Hugging Face's logo

LLM360
/
Crystal

Text Generation
Transformers
PyTorch
English
crystalcoder
llm
code
custom_code
Model card Files Community
2
yukiontheiceberg commited on
Commit
9e73530
•
1 Parent(s): fa1322b

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. config.json +37 -32
  2. generation_config.json +1 -1
  3. model-00001-of-00003.safetensors +3 -0
  4. model-00002-of-00003.safetensors +3 -0
  5. model-00003-of-00003.safetensors +3 -0
  6. model.safetensors.index.json +458 -0
  7. special_tokens_map.json +47 -0
  8. tokenizer_config.json +6 -6
config.json CHANGED
@@ -1,33 +1,38 @@
1
  {
2
- "attn_pdrop": 0.0,
3
- "scale_attn_weights": true,
4
- "resid_pdrop": 0.0,
5
- "mup_embeddings_scale": 14.6,
6
- "n_inner": 10922,
7
- "n_embd": 4096,
8
- "layer_norm_epsilon": 1e-05,
9
- "n_positions": 2048,
10
- "activation_function": "swiglu",
11
- "n_head": 32,
12
- "n_layer": 32,
13
- "mup_output_alpha": 2.22,
14
- "mup_width_scale": 0.0625,
15
- "position_embedding_type": "rotary",
16
- "rotary_dim": 32,
17
- "mup_scale_qk_dot_by_d": true,
18
- "tie_word_embeddings": true,
19
- "vocab_size": 32032,
20
- "embd_pdrop": 0.0,
21
- "model_type": "crystalcoder",
22
- "use_cache": true,
23
- "bos_token_id": 1,
24
- "eos_token_id": 2,
25
- "auto_map": {
26
- "AutoConfig": "configuration_crystalcoder.CrystalCoderConfig",
27
- "AutoModel": "modeling_crystalcoder.CrystalCoderModel",
28
- "AutoModelForCausalLM": "modeling_crystalcoder.CrystalCoderLMHeadModel"
29
- },
30
- "architectures": [
31
- "CrystalCoderLMHeadModel"
32
- ]
33
- }
 
 
 
 
 
 
1
  {
2
+ "_name_or_path": "LLM360/CrystalCoder",
3
+ "activation_function": "swiglu",
4
+ "architectures": [
5
+ "CrystalCoderLMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.0,
8
+ "auto_map": {
9
+ "AutoConfig": "LLM360/CrystalCoder--configuration_crystalcoder.CrystalCoderConfig",
10
+ "AutoModel": "LLM360/CrystalCoder--modeling_crystalcoder.CrystalCoderModel",
11
+ "AutoModelForCausalLM": "LLM360/CrystalCoder--modeling_crystalcoder.CrystalCoderLMHeadModel"
12
+ },
13
+ "bos_token_id": 1,
14
+ "embd_pdrop": 0.0,
15
+ "eos_token_id": 2,
16
+ "initializer_range": 0.02,
17
+ "layer_norm_epsilon": 1e-05,
18
+ "model_type": "crystalcoder",
19
+ "mup_embeddings_scale": 14.6,
20
+ "mup_output_alpha": 2.22,
21
+ "mup_scale_qk_dot_by_d": true,
22
+ "mup_width_scale": 0.0625,
23
+ "n_embd": 4096,
24
+ "n_head": 32,
25
+ "n_inner": 10922,
26
+ "n_layer": 32,
27
+ "n_positions": 2048,
28
+ "position_embedding_type": "rotary",
29
+ "reorder_and_upcast_attn": false,
30
+ "resid_pdrop": 0.0,
31
+ "rotary_dim": 32,
32
+ "scale_attn_by_inverse_layer_idx": false,
33
+ "scale_attn_weights": true,
34
+ "torch_dtype": "bfloat16",
35
+ "transformers_version": "4.35.2",
36
+ "use_cache": true,
37
+ "vocab_size": 32032
38
+ }
generation_config.json CHANGED
@@ -2,5 +2,5 @@
2
  "_from_model_config": true,
3
  "bos_token_id": 1,
4
  "eos_token_id": 2,
5
- "transformers_version": "4.33.0.dev0"
6
  }
 
2
  "_from_model_config": true,
3
  "bos_token_id": 1,
4
  "eos_token_id": 2,
5
+ "transformers_version": "4.35.2"
6
  }
model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4e5c6dc69f4715ab32c5990c8700f96a1f304b6d3102fdb4954d90d4ec80a3eb
3
+ size 4916498172
model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31921aef135ba5f555d6543c280315ebf47073aa1e9abfaab4a43e075221270a
3
+ size 4922563684
model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02f6a402eb41a0073b6dd46caa4e8603efc0c51c8e40e8d54593802cd470bc3e
3
+ size 3311543728
model.safetensors.index.json ADDED
@@ -0,0 +1,458 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 13150557440
4
+ },
5
+ "weight_map": {
6
+ "transformer.h.0.attn.c_attn.bias": "model-00001-of-00003.safetensors",
7
+ "transformer.h.0.attn.c_attn.weight": "model-00001-of-00003.safetensors",
8
+ "transformer.h.0.attn.c_proj.bias": "model-00001-of-00003.safetensors",
9
+ "transformer.h.0.attn.c_proj.weight": "model-00001-of-00003.safetensors",
10
+ "transformer.h.0.ln_1.bias": "model-00001-of-00003.safetensors",
11
+ "transformer.h.0.ln_1.weight": "model-00001-of-00003.safetensors",
12
+ "transformer.h.0.ln_2.bias": "model-00001-of-00003.safetensors",
13
+ "transformer.h.0.ln_2.weight": "model-00001-of-00003.safetensors",
14
+ "transformer.h.0.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
15
+ "transformer.h.0.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
16
+ "transformer.h.0.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
17
+ "transformer.h.0.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
18
+ "transformer.h.0.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
19
+ "transformer.h.0.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
20
+ "transformer.h.1.attn.c_attn.bias": "model-00001-of-00003.safetensors",
21
+ "transformer.h.1.attn.c_attn.weight": "model-00001-of-00003.safetensors",
22
+ "transformer.h.1.attn.c_proj.bias": "model-00001-of-00003.safetensors",
23
+ "transformer.h.1.attn.c_proj.weight": "model-00001-of-00003.safetensors",
24
+ "transformer.h.1.ln_1.bias": "model-00001-of-00003.safetensors",
25
+ "transformer.h.1.ln_1.weight": "model-00001-of-00003.safetensors",
26
+ "transformer.h.1.ln_2.bias": "model-00001-of-00003.safetensors",
27
+ "transformer.h.1.ln_2.weight": "model-00001-of-00003.safetensors",
28
+ "transformer.h.1.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
29
+ "transformer.h.1.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
30
+ "transformer.h.1.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
31
+ "transformer.h.1.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
32
+ "transformer.h.1.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
33
+ "transformer.h.1.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
34
+ "transformer.h.10.attn.c_attn.bias": "model-00001-of-00003.safetensors",
35
+ "transformer.h.10.attn.c_attn.weight": "model-00001-of-00003.safetensors",
36
+ "transformer.h.10.attn.c_proj.bias": "model-00001-of-00003.safetensors",
37
+ "transformer.h.10.attn.c_proj.weight": "model-00001-of-00003.safetensors",
38
+ "transformer.h.10.ln_1.bias": "model-00001-of-00003.safetensors",
39
+ "transformer.h.10.ln_1.weight": "model-00001-of-00003.safetensors",
40
+ "transformer.h.10.ln_2.bias": "model-00001-of-00003.safetensors",
41
+ "transformer.h.10.ln_2.weight": "model-00001-of-00003.safetensors",
42
+ "transformer.h.10.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
43
+ "transformer.h.10.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
44
+ "transformer.h.10.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
45
+ "transformer.h.10.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
46
+ "transformer.h.10.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
47
+ "transformer.h.10.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
48
+ "transformer.h.11.attn.c_attn.bias": "model-00001-of-00003.safetensors",
49
+ "transformer.h.11.attn.c_attn.weight": "model-00001-of-00003.safetensors",
50
+ "transformer.h.11.attn.c_proj.bias": "model-00001-of-00003.safetensors",
51
+ "transformer.h.11.attn.c_proj.weight": "model-00001-of-00003.safetensors",
52
+ "transformer.h.11.ln_1.bias": "model-00001-of-00003.safetensors",
53
+ "transformer.h.11.ln_1.weight": "model-00001-of-00003.safetensors",
54
+ "transformer.h.11.ln_2.bias": "model-00001-of-00003.safetensors",
55
+ "transformer.h.11.ln_2.weight": "model-00001-of-00003.safetensors",
56
+ "transformer.h.11.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
57
+ "transformer.h.11.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
58
+ "transformer.h.11.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
59
+ "transformer.h.11.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
60
+ "transformer.h.11.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
61
+ "transformer.h.11.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
62
+ "transformer.h.12.attn.c_attn.bias": "model-00002-of-00003.safetensors",
63
+ "transformer.h.12.attn.c_attn.weight": "model-00002-of-00003.safetensors",
64
+ "transformer.h.12.attn.c_proj.bias": "model-00002-of-00003.safetensors",
65
+ "transformer.h.12.attn.c_proj.weight": "model-00002-of-00003.safetensors",
66
+ "transformer.h.12.ln_1.bias": "model-00002-of-00003.safetensors",
67
+ "transformer.h.12.ln_1.weight": "model-00002-of-00003.safetensors",
68
+ "transformer.h.12.ln_2.bias": "model-00002-of-00003.safetensors",
69
+ "transformer.h.12.ln_2.weight": "model-00002-of-00003.safetensors",
70
+ "transformer.h.12.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
71
+ "transformer.h.12.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
72
+ "transformer.h.12.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
73
+ "transformer.h.12.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
74
+ "transformer.h.12.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
75
+ "transformer.h.12.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
76
+ "transformer.h.13.attn.c_attn.bias": "model-00002-of-00003.safetensors",
77
+ "transformer.h.13.attn.c_attn.weight": "model-00002-of-00003.safetensors",
78
+ "transformer.h.13.attn.c_proj.bias": "model-00002-of-00003.safetensors",
79
+ "transformer.h.13.attn.c_proj.weight": "model-00002-of-00003.safetensors",
80
+ "transformer.h.13.ln_1.bias": "model-00002-of-00003.safetensors",
81
+ "transformer.h.13.ln_1.weight": "model-00002-of-00003.safetensors",
82
+ "transformer.h.13.ln_2.bias": "model-00002-of-00003.safetensors",
83
+ "transformer.h.13.ln_2.weight": "model-00002-of-00003.safetensors",
84
+ "transformer.h.13.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
85
+ "transformer.h.13.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
86
+ "transformer.h.13.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
87
+ "transformer.h.13.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
88
+ "transformer.h.13.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
89
+ "transformer.h.13.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
90
+ "transformer.h.14.attn.c_attn.bias": "model-00002-of-00003.safetensors",
91
+ "transformer.h.14.attn.c_attn.weight": "model-00002-of-00003.safetensors",
92
+ "transformer.h.14.attn.c_proj.bias": "model-00002-of-00003.safetensors",
93
+ "transformer.h.14.attn.c_proj.weight": "model-00002-of-00003.safetensors",
94
+ "transformer.h.14.ln_1.bias": "model-00002-of-00003.safetensors",
95
+ "transformer.h.14.ln_1.weight": "model-00002-of-00003.safetensors",
96
+ "transformer.h.14.ln_2.bias": "model-00002-of-00003.safetensors",
97
+ "transformer.h.14.ln_2.weight": "model-00002-of-00003.safetensors",
98
+ "transformer.h.14.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
99
+ "transformer.h.14.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
100
+ "transformer.h.14.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
101
+ "transformer.h.14.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
102
+ "transformer.h.14.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
103
+ "transformer.h.14.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
104
+ "transformer.h.15.attn.c_attn.bias": "model-00002-of-00003.safetensors",
105
+ "transformer.h.15.attn.c_attn.weight": "model-00002-of-00003.safetensors",
106
+ "transformer.h.15.attn.c_proj.bias": "model-00002-of-00003.safetensors",
107
+ "transformer.h.15.attn.c_proj.weight": "model-00002-of-00003.safetensors",
108
+ "transformer.h.15.ln_1.bias": "model-00002-of-00003.safetensors",
109
+ "transformer.h.15.ln_1.weight": "model-00002-of-00003.safetensors",
110
+ "transformer.h.15.ln_2.bias": "model-00002-of-00003.safetensors",
111
+ "transformer.h.15.ln_2.weight": "model-00002-of-00003.safetensors",
112
+ "transformer.h.15.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
113
+ "transformer.h.15.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
114
+ "transformer.h.15.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
115
+ "transformer.h.15.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
116
+ "transformer.h.15.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
117
+ "transformer.h.15.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
118
+ "transformer.h.16.attn.c_attn.bias": "model-00002-of-00003.safetensors",
119
+ "transformer.h.16.attn.c_attn.weight": "model-00002-of-00003.safetensors",
120
+ "transformer.h.16.attn.c_proj.bias": "model-00002-of-00003.safetensors",
121
+ "transformer.h.16.attn.c_proj.weight": "model-00002-of-00003.safetensors",
122
+ "transformer.h.16.ln_1.bias": "model-00002-of-00003.safetensors",
123
+ "transformer.h.16.ln_1.weight": "model-00002-of-00003.safetensors",
124
+ "transformer.h.16.ln_2.bias": "model-00002-of-00003.safetensors",
125
+ "transformer.h.16.ln_2.weight": "model-00002-of-00003.safetensors",
126
+ "transformer.h.16.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
127
+ "transformer.h.16.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
128
+ "transformer.h.16.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
129
+ "transformer.h.16.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
130
+ "transformer.h.16.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
131
+ "transformer.h.16.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
132
+ "transformer.h.17.attn.c_attn.bias": "model-00002-of-00003.safetensors",
133
+ "transformer.h.17.attn.c_attn.weight": "model-00002-of-00003.safetensors",
134
+ "transformer.h.17.attn.c_proj.bias": "model-00002-of-00003.safetensors",
135
+ "transformer.h.17.attn.c_proj.weight": "model-00002-of-00003.safetensors",
136
+ "transformer.h.17.ln_1.bias": "model-00002-of-00003.safetensors",
137
+ "transformer.h.17.ln_1.weight": "model-00002-of-00003.safetensors",
138
+ "transformer.h.17.ln_2.bias": "model-00002-of-00003.safetensors",
139
+ "transformer.h.17.ln_2.weight": "model-00002-of-00003.safetensors",
140
+ "transformer.h.17.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
141
+ "transformer.h.17.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
142
+ "transformer.h.17.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
143
+ "transformer.h.17.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
144
+ "transformer.h.17.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
145
+ "transformer.h.17.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
146
+ "transformer.h.18.attn.c_attn.bias": "model-00002-of-00003.safetensors",
147
+ "transformer.h.18.attn.c_attn.weight": "model-00002-of-00003.safetensors",
148
+ "transformer.h.18.attn.c_proj.bias": "model-00002-of-00003.safetensors",
149
+ "transformer.h.18.attn.c_proj.weight": "model-00002-of-00003.safetensors",
150
+ "transformer.h.18.ln_1.bias": "model-00002-of-00003.safetensors",
151
+ "transformer.h.18.ln_1.weight": "model-00002-of-00003.safetensors",
152
+ "transformer.h.18.ln_2.bias": "model-00002-of-00003.safetensors",
153
+ "transformer.h.18.ln_2.weight": "model-00002-of-00003.safetensors",
154
+ "transformer.h.18.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
155
+ "transformer.h.18.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
156
+ "transformer.h.18.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
157
+ "transformer.h.18.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
158
+ "transformer.h.18.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
159
+ "transformer.h.18.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
160
+ "transformer.h.19.attn.c_attn.bias": "model-00002-of-00003.safetensors",
161
+ "transformer.h.19.attn.c_attn.weight": "model-00002-of-00003.safetensors",
162
+ "transformer.h.19.attn.c_proj.bias": "model-00002-of-00003.safetensors",
163
+ "transformer.h.19.attn.c_proj.weight": "model-00002-of-00003.safetensors",
164
+ "transformer.h.19.ln_1.bias": "model-00002-of-00003.safetensors",
165
+ "transformer.h.19.ln_1.weight": "model-00002-of-00003.safetensors",
166
+ "transformer.h.19.ln_2.bias": "model-00002-of-00003.safetensors",
167
+ "transformer.h.19.ln_2.weight": "model-00002-of-00003.safetensors",
168
+ "transformer.h.19.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
169
+ "transformer.h.19.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
170
+ "transformer.h.19.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
171
+ "transformer.h.19.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
172
+ "transformer.h.19.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
173
+ "transformer.h.19.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
174
+ "transformer.h.2.attn.c_attn.bias": "model-00001-of-00003.safetensors",
175
+ "transformer.h.2.attn.c_attn.weight": "model-00001-of-00003.safetensors",
176
+ "transformer.h.2.attn.c_proj.bias": "model-00001-of-00003.safetensors",
177
+ "transformer.h.2.attn.c_proj.weight": "model-00001-of-00003.safetensors",
178
+ "transformer.h.2.ln_1.bias": "model-00001-of-00003.safetensors",
179
+ "transformer.h.2.ln_1.weight": "model-00001-of-00003.safetensors",
180
+ "transformer.h.2.ln_2.bias": "model-00001-of-00003.safetensors",
181
+ "transformer.h.2.ln_2.weight": "model-00001-of-00003.safetensors",
182
+ "transformer.h.2.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
183
+ "transformer.h.2.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
184
+ "transformer.h.2.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
185
+ "transformer.h.2.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
186
+ "transformer.h.2.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
187
+ "transformer.h.2.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
188
+ "transformer.h.20.attn.c_attn.bias": "model-00002-of-00003.safetensors",
189
+ "transformer.h.20.attn.c_attn.weight": "model-00002-of-00003.safetensors",
190
+ "transformer.h.20.attn.c_proj.bias": "model-00002-of-00003.safetensors",
191
+ "transformer.h.20.attn.c_proj.weight": "model-00002-of-00003.safetensors",
192
+ "transformer.h.20.ln_1.bias": "model-00002-of-00003.safetensors",
193
+ "transformer.h.20.ln_1.weight": "model-00002-of-00003.safetensors",
194
+ "transformer.h.20.ln_2.bias": "model-00002-of-00003.safetensors",
195
+ "transformer.h.20.ln_2.weight": "model-00002-of-00003.safetensors",
196
+ "transformer.h.20.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
197
+ "transformer.h.20.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
198
+ "transformer.h.20.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
199
+ "transformer.h.20.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
200
+ "transformer.h.20.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
201
+ "transformer.h.20.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
202
+ "transformer.h.21.attn.c_attn.bias": "model-00002-of-00003.safetensors",
203
+ "transformer.h.21.attn.c_attn.weight": "model-00002-of-00003.safetensors",
204
+ "transformer.h.21.attn.c_proj.bias": "model-00002-of-00003.safetensors",
205
+ "transformer.h.21.attn.c_proj.weight": "model-00002-of-00003.safetensors",
206
+ "transformer.h.21.ln_1.bias": "model-00002-of-00003.safetensors",
207
+ "transformer.h.21.ln_1.weight": "model-00002-of-00003.safetensors",
208
+ "transformer.h.21.ln_2.bias": "model-00002-of-00003.safetensors",
209
+ "transformer.h.21.ln_2.weight": "model-00002-of-00003.safetensors",
210
+ "transformer.h.21.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
211
+ "transformer.h.21.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
212
+ "transformer.h.21.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
213
+ "transformer.h.21.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
214
+ "transformer.h.21.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
215
+ "transformer.h.21.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
216
+ "transformer.h.22.attn.c_attn.bias": "model-00002-of-00003.safetensors",
217
+ "transformer.h.22.attn.c_attn.weight": "model-00002-of-00003.safetensors",
218
+ "transformer.h.22.attn.c_proj.bias": "model-00002-of-00003.safetensors",
219
+ "transformer.h.22.attn.c_proj.weight": "model-00002-of-00003.safetensors",
220
+ "transformer.h.22.ln_1.bias": "model-00002-of-00003.safetensors",
221
+ "transformer.h.22.ln_1.weight": "model-00002-of-00003.safetensors",
222
+ "transformer.h.22.ln_2.bias": "model-00002-of-00003.safetensors",
223
+ "transformer.h.22.ln_2.weight": "model-00002-of-00003.safetensors",
224
+ "transformer.h.22.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
225
+ "transformer.h.22.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
226
+ "transformer.h.22.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
227
+ "transformer.h.22.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
228
+ "transformer.h.22.mlp.c_proj.bias": "model-00002-of-00003.safetensors",
229
+ "transformer.h.22.mlp.c_proj.weight": "model-00002-of-00003.safetensors",
230
+ "transformer.h.23.attn.c_attn.bias": "model-00002-of-00003.safetensors",
231
+ "transformer.h.23.attn.c_attn.weight": "model-00002-of-00003.safetensors",
232
+ "transformer.h.23.attn.c_proj.bias": "model-00002-of-00003.safetensors",
233
+ "transformer.h.23.attn.c_proj.weight": "model-00002-of-00003.safetensors",
234
+ "transformer.h.23.ln_1.bias": "model-00002-of-00003.safetensors",
235
+ "transformer.h.23.ln_1.weight": "model-00002-of-00003.safetensors",
236
+ "transformer.h.23.ln_2.bias": "model-00002-of-00003.safetensors",
237
+ "transformer.h.23.ln_2.weight": "model-00002-of-00003.safetensors",
238
+ "transformer.h.23.mlp.c_fc.bias": "model-00002-of-00003.safetensors",
239
+ "transformer.h.23.mlp.c_fc.weight": "model-00002-of-00003.safetensors",
240
+ "transformer.h.23.mlp.c_fc2.bias": "model-00002-of-00003.safetensors",
241
+ "transformer.h.23.mlp.c_fc2.weight": "model-00002-of-00003.safetensors",
242
+ "transformer.h.23.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
243
+ "transformer.h.23.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
244
+ "transformer.h.24.attn.c_attn.bias": "model-00003-of-00003.safetensors",
245
+ "transformer.h.24.attn.c_attn.weight": "model-00003-of-00003.safetensors",
246
+ "transformer.h.24.attn.c_proj.bias": "model-00003-of-00003.safetensors",
247
+ "transformer.h.24.attn.c_proj.weight": "model-00003-of-00003.safetensors",
248
+ "transformer.h.24.ln_1.bias": "model-00003-of-00003.safetensors",
249
+ "transformer.h.24.ln_1.weight": "model-00003-of-00003.safetensors",
250
+ "transformer.h.24.ln_2.bias": "model-00003-of-00003.safetensors",
251
+ "transformer.h.24.ln_2.weight": "model-00003-of-00003.safetensors",
252
+ "transformer.h.24.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
253
+ "transformer.h.24.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
254
+ "transformer.h.24.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
255
+ "transformer.h.24.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
256
+ "transformer.h.24.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
257
+ "transformer.h.24.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
258
+ "transformer.h.25.attn.c_attn.bias": "model-00003-of-00003.safetensors",
259
+ "transformer.h.25.attn.c_attn.weight": "model-00003-of-00003.safetensors",
260
+ "transformer.h.25.attn.c_proj.bias": "model-00003-of-00003.safetensors",
261
+ "transformer.h.25.attn.c_proj.weight": "model-00003-of-00003.safetensors",
262
+ "transformer.h.25.ln_1.bias": "model-00003-of-00003.safetensors",
263
+ "transformer.h.25.ln_1.weight": "model-00003-of-00003.safetensors",
264
+ "transformer.h.25.ln_2.bias": "model-00003-of-00003.safetensors",
265
+ "transformer.h.25.ln_2.weight": "model-00003-of-00003.safetensors",
266
+ "transformer.h.25.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
267
+ "transformer.h.25.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
268
+ "transformer.h.25.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
269
+ "transformer.h.25.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
270
+ "transformer.h.25.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
271
+ "transformer.h.25.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
272
+ "transformer.h.26.attn.c_attn.bias": "model-00003-of-00003.safetensors",
273
+ "transformer.h.26.attn.c_attn.weight": "model-00003-of-00003.safetensors",
274
+ "transformer.h.26.attn.c_proj.bias": "model-00003-of-00003.safetensors",
275
+ "transformer.h.26.attn.c_proj.weight": "model-00003-of-00003.safetensors",
276
+ "transformer.h.26.ln_1.bias": "model-00003-of-00003.safetensors",
277
+ "transformer.h.26.ln_1.weight": "model-00003-of-00003.safetensors",
278
+ "transformer.h.26.ln_2.bias": "model-00003-of-00003.safetensors",
279
+ "transformer.h.26.ln_2.weight": "model-00003-of-00003.safetensors",
280
+ "transformer.h.26.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
281
+ "transformer.h.26.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
282
+ "transformer.h.26.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
283
+ "transformer.h.26.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
284
+ "transformer.h.26.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
285
+ "transformer.h.26.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
286
+ "transformer.h.27.attn.c_attn.bias": "model-00003-of-00003.safetensors",
287
+ "transformer.h.27.attn.c_attn.weight": "model-00003-of-00003.safetensors",
288
+ "transformer.h.27.attn.c_proj.bias": "model-00003-of-00003.safetensors",
289
+ "transformer.h.27.attn.c_proj.weight": "model-00003-of-00003.safetensors",
290
+ "transformer.h.27.ln_1.bias": "model-00003-of-00003.safetensors",
291
+ "transformer.h.27.ln_1.weight": "model-00003-of-00003.safetensors",
292
+ "transformer.h.27.ln_2.bias": "model-00003-of-00003.safetensors",
293
+ "transformer.h.27.ln_2.weight": "model-00003-of-00003.safetensors",
294
+ "transformer.h.27.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
295
+ "transformer.h.27.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
296
+ "transformer.h.27.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
297
+ "transformer.h.27.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
298
+ "transformer.h.27.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
299
+ "transformer.h.27.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
300
+ "transformer.h.28.attn.c_attn.bias": "model-00003-of-00003.safetensors",
301
+ "transformer.h.28.attn.c_attn.weight": "model-00003-of-00003.safetensors",
302
+ "transformer.h.28.attn.c_proj.bias": "model-00003-of-00003.safetensors",
303
+ "transformer.h.28.attn.c_proj.weight": "model-00003-of-00003.safetensors",
304
+ "transformer.h.28.ln_1.bias": "model-00003-of-00003.safetensors",
305
+ "transformer.h.28.ln_1.weight": "model-00003-of-00003.safetensors",
306
+ "transformer.h.28.ln_2.bias": "model-00003-of-00003.safetensors",
307
+ "transformer.h.28.ln_2.weight": "model-00003-of-00003.safetensors",
308
+ "transformer.h.28.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
309
+ "transformer.h.28.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
310
+ "transformer.h.28.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
311
+ "transformer.h.28.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
312
+ "transformer.h.28.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
313
+ "transformer.h.28.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
314
+ "transformer.h.29.attn.c_attn.bias": "model-00003-of-00003.safetensors",
315
+ "transformer.h.29.attn.c_attn.weight": "model-00003-of-00003.safetensors",
316
+ "transformer.h.29.attn.c_proj.bias": "model-00003-of-00003.safetensors",
317
+ "transformer.h.29.attn.c_proj.weight": "model-00003-of-00003.safetensors",
318
+ "transformer.h.29.ln_1.bias": "model-00003-of-00003.safetensors",
319
+ "transformer.h.29.ln_1.weight": "model-00003-of-00003.safetensors",
320
+ "transformer.h.29.ln_2.bias": "model-00003-of-00003.safetensors",
321
+ "transformer.h.29.ln_2.weight": "model-00003-of-00003.safetensors",
322
+ "transformer.h.29.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
323
+ "transformer.h.29.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
324
+ "transformer.h.29.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
325
+ "transformer.h.29.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
326
+ "transformer.h.29.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
327
+ "transformer.h.29.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
328
+ "transformer.h.3.attn.c_attn.bias": "model-00001-of-00003.safetensors",
329
+ "transformer.h.3.attn.c_attn.weight": "model-00001-of-00003.safetensors",
330
+ "transformer.h.3.attn.c_proj.bias": "model-00001-of-00003.safetensors",
331
+ "transformer.h.3.attn.c_proj.weight": "model-00001-of-00003.safetensors",
332
+ "transformer.h.3.ln_1.bias": "model-00001-of-00003.safetensors",
333
+ "transformer.h.3.ln_1.weight": "model-00001-of-00003.safetensors",
334
+ "transformer.h.3.ln_2.bias": "model-00001-of-00003.safetensors",
335
+ "transformer.h.3.ln_2.weight": "model-00001-of-00003.safetensors",
336
+ "transformer.h.3.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
337
+ "transformer.h.3.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
338
+ "transformer.h.3.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
339
+ "transformer.h.3.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
340
+ "transformer.h.3.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
341
+ "transformer.h.3.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
342
+ "transformer.h.30.attn.c_attn.bias": "model-00003-of-00003.safetensors",
343
+ "transformer.h.30.attn.c_attn.weight": "model-00003-of-00003.safetensors",
344
+ "transformer.h.30.attn.c_proj.bias": "model-00003-of-00003.safetensors",
345
+ "transformer.h.30.attn.c_proj.weight": "model-00003-of-00003.safetensors",
346
+ "transformer.h.30.ln_1.bias": "model-00003-of-00003.safetensors",
347
+ "transformer.h.30.ln_1.weight": "model-00003-of-00003.safetensors",
348
+ "transformer.h.30.ln_2.bias": "model-00003-of-00003.safetensors",
349
+ "transformer.h.30.ln_2.weight": "model-00003-of-00003.safetensors",
350
+ "transformer.h.30.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
351
+ "transformer.h.30.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
352
+ "transformer.h.30.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
353
+ "transformer.h.30.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
354
+ "transformer.h.30.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
355
+ "transformer.h.30.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
356
+ "transformer.h.31.attn.c_attn.bias": "model-00003-of-00003.safetensors",
357
+ "transformer.h.31.attn.c_attn.weight": "model-00003-of-00003.safetensors",
358
+ "transformer.h.31.attn.c_proj.bias": "model-00003-of-00003.safetensors",
359
+ "transformer.h.31.attn.c_proj.weight": "model-00003-of-00003.safetensors",
360
+ "transformer.h.31.ln_1.bias": "model-00003-of-00003.safetensors",
361
+ "transformer.h.31.ln_1.weight": "model-00003-of-00003.safetensors",
362
+ "transformer.h.31.ln_2.bias": "model-00003-of-00003.safetensors",
363
+ "transformer.h.31.ln_2.weight": "model-00003-of-00003.safetensors",
364
+ "transformer.h.31.mlp.c_fc.bias": "model-00003-of-00003.safetensors",
365
+ "transformer.h.31.mlp.c_fc.weight": "model-00003-of-00003.safetensors",
366
+ "transformer.h.31.mlp.c_fc2.bias": "model-00003-of-00003.safetensors",
367
+ "transformer.h.31.mlp.c_fc2.weight": "model-00003-of-00003.safetensors",
368
+ "transformer.h.31.mlp.c_proj.bias": "model-00003-of-00003.safetensors",
369
+ "transformer.h.31.mlp.c_proj.weight": "model-00003-of-00003.safetensors",
370
+ "transformer.h.4.attn.c_attn.bias": "model-00001-of-00003.safetensors",
371
+ "transformer.h.4.attn.c_attn.weight": "model-00001-of-00003.safetensors",
372
+ "transformer.h.4.attn.c_proj.bias": "model-00001-of-00003.safetensors",
373
+ "transformer.h.4.attn.c_proj.weight": "model-00001-of-00003.safetensors",
374
+ "transformer.h.4.ln_1.bias": "model-00001-of-00003.safetensors",
375
+ "transformer.h.4.ln_1.weight": "model-00001-of-00003.safetensors",
376
+ "transformer.h.4.ln_2.bias": "model-00001-of-00003.safetensors",
377
+ "transformer.h.4.ln_2.weight": "model-00001-of-00003.safetensors",
378
+ "transformer.h.4.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
379
+ "transformer.h.4.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
380
+ "transformer.h.4.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
381
+ "transformer.h.4.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
382
+ "transformer.h.4.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
383
+ "transformer.h.4.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
384
+ "transformer.h.5.attn.c_attn.bias": "model-00001-of-00003.safetensors",
385
+ "transformer.h.5.attn.c_attn.weight": "model-00001-of-00003.safetensors",
386
+ "transformer.h.5.attn.c_proj.bias": "model-00001-of-00003.safetensors",
387
+ "transformer.h.5.attn.c_proj.weight": "model-00001-of-00003.safetensors",
388
+ "transformer.h.5.ln_1.bias": "model-00001-of-00003.safetensors",
389
+ "transformer.h.5.ln_1.weight": "model-00001-of-00003.safetensors",
390
+ "transformer.h.5.ln_2.bias": "model-00001-of-00003.safetensors",
391
+ "transformer.h.5.ln_2.weight": "model-00001-of-00003.safetensors",
392
+ "transformer.h.5.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
393
+ "transformer.h.5.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
394
+ "transformer.h.5.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
395
+ "transformer.h.5.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
396
+ "transformer.h.5.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
397
+ "transformer.h.5.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
398
+ "transformer.h.6.attn.c_attn.bias": "model-00001-of-00003.safetensors",
399
+ "transformer.h.6.attn.c_attn.weight": "model-00001-of-00003.safetensors",
400
+ "transformer.h.6.attn.c_proj.bias": "model-00001-of-00003.safetensors",
401
+ "transformer.h.6.attn.c_proj.weight": "model-00001-of-00003.safetensors",
402
+ "transformer.h.6.ln_1.bias": "model-00001-of-00003.safetensors",
403
+ "transformer.h.6.ln_1.weight": "model-00001-of-00003.safetensors",
404
+ "transformer.h.6.ln_2.bias": "model-00001-of-00003.safetensors",
405
+ "transformer.h.6.ln_2.weight": "model-00001-of-00003.safetensors",
406
+ "transformer.h.6.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
407
+ "transformer.h.6.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
408
+ "transformer.h.6.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
409
+ "transformer.h.6.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
410
+ "transformer.h.6.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
411
+ "transformer.h.6.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
412
+ "transformer.h.7.attn.c_attn.bias": "model-00001-of-00003.safetensors",
413
+ "transformer.h.7.attn.c_attn.weight": "model-00001-of-00003.safetensors",
414
+ "transformer.h.7.attn.c_proj.bias": "model-00001-of-00003.safetensors",
415
+ "transformer.h.7.attn.c_proj.weight": "model-00001-of-00003.safetensors",
416
+ "transformer.h.7.ln_1.bias": "model-00001-of-00003.safetensors",
417
+ "transformer.h.7.ln_1.weight": "model-00001-of-00003.safetensors",
418
+ "transformer.h.7.ln_2.bias": "model-00001-of-00003.safetensors",
419
+ "transformer.h.7.ln_2.weight": "model-00001-of-00003.safetensors",
420
+ "transformer.h.7.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
421
+ "transformer.h.7.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
422
+ "transformer.h.7.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
423
+ "transformer.h.7.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
424
+ "transformer.h.7.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
425
+ "transformer.h.7.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
426
+ "transformer.h.8.attn.c_attn.bias": "model-00001-of-00003.safetensors",
427
+ "transformer.h.8.attn.c_attn.weight": "model-00001-of-00003.safetensors",
428
+ "transformer.h.8.attn.c_proj.bias": "model-00001-of-00003.safetensors",
429
+ "transformer.h.8.attn.c_proj.weight": "model-00001-of-00003.safetensors",
430
+ "transformer.h.8.ln_1.bias": "model-00001-of-00003.safetensors",
431
+ "transformer.h.8.ln_1.weight": "model-00001-of-00003.safetensors",
432
+ "transformer.h.8.ln_2.bias": "model-00001-of-00003.safetensors",
433
+ "transformer.h.8.ln_2.weight": "model-00001-of-00003.safetensors",
434
+ "transformer.h.8.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
435
+ "transformer.h.8.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
436
+ "transformer.h.8.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
437
+ "transformer.h.8.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
438
+ "transformer.h.8.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
439
+ "transformer.h.8.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
440
+ "transformer.h.9.attn.c_attn.bias": "model-00001-of-00003.safetensors",
441
+ "transformer.h.9.attn.c_attn.weight": "model-00001-of-00003.safetensors",
442
+ "transformer.h.9.attn.c_proj.bias": "model-00001-of-00003.safetensors",
443
+ "transformer.h.9.attn.c_proj.weight": "model-00001-of-00003.safetensors",
444
+ "transformer.h.9.ln_1.bias": "model-00001-of-00003.safetensors",
445
+ "transformer.h.9.ln_1.weight": "model-00001-of-00003.safetensors",
446
+ "transformer.h.9.ln_2.bias": "model-00001-of-00003.safetensors",
447
+ "transformer.h.9.ln_2.weight": "model-00001-of-00003.safetensors",
448
+ "transformer.h.9.mlp.c_fc.bias": "model-00001-of-00003.safetensors",
449
+ "transformer.h.9.mlp.c_fc.weight": "model-00001-of-00003.safetensors",
450
+ "transformer.h.9.mlp.c_fc2.bias": "model-00001-of-00003.safetensors",
451
+ "transformer.h.9.mlp.c_fc2.weight": "model-00001-of-00003.safetensors",
452
+ "transformer.h.9.mlp.c_proj.bias": "model-00001-of-00003.safetensors",
453
+ "transformer.h.9.mlp.c_proj.weight": "model-00001-of-00003.safetensors",
454
+ "transformer.ln_f.bias": "model-00003-of-00003.safetensors",
455
+ "transformer.ln_f.weight": "model-00003-of-00003.safetensors",
456
+ "transformer.wte.weight": "model-00001-of-00003.safetensors"
457
+ }
458
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|fim_prefix|>",
4
+ "<|fim_middle|>",
5
+ "<|fim_suffix|>",
6
+ "<|fim_pad|>",
7
+ "<|filename|>",
8
+ "<|gh_stars|>",
9
+ "<|issue_start|>",
10
+ "<|issue_comment|>",
11
+ "<|issue_closed|>",
12
+ "<|jupyter_start|>",
13
+ "<|jupyter_text|>",
14
+ "<|jupyter_code|>",
15
+ "<|jupyter_output|>",
16
+ "<|empty_output|>",
17
+ "<|commit_before|>",
18
+ "<|commit_msg|>",
19
+ "<|commit_after|>",
20
+ "<|reponame|>",
21
+ "<|im_start|>",
22
+ "<|im_end|>",
23
+ "<|sys_start|>",
24
+ "<|sys_end|>"
25
+ ],
26
+ "bos_token": {
27
+ "content": "<s>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ },
33
+ "eos_token": {
34
+ "content": "</s>",
35
+ "lstrip": false,
36
+ "normalized": false,
37
+ "rstrip": false,
38
+ "single_word": false
39
+ },
40
+ "unk_token": {
41
+ "content": "<unk>",
42
+ "lstrip": false,
43
+ "normalized": false,
44
+ "rstrip": false,
45
+ "single_word": false
46
+ }
47
+ }
tokenizer_config.json CHANGED
@@ -225,6 +225,12 @@
225
  "<|sys_start|>",
226
  "<|sys_end|>"
227
  ],
 
 
 
 
 
 
228
  "bos_token": "<s>",
229
  "clean_up_tokenization_spaces": false,
230
  "eos_token": "</s>",
@@ -233,12 +239,6 @@
233
  "pad_token": null,
234
  "padding_side": "right",
235
  "sp_model_kwargs": {},
236
- "auto_map": {
237
- "AutoTokenizer": [
238
- null,
239
- "tokenization_crystalcoder_fast.CrystalCoderTokenizerFast"
240
- ]
241
- },
242
  "tokenizer_class": "CrystalCoderTokenizer",
243
  "unk_token": "<unk>",
244
  "use_default_system_prompt": false
 
225
  "<|sys_start|>",
226
  "<|sys_end|>"
227
  ],
228
+ "auto_map": {
229
+ "AutoTokenizer": [
230
+ null,
231
+ "LLM360/CrystalCoder--tokenization_crystalcoder_fast.CrystalCoderTokenizerFast"
232
+ ]
233
+ },
234
  "bos_token": "<s>",
235
  "clean_up_tokenization_spaces": false,
236
  "eos_token": "</s>",
 
239
  "pad_token": null,
240
  "padding_side": "right",
241
  "sp_model_kwargs": {},
 
 
 
 
 
 
242
  "tokenizer_class": "CrystalCoderTokenizer",
243
  "unk_token": "<unk>",
244
  "use_default_system_prompt": false