Training complete

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: google/pegasus-multi_news
+base_model: google/pegasus-cnn_dailymail
 tags:
 - summarization
 - generated_from_trainer
@@ -13,7 +13,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # pegasus-reviews
 
-This model is a fine-tuned version of [google/pegasus-multi_news](https://huggingface.co/google/pegasus-multi_news) on the None dataset.
+This model is a fine-tuned version of [google/pegasus-cnn_dailymail](https://huggingface.co/google/pegasus-cnn_dailymail) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 4.3085
 
 ## Model description
 
@@ -33,18 +35,26 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 5e-05
-- train_batch_size: 1
-- eval_batch_size: 1
+- train_batch_size: 2
+- eval_batch_size: 2
 - seed: 42
 - gradient_accumulation_steps: 16
-- total_train_batch_size: 16
+- total_train_batch_size: 32
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- num_epochs: 10
+- num_epochs: 6
 
 ### Training results
 
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 0.8541        | 1.0   | 1    | 4.3114          |
+| 0.8178        | 2.0   | 2    | 4.3112          |
+| 0.9366        | 3.0   | 3    | 4.3108          |
+| 0.831         | 4.0   | 4    | 4.3102          |
+| 0.8294        | 5.0   | 5    | 4.3094          |
+| 0.2781        | 5.33  | 6    | 4.3085          |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "google/pegasus-multi_news",
+  "_name_or_path": "google/pegasus-cnn_dailymail",
   "activation_dropout": 0.1,
   "activation_function": "relu",
   "add_bias_logits": false,
@@ -37,7 +37,7 @@
     "LABEL_2": 2
   },
   "length_penalty": 0.8,
-  "max_length": 256,
+  "max_length": 128,
   "max_position_embeddings": 1024,
   "min_length": 32,
   "model_type": "pegasus",

generation_config.json

@@ -5,7 +5,7 @@
   "eos_token_id": 1,
   "forced_eos_token_id": 1,
   "length_penalty": 0.8,
-  "max_length": 256,
+  "max_length": 128,
   "min_length": 32,
   "num_beams": 8,
   "pad_token_id": 0,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d52fa58f93ea63358f4c5e4375da7fc9a003c94ec3fff3268598c595beda54a9
+oid sha256:893850642676d5856416dc92a31e8a49a9aaadc0192973e80e125400140ec5ac
 size 2283795562

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:217b8d6f889215d3c17ad9dd550280fc4931d33e68f4a384ec2ed18700d96fe1
+oid sha256:ee685293c42685d6db0780991a03017d68f1795874ce7bed93b559d13a49b3eb
 size 4472