Training complete

README.md

@@ -19,9 +19,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/mbart-large-50](https://huggingface.co/facebook/mbart-large-50) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2519
-- Bleu: 87.0261
-- Gen Len: 18.4048
+- Loss: 0.0138
+- Bleu: 98.4772
+- Gen Len: 18.5104
 
 ## Model description
 
@@ -41,28 +41,24 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 5.6e-05
-- train_batch_size: 1
-- eval_batch_size: 1
+- train_batch_size: 8
+- eval_batch_size: 8
 - seed: 42
-- gradient_accumulation_steps: 8
-- total_train_batch_size: 8
 - 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_warmup_ratio: 0.1
 - num_epochs: 2
-- mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Bleu    | Gen Len |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|
-| No log        | 1.0   | 472  | 0.2977          | 85.4471 | 18.2429 |
-| 0.9756        | 2.0   | 944  | 0.2519          | 87.0261 | 18.4048 |
+| No log        | 1.0   | 440  | 0.0246          | 98.2861 | 18.5729 |
+| 0.2226        | 2.0   | 880  | 0.0138          | 98.4772 | 18.5104 |
 
 
 ### Framework versions
 
 - Transformers 4.51.2
 - Pytorch 2.10.0+cu128
-- Datasets 4.0.0
+- Datasets 4.6.0
 - Tokenizers 0.21.4

generation_config.json

@@ -3,11 +3,9 @@
   "decoder_start_token_id": 2,
   "early_stopping": true,
   "eos_token_id": 2,
-  "forced_bos_token_id": 250005,
   "forced_eos_token_id": 2,
   "max_length": 200,
   "num_beams": 5,
   "pad_token_id": 1,
-  "transformers_version": "4.51.2",
-  "use_cache": false
+  "transformers_version": "4.51.2"
 }