Hugging Face's logo

Vivek
/
gptneo_piqa

Transformers
JAX
TensorBoard
gpt_neo
Inference Endpoints
Model card Files Metrics Community
Vivek commited on
Commit
16aca1d
·
1 Parent(s): 7b062dd

Piqa:Saving weights of epoch 3 at step 5

Browse files
Files changed (7) hide show
  1. config.json +72 -0
  2. flax_model.msgpack +3 -0
  3. results_tensorboard/events.out.tfevents.1627576737.t1v-n-cb40e504-w-0.15810.3.v2 +3 -0
  4. results_tensorboard/events.out.tfevents.1627576907.t1v-n-cb40e504-w-0.17188.3.v2 +3 -0
  5. results_tensorboard/events.out.tfevents.1627577001.t1v-n-cb40e504-w-0.18487.3.v2 +3 -0
  6. src/__pycache__/model_file.cpython-38.pyc +0 -0
  7. src/gptneo_piqa.py +4 -4
config.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPTNeoForMultipleChoice"
5
+ ],
6
+ "attention_dropout": 0,
7
+ "attention_layers": [
8
+ "global",
9
+ "local",
10
+ "global",
11
+ "local",
12
+ "global",
13
+ "local",
14
+ "global",
15
+ "local",
16
+ "global",
17
+ "local",
18
+ "global",
19
+ "local",
20
+ "global",
21
+ "local",
22
+ "global",
23
+ "local",
24
+ "global",
25
+ "local",
26
+ "global",
27
+ "local",
28
+ "global",
29
+ "local",
30
+ "global",
31
+ "local"
32
+ ],
33
+ "attention_types": [
34
+ [
35
+ [
36
+ "global",
37
+ "local"
38
+ ],
39
+ 12
40
+ ]
41
+ ],
42
+ "bos_token_id": 50256,
43
+ "embed_dropout": 0,
44
+ "eos_token_id": 50256,
45
+ "gradient_checkpointing": false,
46
+ "hidden_size": 2048,
47
+ "initializer_range": 0.02,
48
+ "intermediate_size": null,
49
+ "layer_norm_epsilon": 1e-05,
50
+ "max_position_embeddings": 2048,
51
+ "model_type": "gpt_neo",
52
+ "num_heads": 16,
53
+ "num_layers": 24,
54
+ "resid_dropout": 0,
55
+ "summary_activation": null,
56
+ "summary_first_dropout": 0.1,
57
+ "summary_proj_to_labels": true,
58
+ "summary_type": "cls_index",
59
+ "summary_use_proj": true,
60
+ "task_specific_params": {
61
+ "text-generation": {
62
+ "do_sample": true,
63
+ "max_length": 50,
64
+ "temperature": 0.9
65
+ }
66
+ },
67
+ "tokenizer_class": "GPT2Tokenizer",
68
+ "transformers_version": "4.10.0.dev0",
69
+ "use_cache": true,
70
+ "vocab_size": 50257,
71
+ "window_size": 256
72
+ }
flax_model.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:926f94ac5ef99648eb211da536ce6f12659041259804b0bd8b93b6cde9933c57
3
+ size 5262347427
results_tensorboard/events.out.tfevents.1627576737.t1v-n-cb40e504-w-0.15810.3.v2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a5fe34e7ef880457412f867be4ccb7f04e67e81d70c9be50d12d7a482216535
3
+ size 40
results_tensorboard/events.out.tfevents.1627576907.t1v-n-cb40e504-w-0.17188.3.v2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0542593e98b4630e7d5f56f3950b674d9d4caf7a9bd56edbfa8a027cdebe2266
3
+ size 40
results_tensorboard/events.out.tfevents.1627577001.t1v-n-cb40e504-w-0.18487.3.v2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f631830ee7bccb688c06d53d54159d5cad05dca99449f0d9de7951414532e8f
3
+ size 1720
src/__pycache__/model_file.cpython-38.pyc ADDED
Binary file (9.24 kB). View file
 
src/gptneo_piqa.py CHANGED
@@ -45,9 +45,9 @@ train_dataset=dataset['train'].map(preprocess)
45
  validation_dataset=dataset['validation'].map(preprocess)
46
  test_dataset=dataset['test'].map(preprocess)
47
 
48
- len_train_dataset=16113
49
- len_validation_dataset=1838
50
- len_test_dataset=3084
51
 
52
  train_dataset=train_dataset.select(range(len_train_dataset))
53
  test_dataset=test_dataset.select(range(len_test_dataset))
@@ -56,7 +56,7 @@ validation_dataset=validation_dataset.select(range(len_validation_dataset))
56
  remove_col=train_dataset.column_names
57
 
58
  def tokenize(examples):
59
- tokenized_examples=tokenizer(examples['first_sentence'],examples['second_sentence'],padding='max_length',truncation=True,max_length=256,return_tensors='jax')
60
  tokenized_examples['labels']=int(examples['label'])
61
  return tokenized_examples
62
 
 
45
  validation_dataset=dataset['validation'].map(preprocess)
46
  test_dataset=dataset['test'].map(preprocess)
47
 
48
+ len_train_dataset=100
49
+ len_validation_dataset=100
50
+ len_test_dataset=100
51
 
52
  train_dataset=train_dataset.select(range(len_train_dataset))
53
  test_dataset=test_dataset.select(range(len_test_dataset))
 
56
  remove_col=train_dataset.column_names
57
 
58
  def tokenize(examples):
59
+ tokenized_examples=tokenizer(examples['first_sentence'],examples['second_sentence'],padding='max_length',truncation=True,max_length=512,return_tensors='jax')
60
  tokenized_examples['labels']=int(examples['label'])
61
  return tokenized_examples
62