madlag commited on
Commit
c8c970e
1 Parent(s): 146a421

Initial version of 75% sparsity BERT finetuned on SQuAD1.1.

Browse files
.gitignore ADDED
@@ -0,0 +1 @@
 
1
+ tmp
config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "madlag/bert-base-uncased-squad-v1-sparse0.25",
3
+ "ampere_mask_init": "constant",
4
+ "ampere_mask_scale": 0.0,
5
+ "ampere_pruning_method": "disabled",
6
+ "architectures": [
7
+ "BertForQuestionAnswering"
8
+ ],
9
+ "attention_probs_dropout_prob": 0.1,
10
+ "gradient_checkpointing": false,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 768,
14
+ "in_shuffling_group": 4,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 3072,
17
+ "layer_norm_eps": 1e-12,
18
+ "mask_block_cols": 32,
19
+ "mask_block_rows": 32,
20
+ "mask_init": "constant",
21
+ "mask_scale": 0.0,
22
+ "max_position_embeddings": 512,
23
+ "model_type": "bert",
24
+ "num_attention_heads": 12,
25
+ "num_hidden_layers": 12,
26
+ "out_shuffling_group": 4,
27
+ "pad_token_id": 0,
28
+ "pruning_method": "topK",
29
+ "pruning_submethod": "default",
30
+ "shuffling_method": "disabled",
31
+ "type_vocab_size": 2,
32
+ "vocab_size": 30522
33
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63ff5ba16f2fc8bb58b667c901bf6065fcf147abf6790278e3d243e2452bfbf1
3
+ size 435658487
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tf_model.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7143c4119830543678511f3e3c5058d6d9a1993eafd56defbbf261ee40521ac
3
+ size 438204800
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
1
+ {"do_lower_case": true, "do_basic_tokenize": true, "never_split": null, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "name_or_path": "bert-base-uncased"}
training_args.json ADDED
File without changes
vocab.txt ADDED
The diff for this file is too large to render. See raw diff