End of training

README.md

@@ -32,14 +32,21 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
-- train_batch_size: 2
-- eval_batch_size: 2
+- train_batch_size: 8
+- eval_batch_size: 8
 - seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 8
+- gradient_accumulation_steps: 16
+- total_train_batch_size: 128
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 60
+- num_epochs: 1
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
+|:-------------:|:-----:|:----:|:---------------:|:----:|:-------:|
+| No log        | 0.96  | 21   | 1.3691          | 0.0  | 19.0    |
+
 
 ### Framework versions
 

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:575378cfe82c5fe1b660e337ef925086a3d10e20a89693e354b22a580803ef77
+oid sha256:c3e134b4a856674b2ca06ce11c3cd228a1458bdedaf9bcaad9262f3dacc7d8f1
 size 990236853

tokenizer_config.json

@@ -104,8 +104,12 @@
   "clean_up_tokenization_spaces": true,
   "eos_token": "</s>",
   "extra_ids": 100,
+  "max_length": 1024,
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": "<pad>",
+  "stride": 0,
   "tokenizer_class": "T5Tokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
   "unk_token": "<unk>"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c0167e949de2f51cece562fbcd38f14e93565a982d3bd0717d016d1edf5a79c1
+oid sha256:2da484213889f405fb841ef3ea597caa56d45c6ea603341ac9cf26117859e10b
 size 4155