Hugging Face's logo

RichardErkhov
/
NexaAIDev_-_octopus-planning-awq

Safetensors
phi3
custom_code
4-bit precision
awq
Model card Files Community
RichardErkhov commited on
Commit
e155ba8
·
verified ·
1 Parent(s): dad4980

uploaded model

Browse files
Files changed (1) hide show
  1. config.json +148 -0
config.json ADDED
@@ -0,0 +1,148 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "octopus-planning",
3
+ "architectures": [
4
+ "Phi3ForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "auto_map": {
9
+ "AutoConfig": "configuration_phi3.Phi3Config",
10
+ "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM",
11
+ "AutoModelForSequenceClassification": "microsoft/Phi-3-mini-128k-instruct--modeling_phi3.Phi3ForSequenceClassification",
12
+ "AutoModelForTokenClassification": "microsoft/Phi-3-mini-128k-instruct--modeling_phi3.Phi3ForTokenClassification"
13
+ },
14
+ "bos_token_id": 1,
15
+ "embd_pdrop": 0.0,
16
+ "eos_token_id": 32000,
17
+ "hidden_act": "silu",
18
+ "hidden_size": 3072,
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 8192,
21
+ "max_position_embeddings": 131072,
22
+ "model_type": "phi3",
23
+ "num_attention_heads": 32,
24
+ "num_hidden_layers": 32,
25
+ "num_key_value_heads": 32,
26
+ "original_max_position_embeddings": 4096,
27
+ "pad_token_id": 32000,
28
+ "quantization_config": {
29
+ "bits": 4,
30
+ "group_size": 128,
31
+ "modules_to_not_convert": null,
32
+ "quant_method": "awq",
33
+ "version": "gemm",
34
+ "zero_point": true
35
+ },
36
+ "resid_pdrop": 0.0,
37
+ "rms_norm_eps": 1e-05,
38
+ "rope_scaling": {
39
+ "long_factor": [
40
+ 1.0299999713897705,
41
+ 1.0499999523162842,
42
+ 1.0499999523162842,
43
+ 1.0799999237060547,
44
+ 1.2299998998641968,
45
+ 1.2299998998641968,
46
+ 1.2999999523162842,
47
+ 1.4499999284744263,
48
+ 1.5999999046325684,
49
+ 1.6499998569488525,
50
+ 1.8999998569488525,
51
+ 2.859999895095825,
52
+ 3.68999981880188,
53
+ 5.419999599456787,
54
+ 5.489999771118164,
55
+ 5.489999771118164,
56
+ 9.09000015258789,
57
+ 11.579999923706055,
58
+ 15.65999984741211,
59
+ 15.769999504089355,
60
+ 15.789999961853027,
61
+ 18.360000610351562,
62
+ 21.989999771118164,
63
+ 23.079999923706055,
64
+ 30.009998321533203,
65
+ 32.35000228881836,
66
+ 32.590003967285156,
67
+ 35.56000518798828,
68
+ 39.95000457763672,
69
+ 53.840003967285156,
70
+ 56.20000457763672,
71
+ 57.95000457763672,
72
+ 59.29000473022461,
73
+ 59.77000427246094,
74
+ 59.920005798339844,
75
+ 61.190006256103516,
76
+ 61.96000671386719,
77
+ 62.50000762939453,
78
+ 63.3700065612793,
79
+ 63.48000717163086,
80
+ 63.48000717163086,
81
+ 63.66000747680664,
82
+ 63.850006103515625,
83
+ 64.08000946044922,
84
+ 64.760009765625,
85
+ 64.80001068115234,
86
+ 64.81001281738281,
87
+ 64.81001281738281
88
+ ],
89
+ "short_factor": [
90
+ 1.05,
91
+ 1.05,
92
+ 1.05,
93
+ 1.1,
94
+ 1.1,
95
+ 1.1500000000000001,
96
+ 1.2000000000000002,
97
+ 1.2500000000000002,
98
+ 1.3000000000000003,
99
+ 1.3500000000000003,
100
+ 1.5000000000000004,
101
+ 2.000000000000001,
102
+ 2.000000000000001,
103
+ 2.000000000000001,
104
+ 2.000000000000001,
105
+ 2.000000000000001,
106
+ 2.000000000000001,
107
+ 2.000000000000001,
108
+ 2.000000000000001,
109
+ 2.000000000000001,
110
+ 2.000000000000001,
111
+ 2.000000000000001,
112
+ 2.000000000000001,
113
+ 2.000000000000001,
114
+ 2.000000000000001,
115
+ 2.000000000000001,
116
+ 2.000000000000001,
117
+ 2.000000000000001,
118
+ 2.000000000000001,
119
+ 2.000000000000001,
120
+ 2.000000000000001,
121
+ 2.000000000000001,
122
+ 2.0500000000000007,
123
+ 2.0500000000000007,
124
+ 2.0500000000000007,
125
+ 2.1000000000000005,
126
+ 2.1000000000000005,
127
+ 2.1000000000000005,
128
+ 2.1500000000000004,
129
+ 2.1500000000000004,
130
+ 2.3499999999999996,
131
+ 2.549999999999999,
132
+ 2.5999999999999988,
133
+ 2.5999999999999988,
134
+ 2.7499999999999982,
135
+ 2.849999999999998,
136
+ 2.849999999999998,
137
+ 2.9499999999999975
138
+ ],
139
+ "type": "su"
140
+ },
141
+ "rope_theta": 10000.0,
142
+ "sliding_window": 262144,
143
+ "tie_word_embeddings": false,
144
+ "torch_dtype": "float16",
145
+ "transformers_version": "4.47.1",
146
+ "use_cache": false,
147
+ "vocab_size": 32064
148
+ }