Model save

README.md

@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/byt5-base](https://huggingface.co/google/byt5-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1012
+- Loss: 0.0313
 
 ## Model description
 
@@ -37,25 +37,26 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 5e-05
+- learning_rate: 3e-05
 - train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
 - gradient_accumulation_steps: 2
 - total_train_batch_size: 16
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
-- lr_scheduler_type: linear
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
 - num_epochs: 5
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 0.3238        | 1.0   | 971  | 0.1668          |
-| 0.2426        | 2.0   | 1942 | 0.1217          |
-| 0.2278        | 3.0   | 2913 | 0.1152          |
-| 0.2076        | 4.0   | 3884 | 0.1062          |
-| 0.1946        | 5.0   | 4855 | 0.1012          |
+| 0.2915        | 1.0   | 971  | 0.1431          |
+| 0.1664        | 2.0   | 1942 | 0.0804          |
+| 0.0931        | 3.0   | 2913 | 0.0365          |
+| 0.0768        | 4.0   | 3884 | 0.0319          |
+| 0.076         | 5.0   | 4855 | 0.0313          |
 
 
 ### Framework versions

adapter_config.json

@@ -13,7 +13,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 32,
+  "lora_alpha": 64,
   "lora_bias": false,
   "lora_dropout": 0.05,
   "megatron_config": null,
@@ -21,12 +21,14 @@
   "modules_to_save": null,
   "peft_type": "LORA",
   "qalora_group_size": 16,
-  "r": 16,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "o",
     "q",
-    "v"
+    "v",
+    "k"
   ],
   "task_type": "SEQ_2_SEQ_LM",
   "trainable_token_indices": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3da6f1cf1a94b3131627dda9b89e692d2ccab3ea11b285398a3facfa546f22a4
-size 8864192
+oid sha256:21fe614b7a89d08829e033dca2e81c895db46dfe8d6f5c2d0cd09cf4ede9c6a9
+size 35423448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e0c214c005419021773232efae913c409aa7202e8e7e16875d75f407e83f7a8
+oid sha256:16f042ff3cdbede964873f42ac6dc80e1d4ff2b4c66c11330b0531eb8ee392ef
 size 5969