commit files to HF hub
Browse files- config.json +1 -1
- config_rope.json +175 -0
config.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
"_name_or_path": "PersianStories-4k",
|
3 |
"architectures": [
|
4 |
-
"
|
5 |
],
|
6 |
"attention_dropout": 0.0,
|
7 |
"auto_map": {
|
|
|
1 |
{
|
2 |
"_name_or_path": "PersianStories-4k",
|
3 |
"architectures": [
|
4 |
+
"PersianStoriesForCausalLM"
|
5 |
],
|
6 |
"attention_dropout": 0.0,
|
7 |
"auto_map": {
|
config_rope.json
ADDED
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "PersianStories-4k",
|
3 |
+
"architectures": [
|
4 |
+
"Phi3ForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_dropout": 0.0,
|
7 |
+
"auto_map": {
|
8 |
+
"AutoConfig": "configuration.PersianStoriesConfig",
|
9 |
+
"AutoModelForCausalLM": "modeling.PersianStoriesForCausalLM"
|
10 |
+
},
|
11 |
+
"bos_token_id": 1,
|
12 |
+
"embd_pdrop": 0.0,
|
13 |
+
"eos_token_id": 32000,
|
14 |
+
"hidden_act": "silu",
|
15 |
+
"hidden_size": 2048,
|
16 |
+
"initializer_range": 0.02,
|
17 |
+
"intermediate_size": 8192,
|
18 |
+
"max_position_embeddings": 4096,
|
19 |
+
"model_type": "phi3",
|
20 |
+
"num_attention_heads": 16,
|
21 |
+
"num_hidden_layers": 24,
|
22 |
+
"num_key_value_heads": 4,
|
23 |
+
"original_max_position_embeddings": 4096,
|
24 |
+
"pad_token_id": 32000,
|
25 |
+
"resid_pdrop": 0.0,
|
26 |
+
"rms_norm_eps": 1e-05,
|
27 |
+
"rope_position_scale": 1.0,
|
28 |
+
"rope_scaling": {
|
29 |
+
"long_factor": [
|
30 |
+
1.0,
|
31 |
+
1.01,
|
32 |
+
1.01,
|
33 |
+
1.02,
|
34 |
+
1.04,
|
35 |
+
1.04,
|
36 |
+
1.04,
|
37 |
+
1.05,
|
38 |
+
1.05,
|
39 |
+
1.06,
|
40 |
+
1.07,
|
41 |
+
1.08,
|
42 |
+
1.08,
|
43 |
+
1.08,
|
44 |
+
1.08,
|
45 |
+
1.08,
|
46 |
+
1.08,
|
47 |
+
1.08,
|
48 |
+
1.09,
|
49 |
+
1.09,
|
50 |
+
1.2,
|
51 |
+
2.31,
|
52 |
+
3.76,
|
53 |
+
9.38,
|
54 |
+
10.1,
|
55 |
+
10.8,
|
56 |
+
18.1,
|
57 |
+
25.2,
|
58 |
+
25.3,
|
59 |
+
26.1,
|
60 |
+
26.6,
|
61 |
+
30.2,
|
62 |
+
33.0,
|
63 |
+
41.5,
|
64 |
+
44.4,
|
65 |
+
44.8,
|
66 |
+
50.2,
|
67 |
+
51.9,
|
68 |
+
59.3,
|
69 |
+
62.7,
|
70 |
+
66.1,
|
71 |
+
66.3,
|
72 |
+
85.8,
|
73 |
+
89.3,
|
74 |
+
90.0,
|
75 |
+
99.9,
|
76 |
+
107.0,
|
77 |
+
110.0,
|
78 |
+
111.0,
|
79 |
+
117.0,
|
80 |
+
118.0,
|
81 |
+
121.0,
|
82 |
+
122.0,
|
83 |
+
127.0,
|
84 |
+
127.0,
|
85 |
+
128.0,
|
86 |
+
128.0,
|
87 |
+
128.0,
|
88 |
+
128.0,
|
89 |
+
128.0,
|
90 |
+
128.0,
|
91 |
+
129.0,
|
92 |
+
129.0,
|
93 |
+
129.0
|
94 |
+
],
|
95 |
+
"long_mscale": 1.1902380714238083,
|
96 |
+
"original_max_position_embeddings": 8192,
|
97 |
+
"short_factor": [
|
98 |
+
1.02,
|
99 |
+
1.02,
|
100 |
+
1.05,
|
101 |
+
1.05,
|
102 |
+
1.06,
|
103 |
+
1.08,
|
104 |
+
1.08,
|
105 |
+
1.08,
|
106 |
+
1.08,
|
107 |
+
1.12,
|
108 |
+
1.1800000000000002,
|
109 |
+
1.1900000000000002,
|
110 |
+
1.1900000000000002,
|
111 |
+
1.2100000000000002,
|
112 |
+
1.2300000000000002,
|
113 |
+
1.2400000000000002,
|
114 |
+
1.2400000000000002,
|
115 |
+
1.2500000000000002,
|
116 |
+
1.3000000000000003,
|
117 |
+
1.3100000000000003,
|
118 |
+
1.4600000000000004,
|
119 |
+
1.5100000000000005,
|
120 |
+
1.7000000000000006,
|
121 |
+
1.9300000000000008,
|
122 |
+
2.080000000000001,
|
123 |
+
2.4399999999999933,
|
124 |
+
3.2199999999999767,
|
125 |
+
3.4499999999999718,
|
126 |
+
3.579999999999969,
|
127 |
+
4.669999999999946,
|
128 |
+
4.779999999999943,
|
129 |
+
5.999999999999917,
|
130 |
+
6.009999999999917,
|
131 |
+
6.4199999999999084,
|
132 |
+
6.619999999999904,
|
133 |
+
7.189999999999892,
|
134 |
+
7.3099999999998895,
|
135 |
+
7.339999999999889,
|
136 |
+
7.479999999999886,
|
137 |
+
9.749999999999837,
|
138 |
+
10.919999999999812,
|
139 |
+
11.219999999999805,
|
140 |
+
11.749999999999794,
|
141 |
+
11.979999999999789,
|
142 |
+
13.239999999999762,
|
143 |
+
13.579999999999755,
|
144 |
+
13.669999999999753,
|
145 |
+
13.82999999999975,
|
146 |
+
14.009999999999746,
|
147 |
+
14.679999999999731,
|
148 |
+
14.889999999999727,
|
149 |
+
15.769999999999708,
|
150 |
+
15.769999999999708,
|
151 |
+
15.819999999999707,
|
152 |
+
15.839999999999707,
|
153 |
+
15.919999999999705,
|
154 |
+
16.029999999999703,
|
155 |
+
16.12999999999972,
|
156 |
+
16.44999999999977,
|
157 |
+
16.44999999999977,
|
158 |
+
16.77999999999982,
|
159 |
+
16.83999999999983,
|
160 |
+
16.83999999999983,
|
161 |
+
16.889999999999837
|
162 |
+
],
|
163 |
+
"short_mscale": 1.0,
|
164 |
+
"type": "su"
|
165 |
+
},
|
166 |
+
"rope_theta": 10000.0,
|
167 |
+
"sliding_window": 2047,
|
168 |
+
"tie_word_embeddings": false,
|
169 |
+
"torch_dtype": "bfloat16",
|
170 |
+
"transformers_version": "4.40.2",
|
171 |
+
"use_cache": true,
|
172 |
+
"attention_bias": false,
|
173 |
+
"vocab_size": 32064
|
174 |
+
}
|
175 |
+
|