Yirui091/lab1_finetuning

README.md

@@ -9,7 +9,7 @@ datasets:
 metrics:
 - bleu
 model-index:
-- name: marian-finetuned-kde4-en-to-fr
+- name: lab1_finetuning
   results:
   - task:
       name: Sequence-to-sequence Language Modeling
@@ -23,19 +23,19 @@ model-index:
     metrics:
     - name: Bleu
       type: bleu
-      value: 48.93253657516627
+      value: 48.52734685570538
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# marian-finetuned-kde4-en-to-fr
+# lab1_finetuning
 
 This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on the kde4 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0256
-- Model Preparation Time: 0.0267
-- Bleu: 48.9325
+- Loss: 1.0255
+- Model Preparation Time: 0.003
+- Bleu: 48.5273
 
 ## Model description
 
@@ -58,7 +58,7 @@ The following hyperparameters were used during training:
 - train_batch_size: 16
 - eval_batch_size: 32
 - seed: 42
-- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - training_steps: 5000
 - mixed_precision_training: Native AMP
@@ -69,7 +69,7 @@ The following hyperparameters were used during training:
 
 ### Framework versions
 
-- Transformers 4.57.0
+- Transformers 4.45.2
 - Pytorch 2.10.0+cu128
 - Datasets 2.18.0
-- Tokenizers 0.22.2
+- Tokenizers 0.20.3

generation_config.json

@@ -4,14 +4,13 @@
       59513
     ]
   ],
+  "bos_token_id": 0,
   "decoder_start_token_id": 59513,
-  "eos_token_id": [
-    0
-  ],
+  "eos_token_id": 0,
   "forced_eos_token_id": 0,
   "max_length": 512,
   "num_beams": 4,
   "pad_token_id": 59513,
   "renormalize_logits": true,
-  "transformers_version": "4.57.0"
+  "transformers_version": "4.45.2"
 }