Fine-tuned with denoising, version mbart_denoised_v1

README.md

@@ -1,6 +1,6 @@
 ---
 library_name: transformers
-base_model: facebook/mbart-large-50-many-to-one-mmt
+base_model: facebook/mbart-large-50-one-to-many-mmt
 tags:
 - generated_from_trainer
 model-index:
@@ -13,9 +13,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # e2m_denoise_project
 
-This model is a fine-tuned version of [facebook/mbart-large-50-many-to-one-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-one-mmt) on the None dataset.
+This model is a fine-tuned version of [facebook/mbart-large-50-one-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-one-to-many-mmt) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1636
+- Loss: 0.3997
 
 ## Model description
 
@@ -46,8 +46,8 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 2.6348        | 1.0   | 125  | 0.1741          |
-| 0.1862        | 2.0   | 250  | 0.1636          |
+| 2.8775        | 1.0   | 125  | 0.4854          |
+| 0.4479        | 2.0   | 250  | 0.3997          |
 
 
 ### Framework versions

generation_config.json

@@ -2,8 +2,8 @@
   "_from_model_config": true,
   "bos_token_id": 0,
   "decoder_start_token_id": 2,
+  "early_stopping": true,
   "eos_token_id": 2,
-  "forced_bos_token_id": 250004,
   "forced_eos_token_id": 2,
   "max_length": 200,
   "num_beams": 5,