Hugging Face's logo

catplusplus
/
Qwen3-Embedding-4B-NVFP4

Safetensors
qwen3
compressed-tensors
Model card Files
xet
 Community
catplusplus commited on
Commit
699806b
·
verified ·
1 Parent(s): 489a769

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. .gitattributes +1 -0
  2. chat_template.jinja +54 -0
  3. config.json +267 -0
  4. model.safetensors +3 -0
  5. recipe.yaml +7 -0
  6. tokenizer.json +3 -0
  7. tokenizer_config.json +14 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,267 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3Model"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2560,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 9728,
15
+ "layer_types": [
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention"
52
+ ],
53
+ "max_position_embeddings": 40960,
54
+ "max_window_layers": 36,
55
+ "model_type": "qwen3",
56
+ "num_attention_heads": 32,
57
+ "num_hidden_layers": 36,
58
+ "num_key_value_heads": 8,
59
+ "pad_token_id": null,
60
+ "quantization_config": {
61
+ "config_groups": {
62
+ "group_0": {
63
+ "format": "nvfp4-pack-quantized",
64
+ "input_activations": {
65
+ "actorder": null,
66
+ "block_structure": null,
67
+ "dynamic": "local",
68
+ "group_size": 16,
69
+ "num_bits": 4,
70
+ "observer": "static_minmax",
71
+ "observer_kwargs": {},
72
+ "scale_dtype": "torch.float8_e4m3fn",
73
+ "strategy": "tensor_group",
74
+ "symmetric": true,
75
+ "type": "float",
76
+ "zp_dtype": null
77
+ },
78
+ "output_activations": null,
79
+ "targets": [
80
+ "re:.*mlp.gate_proj$",
81
+ "re:.*mlp.up_proj$",
82
+ "re:.*mlp.down_proj$"
83
+ ],
84
+ "weights": {
85
+ "actorder": null,
86
+ "block_structure": null,
87
+ "dynamic": false,
88
+ "group_size": 16,
89
+ "num_bits": 4,
90
+ "observer": "memoryless_minmax",
91
+ "observer_kwargs": {},
92
+ "scale_dtype": "torch.float8_e4m3fn",
93
+ "strategy": "tensor_group",
94
+ "symmetric": true,
95
+ "type": "float",
96
+ "zp_dtype": null
97
+ }
98
+ }
99
+ },
100
+ "format": "nvfp4-pack-quantized",
101
+ "global_compression_ratio": null,
102
+ "ignore": [
103
+ "re:.*self_attn.*",
104
+ "layers.0.self_attn.q_proj",
105
+ "layers.0.self_attn.k_proj",
106
+ "layers.0.self_attn.v_proj",
107
+ "layers.0.self_attn.o_proj",
108
+ "layers.1.self_attn.q_proj",
109
+ "layers.1.self_attn.k_proj",
110
+ "layers.1.self_attn.v_proj",
111
+ "layers.1.self_attn.o_proj",
112
+ "layers.2.self_attn.q_proj",
113
+ "layers.2.self_attn.k_proj",
114
+ "layers.2.self_attn.v_proj",
115
+ "layers.2.self_attn.o_proj",
116
+ "layers.3.self_attn.q_proj",
117
+ "layers.3.self_attn.k_proj",
118
+ "layers.3.self_attn.v_proj",
119
+ "layers.3.self_attn.o_proj",
120
+ "layers.4.self_attn.q_proj",
121
+ "layers.4.self_attn.k_proj",
122
+ "layers.4.self_attn.v_proj",
123
+ "layers.4.self_attn.o_proj",
124
+ "layers.5.self_attn.q_proj",
125
+ "layers.5.self_attn.k_proj",
126
+ "layers.5.self_attn.v_proj",
127
+ "layers.5.self_attn.o_proj",
128
+ "layers.6.self_attn.q_proj",
129
+ "layers.6.self_attn.k_proj",
130
+ "layers.6.self_attn.v_proj",
131
+ "layers.6.self_attn.o_proj",
132
+ "layers.7.self_attn.q_proj",
133
+ "layers.7.self_attn.k_proj",
134
+ "layers.7.self_attn.v_proj",
135
+ "layers.7.self_attn.o_proj",
136
+ "layers.8.self_attn.q_proj",
137
+ "layers.8.self_attn.k_proj",
138
+ "layers.8.self_attn.v_proj",
139
+ "layers.8.self_attn.o_proj",
140
+ "layers.9.self_attn.q_proj",
141
+ "layers.9.self_attn.k_proj",
142
+ "layers.9.self_attn.v_proj",
143
+ "layers.9.self_attn.o_proj",
144
+ "layers.10.self_attn.q_proj",
145
+ "layers.10.self_attn.k_proj",
146
+ "layers.10.self_attn.v_proj",
147
+ "layers.10.self_attn.o_proj",
148
+ "layers.11.self_attn.q_proj",
149
+ "layers.11.self_attn.k_proj",
150
+ "layers.11.self_attn.v_proj",
151
+ "layers.11.self_attn.o_proj",
152
+ "layers.12.self_attn.q_proj",
153
+ "layers.12.self_attn.k_proj",
154
+ "layers.12.self_attn.v_proj",
155
+ "layers.12.self_attn.o_proj",
156
+ "layers.13.self_attn.q_proj",
157
+ "layers.13.self_attn.k_proj",
158
+ "layers.13.self_attn.v_proj",
159
+ "layers.13.self_attn.o_proj",
160
+ "layers.14.self_attn.q_proj",
161
+ "layers.14.self_attn.k_proj",
162
+ "layers.14.self_attn.v_proj",
163
+ "layers.14.self_attn.o_proj",
164
+ "layers.15.self_attn.q_proj",
165
+ "layers.15.self_attn.k_proj",
166
+ "layers.15.self_attn.v_proj",
167
+ "layers.15.self_attn.o_proj",
168
+ "layers.16.self_attn.q_proj",
169
+ "layers.16.self_attn.k_proj",
170
+ "layers.16.self_attn.v_proj",
171
+ "layers.16.self_attn.o_proj",
172
+ "layers.17.self_attn.q_proj",
173
+ "layers.17.self_attn.k_proj",
174
+ "layers.17.self_attn.v_proj",
175
+ "layers.17.self_attn.o_proj",
176
+ "layers.18.self_attn.q_proj",
177
+ "layers.18.self_attn.k_proj",
178
+ "layers.18.self_attn.v_proj",
179
+ "layers.18.self_attn.o_proj",
180
+ "layers.19.self_attn.q_proj",
181
+ "layers.19.self_attn.k_proj",
182
+ "layers.19.self_attn.v_proj",
183
+ "layers.19.self_attn.o_proj",
184
+ "layers.20.self_attn.q_proj",
185
+ "layers.20.self_attn.k_proj",
186
+ "layers.20.self_attn.v_proj",
187
+ "layers.20.self_attn.o_proj",
188
+ "layers.21.self_attn.q_proj",
189
+ "layers.21.self_attn.k_proj",
190
+ "layers.21.self_attn.v_proj",
191
+ "layers.21.self_attn.o_proj",
192
+ "layers.22.self_attn.q_proj",
193
+ "layers.22.self_attn.k_proj",
194
+ "layers.22.self_attn.v_proj",
195
+ "layers.22.self_attn.o_proj",
196
+ "layers.23.self_attn.q_proj",
197
+ "layers.23.self_attn.k_proj",
198
+ "layers.23.self_attn.v_proj",
199
+ "layers.23.self_attn.o_proj",
200
+ "layers.24.self_attn.q_proj",
201
+ "layers.24.self_attn.k_proj",
202
+ "layers.24.self_attn.v_proj",
203
+ "layers.24.self_attn.o_proj",
204
+ "layers.25.self_attn.q_proj",
205
+ "layers.25.self_attn.k_proj",
206
+ "layers.25.self_attn.v_proj",
207
+ "layers.25.self_attn.o_proj",
208
+ "layers.26.self_attn.q_proj",
209
+ "layers.26.self_attn.k_proj",
210
+ "layers.26.self_attn.v_proj",
211
+ "layers.26.self_attn.o_proj",
212
+ "layers.27.self_attn.q_proj",
213
+ "layers.27.self_attn.k_proj",
214
+ "layers.27.self_attn.v_proj",
215
+ "layers.27.self_attn.o_proj",
216
+ "layers.28.self_attn.q_proj",
217
+ "layers.28.self_attn.k_proj",
218
+ "layers.28.self_attn.v_proj",
219
+ "layers.28.self_attn.o_proj",
220
+ "layers.29.self_attn.q_proj",
221
+ "layers.29.self_attn.k_proj",
222
+ "layers.29.self_attn.v_proj",
223
+ "layers.29.self_attn.o_proj",
224
+ "layers.30.self_attn.q_proj",
225
+ "layers.30.self_attn.k_proj",
226
+ "layers.30.self_attn.v_proj",
227
+ "layers.30.self_attn.o_proj",
228
+ "layers.31.self_attn.q_proj",
229
+ "layers.31.self_attn.k_proj",
230
+ "layers.31.self_attn.v_proj",
231
+ "layers.31.self_attn.o_proj",
232
+ "layers.32.self_attn.q_proj",
233
+ "layers.32.self_attn.k_proj",
234
+ "layers.32.self_attn.v_proj",
235
+ "layers.32.self_attn.o_proj",
236
+ "layers.33.self_attn.q_proj",
237
+ "layers.33.self_attn.k_proj",
238
+ "layers.33.self_attn.v_proj",
239
+ "layers.33.self_attn.o_proj",
240
+ "layers.34.self_attn.q_proj",
241
+ "layers.34.self_attn.k_proj",
242
+ "layers.34.self_attn.v_proj",
243
+ "layers.34.self_attn.o_proj",
244
+ "layers.35.self_attn.q_proj",
245
+ "layers.35.self_attn.k_proj",
246
+ "layers.35.self_attn.v_proj",
247
+ "layers.35.self_attn.o_proj"
248
+ ],
249
+ "kv_cache_scheme": null,
250
+ "quant_method": "compressed-tensors",
251
+ "quantization_status": "compressed",
252
+ "sparsity_config": {},
253
+ "transform_config": {},
254
+ "version": "0.15.1.dev14+g01a1c9a"
255
+ },
256
+ "rms_norm_eps": 1e-06,
257
+ "rope_parameters": {
258
+ "rope_theta": 1000000,
259
+ "rope_type": "default"
260
+ },
261
+ "sliding_window": null,
262
+ "tie_word_embeddings": true,
263
+ "transformers_version": "5.2.0",
264
+ "use_cache": true,
265
+ "use_sliding_window": false,
266
+ "vocab_size": 151665
267
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4c16d5ce03db20422350b813c41044a949d142b25c09f4ae7917c4f4c102450
3
+ size 4177330760
recipe.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ default_stage:
2
+ default_modifiers:
3
+ QuantizationModifier:
4
+ targets: ['re:.*mlp.gate_proj$', 're:.*mlp.up_proj$', 're:.*mlp.down_proj$']
5
+ ignore: []
6
+ scheme: NVFP4
7
+ bypass_divisibility_checks: false
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5313a7bef3012f7dab2e49f82ee0dead33fb9c864c024c673a731606409560f
3
+ size 11423041
tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "is_local": true,
9
+ "model_max_length": 131072,
10
+ "pad_token": "<|endoftext|>",
11
+ "split_special_tokens": false,
12
+ "tokenizer_class": "Qwen2Tokenizer",
13
+ "unk_token": null
14
+ }