Hugging Face's logo

abanm
/
Dubs_O

Text Generation
Transformers
Safetensors
English
phi3
nlp
code
conversational
custom_code
text-generation-inference
Inference Endpoints
4-bit precision
bitsandbytes
Model card Files Community
abanm commited on
Commit
c44109a
·
verified ·
1 Parent(s): 6adf3d1

Upload config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. config.json +154 -0
config.json ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "abanm/Dubs_Old",
3
+ "architectures": [
4
+ "Phi3ForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "auto_map": {
9
+ "AutoConfig": "microsoft/Phi-3-mini-128k-instruct--configuration_phi3.Phi3Config",
10
+ "AutoModelForCausalLM": "microsoft/Phi-3-mini-128k-instruct--modeling_phi3.Phi3ForCausalLM"
11
+ },
12
+ "bos_token_id": 1,
13
+ "description": "Finetuned on RayBernard/leetcode, MAsad789565/Coding_GPT4_Data,abanm/preprocessed-leetcode-multi-shot",
14
+ "embd_pdrop": 0.0,
15
+ "eos_token_id": 32000,
16
+ "hidden_act": "silu",
17
+ "hidden_size": 3072,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 8192,
20
+ "max_position_embeddings": 131072,
21
+ "model_type": "phi3",
22
+ "num_attention_heads": 32,
23
+ "num_hidden_layers": 32,
24
+ "num_key_value_heads": 32,
25
+ "original_max_position_embeddings": 4096,
26
+ "pad_token_id": 32000,
27
+ "quantization_config": {
28
+ "_load_in_4bit": true,
29
+ "_load_in_8bit": false,
30
+ "bnb_4bit_compute_dtype": "float32",
31
+ "bnb_4bit_quant_storage": "uint8",
32
+ "bnb_4bit_quant_type": "fp4",
33
+ "bnb_4bit_use_double_quant": false,
34
+ "llm_int8_enable_fp32_cpu_offload": false,
35
+ "llm_int8_has_fp16_weight": false,
36
+ "llm_int8_skip_modules": null,
37
+ "llm_int8_threshold": 6.0,
38
+ "load_in_4bit": true,
39
+ "load_in_8bit": false,
40
+ "quant_method": "bitsandbytes"
41
+ },
42
+ "resid_pdrop": 0.0,
43
+ "rms_norm_eps": 1e-05,
44
+ "rope_scaling": {
45
+ "long_factor": [
46
+ 1.0700000524520874,
47
+ 1.1200000047683716,
48
+ 1.149999976158142,
49
+ 1.4199999570846558,
50
+ 1.5699999332427979,
51
+ 1.7999999523162842,
52
+ 2.129999876022339,
53
+ 2.129999876022339,
54
+ 3.009999990463257,
55
+ 5.910000324249268,
56
+ 6.950000286102295,
57
+ 9.070000648498535,
58
+ 9.930000305175781,
59
+ 10.710000038146973,
60
+ 11.130000114440918,
61
+ 14.609999656677246,
62
+ 15.409998893737793,
63
+ 19.809999465942383,
64
+ 37.279998779296875,
65
+ 38.279998779296875,
66
+ 38.599998474121094,
67
+ 40.12000274658203,
68
+ 46.20000457763672,
69
+ 50.940006256103516,
70
+ 53.66000747680664,
71
+ 54.9373893737793,
72
+ 56.89738845825195,
73
+ 57.28738784790039,
74
+ 59.98738479614258,
75
+ 60.86738586425781,
76
+ 60.887386322021484,
77
+ 61.71739196777344,
78
+ 62.91739273071289,
79
+ 62.957393646240234,
80
+ 63.41739273071289,
81
+ 63.8173942565918,
82
+ 63.83739471435547,
83
+ 63.897396087646484,
84
+ 63.93739700317383,
85
+ 64.06739807128906,
86
+ 64.11434936523438,
87
+ 64.12435150146484,
88
+ 64.15435028076172,
89
+ 64.19435119628906,
90
+ 64.24435424804688,
91
+ 64.57435607910156,
92
+ 64.69000244140625,
93
+ 64.76000213623047
94
+ ],
95
+ "short_factor": [
96
+ 1.1,
97
+ 1.1,
98
+ 1.1,
99
+ 1.3000000000000003,
100
+ 1.3500000000000003,
101
+ 1.3500000000000003,
102
+ 1.4000000000000004,
103
+ 1.5500000000000005,
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
+ 2.0500000000000007,
122
+ 2.0500000000000007,
123
+ 2.0500000000000007,
124
+ 2.0500000000000007,
125
+ 2.0500000000000007,
126
+ 2.0500000000000007,
127
+ 2.1000000000000005,
128
+ 2.1000000000000005,
129
+ 2.1500000000000004,
130
+ 2.25,
131
+ 2.25,
132
+ 2.25,
133
+ 2.25,
134
+ 2.25,
135
+ 2.3999999999999995,
136
+ 2.4499999999999993,
137
+ 2.499999999999999,
138
+ 2.6999999999999984,
139
+ 2.6999999999999984,
140
+ 2.7499999999999982,
141
+ 2.799999999999998,
142
+ 2.8999999999999977,
143
+ 3.049999999999997
144
+ ],
145
+ "type": "longrope"
146
+ },
147
+ "rope_theta": 10000.0,
148
+ "sliding_window": 262144,
149
+ "tie_word_embeddings": false,
150
+ "torch_dtype": "bfloat16",
151
+ "transformers_version": "4.47.1",
152
+ "use_cache": true,
153
+ "vocab_size": 32064
154
+ }