | 240  | 3.6184          |
-| 3.5714        | 35.86 | 260  | 3.6184          |
-| 3.5642        | 38.62 | 280  | 3.6184          |
-| 3.5559        | 41.38 | 300  | 3.6184          |
-| 3.5775        | 44.14 | 320  | 3.6184          |
-| 3.566         | 46.9  | 340  | 3.6184          |
-| 3.5622        | 49.66 | 360  | 3.6184          |
-| 3.5823        | 52.41 | 380  | 3.6184          |
-| 3.5603        | 55.17 | 400  | 3.6184          |
+| 3.4167        | 2.76  | 20   | 3.3131          |
+| 2.822         | 5.52  | 40   | 2.6244          |
+| 1.8781        | 8.28  | 60   | 1.7491          |
+| 1.1619        | 11.03 | 80   | 1.3205          |
+| 0.7079        | 13.79 | 100  | 1.0339          |
+| 0.4644        | 16.55 | 120  | 0.8472          |
+| 0.3597        | 19.31 | 140  | 0.7265          |
+| 0.3092        | 22.07 | 160  | 0.6580          |
+| 0.2789        | 24.83 | 180  | 0.6310          |
+| 0.2646        | 27.59 | 200  | 0.6240          |
 
 
 ### Framework versions
 
-- Transformers 4.33.3
+- Transformers 4.34.0
 - Pytorch 2.0.1+cu118
 - Datasets 2.14.5
-- Tokenizers 0.13.3
+- Tokenizers 0.14.0

adapter_config.json

@@ -1,6 +1,6 @@
 {
   "auto_mapping": null,
-  "base_model_name_or_path": null,
+  "base_model_name_or_path": "facebook/opt-iml-1.3b",
   "bias": "none",
   "fan_in_fan_out": false,
   "inference_mode": true,
@@ -11,7 +11,7 @@
   "lora_dropout": 0.05,
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 16,
+  "r": 8,
   "revision": null,
   "target_modules": [
     "q_proj",

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0147be7d3e29377510a7af7dc836a0d8100b1d61ff50165dce733054100462c
-size 4027
+oid sha256:b0b30dc83d5418e230308d1c9ae61ca03fa0aff3b748099e378fa7016b0e27ee
+size 4091