knvarad commited on
Commit
4c2b687
1 Parent(s): 7c70c09

Upload TFT5ForConditionalGeneration

Browse files
Files changed (4) hide show
  1. README.md +8 -8
  2. config.json +1 -1
  3. generation_config.json +13 -0
  4. tf_model.h5 +2 -2
README.md CHANGED
@@ -14,9 +14,9 @@ probably proofread and complete it, then remove this comment. -->
14
 
15
  This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
- - Train Loss: 3.8958
18
- - Validation Loss: 3.5264
19
- - Train Rougel: tf.Tensor(0.13228828, shape=(), dtype=float32)
20
  - Epoch: 0
21
 
22
  ## Model description
@@ -36,19 +36,19 @@ More information needed
36
  ### Training hyperparameters
37
 
38
  The following hyperparameters were used during training:
39
- - optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': False, 'is_legacy_optimizer': False, 'learning_rate': 2e-05, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
40
  - training_precision: float32
41
 
42
  ### Training results
43
 
44
  | Train Loss | Validation Loss | Train Rougel | Epoch |
45
  |:----------:|:---------------:|:----------------------------------------------:|:-----:|
46
- | 3.8958 | 3.5264 | tf.Tensor(0.13228828, shape=(), dtype=float32) | 0 |
47
 
48
 
49
  ### Framework versions
50
 
51
- - Transformers 4.20.0
52
- - TensorFlow 2.13.0
53
- - Datasets 2.14.4
54
  - Tokenizers 0.12.1
 
14
 
15
  This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
+ - Train Loss: 3.8679
18
+ - Validation Loss: 3.5523
19
+ - Train Rougel: tf.Tensor(0.11994212, shape=(), dtype=float32)
20
  - Epoch: 0
21
 
22
  ## Model description
 
36
  ### Training hyperparameters
37
 
38
  The following hyperparameters were used during training:
39
+ - optimizer: {'name': 'Adam', 'learning_rate': 2e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
40
  - training_precision: float32
41
 
42
  ### Training results
43
 
44
  | Train Loss | Validation Loss | Train Rougel | Epoch |
45
  |:----------:|:---------------:|:----------------------------------------------:|:-----:|
46
+ | 3.8679 | 3.5523 | tf.Tensor(0.11994212, shape=(), dtype=float32) | 0 |
47
 
48
 
49
  ### Framework versions
50
 
51
+ - Transformers 4.26.1
52
+ - TensorFlow 2.10.1
53
+ - Datasets 2.13.1
54
  - Tokenizers 0.12.1
config.json CHANGED
@@ -60,7 +60,7 @@
60
  "prefix": "translate English to Romanian: "
61
  }
62
  },
63
- "transformers_version": "4.20.0",
64
  "use_cache": true,
65
  "vocab_size": 32128
66
  }
 
60
  "prefix": "translate English to Romanian: "
61
  }
62
  },
63
+ "transformers_version": "4.26.1",
64
  "use_cache": true,
65
  "vocab_size": 32128
66
  }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "decoder_start_token_id": 0,
4
+ "early_stopping": true,
5
+ "eos_token_id": 1,
6
+ "length_penalty": 2.0,
7
+ "max_length": 200,
8
+ "min_length": 30,
9
+ "no_repeat_ngram_size": 3,
10
+ "num_beams": 4,
11
+ "pad_token_id": 0,
12
+ "transformers_version": "4.26.1"
13
+ }
tf_model.h5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5bc56c3d9f36c7dc846c294bf88c4304fe2b6af29f722f3ba8b635bb41ce70ef
3
- size 242301696
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4e5c73b02d0ffc722ade34ff3f9abd341d9856bbc11caecb4c99d1ea9c5be19c
3
+ size 373902664