Upload folder using huggingface_hub
Browse files- config.json +2 -1
- onnx/decoder_model.onnx +1 -1
- onnx/decoder_model_merged.onnx +1 -1
- onnx/decoder_model_merged_quantized.onnx +2 -2
- onnx/decoder_model_quantized.onnx +2 -2
- onnx/decoder_with_past_model.onnx +1 -1
- onnx/decoder_with_past_model_quantized.onnx +2 -2
- quantize_config.json +142 -0
- tokenizer_config.json +7 -0
config.json
CHANGED
@@ -3,6 +3,7 @@
|
|
3 |
"architectures": [
|
4 |
"T5EncoderModel"
|
5 |
],
|
|
|
6 |
"d_ff": 4096,
|
7 |
"d_kv": 64,
|
8 |
"d_model": 1024,
|
@@ -53,7 +54,7 @@
|
|
53 |
"prefix": "translate English to Romanian: "
|
54 |
}
|
55 |
},
|
56 |
-
"transformers_version": "4.
|
57 |
"use_cache": true,
|
58 |
"vocab_size": 32128
|
59 |
}
|
|
|
3 |
"architectures": [
|
4 |
"T5EncoderModel"
|
5 |
],
|
6 |
+
"classifier_dropout": 0.0,
|
7 |
"d_ff": 4096,
|
8 |
"d_kv": 64,
|
9 |
"d_model": 1024,
|
|
|
54 |
"prefix": "translate English to Romanian: "
|
55 |
}
|
56 |
},
|
57 |
+
"transformers_version": "4.34.0.dev0",
|
58 |
"use_cache": true,
|
59 |
"vocab_size": 32128
|
60 |
}
|
onnx/decoder_model.onnx
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1742842287
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:86c479039901444079169bc6eb59dbb2fb63186e7c45bf0c7b3e1e007a99922b
|
3 |
size 1742842287
|
onnx/decoder_model_merged.onnx
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1743425095
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8eddc9dc31011b9615bb7475a5e09605c8334dbe6fbef57f9ccb1d2101fd069e
|
3 |
size 1743425095
|
onnx/decoder_model_merged_quantized.onnx
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:50f43f5a9e52b4ed68c95a1a2686039ccb3a18ea291e48c04a81431894db8544
|
3 |
+
size 439300935
|
onnx/decoder_model_quantized.onnx
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:65fd7d4a9ebe6287bd98b15b40b4adf57a3340abcf549f41d93013c5f154a66e
|
3 |
+
size 438286627
|
onnx/decoder_with_past_model.onnx
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1541451703
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:51eb107fedf43e51c3ffa6e67d0f58f316fee2f156c4c1d6238e0aaff34df597
|
3 |
size 1541451703
|
onnx/decoder_with_past_model_quantized.onnx
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:76554462422691e1f974455b185fb17722a70bcfd1e477176131ba051925ebd8
|
3 |
+
size 387577779
|
quantize_config.json
ADDED
@@ -0,0 +1,142 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"per_channel": true,
|
3 |
+
"reduce_range": true,
|
4 |
+
"per_model_config": {
|
5 |
+
"decoder_model": {
|
6 |
+
"op_types": [
|
7 |
+
"Constant",
|
8 |
+
"Where",
|
9 |
+
"Sqrt",
|
10 |
+
"Neg",
|
11 |
+
"Add",
|
12 |
+
"Min",
|
13 |
+
"Softmax",
|
14 |
+
"ReduceMean",
|
15 |
+
"Identity",
|
16 |
+
"Reshape",
|
17 |
+
"LessOrEqual",
|
18 |
+
"Unsqueeze",
|
19 |
+
"ConstantOfShape",
|
20 |
+
"Relu",
|
21 |
+
"Pow",
|
22 |
+
"Log",
|
23 |
+
"Div",
|
24 |
+
"Shape",
|
25 |
+
"Range",
|
26 |
+
"Expand",
|
27 |
+
"Mul",
|
28 |
+
"Concat",
|
29 |
+
"Gather",
|
30 |
+
"Tile",
|
31 |
+
"Less",
|
32 |
+
"Sub",
|
33 |
+
"MatMul",
|
34 |
+
"Transpose",
|
35 |
+
"Cast"
|
36 |
+
],
|
37 |
+
"weight_type": "QInt8"
|
38 |
+
},
|
39 |
+
"encoder_model": {
|
40 |
+
"op_types": [
|
41 |
+
"Constant",
|
42 |
+
"Where",
|
43 |
+
"Sqrt",
|
44 |
+
"Add",
|
45 |
+
"Min",
|
46 |
+
"Softmax",
|
47 |
+
"ReduceMean",
|
48 |
+
"Reshape",
|
49 |
+
"Abs",
|
50 |
+
"Unsqueeze",
|
51 |
+
"ConstantOfShape",
|
52 |
+
"Relu",
|
53 |
+
"Pow",
|
54 |
+
"Log",
|
55 |
+
"Div",
|
56 |
+
"Shape",
|
57 |
+
"Range",
|
58 |
+
"Mul",
|
59 |
+
"Concat",
|
60 |
+
"Gather",
|
61 |
+
"Less",
|
62 |
+
"Sub",
|
63 |
+
"MatMul",
|
64 |
+
"Transpose",
|
65 |
+
"Cast",
|
66 |
+
"Greater"
|
67 |
+
],
|
68 |
+
"weight_type": "QInt8"
|
69 |
+
},
|
70 |
+
"decoder_with_past_model": {
|
71 |
+
"op_types": [
|
72 |
+
"Constant",
|
73 |
+
"Where",
|
74 |
+
"Sqrt",
|
75 |
+
"Neg",
|
76 |
+
"Add",
|
77 |
+
"Min",
|
78 |
+
"Softmax",
|
79 |
+
"ReduceMean",
|
80 |
+
"Identity",
|
81 |
+
"Reshape",
|
82 |
+
"LessOrEqual",
|
83 |
+
"Unsqueeze",
|
84 |
+
"ConstantOfShape",
|
85 |
+
"Relu",
|
86 |
+
"Pow",
|
87 |
+
"Log",
|
88 |
+
"Div",
|
89 |
+
"Slice",
|
90 |
+
"Shape",
|
91 |
+
"Range",
|
92 |
+
"Expand",
|
93 |
+
"Mul",
|
94 |
+
"Concat",
|
95 |
+
"Gather",
|
96 |
+
"Tile",
|
97 |
+
"Less",
|
98 |
+
"Sub",
|
99 |
+
"MatMul",
|
100 |
+
"Transpose",
|
101 |
+
"Cast"
|
102 |
+
],
|
103 |
+
"weight_type": "QInt8"
|
104 |
+
},
|
105 |
+
"decoder_model_merged": {
|
106 |
+
"op_types": [
|
107 |
+
"Constant",
|
108 |
+
"Where",
|
109 |
+
"Sqrt",
|
110 |
+
"Neg",
|
111 |
+
"Add",
|
112 |
+
"Min",
|
113 |
+
"Softmax",
|
114 |
+
"ReduceMean",
|
115 |
+
"Identity",
|
116 |
+
"Reshape",
|
117 |
+
"LessOrEqual",
|
118 |
+
"Unsqueeze",
|
119 |
+
"ConstantOfShape",
|
120 |
+
"Relu",
|
121 |
+
"Pow",
|
122 |
+
"Log",
|
123 |
+
"Div",
|
124 |
+
"Slice",
|
125 |
+
"Shape",
|
126 |
+
"Range",
|
127 |
+
"Expand",
|
128 |
+
"Mul",
|
129 |
+
"Concat",
|
130 |
+
"Gather",
|
131 |
+
"Tile",
|
132 |
+
"Less",
|
133 |
+
"Sub",
|
134 |
+
"MatMul",
|
135 |
+
"Transpose",
|
136 |
+
"Cast",
|
137 |
+
"If"
|
138 |
+
],
|
139 |
+
"weight_type": "QInt8"
|
140 |
+
}
|
141 |
+
}
|
142 |
+
}
|
tokenizer_config.json
CHANGED
@@ -104,8 +104,15 @@
|
|
104 |
"clean_up_tokenization_spaces": true,
|
105 |
"eos_token": "</s>",
|
106 |
"extra_ids": 100,
|
|
|
107 |
"model_max_length": 512,
|
|
|
108 |
"pad_token": "<pad>",
|
|
|
|
|
|
|
109 |
"tokenizer_class": "T5Tokenizer",
|
|
|
|
|
110 |
"unk_token": "<unk>"
|
111 |
}
|
|
|
104 |
"clean_up_tokenization_spaces": true,
|
105 |
"eos_token": "</s>",
|
106 |
"extra_ids": 100,
|
107 |
+
"max_length": 512,
|
108 |
"model_max_length": 512,
|
109 |
+
"pad_to_multiple_of": null,
|
110 |
"pad_token": "<pad>",
|
111 |
+
"pad_token_type_id": 0,
|
112 |
+
"padding_side": "right",
|
113 |
+
"stride": 0,
|
114 |
"tokenizer_class": "T5Tokenizer",
|
115 |
+
"truncation_side": "right",
|
116 |
+
"truncation_strategy": "longest_first",
|
117 |
"unk_token": "<unk>"
|
118 |
}
|