Pretrained with noised Malayalam for denoising version mbart_mldenoised_v1

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/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1827
+- Loss: 0.1520
 
 ## Model description
 
@@ -46,13 +46,13 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 2.2931        | 1.0   | 125  | 0.1883          |
-| 0.1661        | 2.0   | 250  | 0.1827          |
+| 0.1797        | 1.0   | 625  | 0.1577          |
+| 0.1154        | 2.0   | 1250 | 0.1520          |
 
 
 ### Framework versions
 
-- Transformers 4.46.3
+- Transformers 4.47.0
 - Pytorch 2.5.1+cu121
 - Datasets 3.1.0
-- Tokenizers 0.20.3
+- Tokenizers 0.21.0

generation_config.json

@@ -8,5 +8,5 @@
   "max_length": 200,
   "num_beams": 5,
   "pad_token_id": 1,
-  "transformers_version": "4.46.3"
+  "transformers_version": "4.47.0"
 }

tokenizer_config.json

@@ -515,6 +515,7 @@
   "clean_up_tokenization_spaces": false,
   "cls_token": "<s>",
   "eos_token": "</s>",
+  "extra_special_tokens": {},
   "language_codes": "ML50",
   "mask_token": "<mask>",
   "model_max_length": 1000000000000000019884624838656,