Upload TFT5ForConditionalGeneration

README.md

@@ -1,31 +1,19 @@
 ---
-license: apache-2.0
-base_model: t5-small
 tags:
-- generated_from_trainer
-metrics:
-- rouge
+- generated_from_keras_callback
 model-index:
 - name: t5-small-MedicoSummarizer
   results: []
-language:
-- en
 ---
 
-<!-- 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. -->
+<!-- This model card has been generated automatically according to the information Keras had access to. You should
+probably proofread and complete it, then remove this comment. -->
 
 # t5-small-MedicoSummarizer
 
-This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on processed 25,000 PubMed articles.
-It achieves the following results on the evaluation set of 1000 articles for every epoch :
-- Epochs: 10
-- Loss: 2.8533
-- Rouge1: 0.3234
-- Rouge2: 0.0787
-- Rougel: 0.1967
-- Rougelsum: 0.1965
-- Gen Len: 123.98
+This model was trained from scratch on an unknown dataset.
+It achieves the following results on the evaluation set:
+
 
 ## Model description
 
@@ -41,39 +29,19 @@ More information needed
 
 ## Training procedure
 
-The inference engine doesn't do justice to its operation as the inference engine API doesn't work good for trainer checkpoints as the context limit is low in default for T5 which you can change while using it on backend of your application ! So, you should rather load it on the pipeline and just try it !
-
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 2e-05
-- train_batch_size: 16
-- eval_batch_size: 16
-- seed: 42
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- num_epochs: 10
-- mixed_precision_training: Native AMP
+- optimizer: None
+- training_precision: float32
 
 ### Training results
 
-| Training Loss | Epoch | Step  | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
-|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:---------:|:-------:|
-| 3.2353        | 1.0   | 1563  | 2.9967          | 0.3034 | 0.0717 | 0.1837 | 0.1836    | 117.308 |
-| 3.1623        | 2.0   | 3126  | 2.9421          | 0.3178 | 0.0763 | 0.1941 | 0.1941    | 121.529 |
-| 3.1149        | 3.0   | 4689  | 2.9152          | 0.3223 | 0.078  | 0.1964 | 0.1964    | 123.223 |
-| 3.1038        | 4.0   | 6252  | 2.8929          | 0.3245 | 0.0793 | 0.1979 | 0.1978    | 123.491 |
-| 3.0728        | 5.0   | 7815  | 2.8802          | 0.3227 | 0.0777 | 0.1973 | 0.1972    | 123.6   |
-| 3.0592        | 6.0   | 9378  | 2.8714          | 0.3213 | 0.0788 | 0.1966 | 0.1965    | 123.604 |
-| 3.0448        | 7.0   | 10941 | 2.8635          | 0.3211 | 0.0776 | 0.1959 | 0.1957    | 123.632 |
-| 3.0416        | 8.0   | 12504 | 2.8561          | 0.3204 | 0.0777 | 0.1957 | 0.1955    | 123.851 |
-| 3.0324        | 9.0   | 14067 | 2.8548          | 0.3237 | 0.0788 | 0.1965 | 0.1963    | 123.934 |
-| 3.0375        | 10.0  | 15630 | 2.8533          | 0.3234 | 0.0787 | 0.1967 | 0.1965    | 123.98  |
 
 
 ### Framework versions
 
 - Transformers 4.35.2
-- Pytorch 2.1.0+cu121
-- Datasets 2.15.0
-- Tokenizers 0.15.0
+- TensorFlow 2.15.0
+- Datasets 2.16.1
+- Tokenizers 0.15.0

generation_config.json

@@ -1,4 +1,5 @@
 {
+  "_from_model_config": true,
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,

tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7a96117a913213a3de1eb94c8a1f613f60daa1904924329d9824e8949a8a722
+size 373902664