Upload folder using huggingface_hub
Browse files
qfloat8_text_encoder_2/config.json
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "black-forest-labs/FLUX.1-dev",
|
3 |
+
"architectures": [
|
4 |
+
"T5EncoderModel"
|
5 |
+
],
|
6 |
+
"classifier_dropout": 0.0,
|
7 |
+
"d_ff": 10240,
|
8 |
+
"d_kv": 64,
|
9 |
+
"d_model": 4096,
|
10 |
+
"decoder_start_token_id": 0,
|
11 |
+
"dense_act_fn": "gelu_new",
|
12 |
+
"dropout_rate": 0.1,
|
13 |
+
"eos_token_id": 1,
|
14 |
+
"feed_forward_proj": "gated-gelu",
|
15 |
+
"initializer_factor": 1.0,
|
16 |
+
"is_encoder_decoder": true,
|
17 |
+
"is_gated_act": true,
|
18 |
+
"layer_norm_epsilon": 1e-06,
|
19 |
+
"model_type": "t5",
|
20 |
+
"num_decoder_layers": 24,
|
21 |
+
"num_heads": 64,
|
22 |
+
"num_layers": 24,
|
23 |
+
"output_past": true,
|
24 |
+
"pad_token_id": 0,
|
25 |
+
"relative_attention_max_distance": 128,
|
26 |
+
"relative_attention_num_buckets": 32,
|
27 |
+
"tie_word_embeddings": false,
|
28 |
+
"torch_dtype": "bfloat16",
|
29 |
+
"transformers_version": "4.44.2",
|
30 |
+
"use_cache": true,
|
31 |
+
"vocab_size": 32128
|
32 |
+
}
|
qfloat8_text_encoder_2/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8490f7a22615c20651a63dbe7b4241929826a4de20292dc8e63bfc3c61e3654f
|
3 |
+
size 4896163040
|
qfloat8_text_encoder_2/quanto_qmap.json
ADDED
@@ -0,0 +1,674 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"encoder.block.0.layer.0.SelfAttention.q": {
|
3 |
+
"weights": "qfloat8_e4m3fn",
|
4 |
+
"activations": "none"
|
5 |
+
},
|
6 |
+
"encoder.block.0.layer.0.SelfAttention.k": {
|
7 |
+
"weights": "qfloat8_e4m3fn",
|
8 |
+
"activations": "none"
|
9 |
+
},
|
10 |
+
"encoder.block.0.layer.0.SelfAttention.v": {
|
11 |
+
"weights": "qfloat8_e4m3fn",
|
12 |
+
"activations": "none"
|
13 |
+
},
|
14 |
+
"encoder.block.0.layer.0.SelfAttention.o": {
|
15 |
+
"weights": "qfloat8_e4m3fn",
|
16 |
+
"activations": "none"
|
17 |
+
},
|
18 |
+
"encoder.block.0.layer.1.DenseReluDense.wi_0": {
|
19 |
+
"weights": "qfloat8_e4m3fn",
|
20 |
+
"activations": "none"
|
21 |
+
},
|
22 |
+
"encoder.block.0.layer.1.DenseReluDense.wi_1": {
|
23 |
+
"weights": "qfloat8_e4m3fn",
|
24 |
+
"activations": "none"
|
25 |
+
},
|
26 |
+
"encoder.block.0.layer.1.DenseReluDense.wo": {
|
27 |
+
"weights": "qfloat8_e4m3fn",
|
28 |
+
"activations": "none"
|
29 |
+
},
|
30 |
+
"encoder.block.1.layer.0.SelfAttention.q": {
|
31 |
+
"weights": "qfloat8_e4m3fn",
|
32 |
+
"activations": "none"
|
33 |
+
},
|
34 |
+
"encoder.block.1.layer.0.SelfAttention.k": {
|
35 |
+
"weights": "qfloat8_e4m3fn",
|
36 |
+
"activations": "none"
|
37 |
+
},
|
38 |
+
"encoder.block.1.layer.0.SelfAttention.v": {
|
39 |
+
"weights": "qfloat8_e4m3fn",
|
40 |
+
"activations": "none"
|
41 |
+
},
|
42 |
+
"encoder.block.1.layer.0.SelfAttention.o": {
|
43 |
+
"weights": "qfloat8_e4m3fn",
|
44 |
+
"activations": "none"
|
45 |
+
},
|
46 |
+
"encoder.block.1.layer.1.DenseReluDense.wi_0": {
|
47 |
+
"weights": "qfloat8_e4m3fn",
|
48 |
+
"activations": "none"
|
49 |
+
},
|
50 |
+
"encoder.block.1.layer.1.DenseReluDense.wi_1": {
|
51 |
+
"weights": "qfloat8_e4m3fn",
|
52 |
+
"activations": "none"
|
53 |
+
},
|
54 |
+
"encoder.block.1.layer.1.DenseReluDense.wo": {
|
55 |
+
"weights": "qfloat8_e4m3fn",
|
56 |
+
"activations": "none"
|
57 |
+
},
|
58 |
+
"encoder.block.2.layer.0.SelfAttention.q": {
|
59 |
+
"weights": "qfloat8_e4m3fn",
|
60 |
+
"activations": "none"
|
61 |
+
},
|
62 |
+
"encoder.block.2.layer.0.SelfAttention.k": {
|
63 |
+
"weights": "qfloat8_e4m3fn",
|
64 |
+
"activations": "none"
|
65 |
+
},
|
66 |
+
"encoder.block.2.layer.0.SelfAttention.v": {
|
67 |
+
"weights": "qfloat8_e4m3fn",
|
68 |
+
"activations": "none"
|
69 |
+
},
|
70 |
+
"encoder.block.2.layer.0.SelfAttention.o": {
|
71 |
+
"weights": "qfloat8_e4m3fn",
|
72 |
+
"activations": "none"
|
73 |
+
},
|
74 |
+
"encoder.block.2.layer.1.DenseReluDense.wi_0": {
|
75 |
+
"weights": "qfloat8_e4m3fn",
|
76 |
+
"activations": "none"
|
77 |
+
},
|
78 |
+
"encoder.block.2.layer.1.DenseReluDense.wi_1": {
|
79 |
+
"weights": "qfloat8_e4m3fn",
|
80 |
+
"activations": "none"
|
81 |
+
},
|
82 |
+
"encoder.block.2.layer.1.DenseReluDense.wo": {
|
83 |
+
"weights": "qfloat8_e4m3fn",
|
84 |
+
"activations": "none"
|
85 |
+
},
|
86 |
+
"encoder.block.3.layer.0.SelfAttention.q": {
|
87 |
+
"weights": "qfloat8_e4m3fn",
|
88 |
+
"activations": "none"
|
89 |
+
},
|
90 |
+
"encoder.block.3.layer.0.SelfAttention.k": {
|
91 |
+
"weights": "qfloat8_e4m3fn",
|
92 |
+
"activations": "none"
|
93 |
+
},
|
94 |
+
"encoder.block.3.layer.0.SelfAttention.v": {
|
95 |
+
"weights": "qfloat8_e4m3fn",
|
96 |
+
"activations": "none"
|
97 |
+
},
|
98 |
+
"encoder.block.3.layer.0.SelfAttention.o": {
|
99 |
+
"weights": "qfloat8_e4m3fn",
|
100 |
+
"activations": "none"
|
101 |
+
},
|
102 |
+
"encoder.block.3.layer.1.DenseReluDense.wi_0": {
|
103 |
+
"weights": "qfloat8_e4m3fn",
|
104 |
+
"activations": "none"
|
105 |
+
},
|
106 |
+
"encoder.block.3.layer.1.DenseReluDense.wi_1": {
|
107 |
+
"weights": "qfloat8_e4m3fn",
|
108 |
+
"activations": "none"
|
109 |
+
},
|
110 |
+
"encoder.block.3.layer.1.DenseReluDense.wo": {
|
111 |
+
"weights": "qfloat8_e4m3fn",
|
112 |
+
"activations": "none"
|
113 |
+
},
|
114 |
+
"encoder.block.4.layer.0.SelfAttention.q": {
|
115 |
+
"weights": "qfloat8_e4m3fn",
|
116 |
+
"activations": "none"
|
117 |
+
},
|
118 |
+
"encoder.block.4.layer.0.SelfAttention.k": {
|
119 |
+
"weights": "qfloat8_e4m3fn",
|
120 |
+
"activations": "none"
|
121 |
+
},
|
122 |
+
"encoder.block.4.layer.0.SelfAttention.v": {
|
123 |
+
"weights": "qfloat8_e4m3fn",
|
124 |
+
"activations": "none"
|
125 |
+
},
|
126 |
+
"encoder.block.4.layer.0.SelfAttention.o": {
|
127 |
+
"weights": "qfloat8_e4m3fn",
|
128 |
+
"activations": "none"
|
129 |
+
},
|
130 |
+
"encoder.block.4.layer.1.DenseReluDense.wi_0": {
|
131 |
+
"weights": "qfloat8_e4m3fn",
|
132 |
+
"activations": "none"
|
133 |
+
},
|
134 |
+
"encoder.block.4.layer.1.DenseReluDense.wi_1": {
|
135 |
+
"weights": "qfloat8_e4m3fn",
|
136 |
+
"activations": "none"
|
137 |
+
},
|
138 |
+
"encoder.block.4.layer.1.DenseReluDense.wo": {
|
139 |
+
"weights": "qfloat8_e4m3fn",
|
140 |
+
"activations": "none"
|
141 |
+
},
|
142 |
+
"encoder.block.5.layer.0.SelfAttention.q": {
|
143 |
+
"weights": "qfloat8_e4m3fn",
|
144 |
+
"activations": "none"
|
145 |
+
},
|
146 |
+
"encoder.block.5.layer.0.SelfAttention.k": {
|
147 |
+
"weights": "qfloat8_e4m3fn",
|
148 |
+
"activations": "none"
|
149 |
+
},
|
150 |
+
"encoder.block.5.layer.0.SelfAttention.v": {
|
151 |
+
"weights": "qfloat8_e4m3fn",
|
152 |
+
"activations": "none"
|
153 |
+
},
|
154 |
+
"encoder.block.5.layer.0.SelfAttention.o": {
|
155 |
+
"weights": "qfloat8_e4m3fn",
|
156 |
+
"activations": "none"
|
157 |
+
},
|
158 |
+
"encoder.block.5.layer.1.DenseReluDense.wi_0": {
|
159 |
+
"weights": "qfloat8_e4m3fn",
|
160 |
+
"activations": "none"
|
161 |
+
},
|
162 |
+
"encoder.block.5.layer.1.DenseReluDense.wi_1": {
|
163 |
+
"weights": "qfloat8_e4m3fn",
|
164 |
+
"activations": "none"
|
165 |
+
},
|
166 |
+
"encoder.block.5.layer.1.DenseReluDense.wo": {
|
167 |
+
"weights": "qfloat8_e4m3fn",
|
168 |
+
"activations": "none"
|
169 |
+
},
|
170 |
+
"encoder.block.6.layer.0.SelfAttention.q": {
|
171 |
+
"weights": "qfloat8_e4m3fn",
|
172 |
+
"activations": "none"
|
173 |
+
},
|
174 |
+
"encoder.block.6.layer.0.SelfAttention.k": {
|
175 |
+
"weights": "qfloat8_e4m3fn",
|
176 |
+
"activations": "none"
|
177 |
+
},
|
178 |
+
"encoder.block.6.layer.0.SelfAttention.v": {
|
179 |
+
"weights": "qfloat8_e4m3fn",
|
180 |
+
"activations": "none"
|
181 |
+
},
|
182 |
+
"encoder.block.6.layer.0.SelfAttention.o": {
|
183 |
+
"weights": "qfloat8_e4m3fn",
|
184 |
+
"activations": "none"
|
185 |
+
},
|
186 |
+
"encoder.block.6.layer.1.DenseReluDense.wi_0": {
|
187 |
+
"weights": "qfloat8_e4m3fn",
|
188 |
+
"activations": "none"
|
189 |
+
},
|
190 |
+
"encoder.block.6.layer.1.DenseReluDense.wi_1": {
|
191 |
+
"weights": "qfloat8_e4m3fn",
|
192 |
+
"activations": "none"
|
193 |
+
},
|
194 |
+
"encoder.block.6.layer.1.DenseReluDense.wo": {
|
195 |
+
"weights": "qfloat8_e4m3fn",
|
196 |
+
"activations": "none"
|
197 |
+
},
|
198 |
+
"encoder.block.7.layer.0.SelfAttention.q": {
|
199 |
+
"weights": "qfloat8_e4m3fn",
|
200 |
+
"activations": "none"
|
201 |
+
},
|
202 |
+
"encoder.block.7.layer.0.SelfAttention.k": {
|
203 |
+
"weights": "qfloat8_e4m3fn",
|
204 |
+
"activations": "none"
|
205 |
+
},
|
206 |
+
"encoder.block.7.layer.0.SelfAttention.v": {
|
207 |
+
"weights": "qfloat8_e4m3fn",
|
208 |
+
"activations": "none"
|
209 |
+
},
|
210 |
+
"encoder.block.7.layer.0.SelfAttention.o": {
|
211 |
+
"weights": "qfloat8_e4m3fn",
|
212 |
+
"activations": "none"
|
213 |
+
},
|
214 |
+
"encoder.block.7.layer.1.DenseReluDense.wi_0": {
|
215 |
+
"weights": "qfloat8_e4m3fn",
|
216 |
+
"activations": "none"
|
217 |
+
},
|
218 |
+
"encoder.block.7.layer.1.DenseReluDense.wi_1": {
|
219 |
+
"weights": "qfloat8_e4m3fn",
|
220 |
+
"activations": "none"
|
221 |
+
},
|
222 |
+
"encoder.block.7.layer.1.DenseReluDense.wo": {
|
223 |
+
"weights": "qfloat8_e4m3fn",
|
224 |
+
"activations": "none"
|
225 |
+
},
|
226 |
+
"encoder.block.8.layer.0.SelfAttention.q": {
|
227 |
+
"weights": "qfloat8_e4m3fn",
|
228 |
+
"activations": "none"
|
229 |
+
},
|
230 |
+
"encoder.block.8.layer.0.SelfAttention.k": {
|
231 |
+
"weights": "qfloat8_e4m3fn",
|
232 |
+
"activations": "none"
|
233 |
+
},
|
234 |
+
"encoder.block.8.layer.0.SelfAttention.v": {
|
235 |
+
"weights": "qfloat8_e4m3fn",
|
236 |
+
"activations": "none"
|
237 |
+
},
|
238 |
+
"encoder.block.8.layer.0.SelfAttention.o": {
|
239 |
+
"weights": "qfloat8_e4m3fn",
|
240 |
+
"activations": "none"
|
241 |
+
},
|
242 |
+
"encoder.block.8.layer.1.DenseReluDense.wi_0": {
|
243 |
+
"weights": "qfloat8_e4m3fn",
|
244 |
+
"activations": "none"
|
245 |
+
},
|
246 |
+
"encoder.block.8.layer.1.DenseReluDense.wi_1": {
|
247 |
+
"weights": "qfloat8_e4m3fn",
|
248 |
+
"activations": "none"
|
249 |
+
},
|
250 |
+
"encoder.block.8.layer.1.DenseReluDense.wo": {
|
251 |
+
"weights": "qfloat8_e4m3fn",
|
252 |
+
"activations": "none"
|
253 |
+
},
|
254 |
+
"encoder.block.9.layer.0.SelfAttention.q": {
|
255 |
+
"weights": "qfloat8_e4m3fn",
|
256 |
+
"activations": "none"
|
257 |
+
},
|
258 |
+
"encoder.block.9.layer.0.SelfAttention.k": {
|
259 |
+
"weights": "qfloat8_e4m3fn",
|
260 |
+
"activations": "none"
|
261 |
+
},
|
262 |
+
"encoder.block.9.layer.0.SelfAttention.v": {
|
263 |
+
"weights": "qfloat8_e4m3fn",
|
264 |
+
"activations": "none"
|
265 |
+
},
|
266 |
+
"encoder.block.9.layer.0.SelfAttention.o": {
|
267 |
+
"weights": "qfloat8_e4m3fn",
|
268 |
+
"activations": "none"
|
269 |
+
},
|
270 |
+
"encoder.block.9.layer.1.DenseReluDense.wi_0": {
|
271 |
+
"weights": "qfloat8_e4m3fn",
|
272 |
+
"activations": "none"
|
273 |
+
},
|
274 |
+
"encoder.block.9.layer.1.DenseReluDense.wi_1": {
|
275 |
+
"weights": "qfloat8_e4m3fn",
|
276 |
+
"activations": "none"
|
277 |
+
},
|
278 |
+
"encoder.block.9.layer.1.DenseReluDense.wo": {
|
279 |
+
"weights": "qfloat8_e4m3fn",
|
280 |
+
"activations": "none"
|
281 |
+
},
|
282 |
+
"encoder.block.10.layer.0.SelfAttention.q": {
|
283 |
+
"weights": "qfloat8_e4m3fn",
|
284 |
+
"activations": "none"
|
285 |
+
},
|
286 |
+
"encoder.block.10.layer.0.SelfAttention.k": {
|
287 |
+
"weights": "qfloat8_e4m3fn",
|
288 |
+
"activations": "none"
|
289 |
+
},
|
290 |
+
"encoder.block.10.layer.0.SelfAttention.v": {
|
291 |
+
"weights": "qfloat8_e4m3fn",
|
292 |
+
"activations": "none"
|
293 |
+
},
|
294 |
+
"encoder.block.10.layer.0.SelfAttention.o": {
|
295 |
+
"weights": "qfloat8_e4m3fn",
|
296 |
+
"activations": "none"
|
297 |
+
},
|
298 |
+
"encoder.block.10.layer.1.DenseReluDense.wi_0": {
|
299 |
+
"weights": "qfloat8_e4m3fn",
|
300 |
+
"activations": "none"
|
301 |
+
},
|
302 |
+
"encoder.block.10.layer.1.DenseReluDense.wi_1": {
|
303 |
+
"weights": "qfloat8_e4m3fn",
|
304 |
+
"activations": "none"
|
305 |
+
},
|
306 |
+
"encoder.block.10.layer.1.DenseReluDense.wo": {
|
307 |
+
"weights": "qfloat8_e4m3fn",
|
308 |
+
"activations": "none"
|
309 |
+
},
|
310 |
+
"encoder.block.11.layer.0.SelfAttention.q": {
|
311 |
+
"weights": "qfloat8_e4m3fn",
|
312 |
+
"activations": "none"
|
313 |
+
},
|
314 |
+
"encoder.block.11.layer.0.SelfAttention.k": {
|
315 |
+
"weights": "qfloat8_e4m3fn",
|
316 |
+
"activations": "none"
|
317 |
+
},
|
318 |
+
"encoder.block.11.layer.0.SelfAttention.v": {
|
319 |
+
"weights": "qfloat8_e4m3fn",
|
320 |
+
"activations": "none"
|
321 |
+
},
|
322 |
+
"encoder.block.11.layer.0.SelfAttention.o": {
|
323 |
+
"weights": "qfloat8_e4m3fn",
|
324 |
+
"activations": "none"
|
325 |
+
},
|
326 |
+
"encoder.block.11.layer.1.DenseReluDense.wi_0": {
|
327 |
+
"weights": "qfloat8_e4m3fn",
|
328 |
+
"activations": "none"
|
329 |
+
},
|
330 |
+
"encoder.block.11.layer.1.DenseReluDense.wi_1": {
|
331 |
+
"weights": "qfloat8_e4m3fn",
|
332 |
+
"activations": "none"
|
333 |
+
},
|
334 |
+
"encoder.block.11.layer.1.DenseReluDense.wo": {
|
335 |
+
"weights": "qfloat8_e4m3fn",
|
336 |
+
"activations": "none"
|
337 |
+
},
|
338 |
+
"encoder.block.12.layer.0.SelfAttention.q": {
|
339 |
+
"weights": "qfloat8_e4m3fn",
|
340 |
+
"activations": "none"
|
341 |
+
},
|
342 |
+
"encoder.block.12.layer.0.SelfAttention.k": {
|
343 |
+
"weights": "qfloat8_e4m3fn",
|
344 |
+
"activations": "none"
|
345 |
+
},
|
346 |
+
"encoder.block.12.layer.0.SelfAttention.v": {
|
347 |
+
"weights": "qfloat8_e4m3fn",
|
348 |
+
"activations": "none"
|
349 |
+
},
|
350 |
+
"encoder.block.12.layer.0.SelfAttention.o": {
|
351 |
+
"weights": "qfloat8_e4m3fn",
|
352 |
+
"activations": "none"
|
353 |
+
},
|
354 |
+
"encoder.block.12.layer.1.DenseReluDense.wi_0": {
|
355 |
+
"weights": "qfloat8_e4m3fn",
|
356 |
+
"activations": "none"
|
357 |
+
},
|
358 |
+
"encoder.block.12.layer.1.DenseReluDense.wi_1": {
|
359 |
+
"weights": "qfloat8_e4m3fn",
|
360 |
+
"activations": "none"
|
361 |
+
},
|
362 |
+
"encoder.block.12.layer.1.DenseReluDense.wo": {
|
363 |
+
"weights": "qfloat8_e4m3fn",
|
364 |
+
"activations": "none"
|
365 |
+
},
|
366 |
+
"encoder.block.13.layer.0.SelfAttention.q": {
|
367 |
+
"weights": "qfloat8_e4m3fn",
|
368 |
+
"activations": "none"
|
369 |
+
},
|
370 |
+
"encoder.block.13.layer.0.SelfAttention.k": {
|
371 |
+
"weights": "qfloat8_e4m3fn",
|
372 |
+
"activations": "none"
|
373 |
+
},
|
374 |
+
"encoder.block.13.layer.0.SelfAttention.v": {
|
375 |
+
"weights": "qfloat8_e4m3fn",
|
376 |
+
"activations": "none"
|
377 |
+
},
|
378 |
+
"encoder.block.13.layer.0.SelfAttention.o": {
|
379 |
+
"weights": "qfloat8_e4m3fn",
|
380 |
+
"activations": "none"
|
381 |
+
},
|
382 |
+
"encoder.block.13.layer.1.DenseReluDense.wi_0": {
|
383 |
+
"weights": "qfloat8_e4m3fn",
|
384 |
+
"activations": "none"
|
385 |
+
},
|
386 |
+
"encoder.block.13.layer.1.DenseReluDense.wi_1": {
|
387 |
+
"weights": "qfloat8_e4m3fn",
|
388 |
+
"activations": "none"
|
389 |
+
},
|
390 |
+
"encoder.block.13.layer.1.DenseReluDense.wo": {
|
391 |
+
"weights": "qfloat8_e4m3fn",
|
392 |
+
"activations": "none"
|
393 |
+
},
|
394 |
+
"encoder.block.14.layer.0.SelfAttention.q": {
|
395 |
+
"weights": "qfloat8_e4m3fn",
|
396 |
+
"activations": "none"
|
397 |
+
},
|
398 |
+
"encoder.block.14.layer.0.SelfAttention.k": {
|
399 |
+
"weights": "qfloat8_e4m3fn",
|
400 |
+
"activations": "none"
|
401 |
+
},
|
402 |
+
"encoder.block.14.layer.0.SelfAttention.v": {
|
403 |
+
"weights": "qfloat8_e4m3fn",
|
404 |
+
"activations": "none"
|
405 |
+
},
|
406 |
+
"encoder.block.14.layer.0.SelfAttention.o": {
|
407 |
+
"weights": "qfloat8_e4m3fn",
|
408 |
+
"activations": "none"
|
409 |
+
},
|
410 |
+
"encoder.block.14.layer.1.DenseReluDense.wi_0": {
|
411 |
+
"weights": "qfloat8_e4m3fn",
|
412 |
+
"activations": "none"
|
413 |
+
},
|
414 |
+
"encoder.block.14.layer.1.DenseReluDense.wi_1": {
|
415 |
+
"weights": "qfloat8_e4m3fn",
|
416 |
+
"activations": "none"
|
417 |
+
},
|
418 |
+
"encoder.block.14.layer.1.DenseReluDense.wo": {
|
419 |
+
"weights": "qfloat8_e4m3fn",
|
420 |
+
"activations": "none"
|
421 |
+
},
|
422 |
+
"encoder.block.15.layer.0.SelfAttention.q": {
|
423 |
+
"weights": "qfloat8_e4m3fn",
|
424 |
+
"activations": "none"
|
425 |
+
},
|
426 |
+
"encoder.block.15.layer.0.SelfAttention.k": {
|
427 |
+
"weights": "qfloat8_e4m3fn",
|
428 |
+
"activations": "none"
|
429 |
+
},
|
430 |
+
"encoder.block.15.layer.0.SelfAttention.v": {
|
431 |
+
"weights": "qfloat8_e4m3fn",
|
432 |
+
"activations": "none"
|
433 |
+
},
|
434 |
+
"encoder.block.15.layer.0.SelfAttention.o": {
|
435 |
+
"weights": "qfloat8_e4m3fn",
|
436 |
+
"activations": "none"
|
437 |
+
},
|
438 |
+
"encoder.block.15.layer.1.DenseReluDense.wi_0": {
|
439 |
+
"weights": "qfloat8_e4m3fn",
|
440 |
+
"activations": "none"
|
441 |
+
},
|
442 |
+
"encoder.block.15.layer.1.DenseReluDense.wi_1": {
|
443 |
+
"weights": "qfloat8_e4m3fn",
|
444 |
+
"activations": "none"
|
445 |
+
},
|
446 |
+
"encoder.block.15.layer.1.DenseReluDense.wo": {
|
447 |
+
"weights": "qfloat8_e4m3fn",
|
448 |
+
"activations": "none"
|
449 |
+
},
|
450 |
+
"encoder.block.16.layer.0.SelfAttention.q": {
|
451 |
+
"weights": "qfloat8_e4m3fn",
|
452 |
+
"activations": "none"
|
453 |
+
},
|
454 |
+
"encoder.block.16.layer.0.SelfAttention.k": {
|
455 |
+
"weights": "qfloat8_e4m3fn",
|
456 |
+
"activations": "none"
|
457 |
+
},
|
458 |
+
"encoder.block.16.layer.0.SelfAttention.v": {
|
459 |
+
"weights": "qfloat8_e4m3fn",
|
460 |
+
"activations": "none"
|
461 |
+
},
|
462 |
+
"encoder.block.16.layer.0.SelfAttention.o": {
|
463 |
+
"weights": "qfloat8_e4m3fn",
|
464 |
+
"activations": "none"
|
465 |
+
},
|
466 |
+
"encoder.block.16.layer.1.DenseReluDense.wi_0": {
|
467 |
+
"weights": "qfloat8_e4m3fn",
|
468 |
+
"activations": "none"
|
469 |
+
},
|
470 |
+
"encoder.block.16.layer.1.DenseReluDense.wi_1": {
|
471 |
+
"weights": "qfloat8_e4m3fn",
|
472 |
+
"activations": "none"
|
473 |
+
},
|
474 |
+
"encoder.block.16.layer.1.DenseReluDense.wo": {
|
475 |
+
"weights": "qfloat8_e4m3fn",
|
476 |
+
"activations": "none"
|
477 |
+
},
|
478 |
+
"encoder.block.17.layer.0.SelfAttention.q": {
|
479 |
+
"weights": "qfloat8_e4m3fn",
|
480 |
+
"activations": "none"
|
481 |
+
},
|
482 |
+
"encoder.block.17.layer.0.SelfAttention.k": {
|
483 |
+
"weights": "qfloat8_e4m3fn",
|
484 |
+
"activations": "none"
|
485 |
+
},
|
486 |
+
"encoder.block.17.layer.0.SelfAttention.v": {
|
487 |
+
"weights": "qfloat8_e4m3fn",
|
488 |
+
"activations": "none"
|
489 |
+
},
|
490 |
+
"encoder.block.17.layer.0.SelfAttention.o": {
|
491 |
+
"weights": "qfloat8_e4m3fn",
|
492 |
+
"activations": "none"
|
493 |
+
},
|
494 |
+
"encoder.block.17.layer.1.DenseReluDense.wi_0": {
|
495 |
+
"weights": "qfloat8_e4m3fn",
|
496 |
+
"activations": "none"
|
497 |
+
},
|
498 |
+
"encoder.block.17.layer.1.DenseReluDense.wi_1": {
|
499 |
+
"weights": "qfloat8_e4m3fn",
|
500 |
+
"activations": "none"
|
501 |
+
},
|
502 |
+
"encoder.block.17.layer.1.DenseReluDense.wo": {
|
503 |
+
"weights": "qfloat8_e4m3fn",
|
504 |
+
"activations": "none"
|
505 |
+
},
|
506 |
+
"encoder.block.18.layer.0.SelfAttention.q": {
|
507 |
+
"weights": "qfloat8_e4m3fn",
|
508 |
+
"activations": "none"
|
509 |
+
},
|
510 |
+
"encoder.block.18.layer.0.SelfAttention.k": {
|
511 |
+
"weights": "qfloat8_e4m3fn",
|
512 |
+
"activations": "none"
|
513 |
+
},
|
514 |
+
"encoder.block.18.layer.0.SelfAttention.v": {
|
515 |
+
"weights": "qfloat8_e4m3fn",
|
516 |
+
"activations": "none"
|
517 |
+
},
|
518 |
+
"encoder.block.18.layer.0.SelfAttention.o": {
|
519 |
+
"weights": "qfloat8_e4m3fn",
|
520 |
+
"activations": "none"
|
521 |
+
},
|
522 |
+
"encoder.block.18.layer.1.DenseReluDense.wi_0": {
|
523 |
+
"weights": "qfloat8_e4m3fn",
|
524 |
+
"activations": "none"
|
525 |
+
},
|
526 |
+
"encoder.block.18.layer.1.DenseReluDense.wi_1": {
|
527 |
+
"weights": "qfloat8_e4m3fn",
|
528 |
+
"activations": "none"
|
529 |
+
},
|
530 |
+
"encoder.block.18.layer.1.DenseReluDense.wo": {
|
531 |
+
"weights": "qfloat8_e4m3fn",
|
532 |
+
"activations": "none"
|
533 |
+
},
|
534 |
+
"encoder.block.19.layer.0.SelfAttention.q": {
|
535 |
+
"weights": "qfloat8_e4m3fn",
|
536 |
+
"activations": "none"
|
537 |
+
},
|
538 |
+
"encoder.block.19.layer.0.SelfAttention.k": {
|
539 |
+
"weights": "qfloat8_e4m3fn",
|
540 |
+
"activations": "none"
|
541 |
+
},
|
542 |
+
"encoder.block.19.layer.0.SelfAttention.v": {
|
543 |
+
"weights": "qfloat8_e4m3fn",
|
544 |
+
"activations": "none"
|
545 |
+
},
|
546 |
+
"encoder.block.19.layer.0.SelfAttention.o": {
|
547 |
+
"weights": "qfloat8_e4m3fn",
|
548 |
+
"activations": "none"
|
549 |
+
},
|
550 |
+
"encoder.block.19.layer.1.DenseReluDense.wi_0": {
|
551 |
+
"weights": "qfloat8_e4m3fn",
|
552 |
+
"activations": "none"
|
553 |
+
},
|
554 |
+
"encoder.block.19.layer.1.DenseReluDense.wi_1": {
|
555 |
+
"weights": "qfloat8_e4m3fn",
|
556 |
+
"activations": "none"
|
557 |
+
},
|
558 |
+
"encoder.block.19.layer.1.DenseReluDense.wo": {
|
559 |
+
"weights": "qfloat8_e4m3fn",
|
560 |
+
"activations": "none"
|
561 |
+
},
|
562 |
+
"encoder.block.20.layer.0.SelfAttention.q": {
|
563 |
+
"weights": "qfloat8_e4m3fn",
|
564 |
+
"activations": "none"
|
565 |
+
},
|
566 |
+
"encoder.block.20.layer.0.SelfAttention.k": {
|
567 |
+
"weights": "qfloat8_e4m3fn",
|
568 |
+
"activations": "none"
|
569 |
+
},
|
570 |
+
"encoder.block.20.layer.0.SelfAttention.v": {
|
571 |
+
"weights": "qfloat8_e4m3fn",
|
572 |
+
"activations": "none"
|
573 |
+
},
|
574 |
+
"encoder.block.20.layer.0.SelfAttention.o": {
|
575 |
+
"weights": "qfloat8_e4m3fn",
|
576 |
+
"activations": "none"
|
577 |
+
},
|
578 |
+
"encoder.block.20.layer.1.DenseReluDense.wi_0": {
|
579 |
+
"weights": "qfloat8_e4m3fn",
|
580 |
+
"activations": "none"
|
581 |
+
},
|
582 |
+
"encoder.block.20.layer.1.DenseReluDense.wi_1": {
|
583 |
+
"weights": "qfloat8_e4m3fn",
|
584 |
+
"activations": "none"
|
585 |
+
},
|
586 |
+
"encoder.block.20.layer.1.DenseReluDense.wo": {
|
587 |
+
"weights": "qfloat8_e4m3fn",
|
588 |
+
"activations": "none"
|
589 |
+
},
|
590 |
+
"encoder.block.21.layer.0.SelfAttention.q": {
|
591 |
+
"weights": "qfloat8_e4m3fn",
|
592 |
+
"activations": "none"
|
593 |
+
},
|
594 |
+
"encoder.block.21.layer.0.SelfAttention.k": {
|
595 |
+
"weights": "qfloat8_e4m3fn",
|
596 |
+
"activations": "none"
|
597 |
+
},
|
598 |
+
"encoder.block.21.layer.0.SelfAttention.v": {
|
599 |
+
"weights": "qfloat8_e4m3fn",
|
600 |
+
"activations": "none"
|
601 |
+
},
|
602 |
+
"encoder.block.21.layer.0.SelfAttention.o": {
|
603 |
+
"weights": "qfloat8_e4m3fn",
|
604 |
+
"activations": "none"
|
605 |
+
},
|
606 |
+
"encoder.block.21.layer.1.DenseReluDense.wi_0": {
|
607 |
+
"weights": "qfloat8_e4m3fn",
|
608 |
+
"activations": "none"
|
609 |
+
},
|
610 |
+
"encoder.block.21.layer.1.DenseReluDense.wi_1": {
|
611 |
+
"weights": "qfloat8_e4m3fn",
|
612 |
+
"activations": "none"
|
613 |
+
},
|
614 |
+
"encoder.block.21.layer.1.DenseReluDense.wo": {
|
615 |
+
"weights": "qfloat8_e4m3fn",
|
616 |
+
"activations": "none"
|
617 |
+
},
|
618 |
+
"encoder.block.22.layer.0.SelfAttention.q": {
|
619 |
+
"weights": "qfloat8_e4m3fn",
|
620 |
+
"activations": "none"
|
621 |
+
},
|
622 |
+
"encoder.block.22.layer.0.SelfAttention.k": {
|
623 |
+
"weights": "qfloat8_e4m3fn",
|
624 |
+
"activations": "none"
|
625 |
+
},
|
626 |
+
"encoder.block.22.layer.0.SelfAttention.v": {
|
627 |
+
"weights": "qfloat8_e4m3fn",
|
628 |
+
"activations": "none"
|
629 |
+
},
|
630 |
+
"encoder.block.22.layer.0.SelfAttention.o": {
|
631 |
+
"weights": "qfloat8_e4m3fn",
|
632 |
+
"activations": "none"
|
633 |
+
},
|
634 |
+
"encoder.block.22.layer.1.DenseReluDense.wi_0": {
|
635 |
+
"weights": "qfloat8_e4m3fn",
|
636 |
+
"activations": "none"
|
637 |
+
},
|
638 |
+
"encoder.block.22.layer.1.DenseReluDense.wi_1": {
|
639 |
+
"weights": "qfloat8_e4m3fn",
|
640 |
+
"activations": "none"
|
641 |
+
},
|
642 |
+
"encoder.block.22.layer.1.DenseReluDense.wo": {
|
643 |
+
"weights": "qfloat8_e4m3fn",
|
644 |
+
"activations": "none"
|
645 |
+
},
|
646 |
+
"encoder.block.23.layer.0.SelfAttention.q": {
|
647 |
+
"weights": "qfloat8_e4m3fn",
|
648 |
+
"activations": "none"
|
649 |
+
},
|
650 |
+
"encoder.block.23.layer.0.SelfAttention.k": {
|
651 |
+
"weights": "qfloat8_e4m3fn",
|
652 |
+
"activations": "none"
|
653 |
+
},
|
654 |
+
"encoder.block.23.layer.0.SelfAttention.v": {
|
655 |
+
"weights": "qfloat8_e4m3fn",
|
656 |
+
"activations": "none"
|
657 |
+
},
|
658 |
+
"encoder.block.23.layer.0.SelfAttention.o": {
|
659 |
+
"weights": "qfloat8_e4m3fn",
|
660 |
+
"activations": "none"
|
661 |
+
},
|
662 |
+
"encoder.block.23.layer.1.DenseReluDense.wi_0": {
|
663 |
+
"weights": "qfloat8_e4m3fn",
|
664 |
+
"activations": "none"
|
665 |
+
},
|
666 |
+
"encoder.block.23.layer.1.DenseReluDense.wi_1": {
|
667 |
+
"weights": "qfloat8_e4m3fn",
|
668 |
+
"activations": "none"
|
669 |
+
},
|
670 |
+
"encoder.block.23.layer.1.DenseReluDense.wo": {
|
671 |
+
"weights": "qfloat8_e4m3fn",
|
672 |
+
"activations": "none"
|
673 |
+
}
|
674 |
+
}
|