jmeneu/Fine-tuning-Mistral

README.md

@@ -19,8 +19,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the imdb dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6942
-- Accuracy: {'accuracy': 0.50984}
+- Loss: 7.4068
+- Accuracy: {'accuracy': 0.9338}
 
 ## Model description
 
@@ -39,28 +39,27 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.001
-- train_batch_size: 2
-- eval_batch_size: 2
+- learning_rate: 0.01
+- train_batch_size: 4
+- eval_batch_size: 4
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- num_epochs: 5
+- lr_scheduler_type: constant
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 3
 
 ### Training results
 
-| Training Loss | Epoch | Step  | Validation Loss | Accuracy              |
-|:-------------:|:-----:|:-----:|:---------------:|:---------------------:|
-| 1.5482        | 1.0   | 12500 | 1.3653          | {'accuracy': 0.58188} |
-| 0.7082        | 2.0   | 25000 | 0.6882          | {'accuracy': 0.55924} |
-| 0.7127        | 3.0   | 37500 | 0.6934          | {'accuracy': 0.50536} |
-| 0.6993        | 4.0   | 50000 | 0.6948          | {'accuracy': 0.5022}  |
-| 0.693         | 5.0   | 62500 | 0.6942          | {'accuracy': 0.50984} |
+| Training Loss | Epoch | Step  | Validation Loss | Accuracy             |
+|:-------------:|:-----:|:-----:|:---------------:|:--------------------:|
+| 57.25         | 1.0   | 5000  | 9.4656          | {'accuracy': 0.9292} |
+| 0.0           | 2.0   | 10000 | 8.0567          | {'accuracy': 0.9384} |
+| 0.0           | 3.0   | 15000 | 7.4068          | {'accuracy': 0.9338} |
 
 
 ### Framework versions
 
 - Transformers 4.35.2
-- Pytorch 2.1.0+cu118
-- Datasets 2.15.0
+- Pytorch 2.1.1+cu121
+- Datasets 2.4.0
 - Tokenizers 0.15.0

adapter_config.json

@@ -1,7 +1,7 @@
 {
   "alpha_pattern": {},
   "auto_mapping": null,
-  "base_model_name_or_path": "mistralai/Mistral-7B-v0.1",
+  "base_model_name_or_path": null,
   "bias": "none",
   "fan_in_fan_out": false,
   "inference_mode": true,
@@ -16,8 +16,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "v_proj"
+    "v_proj",
+    "q_proj"
   ],
   "task_type": "SEQ_CLS"
 }

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cdb0d2945402fe5d8ffa85d3bb6bdf0cede8ac44e1e5adf7580f1e009bd3d9f7
-size 3441168
+oid sha256:09445afa535bdc15e772b6f9c07ae2bd3acbe0e81831a622c42bb8eb8f2ae67f
+size 3443360

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[PAD]": 32000
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:726aff8defb984a7f4487d0b3046690bdfdfdd7c864269e7a5e6a9ed02458b1a
-size 4664
+oid sha256:30a1d1b03f6a1d8817ca86a84f3165ab5dd92d5b25acb127f1e184f2fd26537d
+size 4600