Hugging Face's logo

EleutherAI
/
gpt-neox-20b

Text Generation
Transformers
PyTorch
Safetensors
English
gpt_neox
causal-lm
Inference Endpoints
text-generation-inference
4 papers
Model card Files Community
25

Upload FlaxGPTNeoXForCausalLM

#24
by heegyu - opened
base: refs/heads/main
←
from: refs/pr/24
Discussion Files changed
+567
-1
Files changed (8) hide show
  1. config.json +7 -1
  2. flax_model-00001-of-00005.msgpack +3 -0
  3. flax_model-00002-of-00005.msgpack +3 -0
  4. flax_model-00003-of-00005.msgpack +3 -0
  5. flax_model-00004-of-00005.msgpack +3 -0
  6. flax_model-00005-of-00005.msgpack +3 -0
  7. flax_model.msgpack.index.json +539 -0
  8. generation_config.json +6 -0
config.json CHANGED
@@ -1,11 +1,15 @@
1
  {
 
2
  "architectures": [
3
  "GPTNeoXForCausalLM"
4
  ],
 
5
  "attention_probs_dropout_prob": 0,
6
  "bos_token_id": 0,
 
7
  "eos_token_id": 0,
8
  "hidden_act": "gelu_fast",
 
9
  "hidden_dropout_prob": 0,
10
  "hidden_size": 6144,
11
  "initializer_range": 0.02,
@@ -15,11 +19,13 @@
15
  "model_type": "gpt_neox",
16
  "num_attention_heads": 64,
17
  "num_hidden_layers": 44,
 
18
  "rotary_emb_base": 10000,
19
  "rotary_pct": 0.25,
20
  "tie_word_embeddings": false,
21
  "torch_dtype": "float16",
22
- "transformers_version": "4.19.0.dev0",
23
  "use_cache": true,
 
24
  "vocab_size": 50432
25
  }
1
  {
2
+ "_name_or_path": "EleutherAI/gpt-neox-20b",
3
  "architectures": [
4
  "GPTNeoXForCausalLM"
5
  ],
6
+ "attention_dropout": 0.0,
7
  "attention_probs_dropout_prob": 0,
8
  "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
  "eos_token_id": 0,
11
  "hidden_act": "gelu_fast",
12
+ "hidden_dropout": 0.0,
13
  "hidden_dropout_prob": 0,
14
  "hidden_size": 6144,
15
  "initializer_range": 0.02,
19
  "model_type": "gpt_neox",
20
  "num_attention_heads": 64,
21
  "num_hidden_layers": 44,
22
+ "rope_scaling": null,
23
  "rotary_emb_base": 10000,
24
  "rotary_pct": 0.25,
25
  "tie_word_embeddings": false,
26
  "torch_dtype": "float16",
27
+ "transformers_version": "4.34.0.dev0",
28
  "use_cache": true,
29
+ "use_parallel_residual": true,
30
  "vocab_size": 50432
31
  }
flax_model-00001-of-00005.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94bb6f2b8d73e5f9968a9002d8408e24e7d3e3042e5603024a59eb8b57d69c4a
3
+ size 9983095596
flax_model-00002-of-00005.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0daa9108c1162dcbc7a1cd7c20aa1cf4fb9a5463ab52c2a8825a39d2858cac05
3
+ size 9967428516
flax_model-00003-of-00005.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ab58bdc5287fbbd376b9c13bf358dd78e632aa01168a020a2618609f3689f4c
3
+ size 9967428516
flax_model-00004-of-00005.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85dd8b6c95faf89cba3c5b58934514f12d6edd872d71e82ec1da7f4c1ae62fb1
3
+ size 9967428516
flax_model-00005-of-00005.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:638873c1c823f2ae9c5ab693e2e6319e0731e8d51342e59fb03794f154a4b2bf
3
+ size 1223774498
flax_model.msgpack.index.json ADDED
@@ -0,0 +1,539 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 41109135360
4
+ },
5
+ "weight_map": {
6
+ "embed_out/kernel": "flax_model-00005-of-00005.msgpack",
7
+ "gpt_neox/embed_in/embedding": "flax_model-00001-of-00005.msgpack",
8
+ "gpt_neox/final_layer_norm/bias": "flax_model-00005-of-00005.msgpack",
9
+ "gpt_neox/final_layer_norm/scale": "flax_model-00005-of-00005.msgpack",
10
+ "gpt_neox/layers/0/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
11
+ "gpt_neox/layers/0/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
12
+ "gpt_neox/layers/0/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
13
+ "gpt_neox/layers/0/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
14
+ "gpt_neox/layers/0/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
15
+ "gpt_neox/layers/0/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
16
+ "gpt_neox/layers/0/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
17
+ "gpt_neox/layers/0/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
18
+ "gpt_neox/layers/0/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
19
+ "gpt_neox/layers/0/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
20
+ "gpt_neox/layers/0/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
21
+ "gpt_neox/layers/0/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
22
+ "gpt_neox/layers/1/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
23
+ "gpt_neox/layers/1/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
24
+ "gpt_neox/layers/1/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
25
+ "gpt_neox/layers/1/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
26
+ "gpt_neox/layers/1/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
27
+ "gpt_neox/layers/1/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
28
+ "gpt_neox/layers/1/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
29
+ "gpt_neox/layers/1/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
30
+ "gpt_neox/layers/1/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
31
+ "gpt_neox/layers/1/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
32
+ "gpt_neox/layers/1/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
33
+ "gpt_neox/layers/1/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
34
+ "gpt_neox/layers/10/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
35
+ "gpt_neox/layers/10/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
36
+ "gpt_neox/layers/10/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
37
+ "gpt_neox/layers/10/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
38
+ "gpt_neox/layers/10/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
39
+ "gpt_neox/layers/10/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
40
+ "gpt_neox/layers/10/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
41
+ "gpt_neox/layers/10/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
42
+ "gpt_neox/layers/10/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
43
+ "gpt_neox/layers/10/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
44
+ "gpt_neox/layers/10/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
45
+ "gpt_neox/layers/10/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
46
+ "gpt_neox/layers/11/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
47
+ "gpt_neox/layers/11/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
48
+ "gpt_neox/layers/11/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
49
+ "gpt_neox/layers/11/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
50
+ "gpt_neox/layers/11/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
51
+ "gpt_neox/layers/11/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
52
+ "gpt_neox/layers/11/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
53
+ "gpt_neox/layers/11/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
54
+ "gpt_neox/layers/11/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
55
+ "gpt_neox/layers/11/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
56
+ "gpt_neox/layers/11/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
57
+ "gpt_neox/layers/11/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
58
+ "gpt_neox/layers/12/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
59
+ "gpt_neox/layers/12/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
60
+ "gpt_neox/layers/12/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
61
+ "gpt_neox/layers/12/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
62
+ "gpt_neox/layers/12/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
63
+ "gpt_neox/layers/12/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
64
+ "gpt_neox/layers/12/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
65
+ "gpt_neox/layers/12/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
66
+ "gpt_neox/layers/12/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
67
+ "gpt_neox/layers/12/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
68
+ "gpt_neox/layers/12/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
69
+ "gpt_neox/layers/12/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
70
+ "gpt_neox/layers/13/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
71
+ "gpt_neox/layers/13/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
72
+ "gpt_neox/layers/13/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
73
+ "gpt_neox/layers/13/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
74
+ "gpt_neox/layers/13/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
75
+ "gpt_neox/layers/13/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
76
+ "gpt_neox/layers/13/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
77
+ "gpt_neox/layers/13/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
78
+ "gpt_neox/layers/13/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
79
+ "gpt_neox/layers/13/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
80
+ "gpt_neox/layers/13/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
81
+ "gpt_neox/layers/13/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
82
+ "gpt_neox/layers/14/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
83
+ "gpt_neox/layers/14/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
84
+ "gpt_neox/layers/14/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
85
+ "gpt_neox/layers/14/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
86
+ "gpt_neox/layers/14/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
87
+ "gpt_neox/layers/14/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
88
+ "gpt_neox/layers/14/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
89
+ "gpt_neox/layers/14/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
90
+ "gpt_neox/layers/14/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
91
+ "gpt_neox/layers/14/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
92
+ "gpt_neox/layers/14/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
93
+ "gpt_neox/layers/14/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
94
+ "gpt_neox/layers/15/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
95
+ "gpt_neox/layers/15/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
96
+ "gpt_neox/layers/15/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
97
+ "gpt_neox/layers/15/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
98
+ "gpt_neox/layers/15/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
99
+ "gpt_neox/layers/15/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
100
+ "gpt_neox/layers/15/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
101
+ "gpt_neox/layers/15/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
102
+ "gpt_neox/layers/15/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
103
+ "gpt_neox/layers/15/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
104
+ "gpt_neox/layers/15/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
105
+ "gpt_neox/layers/15/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
106
+ "gpt_neox/layers/16/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
107
+ "gpt_neox/layers/16/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
108
+ "gpt_neox/layers/16/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
109
+ "gpt_neox/layers/16/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
110
+ "gpt_neox/layers/16/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
111
+ "gpt_neox/layers/16/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
112
+ "gpt_neox/layers/16/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
113
+ "gpt_neox/layers/16/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
114
+ "gpt_neox/layers/16/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
115
+ "gpt_neox/layers/16/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
116
+ "gpt_neox/layers/16/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
117
+ "gpt_neox/layers/16/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
118
+ "gpt_neox/layers/17/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
119
+ "gpt_neox/layers/17/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
120
+ "gpt_neox/layers/17/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
121
+ "gpt_neox/layers/17/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
122
+ "gpt_neox/layers/17/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
123
+ "gpt_neox/layers/17/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
124
+ "gpt_neox/layers/17/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
125
+ "gpt_neox/layers/17/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
126
+ "gpt_neox/layers/17/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
127
+ "gpt_neox/layers/17/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
128
+ "gpt_neox/layers/17/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
129
+ "gpt_neox/layers/17/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
130
+ "gpt_neox/layers/18/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
131
+ "gpt_neox/layers/18/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
132
+ "gpt_neox/layers/18/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
133
+ "gpt_neox/layers/18/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
134
+ "gpt_neox/layers/18/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
135
+ "gpt_neox/layers/18/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
136
+ "gpt_neox/layers/18/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
137
+ "gpt_neox/layers/18/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
138
+ "gpt_neox/layers/18/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
139
+ "gpt_neox/layers/18/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
140
+ "gpt_neox/layers/18/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
141
+ "gpt_neox/layers/18/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
142
+ "gpt_neox/layers/19/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
143
+ "gpt_neox/layers/19/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
144
+ "gpt_neox/layers/19/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
145
+ "gpt_neox/layers/19/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
146
+ "gpt_neox/layers/19/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
147
+ "gpt_neox/layers/19/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
148
+ "gpt_neox/layers/19/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
149
+ "gpt_neox/layers/19/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
150
+ "gpt_neox/layers/19/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
151
+ "gpt_neox/layers/19/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
152
+ "gpt_neox/layers/19/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
153
+ "gpt_neox/layers/19/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
154
+ "gpt_neox/layers/2/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
155
+ "gpt_neox/layers/2/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
156
+ "gpt_neox/layers/2/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
157
+ "gpt_neox/layers/2/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
158
+ "gpt_neox/layers/2/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
159
+ "gpt_neox/layers/2/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
160
+ "gpt_neox/layers/2/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
161
+ "gpt_neox/layers/2/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
162
+ "gpt_neox/layers/2/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
163
+ "gpt_neox/layers/2/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
164
+ "gpt_neox/layers/2/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
165
+ "gpt_neox/layers/2/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
166
+ "gpt_neox/layers/20/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
167
+ "gpt_neox/layers/20/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
168
+ "gpt_neox/layers/20/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
169
+ "gpt_neox/layers/20/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
170
+ "gpt_neox/layers/20/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
171
+ "gpt_neox/layers/20/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
172
+ "gpt_neox/layers/20/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00005.msgpack",
173
+ "gpt_neox/layers/20/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00005.msgpack",
174
+ "gpt_neox/layers/20/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00005.msgpack",
175
+ "gpt_neox/layers/20/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00005.msgpack",
176
+ "gpt_neox/layers/20/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
177
+ "gpt_neox/layers/20/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
178
+ "gpt_neox/layers/21/attention/dense/bias": "flax_model-00002-of-00005.msgpack",
179
+ "gpt_neox/layers/21/attention/dense/kernel": "flax_model-00002-of-00005.msgpack",
180
+ "gpt_neox/layers/21/attention/query_key_value/bias": "flax_model-00002-of-00005.msgpack",
181
+ "gpt_neox/layers/21/attention/query_key_value/kernel": "flax_model-00002-of-00005.msgpack",
182
+ "gpt_neox/layers/21/input_layernorm/bias": "flax_model-00002-of-00005.msgpack",
183
+ "gpt_neox/layers/21/input_layernorm/scale": "flax_model-00002-of-00005.msgpack",
184
+ "gpt_neox/layers/21/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
185
+ "gpt_neox/layers/21/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
186
+ "gpt_neox/layers/21/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
187
+ "gpt_neox/layers/21/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
188
+ "gpt_neox/layers/21/post_attention_layernorm/bias": "flax_model-00002-of-00005.msgpack",
189
+ "gpt_neox/layers/21/post_attention_layernorm/scale": "flax_model-00002-of-00005.msgpack",
190
+ "gpt_neox/layers/22/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
191
+ "gpt_neox/layers/22/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
192
+ "gpt_neox/layers/22/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
193
+ "gpt_neox/layers/22/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
194
+ "gpt_neox/layers/22/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
195
+ "gpt_neox/layers/22/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
196
+ "gpt_neox/layers/22/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
197
+ "gpt_neox/layers/22/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
198
+ "gpt_neox/layers/22/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
199
+ "gpt_neox/layers/22/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
200
+ "gpt_neox/layers/22/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
201
+ "gpt_neox/layers/22/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
202
+ "gpt_neox/layers/23/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
203
+ "gpt_neox/layers/23/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
204
+ "gpt_neox/layers/23/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
205
+ "gpt_neox/layers/23/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
206
+ "gpt_neox/layers/23/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
207
+ "gpt_neox/layers/23/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
208
+ "gpt_neox/layers/23/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
209
+ "gpt_neox/layers/23/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
210
+ "gpt_neox/layers/23/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
211
+ "gpt_neox/layers/23/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
212
+ "gpt_neox/layers/23/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
213
+ "gpt_neox/layers/23/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
214
+ "gpt_neox/layers/24/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
215
+ "gpt_neox/layers/24/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
216
+ "gpt_neox/layers/24/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
217
+ "gpt_neox/layers/24/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
218
+ "gpt_neox/layers/24/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
219
+ "gpt_neox/layers/24/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
220
+ "gpt_neox/layers/24/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
221
+ "gpt_neox/layers/24/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
222
+ "gpt_neox/layers/24/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
223
+ "gpt_neox/layers/24/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
224
+ "gpt_neox/layers/24/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
225
+ "gpt_neox/layers/24/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
226
+ "gpt_neox/layers/25/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
227
+ "gpt_neox/layers/25/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
228
+ "gpt_neox/layers/25/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
229
+ "gpt_neox/layers/25/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
230
+ "gpt_neox/layers/25/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
231
+ "gpt_neox/layers/25/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
232
+ "gpt_neox/layers/25/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
233
+ "gpt_neox/layers/25/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
234
+ "gpt_neox/layers/25/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
235
+ "gpt_neox/layers/25/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
236
+ "gpt_neox/layers/25/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
237
+ "gpt_neox/layers/25/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
238
+ "gpt_neox/layers/26/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
239
+ "gpt_neox/layers/26/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
240
+ "gpt_neox/layers/26/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
241
+ "gpt_neox/layers/26/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
242
+ "gpt_neox/layers/26/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
243
+ "gpt_neox/layers/26/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
244
+ "gpt_neox/layers/26/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
245
+ "gpt_neox/layers/26/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
246
+ "gpt_neox/layers/26/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
247
+ "gpt_neox/layers/26/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
248
+ "gpt_neox/layers/26/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
249
+ "gpt_neox/layers/26/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
250
+ "gpt_neox/layers/27/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
251
+ "gpt_neox/layers/27/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
252
+ "gpt_neox/layers/27/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
253
+ "gpt_neox/layers/27/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
254
+ "gpt_neox/layers/27/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
255
+ "gpt_neox/layers/27/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
256
+ "gpt_neox/layers/27/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
257
+ "gpt_neox/layers/27/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
258
+ "gpt_neox/layers/27/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
259
+ "gpt_neox/layers/27/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
260
+ "gpt_neox/layers/27/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
261
+ "gpt_neox/layers/27/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
262
+ "gpt_neox/layers/28/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
263
+ "gpt_neox/layers/28/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
264
+ "gpt_neox/layers/28/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
265
+ "gpt_neox/layers/28/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
266
+ "gpt_neox/layers/28/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
267
+ "gpt_neox/layers/28/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
268
+ "gpt_neox/layers/28/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
269
+ "gpt_neox/layers/28/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
270
+ "gpt_neox/layers/28/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
271
+ "gpt_neox/layers/28/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
272
+ "gpt_neox/layers/28/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
273
+ "gpt_neox/layers/28/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
274
+ "gpt_neox/layers/29/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
275
+ "gpt_neox/layers/29/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
276
+ "gpt_neox/layers/29/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
277
+ "gpt_neox/layers/29/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
278
+ "gpt_neox/layers/29/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
279
+ "gpt_neox/layers/29/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
280
+ "gpt_neox/layers/29/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
281
+ "gpt_neox/layers/29/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
282
+ "gpt_neox/layers/29/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
283
+ "gpt_neox/layers/29/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
284
+ "gpt_neox/layers/29/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
285
+ "gpt_neox/layers/29/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
286
+ "gpt_neox/layers/3/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
287
+ "gpt_neox/layers/3/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
288
+ "gpt_neox/layers/3/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
289
+ "gpt_neox/layers/3/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
290
+ "gpt_neox/layers/3/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
291
+ "gpt_neox/layers/3/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
292
+ "gpt_neox/layers/3/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
293
+ "gpt_neox/layers/3/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
294
+ "gpt_neox/layers/3/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
295
+ "gpt_neox/layers/3/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
296
+ "gpt_neox/layers/3/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
297
+ "gpt_neox/layers/3/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
298
+ "gpt_neox/layers/30/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
299
+ "gpt_neox/layers/30/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
300
+ "gpt_neox/layers/30/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
301
+ "gpt_neox/layers/30/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
302
+ "gpt_neox/layers/30/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
303
+ "gpt_neox/layers/30/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
304
+ "gpt_neox/layers/30/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
305
+ "gpt_neox/layers/30/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
306
+ "gpt_neox/layers/30/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
307
+ "gpt_neox/layers/30/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
308
+ "gpt_neox/layers/30/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
309
+ "gpt_neox/layers/30/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
310
+ "gpt_neox/layers/31/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
311
+ "gpt_neox/layers/31/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
312
+ "gpt_neox/layers/31/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
313
+ "gpt_neox/layers/31/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
314
+ "gpt_neox/layers/31/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
315
+ "gpt_neox/layers/31/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
316
+ "gpt_neox/layers/31/mlp/dense_4h_to_h/bias": "flax_model-00003-of-00005.msgpack",
317
+ "gpt_neox/layers/31/mlp/dense_4h_to_h/kernel": "flax_model-00003-of-00005.msgpack",
318
+ "gpt_neox/layers/31/mlp/dense_h_to_4h/bias": "flax_model-00003-of-00005.msgpack",
319
+ "gpt_neox/layers/31/mlp/dense_h_to_4h/kernel": "flax_model-00003-of-00005.msgpack",
320
+ "gpt_neox/layers/31/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
321
+ "gpt_neox/layers/31/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
322
+ "gpt_neox/layers/32/attention/dense/bias": "flax_model-00003-of-00005.msgpack",
323
+ "gpt_neox/layers/32/attention/dense/kernel": "flax_model-00003-of-00005.msgpack",
324
+ "gpt_neox/layers/32/attention/query_key_value/bias": "flax_model-00003-of-00005.msgpack",
325
+ "gpt_neox/layers/32/attention/query_key_value/kernel": "flax_model-00003-of-00005.msgpack",
326
+ "gpt_neox/layers/32/input_layernorm/bias": "flax_model-00003-of-00005.msgpack",
327
+ "gpt_neox/layers/32/input_layernorm/scale": "flax_model-00003-of-00005.msgpack",
328
+ "gpt_neox/layers/32/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
329
+ "gpt_neox/layers/32/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
330
+ "gpt_neox/layers/32/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
331
+ "gpt_neox/layers/32/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
332
+ "gpt_neox/layers/32/post_attention_layernorm/bias": "flax_model-00003-of-00005.msgpack",
333
+ "gpt_neox/layers/32/post_attention_layernorm/scale": "flax_model-00003-of-00005.msgpack",
334
+ "gpt_neox/layers/33/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
335
+ "gpt_neox/layers/33/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
336
+ "gpt_neox/layers/33/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
337
+ "gpt_neox/layers/33/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
338
+ "gpt_neox/layers/33/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
339
+ "gpt_neox/layers/33/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
340
+ "gpt_neox/layers/33/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
341
+ "gpt_neox/layers/33/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
342
+ "gpt_neox/layers/33/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
343
+ "gpt_neox/layers/33/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
344
+ "gpt_neox/layers/33/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
345
+ "gpt_neox/layers/33/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
346
+ "gpt_neox/layers/34/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
347
+ "gpt_neox/layers/34/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
348
+ "gpt_neox/layers/34/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
349
+ "gpt_neox/layers/34/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
350
+ "gpt_neox/layers/34/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
351
+ "gpt_neox/layers/34/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
352
+ "gpt_neox/layers/34/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
353
+ "gpt_neox/layers/34/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
354
+ "gpt_neox/layers/34/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
355
+ "gpt_neox/layers/34/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
356
+ "gpt_neox/layers/34/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
357
+ "gpt_neox/layers/34/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
358
+ "gpt_neox/layers/35/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
359
+ "gpt_neox/layers/35/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
360
+ "gpt_neox/layers/35/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
361
+ "gpt_neox/layers/35/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
362
+ "gpt_neox/layers/35/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
363
+ "gpt_neox/layers/35/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
364
+ "gpt_neox/layers/35/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
365
+ "gpt_neox/layers/35/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
366
+ "gpt_neox/layers/35/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
367
+ "gpt_neox/layers/35/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
368
+ "gpt_neox/layers/35/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
369
+ "gpt_neox/layers/35/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
370
+ "gpt_neox/layers/36/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
371
+ "gpt_neox/layers/36/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
372
+ "gpt_neox/layers/36/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
373
+ "gpt_neox/layers/36/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
374
+ "gpt_neox/layers/36/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
375
+ "gpt_neox/layers/36/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
376
+ "gpt_neox/layers/36/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
377
+ "gpt_neox/layers/36/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
378
+ "gpt_neox/layers/36/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
379
+ "gpt_neox/layers/36/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
380
+ "gpt_neox/layers/36/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
381
+ "gpt_neox/layers/36/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
382
+ "gpt_neox/layers/37/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
383
+ "gpt_neox/layers/37/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
384
+ "gpt_neox/layers/37/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
385
+ "gpt_neox/layers/37/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
386
+ "gpt_neox/layers/37/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
387
+ "gpt_neox/layers/37/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
388
+ "gpt_neox/layers/37/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
389
+ "gpt_neox/layers/37/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
390
+ "gpt_neox/layers/37/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
391
+ "gpt_neox/layers/37/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
392
+ "gpt_neox/layers/37/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
393
+ "gpt_neox/layers/37/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
394
+ "gpt_neox/layers/38/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
395
+ "gpt_neox/layers/38/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
396
+ "gpt_neox/layers/38/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
397
+ "gpt_neox/layers/38/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
398
+ "gpt_neox/layers/38/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
399
+ "gpt_neox/layers/38/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
400
+ "gpt_neox/layers/38/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
401
+ "gpt_neox/layers/38/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
402
+ "gpt_neox/layers/38/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
403
+ "gpt_neox/layers/38/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
404
+ "gpt_neox/layers/38/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
405
+ "gpt_neox/layers/38/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
406
+ "gpt_neox/layers/39/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
407
+ "gpt_neox/layers/39/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
408
+ "gpt_neox/layers/39/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
409
+ "gpt_neox/layers/39/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
410
+ "gpt_neox/layers/39/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
411
+ "gpt_neox/layers/39/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
412
+ "gpt_neox/layers/39/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
413
+ "gpt_neox/layers/39/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
414
+ "gpt_neox/layers/39/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
415
+ "gpt_neox/layers/39/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
416
+ "gpt_neox/layers/39/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
417
+ "gpt_neox/layers/39/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
418
+ "gpt_neox/layers/4/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
419
+ "gpt_neox/layers/4/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
420
+ "gpt_neox/layers/4/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
421
+ "gpt_neox/layers/4/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
422
+ "gpt_neox/layers/4/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
423
+ "gpt_neox/layers/4/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
424
+ "gpt_neox/layers/4/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
425
+ "gpt_neox/layers/4/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
426
+ "gpt_neox/layers/4/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
427
+ "gpt_neox/layers/4/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
428
+ "gpt_neox/layers/4/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
429
+ "gpt_neox/layers/4/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
430
+ "gpt_neox/layers/40/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
431
+ "gpt_neox/layers/40/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
432
+ "gpt_neox/layers/40/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
433
+ "gpt_neox/layers/40/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
434
+ "gpt_neox/layers/40/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
435
+ "gpt_neox/layers/40/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
436
+ "gpt_neox/layers/40/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
437
+ "gpt_neox/layers/40/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
438
+ "gpt_neox/layers/40/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
439
+ "gpt_neox/layers/40/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
440
+ "gpt_neox/layers/40/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
441
+ "gpt_neox/layers/40/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
442
+ "gpt_neox/layers/41/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
443
+ "gpt_neox/layers/41/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
444
+ "gpt_neox/layers/41/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
445
+ "gpt_neox/layers/41/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
446
+ "gpt_neox/layers/41/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
447
+ "gpt_neox/layers/41/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
448
+ "gpt_neox/layers/41/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
449
+ "gpt_neox/layers/41/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
450
+ "gpt_neox/layers/41/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
451
+ "gpt_neox/layers/41/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
452
+ "gpt_neox/layers/41/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
453
+ "gpt_neox/layers/41/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
454
+ "gpt_neox/layers/42/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
455
+ "gpt_neox/layers/42/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
456
+ "gpt_neox/layers/42/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
457
+ "gpt_neox/layers/42/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
458
+ "gpt_neox/layers/42/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
459
+ "gpt_neox/layers/42/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
460
+ "gpt_neox/layers/42/mlp/dense_4h_to_h/bias": "flax_model-00004-of-00005.msgpack",
461
+ "gpt_neox/layers/42/mlp/dense_4h_to_h/kernel": "flax_model-00004-of-00005.msgpack",
462
+ "gpt_neox/layers/42/mlp/dense_h_to_4h/bias": "flax_model-00004-of-00005.msgpack",
463
+ "gpt_neox/layers/42/mlp/dense_h_to_4h/kernel": "flax_model-00004-of-00005.msgpack",
464
+ "gpt_neox/layers/42/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
465
+ "gpt_neox/layers/42/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
466
+ "gpt_neox/layers/43/attention/dense/bias": "flax_model-00004-of-00005.msgpack",
467
+ "gpt_neox/layers/43/attention/dense/kernel": "flax_model-00004-of-00005.msgpack",
468
+ "gpt_neox/layers/43/attention/query_key_value/bias": "flax_model-00004-of-00005.msgpack",
469
+ "gpt_neox/layers/43/attention/query_key_value/kernel": "flax_model-00004-of-00005.msgpack",
470
+ "gpt_neox/layers/43/input_layernorm/bias": "flax_model-00004-of-00005.msgpack",
471
+ "gpt_neox/layers/43/input_layernorm/scale": "flax_model-00004-of-00005.msgpack",
472
+ "gpt_neox/layers/43/mlp/dense_4h_to_h/bias": "flax_model-00005-of-00005.msgpack",
473
+ "gpt_neox/layers/43/mlp/dense_4h_to_h/kernel": "flax_model-00005-of-00005.msgpack",
474
+ "gpt_neox/layers/43/mlp/dense_h_to_4h/bias": "flax_model-00005-of-00005.msgpack",
475
+ "gpt_neox/layers/43/mlp/dense_h_to_4h/kernel": "flax_model-00005-of-00005.msgpack",
476
+ "gpt_neox/layers/43/post_attention_layernorm/bias": "flax_model-00004-of-00005.msgpack",
477
+ "gpt_neox/layers/43/post_attention_layernorm/scale": "flax_model-00004-of-00005.msgpack",
478
+ "gpt_neox/layers/5/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
479
+ "gpt_neox/layers/5/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
480
+ "gpt_neox/layers/5/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
481
+ "gpt_neox/layers/5/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
482
+ "gpt_neox/layers/5/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
483
+ "gpt_neox/layers/5/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
484
+ "gpt_neox/layers/5/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
485
+ "gpt_neox/layers/5/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
486
+ "gpt_neox/layers/5/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
487
+ "gpt_neox/layers/5/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
488
+ "gpt_neox/layers/5/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
489
+ "gpt_neox/layers/5/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
490
+ "gpt_neox/layers/6/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
491
+ "gpt_neox/layers/6/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
492
+ "gpt_neox/layers/6/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
493
+ "gpt_neox/layers/6/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
494
+ "gpt_neox/layers/6/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
495
+ "gpt_neox/layers/6/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
496
+ "gpt_neox/layers/6/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
497
+ "gpt_neox/layers/6/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
498
+ "gpt_neox/layers/6/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
499
+ "gpt_neox/layers/6/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
500
+ "gpt_neox/layers/6/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
501
+ "gpt_neox/layers/6/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
502
+ "gpt_neox/layers/7/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
503
+ "gpt_neox/layers/7/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
504
+ "gpt_neox/layers/7/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
505
+ "gpt_neox/layers/7/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
506
+ "gpt_neox/layers/7/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
507
+ "gpt_neox/layers/7/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
508
+ "gpt_neox/layers/7/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
509
+ "gpt_neox/layers/7/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
510
+ "gpt_neox/layers/7/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
511
+ "gpt_neox/layers/7/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
512
+ "gpt_neox/layers/7/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
513
+ "gpt_neox/layers/7/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
514
+ "gpt_neox/layers/8/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
515
+ "gpt_neox/layers/8/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
516
+ "gpt_neox/layers/8/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
517
+ "gpt_neox/layers/8/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
518
+ "gpt_neox/layers/8/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
519
+ "gpt_neox/layers/8/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
520
+ "gpt_neox/layers/8/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
521
+ "gpt_neox/layers/8/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
522
+ "gpt_neox/layers/8/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
523
+ "gpt_neox/layers/8/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
524
+ "gpt_neox/layers/8/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
525
+ "gpt_neox/layers/8/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack",
526
+ "gpt_neox/layers/9/attention/dense/bias": "flax_model-00001-of-00005.msgpack",
527
+ "gpt_neox/layers/9/attention/dense/kernel": "flax_model-00001-of-00005.msgpack",
528
+ "gpt_neox/layers/9/attention/query_key_value/bias": "flax_model-00001-of-00005.msgpack",
529
+ "gpt_neox/layers/9/attention/query_key_value/kernel": "flax_model-00001-of-00005.msgpack",
530
+ "gpt_neox/layers/9/input_layernorm/bias": "flax_model-00001-of-00005.msgpack",
531
+ "gpt_neox/layers/9/input_layernorm/scale": "flax_model-00001-of-00005.msgpack",
532
+ "gpt_neox/layers/9/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00005.msgpack",
533
+ "gpt_neox/layers/9/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00005.msgpack",
534
+ "gpt_neox/layers/9/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00005.msgpack",
535
+ "gpt_neox/layers/9/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00005.msgpack",
536
+ "gpt_neox/layers/9/post_attention_layernorm/bias": "flax_model-00001-of-00005.msgpack",
537
+ "gpt_neox/layers/9/post_attention_layernorm/scale": "flax_model-00001-of-00005.msgpack"
538
+ }
539
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.34.0.dev0"
6
+ }