Jur1cek commited on
Commit
39436d4
·
verified ·
1 Parent(s): 8c91cb8

Upload 15 files

Browse files
.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 +1,222 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set image_count = namespace(value=0) %}
2
+ {%- set video_count = namespace(value=0) %}
3
+ {%- macro render_content(content, do_vision_count, is_system_content=false) %}
4
+ {%- if content is string %}
5
+ {{- content }}
6
+ {%- elif content is iterable and content is not mapping %}
7
+ {%- for item in content %}
8
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
9
+ {%- if is_system_content %}
10
+ {{- raise_exception('System message cannot contain images.') }}
11
+ {%- endif %}
12
+ {%- if do_vision_count %}
13
+ {%- set image_count.value = image_count.value + 1 %}
14
+ {%- endif %}
15
+ {%- if add_vision_id is defined and add_vision_id %}
16
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
17
+ {%- endif %}
18
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
19
+ {%- elif 'video' in item or item.type == 'video' %}
20
+ {%- if is_system_content %}
21
+ {{- raise_exception('System message cannot contain videos.') }}
22
+ {%- endif %}
23
+ {%- if do_vision_count %}
24
+ {%- set video_count.value = video_count.value + 1 %}
25
+ {%- endif %}
26
+ {%- if add_vision_id is defined and add_vision_id %}
27
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
28
+ {%- endif %}
29
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
30
+ {%- elif 'text' in item %}
31
+ {{- item.text }}
32
+ {%- else %}
33
+ {{- raise_exception('Unexpected item type in content.') }}
34
+ {%- endif %}
35
+ {%- endfor %}
36
+ {%- elif content is none or content is undefined %}
37
+ {{- '' }}
38
+ {%- else %}
39
+ {{- raise_exception('Unexpected content type.') }}
40
+ {%- endif %}
41
+ {%- endmacro %}
42
+ {%- set ns_flags = namespace(enable_thinking=true) %}
43
+ {%- if enable_thinking is defined %}
44
+ {%- set ns_flags.enable_thinking = enable_thinking %}
45
+ {%- endif %}
46
+ {%- set preserve_thinking = preserve_thinking | default(true) %}
47
+ {%- if not messages %}
48
+ {{- raise_exception('No messages provided.') }}
49
+ {%- endif %}
50
+ {%- if add_generation_prompt is defined and add_generation_prompt and continue_final_message is defined and continue_final_message %}
51
+ {{- raise_exception('add_generation_prompt and continue_final_message cannot both be true.') }}
52
+ {%- endif %}
53
+ {%- if tools and tools is iterable and tools is not mapping %}
54
+ {{- '<|im_start|>system\n' }}
55
+ {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
56
+ {%- for tool in tools %}
57
+ {{- "\n" }}
58
+ {{- tool | tojson }}
59
+ {%- endfor %}
60
+ {{- "\n</tools>" }}
61
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
62
+ {%- if messages[0].role == 'system' or messages[0].role == 'developer' %}
63
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
64
+ {%- if '<|think_off|>' in content %}
65
+ {%- set ns_flags.enable_thinking = false %}
66
+ {%- set content = content.replace('<|think_off|>', '') %}
67
+ {%- endif %}
68
+ {%- if '<|think_on|>' in content %}
69
+ {%- set ns_flags.enable_thinking = true %}
70
+ {%- set content = content.replace('<|think_on|>', '') %}
71
+ {%- endif %}
72
+ {%- set content = content.strip() %}
73
+ {%- if content %}
74
+ {{- '\n\n' + content }}
75
+ {%- endif %}
76
+ {%- endif %}
77
+ {{- '<|im_end|>\n' }}
78
+ {%- else %}
79
+ {%- if messages[0].role == 'system' or messages[0].role == 'developer' %}
80
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
81
+ {%- if '<|think_off|>' in content %}
82
+ {%- set ns_flags.enable_thinking = false %}
83
+ {%- set content = content.replace('<|think_off|>', '') %}
84
+ {%- endif %}
85
+ {%- if '<|think_on|>' in content %}
86
+ {%- set ns_flags.enable_thinking = true %}
87
+ {%- set content = content.replace('<|think_on|>', '') %}
88
+ {%- endif %}
89
+ {%- set content = content.strip() %}
90
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
91
+ {%- endif %}
92
+ {%- endif %}
93
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
94
+ {%- for message in messages[::-1] %}
95
+ {%- set index = (messages|length - 1) - loop.index0 %}
96
+ {%- if ns.multi_step_tool and message.role == "user" %}
97
+ {%- set content = render_content(message.content, false)|trim %}
98
+ {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
99
+ {%- set ns.multi_step_tool = false %}
100
+ {%- set ns.last_query_index = index %}
101
+ {%- endif %}
102
+ {%- endif %}
103
+ {%- endfor %}
104
+ {%- if ns.multi_step_tool %}
105
+ {%- set ns.last_query_index = messages|length - 1 %}
106
+ {%- endif %}
107
+ {%- for message in messages %}
108
+ {%- set content = render_content(message.content, true)|trim %}
109
+ {%- set content = content.replace('<|think_off|>', '').replace('<|think_on|>', '') %}
110
+ {%- set content = content.strip() %}
111
+ {%- if message.role == "system" or message.role == "developer" %}
112
+ {%- if not loop.first %}
113
+ {%- set sys_content = render_content(message.content, false, true)|trim %}
114
+ {%- set sys_content = sys_content.replace('<|think_off|>', '').replace('<|think_on|>', '')|trim %}
115
+ {{- '<|im_start|>system\n' + sys_content + '<|im_end|>' + '\n' }}
116
+ {%- endif %}
117
+ {%- elif message.role == "user" %}
118
+ {{- '<|im_start|>' + message.role + '\n' + (content if content else ' ') + '<|im_end|>' + '\n' }}
119
+ {%- elif message.role == "assistant" %}
120
+ {%- set reasoning_content = '' %}
121
+ {%- if message.reasoning_content is string %}
122
+ {%- set reasoning_content = message.reasoning_content %}
123
+ {%- else %}
124
+ {%- set has_think_tag = false %}
125
+ {%- set think_start_token = '<think>' %}
126
+ {%- set think_end_token = '</think>' %}
127
+ {%- if '</think>' in content %}
128
+ {%- set has_think_tag = true %}
129
+ {%- elif '</thinking>' in content %}
130
+ {%- set has_think_tag = true %}
131
+ {%- set think_start_token = '<thinking>' %}
132
+ {%- set think_end_token = '</thinking>' %}
133
+ {%- elif '<think>' in content %}
134
+ {%- set reasoning_content = content.split('<think>')[-1].lstrip('\n') %}
135
+ {%- set content = '' %}
136
+ {%- elif '<thinking>' in content %}
137
+ {%- set reasoning_content = content.split('<thinking>')[-1].lstrip('\n') %}
138
+ {%- set content = '' %}
139
+ {%- endif %}
140
+ {%- if has_think_tag %}
141
+ {%- set reasoning_content = content.split(think_end_token)[0].rstrip('\n').split(think_start_token)[-1].lstrip('\n') %}
142
+ {%- set content = content.split(think_end_token)[-1].lstrip('\n') %}
143
+ {%- endif %}
144
+ {%- endif %}
145
+ {%- set reasoning_content = reasoning_content|trim %}
146
+ {%- set show_think = false %}
147
+ {%- if loop.index0 > ns.last_query_index and reasoning_content|length > 0 %}
148
+ {%- set show_think = true %}
149
+ {%- elif ns_flags.enable_thinking and preserve_thinking and reasoning_content|length > 0 %}
150
+ {%- set show_think = true %}
151
+ {%- endif %}
152
+ {%- if show_think %}
153
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
154
+ {%- else %}
155
+ {{- '<|im_start|>' + message.role + '\n' + content }}
156
+ {%- endif %}
157
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
158
+ {%- for tool_call in message.tool_calls %}
159
+ {%- if tool_call.function is defined %}
160
+ {%- set tool_call = tool_call.function %}
161
+ {%- endif %}
162
+ {%- if loop.first %}
163
+ {%- if content|trim %}
164
+ {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
165
+ {%- else %}
166
+ {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
167
+ {%- endif %}
168
+ {%- else %}
169
+ {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
170
+ {%- endif %}
171
+ {%- if tool_call.arguments is defined and tool_call.arguments is mapping %}
172
+ {%- if tool_call.arguments|length > 0 %}
173
+ {%- for args_name in tool_call.arguments %}
174
+ {%- set args_value = tool_call.arguments[args_name] %}
175
+ {{- '<parameter=' + args_name + '>\n' }}
176
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson %}
177
+ {{- args_value }}
178
+ {{- '\n</parameter>\n' }}
179
+ {%- endfor %}
180
+ {%- endif %}
181
+ {%- elif tool_call.arguments is defined and tool_call.arguments is string %}
182
+ {%- if tool_call.arguments|trim|length > 0 %}
183
+ {#- Note: raw JSON string arguments are emitted as-is and will not match
184
+ the XML parameter format in the tool instructions. Normalize arguments
185
+ to a dict in your serving layer before applying this template. -#}
186
+ {{- tool_call.arguments }}
187
+ {{- '\n' }}
188
+ {%- endif %}
189
+ {%- endif %}
190
+ {{- '</function>\n</tool_call>' }}
191
+ {%- endfor %}
192
+ {%- endif %}
193
+ {%- if not (loop.last and continue_final_message is defined and continue_final_message is true) %}
194
+ {{- '<|im_end|>\n' }}
195
+ {%- endif %}
196
+ {%- elif message.role == "tool" %}
197
+ {%- if not loop.previtem or (loop.previtem.role != "tool" and loop.previtem.role != "assistant") %}
198
+ {{- raise_exception('A tool message must follow an assistant or tool message.') }}
199
+ {%- endif %}
200
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
201
+ {{- '<|im_start|>user' }}
202
+ {%- endif %}
203
+ {{- '\n<tool_response>\n' }}
204
+ {{- content }}
205
+ {{- '\n</tool_response>' }}
206
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
207
+ {{- '<|im_end|>\n' }}
208
+ {%- elif loop.last %}
209
+ {{- '<|im_end|>\n' }}
210
+ {%- endif %}
211
+ {%- else %}
212
+ {{- raise_exception('Unexpected message role.') }}
213
+ {%- endif %}
214
+ {%- endfor %}
215
+ {%- if add_generation_prompt %}
216
+ {{- '<|im_start|>assistant\n' }}
217
+ {%- if ns_flags.enable_thinking is false %}
218
+ {{- '<think>\n\n</think>\n\n' }}
219
+ {%- else %}
220
+ {{- '<think>\n' }}
221
+ {%- endif %}
222
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,999 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3_5ForConditionalGeneration"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "image_token_id": 248056,
7
+ "language_model_only": false,
8
+ "model_type": "qwen3_5",
9
+ "text_config": {
10
+ "attention_bias": false,
11
+ "attention_dropout": 0.0,
12
+ "attn_output_gate": true,
13
+ "bos_token_id": 248044,
14
+ "dtype": "bfloat16",
15
+ "eos_token_id": 248044,
16
+ "full_attention_interval": 4,
17
+ "head_dim": 256,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 5120,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 17408,
22
+ "layer_types": [
23
+ "linear_attention",
24
+ "linear_attention",
25
+ "linear_attention",
26
+ "full_attention",
27
+ "linear_attention",
28
+ "linear_attention",
29
+ "linear_attention",
30
+ "full_attention",
31
+ "linear_attention",
32
+ "linear_attention",
33
+ "linear_attention",
34
+ "full_attention",
35
+ "linear_attention",
36
+ "linear_attention",
37
+ "linear_attention",
38
+ "full_attention",
39
+ "linear_attention",
40
+ "linear_attention",
41
+ "linear_attention",
42
+ "full_attention",
43
+ "linear_attention",
44
+ "linear_attention",
45
+ "linear_attention",
46
+ "full_attention",
47
+ "linear_attention",
48
+ "linear_attention",
49
+ "linear_attention",
50
+ "full_attention",
51
+ "linear_attention",
52
+ "linear_attention",
53
+ "linear_attention",
54
+ "full_attention",
55
+ "linear_attention",
56
+ "linear_attention",
57
+ "linear_attention",
58
+ "full_attention",
59
+ "linear_attention",
60
+ "linear_attention",
61
+ "linear_attention",
62
+ "full_attention",
63
+ "linear_attention",
64
+ "linear_attention",
65
+ "linear_attention",
66
+ "full_attention",
67
+ "linear_attention",
68
+ "linear_attention",
69
+ "linear_attention",
70
+ "full_attention",
71
+ "linear_attention",
72
+ "linear_attention",
73
+ "linear_attention",
74
+ "full_attention",
75
+ "linear_attention",
76
+ "linear_attention",
77
+ "linear_attention",
78
+ "full_attention",
79
+ "linear_attention",
80
+ "linear_attention",
81
+ "linear_attention",
82
+ "full_attention",
83
+ "linear_attention",
84
+ "linear_attention",
85
+ "linear_attention",
86
+ "full_attention"
87
+ ],
88
+ "linear_conv_kernel_dim": 4,
89
+ "linear_key_head_dim": 128,
90
+ "linear_num_key_heads": 16,
91
+ "linear_num_value_heads": 48,
92
+ "linear_value_head_dim": 128,
93
+ "mamba_ssm_dtype": "float32",
94
+ "max_position_embeddings": 262144,
95
+ "model_type": "qwen3_5_text",
96
+ "mtp_num_hidden_layers": 1,
97
+ "mtp_use_dedicated_embeddings": false,
98
+ "num_attention_heads": 24,
99
+ "num_hidden_layers": 64,
100
+ "num_key_value_heads": 4,
101
+ "output_gate_type": "swish",
102
+ "pad_token_id": null,
103
+ "partial_rotary_factor": 0.25,
104
+ "rms_norm_eps": 1e-06,
105
+ "rope_parameters": {
106
+ "mrope_interleaved": true,
107
+ "mrope_section": [
108
+ 11,
109
+ 11,
110
+ 10
111
+ ],
112
+ "partial_rotary_factor": 0.25,
113
+ "rope_theta": 10000000,
114
+ "rope_type": "default"
115
+ },
116
+ "tie_word_embeddings": false,
117
+ "use_cache": true,
118
+ "vocab_size": 248320
119
+ },
120
+ "tie_word_embeddings": false,
121
+ "transformers_version": "5.8.0",
122
+ "video_token_id": 248057,
123
+ "vision_config": {
124
+ "deepstack_visual_indexes": [],
125
+ "depth": 27,
126
+ "dtype": "bfloat16",
127
+ "hidden_act": "gelu_pytorch_tanh",
128
+ "hidden_size": 1152,
129
+ "in_channels": 3,
130
+ "initializer_range": 0.02,
131
+ "intermediate_size": 4304,
132
+ "model_type": "qwen3_5_vision",
133
+ "num_heads": 16,
134
+ "num_position_embeddings": 2304,
135
+ "out_hidden_size": 5120,
136
+ "patch_size": 16,
137
+ "spatial_merge_size": 2,
138
+ "temporal_patch_size": 2
139
+ },
140
+ "vision_end_token_id": 248054,
141
+ "vision_start_token_id": 248053,
142
+ "quantization_config": {
143
+ "quant_method": "compressed-tensors",
144
+ "format": "mixed-precision",
145
+ "config_groups": {
146
+ "group_0": {
147
+ "format": "nvfp4-pack-quantized",
148
+ "weights": {
149
+ "num_bits": 4,
150
+ "type": "float",
151
+ "strategy": "tensor_group",
152
+ "group_size": 16,
153
+ "symmetric": true,
154
+ "dynamic": false,
155
+ "scale_dtype": "torch.float8_e4m3fn",
156
+ "zp_dtype": "torch.float8_e4m3fn",
157
+ "observer": "memoryless_minmax"
158
+ },
159
+ "input_activations": {
160
+ "num_bits": 4,
161
+ "type": "float",
162
+ "strategy": "tensor_group",
163
+ "group_size": 16,
164
+ "symmetric": true,
165
+ "dynamic": "local",
166
+ "observer": "static_minmax",
167
+ "scale_dtype": "torch.float8_e4m3fn",
168
+ "zp_dtype": "torch.float8_e4m3fn"
169
+ },
170
+ "targets": [
171
+ "re:^language_model[.]model[.]layers[.]0[.]mlp[.]down_proj$",
172
+ "re:^language_model[.]model[.]layers[.]0[.]mlp[.]gate_proj$",
173
+ "re:^language_model[.]model[.]layers[.]0[.]mlp[.]gate_up_proj$",
174
+ "re:^language_model[.]model[.]layers[.]0[.]mlp[.]up_proj$",
175
+ "re:^language_model[.]model[.]layers[.]1[.]linear_attn[.]out_proj$",
176
+ "re:^language_model[.]model[.]layers[.]1[.]mlp[.]down_proj$",
177
+ "re:^language_model[.]model[.]layers[.]1[.]mlp[.]gate_proj$",
178
+ "re:^language_model[.]model[.]layers[.]1[.]mlp[.]gate_up_proj$",
179
+ "re:^language_model[.]model[.]layers[.]1[.]mlp[.]up_proj$",
180
+ "re:^language_model[.]model[.]layers[.]10[.]linear_attn[.]out_proj$",
181
+ "re:^language_model[.]model[.]layers[.]10[.]mlp[.]down_proj$",
182
+ "re:^language_model[.]model[.]layers[.]10[.]mlp[.]gate_proj$",
183
+ "re:^language_model[.]model[.]layers[.]10[.]mlp[.]gate_up_proj$",
184
+ "re:^language_model[.]model[.]layers[.]10[.]mlp[.]up_proj$",
185
+ "re:^language_model[.]model[.]layers[.]11[.]mlp[.]down_proj$",
186
+ "re:^language_model[.]model[.]layers[.]11[.]mlp[.]gate_proj$",
187
+ "re:^language_model[.]model[.]layers[.]11[.]mlp[.]gate_up_proj$",
188
+ "re:^language_model[.]model[.]layers[.]11[.]mlp[.]up_proj$",
189
+ "re:^language_model[.]model[.]layers[.]11[.]self_attn[.]o_proj$",
190
+ "re:^language_model[.]model[.]layers[.]12[.]linear_attn[.]out_proj$",
191
+ "re:^language_model[.]model[.]layers[.]12[.]mlp[.]down_proj$",
192
+ "re:^language_model[.]model[.]layers[.]12[.]mlp[.]gate_proj$",
193
+ "re:^language_model[.]model[.]layers[.]12[.]mlp[.]gate_up_proj$",
194
+ "re:^language_model[.]model[.]layers[.]12[.]mlp[.]up_proj$",
195
+ "re:^language_model[.]model[.]layers[.]13[.]linear_attn[.]out_proj$",
196
+ "re:^language_model[.]model[.]layers[.]13[.]mlp[.]down_proj$",
197
+ "re:^language_model[.]model[.]layers[.]13[.]mlp[.]gate_proj$",
198
+ "re:^language_model[.]model[.]layers[.]13[.]mlp[.]gate_up_proj$",
199
+ "re:^language_model[.]model[.]layers[.]13[.]mlp[.]up_proj$",
200
+ "re:^language_model[.]model[.]layers[.]14[.]linear_attn[.]out_proj$",
201
+ "re:^language_model[.]model[.]layers[.]14[.]mlp[.]down_proj$",
202
+ "re:^language_model[.]model[.]layers[.]14[.]mlp[.]gate_proj$",
203
+ "re:^language_model[.]model[.]layers[.]14[.]mlp[.]gate_up_proj$",
204
+ "re:^language_model[.]model[.]layers[.]14[.]mlp[.]up_proj$",
205
+ "re:^language_model[.]model[.]layers[.]15[.]mlp[.]down_proj$",
206
+ "re:^language_model[.]model[.]layers[.]15[.]mlp[.]gate_proj$",
207
+ "re:^language_model[.]model[.]layers[.]15[.]mlp[.]gate_up_proj$",
208
+ "re:^language_model[.]model[.]layers[.]15[.]mlp[.]up_proj$",
209
+ "re:^language_model[.]model[.]layers[.]15[.]self_attn[.]o_proj$",
210
+ "re:^language_model[.]model[.]layers[.]16[.]linear_attn[.]out_proj$",
211
+ "re:^language_model[.]model[.]layers[.]16[.]mlp[.]down_proj$",
212
+ "re:^language_model[.]model[.]layers[.]16[.]mlp[.]gate_proj$",
213
+ "re:^language_model[.]model[.]layers[.]16[.]mlp[.]gate_up_proj$",
214
+ "re:^language_model[.]model[.]layers[.]16[.]mlp[.]up_proj$",
215
+ "re:^language_model[.]model[.]layers[.]17[.]linear_attn[.]out_proj$",
216
+ "re:^language_model[.]model[.]layers[.]17[.]mlp[.]down_proj$",
217
+ "re:^language_model[.]model[.]layers[.]17[.]mlp[.]gate_proj$",
218
+ "re:^language_model[.]model[.]layers[.]17[.]mlp[.]gate_up_proj$",
219
+ "re:^language_model[.]model[.]layers[.]17[.]mlp[.]up_proj$",
220
+ "re:^language_model[.]model[.]layers[.]18[.]linear_attn[.]out_proj$",
221
+ "re:^language_model[.]model[.]layers[.]18[.]mlp[.]down_proj$",
222
+ "re:^language_model[.]model[.]layers[.]19[.]mlp[.]down_proj$",
223
+ "re:^language_model[.]model[.]layers[.]19[.]self_attn[.]o_proj$",
224
+ "re:^language_model[.]model[.]layers[.]2[.]linear_attn[.]out_proj$",
225
+ "re:^language_model[.]model[.]layers[.]2[.]mlp[.]down_proj$",
226
+ "re:^language_model[.]model[.]layers[.]2[.]mlp[.]gate_proj$",
227
+ "re:^language_model[.]model[.]layers[.]2[.]mlp[.]gate_up_proj$",
228
+ "re:^language_model[.]model[.]layers[.]2[.]mlp[.]up_proj$",
229
+ "re:^language_model[.]model[.]layers[.]20[.]linear_attn[.]out_proj$",
230
+ "re:^language_model[.]model[.]layers[.]20[.]mlp[.]down_proj$",
231
+ "re:^language_model[.]model[.]layers[.]21[.]linear_attn[.]out_proj$",
232
+ "re:^language_model[.]model[.]layers[.]21[.]mlp[.]down_proj$",
233
+ "re:^language_model[.]model[.]layers[.]22[.]linear_attn[.]out_proj$",
234
+ "re:^language_model[.]model[.]layers[.]22[.]mlp[.]down_proj$",
235
+ "re:^language_model[.]model[.]layers[.]23[.]mlp[.]down_proj$",
236
+ "re:^language_model[.]model[.]layers[.]23[.]self_attn[.]o_proj$",
237
+ "re:^language_model[.]model[.]layers[.]24[.]linear_attn[.]out_proj$",
238
+ "re:^language_model[.]model[.]layers[.]24[.]mlp[.]down_proj$",
239
+ "re:^language_model[.]model[.]layers[.]25[.]linear_attn[.]out_proj$",
240
+ "re:^language_model[.]model[.]layers[.]25[.]mlp[.]down_proj$",
241
+ "re:^language_model[.]model[.]layers[.]26[.]linear_attn[.]out_proj$",
242
+ "re:^language_model[.]model[.]layers[.]26[.]mlp[.]down_proj$",
243
+ "re:^language_model[.]model[.]layers[.]27[.]mlp[.]down_proj$",
244
+ "re:^language_model[.]model[.]layers[.]27[.]self_attn[.]o_proj$",
245
+ "re:^language_model[.]model[.]layers[.]28[.]linear_attn[.]out_proj$",
246
+ "re:^language_model[.]model[.]layers[.]28[.]mlp[.]down_proj$",
247
+ "re:^language_model[.]model[.]layers[.]29[.]linear_attn[.]out_proj$",
248
+ "re:^language_model[.]model[.]layers[.]29[.]mlp[.]down_proj$",
249
+ "re:^language_model[.]model[.]layers[.]3[.]mlp[.]down_proj$",
250
+ "re:^language_model[.]model[.]layers[.]3[.]mlp[.]gate_proj$",
251
+ "re:^language_model[.]model[.]layers[.]3[.]mlp[.]gate_up_proj$",
252
+ "re:^language_model[.]model[.]layers[.]3[.]mlp[.]up_proj$",
253
+ "re:^language_model[.]model[.]layers[.]3[.]self_attn[.]o_proj$",
254
+ "re:^language_model[.]model[.]layers[.]30[.]linear_attn[.]out_proj$",
255
+ "re:^language_model[.]model[.]layers[.]30[.]mlp[.]down_proj$",
256
+ "re:^language_model[.]model[.]layers[.]31[.]mlp[.]down_proj$",
257
+ "re:^language_model[.]model[.]layers[.]31[.]self_attn[.]o_proj$",
258
+ "re:^language_model[.]model[.]layers[.]32[.]linear_attn[.]out_proj$",
259
+ "re:^language_model[.]model[.]layers[.]32[.]mlp[.]down_proj$",
260
+ "re:^language_model[.]model[.]layers[.]33[.]linear_attn[.]out_proj$",
261
+ "re:^language_model[.]model[.]layers[.]33[.]mlp[.]down_proj$",
262
+ "re:^language_model[.]model[.]layers[.]34[.]linear_attn[.]out_proj$",
263
+ "re:^language_model[.]model[.]layers[.]34[.]mlp[.]down_proj$",
264
+ "re:^language_model[.]model[.]layers[.]35[.]mlp[.]down_proj$",
265
+ "re:^language_model[.]model[.]layers[.]35[.]self_attn[.]o_proj$",
266
+ "re:^language_model[.]model[.]layers[.]36[.]linear_attn[.]out_proj$",
267
+ "re:^language_model[.]model[.]layers[.]36[.]mlp[.]down_proj$",
268
+ "re:^language_model[.]model[.]layers[.]37[.]linear_attn[.]out_proj$",
269
+ "re:^language_model[.]model[.]layers[.]37[.]mlp[.]down_proj$",
270
+ "re:^language_model[.]model[.]layers[.]38[.]linear_attn[.]out_proj$",
271
+ "re:^language_model[.]model[.]layers[.]38[.]mlp[.]down_proj$",
272
+ "re:^language_model[.]model[.]layers[.]39[.]mlp[.]down_proj$",
273
+ "re:^language_model[.]model[.]layers[.]39[.]self_attn[.]o_proj$",
274
+ "re:^language_model[.]model[.]layers[.]4[.]linear_attn[.]out_proj$",
275
+ "re:^language_model[.]model[.]layers[.]4[.]mlp[.]down_proj$",
276
+ "re:^language_model[.]model[.]layers[.]4[.]mlp[.]gate_proj$",
277
+ "re:^language_model[.]model[.]layers[.]4[.]mlp[.]gate_up_proj$",
278
+ "re:^language_model[.]model[.]layers[.]4[.]mlp[.]up_proj$",
279
+ "re:^language_model[.]model[.]layers[.]40[.]linear_attn[.]out_proj$",
280
+ "re:^language_model[.]model[.]layers[.]40[.]mlp[.]down_proj$",
281
+ "re:^language_model[.]model[.]layers[.]41[.]linear_attn[.]out_proj$",
282
+ "re:^language_model[.]model[.]layers[.]41[.]mlp[.]down_proj$",
283
+ "re:^language_model[.]model[.]layers[.]42[.]linear_attn[.]out_proj$",
284
+ "re:^language_model[.]model[.]layers[.]42[.]mlp[.]down_proj$",
285
+ "re:^language_model[.]model[.]layers[.]43[.]mlp[.]down_proj$",
286
+ "re:^language_model[.]model[.]layers[.]43[.]self_attn[.]o_proj$",
287
+ "re:^language_model[.]model[.]layers[.]44[.]linear_attn[.]out_proj$",
288
+ "re:^language_model[.]model[.]layers[.]44[.]mlp[.]down_proj$",
289
+ "re:^language_model[.]model[.]layers[.]45[.]linear_attn[.]out_proj$",
290
+ "re:^language_model[.]model[.]layers[.]45[.]mlp[.]down_proj$",
291
+ "re:^language_model[.]model[.]layers[.]46[.]linear_attn[.]out_proj$",
292
+ "re:^language_model[.]model[.]layers[.]46[.]mlp[.]down_proj$",
293
+ "re:^language_model[.]model[.]layers[.]47[.]mlp[.]down_proj$",
294
+ "re:^language_model[.]model[.]layers[.]47[.]self_attn[.]o_proj$",
295
+ "re:^language_model[.]model[.]layers[.]48[.]linear_attn[.]out_proj$",
296
+ "re:^language_model[.]model[.]layers[.]48[.]mlp[.]down_proj$",
297
+ "re:^language_model[.]model[.]layers[.]49[.]linear_attn[.]out_proj$",
298
+ "re:^language_model[.]model[.]layers[.]49[.]mlp[.]down_proj$",
299
+ "re:^language_model[.]model[.]layers[.]5[.]linear_attn[.]out_proj$",
300
+ "re:^language_model[.]model[.]layers[.]5[.]mlp[.]down_proj$",
301
+ "re:^language_model[.]model[.]layers[.]5[.]mlp[.]gate_proj$",
302
+ "re:^language_model[.]model[.]layers[.]5[.]mlp[.]gate_up_proj$",
303
+ "re:^language_model[.]model[.]layers[.]5[.]mlp[.]up_proj$",
304
+ "re:^language_model[.]model[.]layers[.]50[.]linear_attn[.]out_proj$",
305
+ "re:^language_model[.]model[.]layers[.]51[.]self_attn[.]o_proj$",
306
+ "re:^language_model[.]model[.]layers[.]54[.]linear_attn[.]out_proj$",
307
+ "re:^language_model[.]model[.]layers[.]55[.]self_attn[.]o_proj$",
308
+ "re:^language_model[.]model[.]layers[.]6[.]linear_attn[.]out_proj$",
309
+ "re:^language_model[.]model[.]layers[.]6[.]mlp[.]down_proj$",
310
+ "re:^language_model[.]model[.]layers[.]6[.]mlp[.]gate_proj$",
311
+ "re:^language_model[.]model[.]layers[.]6[.]mlp[.]gate_up_proj$",
312
+ "re:^language_model[.]model[.]layers[.]6[.]mlp[.]up_proj$",
313
+ "re:^language_model[.]model[.]layers[.]7[.]mlp[.]down_proj$",
314
+ "re:^language_model[.]model[.]layers[.]7[.]mlp[.]gate_proj$",
315
+ "re:^language_model[.]model[.]layers[.]7[.]mlp[.]gate_up_proj$",
316
+ "re:^language_model[.]model[.]layers[.]7[.]mlp[.]up_proj$",
317
+ "re:^language_model[.]model[.]layers[.]7[.]self_attn[.]o_proj$",
318
+ "re:^language_model[.]model[.]layers[.]8[.]linear_attn[.]out_proj$",
319
+ "re:^language_model[.]model[.]layers[.]8[.]mlp[.]down_proj$",
320
+ "re:^language_model[.]model[.]layers[.]8[.]mlp[.]gate_proj$",
321
+ "re:^language_model[.]model[.]layers[.]8[.]mlp[.]gate_up_proj$",
322
+ "re:^language_model[.]model[.]layers[.]8[.]mlp[.]up_proj$",
323
+ "re:^language_model[.]model[.]layers[.]9[.]linear_attn[.]out_proj$",
324
+ "re:^language_model[.]model[.]layers[.]9[.]mlp[.]down_proj$",
325
+ "re:^language_model[.]model[.]layers[.]9[.]mlp[.]gate_proj$",
326
+ "re:^language_model[.]model[.]layers[.]9[.]mlp[.]gate_up_proj$",
327
+ "re:^language_model[.]model[.]layers[.]9[.]mlp[.]up_proj$"
328
+ ]
329
+ },
330
+ "group_1": {
331
+ "format": "mxfp8-quantized",
332
+ "weights": {
333
+ "num_bits": 8,
334
+ "type": "float",
335
+ "strategy": "group",
336
+ "group_size": 32,
337
+ "symmetric": true,
338
+ "dynamic": false,
339
+ "scale_dtype": "torch.uint8",
340
+ "zp_dtype": "torch.uint8",
341
+ "observer": "memoryless_minmax"
342
+ },
343
+ "input_activations": {
344
+ "num_bits": 8,
345
+ "type": "float",
346
+ "strategy": "group",
347
+ "group_size": 32,
348
+ "symmetric": true,
349
+ "dynamic": true,
350
+ "scale_dtype": "torch.uint8",
351
+ "zp_dtype": "torch.uint8"
352
+ },
353
+ "targets": [
354
+ "re:^language_model[.]model[.]layers[.]0[.]linear_attn[.]out_proj$",
355
+ "re:^language_model[.]model[.]layers[.]18[.]mlp[.]gate_proj$",
356
+ "re:^language_model[.]model[.]layers[.]18[.]mlp[.]gate_up_proj$",
357
+ "re:^language_model[.]model[.]layers[.]18[.]mlp[.]up_proj$",
358
+ "re:^language_model[.]model[.]layers[.]19[.]mlp[.]gate_proj$",
359
+ "re:^language_model[.]model[.]layers[.]19[.]mlp[.]gate_up_proj$",
360
+ "re:^language_model[.]model[.]layers[.]19[.]mlp[.]up_proj$",
361
+ "re:^language_model[.]model[.]layers[.]20[.]mlp[.]gate_proj$",
362
+ "re:^language_model[.]model[.]layers[.]20[.]mlp[.]gate_up_proj$",
363
+ "re:^language_model[.]model[.]layers[.]20[.]mlp[.]up_proj$",
364
+ "re:^language_model[.]model[.]layers[.]21[.]mlp[.]gate_proj$",
365
+ "re:^language_model[.]model[.]layers[.]21[.]mlp[.]gate_up_proj$",
366
+ "re:^language_model[.]model[.]layers[.]21[.]mlp[.]up_proj$",
367
+ "re:^language_model[.]model[.]layers[.]22[.]mlp[.]gate_proj$",
368
+ "re:^language_model[.]model[.]layers[.]22[.]mlp[.]gate_up_proj$",
369
+ "re:^language_model[.]model[.]layers[.]22[.]mlp[.]up_proj$",
370
+ "re:^language_model[.]model[.]layers[.]23[.]mlp[.]gate_proj$",
371
+ "re:^language_model[.]model[.]layers[.]23[.]mlp[.]gate_up_proj$",
372
+ "re:^language_model[.]model[.]layers[.]23[.]mlp[.]up_proj$",
373
+ "re:^language_model[.]model[.]layers[.]24[.]mlp[.]gate_proj$",
374
+ "re:^language_model[.]model[.]layers[.]24[.]mlp[.]gate_up_proj$",
375
+ "re:^language_model[.]model[.]layers[.]24[.]mlp[.]up_proj$",
376
+ "re:^language_model[.]model[.]layers[.]25[.]mlp[.]gate_proj$",
377
+ "re:^language_model[.]model[.]layers[.]25[.]mlp[.]gate_up_proj$",
378
+ "re:^language_model[.]model[.]layers[.]25[.]mlp[.]up_proj$",
379
+ "re:^language_model[.]model[.]layers[.]26[.]mlp[.]gate_proj$",
380
+ "re:^language_model[.]model[.]layers[.]26[.]mlp[.]gate_up_proj$",
381
+ "re:^language_model[.]model[.]layers[.]26[.]mlp[.]up_proj$",
382
+ "re:^language_model[.]model[.]layers[.]27[.]mlp[.]gate_proj$",
383
+ "re:^language_model[.]model[.]layers[.]27[.]mlp[.]gate_up_proj$",
384
+ "re:^language_model[.]model[.]layers[.]27[.]mlp[.]up_proj$",
385
+ "re:^language_model[.]model[.]layers[.]28[.]mlp[.]gate_proj$",
386
+ "re:^language_model[.]model[.]layers[.]28[.]mlp[.]gate_up_proj$",
387
+ "re:^language_model[.]model[.]layers[.]28[.]mlp[.]up_proj$",
388
+ "re:^language_model[.]model[.]layers[.]29[.]mlp[.]gate_proj$",
389
+ "re:^language_model[.]model[.]layers[.]29[.]mlp[.]gate_up_proj$",
390
+ "re:^language_model[.]model[.]layers[.]29[.]mlp[.]up_proj$",
391
+ "re:^language_model[.]model[.]layers[.]30[.]mlp[.]gate_proj$",
392
+ "re:^language_model[.]model[.]layers[.]30[.]mlp[.]gate_up_proj$",
393
+ "re:^language_model[.]model[.]layers[.]30[.]mlp[.]up_proj$",
394
+ "re:^language_model[.]model[.]layers[.]31[.]mlp[.]gate_proj$",
395
+ "re:^language_model[.]model[.]layers[.]31[.]mlp[.]gate_up_proj$",
396
+ "re:^language_model[.]model[.]layers[.]31[.]mlp[.]up_proj$",
397
+ "re:^language_model[.]model[.]layers[.]32[.]mlp[.]gate_proj$",
398
+ "re:^language_model[.]model[.]layers[.]32[.]mlp[.]gate_up_proj$",
399
+ "re:^language_model[.]model[.]layers[.]32[.]mlp[.]up_proj$",
400
+ "re:^language_model[.]model[.]layers[.]33[.]mlp[.]gate_proj$",
401
+ "re:^language_model[.]model[.]layers[.]33[.]mlp[.]gate_up_proj$",
402
+ "re:^language_model[.]model[.]layers[.]33[.]mlp[.]up_proj$",
403
+ "re:^language_model[.]model[.]layers[.]34[.]mlp[.]gate_proj$",
404
+ "re:^language_model[.]model[.]layers[.]34[.]mlp[.]gate_up_proj$",
405
+ "re:^language_model[.]model[.]layers[.]34[.]mlp[.]up_proj$",
406
+ "re:^language_model[.]model[.]layers[.]35[.]mlp[.]gate_proj$",
407
+ "re:^language_model[.]model[.]layers[.]35[.]mlp[.]gate_up_proj$",
408
+ "re:^language_model[.]model[.]layers[.]35[.]mlp[.]up_proj$",
409
+ "re:^language_model[.]model[.]layers[.]36[.]mlp[.]gate_proj$",
410
+ "re:^language_model[.]model[.]layers[.]36[.]mlp[.]gate_up_proj$",
411
+ "re:^language_model[.]model[.]layers[.]36[.]mlp[.]up_proj$",
412
+ "re:^language_model[.]model[.]layers[.]37[.]mlp[.]gate_proj$",
413
+ "re:^language_model[.]model[.]layers[.]37[.]mlp[.]gate_up_proj$",
414
+ "re:^language_model[.]model[.]layers[.]37[.]mlp[.]up_proj$",
415
+ "re:^language_model[.]model[.]layers[.]38[.]linear_attn[.]in_proj_qkv$",
416
+ "re:^language_model[.]model[.]layers[.]38[.]linear_attn[.]in_proj_qkvz$",
417
+ "re:^language_model[.]model[.]layers[.]38[.]linear_attn[.]in_proj_z$",
418
+ "re:^language_model[.]model[.]layers[.]38[.]mlp[.]gate_proj$",
419
+ "re:^language_model[.]model[.]layers[.]38[.]mlp[.]gate_up_proj$",
420
+ "re:^language_model[.]model[.]layers[.]38[.]mlp[.]up_proj$",
421
+ "re:^language_model[.]model[.]layers[.]39[.]mlp[.]gate_proj$",
422
+ "re:^language_model[.]model[.]layers[.]39[.]mlp[.]gate_up_proj$",
423
+ "re:^language_model[.]model[.]layers[.]39[.]mlp[.]up_proj$",
424
+ "re:^language_model[.]model[.]layers[.]39[.]self_attn[.]k_proj$",
425
+ "re:^language_model[.]model[.]layers[.]39[.]self_attn[.]q_proj$",
426
+ "re:^language_model[.]model[.]layers[.]39[.]self_attn[.]qkv_proj$",
427
+ "re:^language_model[.]model[.]layers[.]39[.]self_attn[.]v_proj$",
428
+ "re:^language_model[.]model[.]layers[.]4[.]linear_attn[.]in_proj_qkv$",
429
+ "re:^language_model[.]model[.]layers[.]4[.]linear_attn[.]in_proj_qkvz$",
430
+ "re:^language_model[.]model[.]layers[.]4[.]linear_attn[.]in_proj_z$",
431
+ "re:^language_model[.]model[.]layers[.]40[.]linear_attn[.]in_proj_qkv$",
432
+ "re:^language_model[.]model[.]layers[.]40[.]linear_attn[.]in_proj_qkvz$",
433
+ "re:^language_model[.]model[.]layers[.]40[.]linear_attn[.]in_proj_z$",
434
+ "re:^language_model[.]model[.]layers[.]40[.]mlp[.]gate_proj$",
435
+ "re:^language_model[.]model[.]layers[.]40[.]mlp[.]gate_up_proj$",
436
+ "re:^language_model[.]model[.]layers[.]40[.]mlp[.]up_proj$",
437
+ "re:^language_model[.]model[.]layers[.]41[.]linear_attn[.]in_proj_qkv$",
438
+ "re:^language_model[.]model[.]layers[.]41[.]linear_attn[.]in_proj_qkvz$",
439
+ "re:^language_model[.]model[.]layers[.]41[.]linear_attn[.]in_proj_z$",
440
+ "re:^language_model[.]model[.]layers[.]41[.]mlp[.]gate_proj$",
441
+ "re:^language_model[.]model[.]layers[.]41[.]mlp[.]gate_up_proj$",
442
+ "re:^language_model[.]model[.]layers[.]41[.]mlp[.]up_proj$",
443
+ "re:^language_model[.]model[.]layers[.]42[.]linear_attn[.]in_proj_qkv$",
444
+ "re:^language_model[.]model[.]layers[.]42[.]linear_attn[.]in_proj_qkvz$",
445
+ "re:^language_model[.]model[.]layers[.]42[.]linear_attn[.]in_proj_z$",
446
+ "re:^language_model[.]model[.]layers[.]42[.]mlp[.]gate_proj$",
447
+ "re:^language_model[.]model[.]layers[.]42[.]mlp[.]gate_up_proj$",
448
+ "re:^language_model[.]model[.]layers[.]42[.]mlp[.]up_proj$",
449
+ "re:^language_model[.]model[.]layers[.]43[.]mlp[.]gate_proj$",
450
+ "re:^language_model[.]model[.]layers[.]43[.]mlp[.]gate_up_proj$",
451
+ "re:^language_model[.]model[.]layers[.]43[.]mlp[.]up_proj$",
452
+ "re:^language_model[.]model[.]layers[.]44[.]linear_attn[.]in_proj_qkv$",
453
+ "re:^language_model[.]model[.]layers[.]44[.]linear_attn[.]in_proj_qkvz$",
454
+ "re:^language_model[.]model[.]layers[.]44[.]linear_attn[.]in_proj_z$",
455
+ "re:^language_model[.]model[.]layers[.]44[.]mlp[.]gate_proj$",
456
+ "re:^language_model[.]model[.]layers[.]44[.]mlp[.]gate_up_proj$",
457
+ "re:^language_model[.]model[.]layers[.]44[.]mlp[.]up_proj$",
458
+ "re:^language_model[.]model[.]layers[.]45[.]linear_attn[.]in_proj_qkv$",
459
+ "re:^language_model[.]model[.]layers[.]45[.]linear_attn[.]in_proj_qkvz$",
460
+ "re:^language_model[.]model[.]layers[.]45[.]linear_attn[.]in_proj_z$",
461
+ "re:^language_model[.]model[.]layers[.]45[.]mlp[.]gate_proj$",
462
+ "re:^language_model[.]model[.]layers[.]45[.]mlp[.]gate_up_proj$",
463
+ "re:^language_model[.]model[.]layers[.]45[.]mlp[.]up_proj$",
464
+ "re:^language_model[.]model[.]layers[.]46[.]mlp[.]gate_proj$",
465
+ "re:^language_model[.]model[.]layers[.]46[.]mlp[.]gate_up_proj$",
466
+ "re:^language_model[.]model[.]layers[.]46[.]mlp[.]up_proj$",
467
+ "re:^language_model[.]model[.]layers[.]47[.]mlp[.]gate_proj$",
468
+ "re:^language_model[.]model[.]layers[.]47[.]mlp[.]gate_up_proj$",
469
+ "re:^language_model[.]model[.]layers[.]47[.]mlp[.]up_proj$",
470
+ "re:^language_model[.]model[.]layers[.]48[.]mlp[.]gate_proj$",
471
+ "re:^language_model[.]model[.]layers[.]48[.]mlp[.]gate_up_proj$",
472
+ "re:^language_model[.]model[.]layers[.]48[.]mlp[.]up_proj$",
473
+ "re:^language_model[.]model[.]layers[.]49[.]mlp[.]gate_proj$",
474
+ "re:^language_model[.]model[.]layers[.]49[.]mlp[.]gate_up_proj$",
475
+ "re:^language_model[.]model[.]layers[.]49[.]mlp[.]up_proj$",
476
+ "re:^language_model[.]model[.]layers[.]50[.]linear_attn[.]in_proj_qkv$",
477
+ "re:^language_model[.]model[.]layers[.]50[.]linear_attn[.]in_proj_qkvz$",
478
+ "re:^language_model[.]model[.]layers[.]50[.]linear_attn[.]in_proj_z$",
479
+ "re:^language_model[.]model[.]layers[.]50[.]mlp[.]down_proj$",
480
+ "re:^language_model[.]model[.]layers[.]50[.]mlp[.]gate_proj$",
481
+ "re:^language_model[.]model[.]layers[.]50[.]mlp[.]gate_up_proj$",
482
+ "re:^language_model[.]model[.]layers[.]50[.]mlp[.]up_proj$",
483
+ "re:^language_model[.]model[.]layers[.]51[.]mlp[.]down_proj$",
484
+ "re:^language_model[.]model[.]layers[.]51[.]mlp[.]gate_proj$",
485
+ "re:^language_model[.]model[.]layers[.]51[.]mlp[.]gate_up_proj$",
486
+ "re:^language_model[.]model[.]layers[.]51[.]mlp[.]up_proj$",
487
+ "re:^language_model[.]model[.]layers[.]51[.]self_attn[.]k_proj$",
488
+ "re:^language_model[.]model[.]layers[.]51[.]self_attn[.]q_proj$",
489
+ "re:^language_model[.]model[.]layers[.]51[.]self_attn[.]qkv_proj$",
490
+ "re:^language_model[.]model[.]layers[.]51[.]self_attn[.]v_proj$",
491
+ "re:^language_model[.]model[.]layers[.]52[.]linear_attn[.]in_proj_qkv$",
492
+ "re:^language_model[.]model[.]layers[.]52[.]linear_attn[.]in_proj_qkvz$",
493
+ "re:^language_model[.]model[.]layers[.]52[.]linear_attn[.]in_proj_z$",
494
+ "re:^language_model[.]model[.]layers[.]52[.]linear_attn[.]out_proj$",
495
+ "re:^language_model[.]model[.]layers[.]52[.]mlp[.]down_proj$",
496
+ "re:^language_model[.]model[.]layers[.]52[.]mlp[.]gate_proj$",
497
+ "re:^language_model[.]model[.]layers[.]52[.]mlp[.]gate_up_proj$",
498
+ "re:^language_model[.]model[.]layers[.]52[.]mlp[.]up_proj$",
499
+ "re:^language_model[.]model[.]layers[.]53[.]linear_attn[.]in_proj_qkv$",
500
+ "re:^language_model[.]model[.]layers[.]53[.]linear_attn[.]in_proj_qkvz$",
501
+ "re:^language_model[.]model[.]layers[.]53[.]linear_attn[.]in_proj_z$",
502
+ "re:^language_model[.]model[.]layers[.]53[.]linear_attn[.]out_proj$",
503
+ "re:^language_model[.]model[.]layers[.]53[.]mlp[.]down_proj$",
504
+ "re:^language_model[.]model[.]layers[.]53[.]mlp[.]gate_proj$",
505
+ "re:^language_model[.]model[.]layers[.]53[.]mlp[.]gate_up_proj$",
506
+ "re:^language_model[.]model[.]layers[.]53[.]mlp[.]up_proj$",
507
+ "re:^language_model[.]model[.]layers[.]54[.]linear_attn[.]in_proj_qkv$",
508
+ "re:^language_model[.]model[.]layers[.]54[.]linear_attn[.]in_proj_qkvz$",
509
+ "re:^language_model[.]model[.]layers[.]54[.]linear_attn[.]in_proj_z$",
510
+ "re:^language_model[.]model[.]layers[.]54[.]mlp[.]down_proj$",
511
+ "re:^language_model[.]model[.]layers[.]54[.]mlp[.]gate_proj$",
512
+ "re:^language_model[.]model[.]layers[.]54[.]mlp[.]gate_up_proj$",
513
+ "re:^language_model[.]model[.]layers[.]54[.]mlp[.]up_proj$",
514
+ "re:^language_model[.]model[.]layers[.]55[.]mlp[.]down_proj$",
515
+ "re:^language_model[.]model[.]layers[.]55[.]mlp[.]gate_proj$",
516
+ "re:^language_model[.]model[.]layers[.]55[.]mlp[.]gate_up_proj$",
517
+ "re:^language_model[.]model[.]layers[.]55[.]mlp[.]up_proj$",
518
+ "re:^language_model[.]model[.]layers[.]55[.]self_attn[.]k_proj$",
519
+ "re:^language_model[.]model[.]layers[.]55[.]self_attn[.]q_proj$",
520
+ "re:^language_model[.]model[.]layers[.]55[.]self_attn[.]qkv_proj$",
521
+ "re:^language_model[.]model[.]layers[.]55[.]self_attn[.]v_proj$",
522
+ "re:^language_model[.]model[.]layers[.]56[.]linear_attn[.]in_proj_qkv$",
523
+ "re:^language_model[.]model[.]layers[.]56[.]linear_attn[.]in_proj_qkvz$",
524
+ "re:^language_model[.]model[.]layers[.]56[.]linear_attn[.]in_proj_z$",
525
+ "re:^language_model[.]model[.]layers[.]56[.]linear_attn[.]out_proj$",
526
+ "re:^language_model[.]model[.]layers[.]56[.]mlp[.]down_proj$",
527
+ "re:^language_model[.]model[.]layers[.]56[.]mlp[.]gate_proj$",
528
+ "re:^language_model[.]model[.]layers[.]56[.]mlp[.]gate_up_proj$",
529
+ "re:^language_model[.]model[.]layers[.]56[.]mlp[.]up_proj$",
530
+ "re:^language_model[.]model[.]layers[.]57[.]linear_attn[.]in_proj_qkv$",
531
+ "re:^language_model[.]model[.]layers[.]57[.]linear_attn[.]in_proj_qkvz$",
532
+ "re:^language_model[.]model[.]layers[.]57[.]linear_attn[.]in_proj_z$",
533
+ "re:^language_model[.]model[.]layers[.]57[.]linear_attn[.]out_proj$",
534
+ "re:^language_model[.]model[.]layers[.]57[.]mlp[.]down_proj$",
535
+ "re:^language_model[.]model[.]layers[.]57[.]mlp[.]gate_proj$",
536
+ "re:^language_model[.]model[.]layers[.]57[.]mlp[.]gate_up_proj$",
537
+ "re:^language_model[.]model[.]layers[.]57[.]mlp[.]up_proj$",
538
+ "re:^language_model[.]model[.]layers[.]58[.]linear_attn[.]in_proj_qkv$",
539
+ "re:^language_model[.]model[.]layers[.]58[.]linear_attn[.]in_proj_qkvz$",
540
+ "re:^language_model[.]model[.]layers[.]58[.]linear_attn[.]in_proj_z$",
541
+ "re:^language_model[.]model[.]layers[.]58[.]linear_attn[.]out_proj$",
542
+ "re:^language_model[.]model[.]layers[.]58[.]mlp[.]gate_proj$",
543
+ "re:^language_model[.]model[.]layers[.]58[.]mlp[.]gate_up_proj$",
544
+ "re:^language_model[.]model[.]layers[.]58[.]mlp[.]up_proj$",
545
+ "re:^language_model[.]model[.]layers[.]59[.]mlp[.]down_proj$",
546
+ "re:^language_model[.]model[.]layers[.]59[.]mlp[.]gate_proj$",
547
+ "re:^language_model[.]model[.]layers[.]59[.]mlp[.]gate_up_proj$",
548
+ "re:^language_model[.]model[.]layers[.]59[.]mlp[.]up_proj$",
549
+ "re:^language_model[.]model[.]layers[.]59[.]self_attn[.]k_proj$",
550
+ "re:^language_model[.]model[.]layers[.]59[.]self_attn[.]o_proj$",
551
+ "re:^language_model[.]model[.]layers[.]59[.]self_attn[.]q_proj$",
552
+ "re:^language_model[.]model[.]layers[.]59[.]self_attn[.]qkv_proj$",
553
+ "re:^language_model[.]model[.]layers[.]59[.]self_attn[.]v_proj$",
554
+ "re:^language_model[.]model[.]layers[.]60[.]linear_attn[.]in_proj_qkv$",
555
+ "re:^language_model[.]model[.]layers[.]60[.]linear_attn[.]in_proj_qkvz$",
556
+ "re:^language_model[.]model[.]layers[.]60[.]linear_attn[.]in_proj_z$",
557
+ "re:^language_model[.]model[.]layers[.]60[.]mlp[.]down_proj$",
558
+ "re:^language_model[.]model[.]layers[.]60[.]mlp[.]gate_proj$",
559
+ "re:^language_model[.]model[.]layers[.]60[.]mlp[.]gate_up_proj$",
560
+ "re:^language_model[.]model[.]layers[.]60[.]mlp[.]up_proj$",
561
+ "re:^language_model[.]model[.]layers[.]61[.]linear_attn[.]in_proj_qkv$",
562
+ "re:^language_model[.]model[.]layers[.]61[.]linear_attn[.]in_proj_qkvz$",
563
+ "re:^language_model[.]model[.]layers[.]61[.]linear_attn[.]in_proj_z$",
564
+ "re:^language_model[.]model[.]layers[.]61[.]linear_attn[.]out_proj$",
565
+ "re:^language_model[.]model[.]layers[.]61[.]mlp[.]down_proj$",
566
+ "re:^language_model[.]model[.]layers[.]61[.]mlp[.]gate_proj$",
567
+ "re:^language_model[.]model[.]layers[.]61[.]mlp[.]gate_up_proj$",
568
+ "re:^language_model[.]model[.]layers[.]61[.]mlp[.]up_proj$",
569
+ "re:^language_model[.]model[.]layers[.]62[.]linear_attn[.]in_proj_qkv$",
570
+ "re:^language_model[.]model[.]layers[.]62[.]linear_attn[.]in_proj_qkvz$",
571
+ "re:^language_model[.]model[.]layers[.]62[.]linear_attn[.]in_proj_z$",
572
+ "re:^language_model[.]model[.]layers[.]62[.]linear_attn[.]out_proj$",
573
+ "re:^language_model[.]model[.]layers[.]62[.]mlp[.]gate_proj$",
574
+ "re:^language_model[.]model[.]layers[.]62[.]mlp[.]gate_up_proj$",
575
+ "re:^language_model[.]model[.]layers[.]62[.]mlp[.]up_proj$",
576
+ "re:^language_model[.]model[.]layers[.]63[.]mlp[.]gate_proj$",
577
+ "re:^language_model[.]model[.]layers[.]63[.]mlp[.]gate_up_proj$",
578
+ "re:^language_model[.]model[.]layers[.]63[.]mlp[.]up_proj$",
579
+ "re:^language_model[.]model[.]layers[.]63[.]self_attn[.]k_proj$",
580
+ "re:^language_model[.]model[.]layers[.]63[.]self_attn[.]o_proj$",
581
+ "re:^language_model[.]model[.]layers[.]63[.]self_attn[.]q_proj$",
582
+ "re:^language_model[.]model[.]layers[.]63[.]self_attn[.]qkv_proj$",
583
+ "re:^language_model[.]model[.]layers[.]63[.]self_attn[.]v_proj$"
584
+ ]
585
+ }
586
+ },
587
+ "ignore": [
588
+ "language_model.lm_head",
589
+ "language_model.model.embed_tokens",
590
+ "language_model.model.layers.0.linear_attn.in_proj_a",
591
+ "language_model.model.layers.0.linear_attn.in_proj_b",
592
+ "language_model.model.layers.0.linear_attn.in_proj_ba",
593
+ "language_model.model.layers.0.linear_attn.in_proj_qkv",
594
+ "language_model.model.layers.0.linear_attn.in_proj_qkvz",
595
+ "language_model.model.layers.0.linear_attn.in_proj_z",
596
+ "language_model.model.layers.1.linear_attn.in_proj_a",
597
+ "language_model.model.layers.1.linear_attn.in_proj_b",
598
+ "language_model.model.layers.1.linear_attn.in_proj_ba",
599
+ "language_model.model.layers.1.linear_attn.in_proj_qkv",
600
+ "language_model.model.layers.1.linear_attn.in_proj_qkvz",
601
+ "language_model.model.layers.1.linear_attn.in_proj_z",
602
+ "language_model.model.layers.10.linear_attn.in_proj_a",
603
+ "language_model.model.layers.10.linear_attn.in_proj_b",
604
+ "language_model.model.layers.10.linear_attn.in_proj_ba",
605
+ "language_model.model.layers.10.linear_attn.in_proj_qkv",
606
+ "language_model.model.layers.10.linear_attn.in_proj_qkvz",
607
+ "language_model.model.layers.10.linear_attn.in_proj_z",
608
+ "language_model.model.layers.11.self_attn.k_proj",
609
+ "language_model.model.layers.11.self_attn.q_proj",
610
+ "language_model.model.layers.11.self_attn.qkv_proj",
611
+ "language_model.model.layers.11.self_attn.v_proj",
612
+ "language_model.model.layers.12.linear_attn.in_proj_a",
613
+ "language_model.model.layers.12.linear_attn.in_proj_b",
614
+ "language_model.model.layers.12.linear_attn.in_proj_ba",
615
+ "language_model.model.layers.12.linear_attn.in_proj_qkv",
616
+ "language_model.model.layers.12.linear_attn.in_proj_qkvz",
617
+ "language_model.model.layers.12.linear_attn.in_proj_z",
618
+ "language_model.model.layers.13.linear_attn.in_proj_a",
619
+ "language_model.model.layers.13.linear_attn.in_proj_b",
620
+ "language_model.model.layers.13.linear_attn.in_proj_ba",
621
+ "language_model.model.layers.13.linear_attn.in_proj_qkv",
622
+ "language_model.model.layers.13.linear_attn.in_proj_qkvz",
623
+ "language_model.model.layers.13.linear_attn.in_proj_z",
624
+ "language_model.model.layers.14.linear_attn.in_proj_a",
625
+ "language_model.model.layers.14.linear_attn.in_proj_b",
626
+ "language_model.model.layers.14.linear_attn.in_proj_ba",
627
+ "language_model.model.layers.14.linear_attn.in_proj_qkv",
628
+ "language_model.model.layers.14.linear_attn.in_proj_qkvz",
629
+ "language_model.model.layers.14.linear_attn.in_proj_z",
630
+ "language_model.model.layers.15.self_attn.k_proj",
631
+ "language_model.model.layers.15.self_attn.q_proj",
632
+ "language_model.model.layers.15.self_attn.qkv_proj",
633
+ "language_model.model.layers.15.self_attn.v_proj",
634
+ "language_model.model.layers.16.linear_attn.in_proj_a",
635
+ "language_model.model.layers.16.linear_attn.in_proj_b",
636
+ "language_model.model.layers.16.linear_attn.in_proj_ba",
637
+ "language_model.model.layers.16.linear_attn.in_proj_qkv",
638
+ "language_model.model.layers.16.linear_attn.in_proj_qkvz",
639
+ "language_model.model.layers.16.linear_attn.in_proj_z",
640
+ "language_model.model.layers.17.linear_attn.in_proj_a",
641
+ "language_model.model.layers.17.linear_attn.in_proj_b",
642
+ "language_model.model.layers.17.linear_attn.in_proj_ba",
643
+ "language_model.model.layers.17.linear_attn.in_proj_qkv",
644
+ "language_model.model.layers.17.linear_attn.in_proj_qkvz",
645
+ "language_model.model.layers.17.linear_attn.in_proj_z",
646
+ "language_model.model.layers.18.linear_attn.in_proj_a",
647
+ "language_model.model.layers.18.linear_attn.in_proj_b",
648
+ "language_model.model.layers.18.linear_attn.in_proj_ba",
649
+ "language_model.model.layers.18.linear_attn.in_proj_qkv",
650
+ "language_model.model.layers.18.linear_attn.in_proj_qkvz",
651
+ "language_model.model.layers.18.linear_attn.in_proj_z",
652
+ "language_model.model.layers.19.self_attn.k_proj",
653
+ "language_model.model.layers.19.self_attn.q_proj",
654
+ "language_model.model.layers.19.self_attn.qkv_proj",
655
+ "language_model.model.layers.19.self_attn.v_proj",
656
+ "language_model.model.layers.2.linear_attn.in_proj_a",
657
+ "language_model.model.layers.2.linear_attn.in_proj_b",
658
+ "language_model.model.layers.2.linear_attn.in_proj_ba",
659
+ "language_model.model.layers.2.linear_attn.in_proj_qkv",
660
+ "language_model.model.layers.2.linear_attn.in_proj_qkvz",
661
+ "language_model.model.layers.2.linear_attn.in_proj_z",
662
+ "language_model.model.layers.20.linear_attn.in_proj_a",
663
+ "language_model.model.layers.20.linear_attn.in_proj_b",
664
+ "language_model.model.layers.20.linear_attn.in_proj_ba",
665
+ "language_model.model.layers.20.linear_attn.in_proj_qkv",
666
+ "language_model.model.layers.20.linear_attn.in_proj_qkvz",
667
+ "language_model.model.layers.20.linear_attn.in_proj_z",
668
+ "language_model.model.layers.21.linear_attn.in_proj_a",
669
+ "language_model.model.layers.21.linear_attn.in_proj_b",
670
+ "language_model.model.layers.21.linear_attn.in_proj_ba",
671
+ "language_model.model.layers.21.linear_attn.in_proj_qkv",
672
+ "language_model.model.layers.21.linear_attn.in_proj_qkvz",
673
+ "language_model.model.layers.21.linear_attn.in_proj_z",
674
+ "language_model.model.layers.22.linear_attn.in_proj_a",
675
+ "language_model.model.layers.22.linear_attn.in_proj_b",
676
+ "language_model.model.layers.22.linear_attn.in_proj_ba",
677
+ "language_model.model.layers.22.linear_attn.in_proj_qkv",
678
+ "language_model.model.layers.22.linear_attn.in_proj_qkvz",
679
+ "language_model.model.layers.22.linear_attn.in_proj_z",
680
+ "language_model.model.layers.23.self_attn.k_proj",
681
+ "language_model.model.layers.23.self_attn.q_proj",
682
+ "language_model.model.layers.23.self_attn.qkv_proj",
683
+ "language_model.model.layers.23.self_attn.v_proj",
684
+ "language_model.model.layers.24.linear_attn.in_proj_a",
685
+ "language_model.model.layers.24.linear_attn.in_proj_b",
686
+ "language_model.model.layers.24.linear_attn.in_proj_ba",
687
+ "language_model.model.layers.24.linear_attn.in_proj_qkv",
688
+ "language_model.model.layers.24.linear_attn.in_proj_qkvz",
689
+ "language_model.model.layers.24.linear_attn.in_proj_z",
690
+ "language_model.model.layers.25.linear_attn.in_proj_a",
691
+ "language_model.model.layers.25.linear_attn.in_proj_b",
692
+ "language_model.model.layers.25.linear_attn.in_proj_ba",
693
+ "language_model.model.layers.25.linear_attn.in_proj_qkv",
694
+ "language_model.model.layers.25.linear_attn.in_proj_qkvz",
695
+ "language_model.model.layers.25.linear_attn.in_proj_z",
696
+ "language_model.model.layers.26.linear_attn.in_proj_a",
697
+ "language_model.model.layers.26.linear_attn.in_proj_b",
698
+ "language_model.model.layers.26.linear_attn.in_proj_ba",
699
+ "language_model.model.layers.26.linear_attn.in_proj_qkv",
700
+ "language_model.model.layers.26.linear_attn.in_proj_qkvz",
701
+ "language_model.model.layers.26.linear_attn.in_proj_z",
702
+ "language_model.model.layers.27.self_attn.k_proj",
703
+ "language_model.model.layers.27.self_attn.q_proj",
704
+ "language_model.model.layers.27.self_attn.qkv_proj",
705
+ "language_model.model.layers.27.self_attn.v_proj",
706
+ "language_model.model.layers.28.linear_attn.in_proj_a",
707
+ "language_model.model.layers.28.linear_attn.in_proj_b",
708
+ "language_model.model.layers.28.linear_attn.in_proj_ba",
709
+ "language_model.model.layers.28.linear_attn.in_proj_qkv",
710
+ "language_model.model.layers.28.linear_attn.in_proj_qkvz",
711
+ "language_model.model.layers.28.linear_attn.in_proj_z",
712
+ "language_model.model.layers.29.linear_attn.in_proj_a",
713
+ "language_model.model.layers.29.linear_attn.in_proj_b",
714
+ "language_model.model.layers.29.linear_attn.in_proj_ba",
715
+ "language_model.model.layers.29.linear_attn.in_proj_qkv",
716
+ "language_model.model.layers.29.linear_attn.in_proj_qkvz",
717
+ "language_model.model.layers.29.linear_attn.in_proj_z",
718
+ "language_model.model.layers.3.self_attn.k_proj",
719
+ "language_model.model.layers.3.self_attn.q_proj",
720
+ "language_model.model.layers.3.self_attn.qkv_proj",
721
+ "language_model.model.layers.3.self_attn.v_proj",
722
+ "language_model.model.layers.30.linear_attn.in_proj_a",
723
+ "language_model.model.layers.30.linear_attn.in_proj_b",
724
+ "language_model.model.layers.30.linear_attn.in_proj_ba",
725
+ "language_model.model.layers.30.linear_attn.in_proj_qkv",
726
+ "language_model.model.layers.30.linear_attn.in_proj_qkvz",
727
+ "language_model.model.layers.30.linear_attn.in_proj_z",
728
+ "language_model.model.layers.31.self_attn.k_proj",
729
+ "language_model.model.layers.31.self_attn.q_proj",
730
+ "language_model.model.layers.31.self_attn.qkv_proj",
731
+ "language_model.model.layers.31.self_attn.v_proj",
732
+ "language_model.model.layers.32.linear_attn.in_proj_a",
733
+ "language_model.model.layers.32.linear_attn.in_proj_b",
734
+ "language_model.model.layers.32.linear_attn.in_proj_ba",
735
+ "language_model.model.layers.32.linear_attn.in_proj_qkv",
736
+ "language_model.model.layers.32.linear_attn.in_proj_qkvz",
737
+ "language_model.model.layers.32.linear_attn.in_proj_z",
738
+ "language_model.model.layers.33.linear_attn.in_proj_a",
739
+ "language_model.model.layers.33.linear_attn.in_proj_b",
740
+ "language_model.model.layers.33.linear_attn.in_proj_ba",
741
+ "language_model.model.layers.33.linear_attn.in_proj_qkv",
742
+ "language_model.model.layers.33.linear_attn.in_proj_qkvz",
743
+ "language_model.model.layers.33.linear_attn.in_proj_z",
744
+ "language_model.model.layers.34.linear_attn.in_proj_a",
745
+ "language_model.model.layers.34.linear_attn.in_proj_b",
746
+ "language_model.model.layers.34.linear_attn.in_proj_ba",
747
+ "language_model.model.layers.34.linear_attn.in_proj_qkv",
748
+ "language_model.model.layers.34.linear_attn.in_proj_qkvz",
749
+ "language_model.model.layers.34.linear_attn.in_proj_z",
750
+ "language_model.model.layers.35.self_attn.k_proj",
751
+ "language_model.model.layers.35.self_attn.q_proj",
752
+ "language_model.model.layers.35.self_attn.qkv_proj",
753
+ "language_model.model.layers.35.self_attn.v_proj",
754
+ "language_model.model.layers.36.linear_attn.in_proj_a",
755
+ "language_model.model.layers.36.linear_attn.in_proj_b",
756
+ "language_model.model.layers.36.linear_attn.in_proj_ba",
757
+ "language_model.model.layers.36.linear_attn.in_proj_qkv",
758
+ "language_model.model.layers.36.linear_attn.in_proj_qkvz",
759
+ "language_model.model.layers.36.linear_attn.in_proj_z",
760
+ "language_model.model.layers.37.linear_attn.in_proj_a",
761
+ "language_model.model.layers.37.linear_attn.in_proj_b",
762
+ "language_model.model.layers.37.linear_attn.in_proj_ba",
763
+ "language_model.model.layers.37.linear_attn.in_proj_qkv",
764
+ "language_model.model.layers.37.linear_attn.in_proj_qkvz",
765
+ "language_model.model.layers.37.linear_attn.in_proj_z",
766
+ "language_model.model.layers.38.linear_attn.in_proj_a",
767
+ "language_model.model.layers.38.linear_attn.in_proj_b",
768
+ "language_model.model.layers.38.linear_attn.in_proj_ba",
769
+ "language_model.model.layers.4.linear_attn.in_proj_a",
770
+ "language_model.model.layers.4.linear_attn.in_proj_b",
771
+ "language_model.model.layers.4.linear_attn.in_proj_ba",
772
+ "language_model.model.layers.40.linear_attn.in_proj_a",
773
+ "language_model.model.layers.40.linear_attn.in_proj_b",
774
+ "language_model.model.layers.40.linear_attn.in_proj_ba",
775
+ "language_model.model.layers.41.linear_attn.in_proj_a",
776
+ "language_model.model.layers.41.linear_attn.in_proj_b",
777
+ "language_model.model.layers.41.linear_attn.in_proj_ba",
778
+ "language_model.model.layers.42.linear_attn.in_proj_a",
779
+ "language_model.model.layers.42.linear_attn.in_proj_b",
780
+ "language_model.model.layers.42.linear_attn.in_proj_ba",
781
+ "language_model.model.layers.43.self_attn.k_proj",
782
+ "language_model.model.layers.43.self_attn.q_proj",
783
+ "language_model.model.layers.43.self_attn.qkv_proj",
784
+ "language_model.model.layers.43.self_attn.v_proj",
785
+ "language_model.model.layers.44.linear_attn.in_proj_a",
786
+ "language_model.model.layers.44.linear_attn.in_proj_b",
787
+ "language_model.model.layers.44.linear_attn.in_proj_ba",
788
+ "language_model.model.layers.45.linear_attn.in_proj_a",
789
+ "language_model.model.layers.45.linear_attn.in_proj_b",
790
+ "language_model.model.layers.45.linear_attn.in_proj_ba",
791
+ "language_model.model.layers.46.linear_attn.in_proj_a",
792
+ "language_model.model.layers.46.linear_attn.in_proj_b",
793
+ "language_model.model.layers.46.linear_attn.in_proj_ba",
794
+ "language_model.model.layers.46.linear_attn.in_proj_qkv",
795
+ "language_model.model.layers.46.linear_attn.in_proj_qkvz",
796
+ "language_model.model.layers.46.linear_attn.in_proj_z",
797
+ "language_model.model.layers.47.self_attn.k_proj",
798
+ "language_model.model.layers.47.self_attn.q_proj",
799
+ "language_model.model.layers.47.self_attn.qkv_proj",
800
+ "language_model.model.layers.47.self_attn.v_proj",
801
+ "language_model.model.layers.48.linear_attn.in_proj_a",
802
+ "language_model.model.layers.48.linear_attn.in_proj_b",
803
+ "language_model.model.layers.48.linear_attn.in_proj_ba",
804
+ "language_model.model.layers.48.linear_attn.in_proj_qkv",
805
+ "language_model.model.layers.48.linear_attn.in_proj_qkvz",
806
+ "language_model.model.layers.48.linear_attn.in_proj_z",
807
+ "language_model.model.layers.49.linear_attn.in_proj_a",
808
+ "language_model.model.layers.49.linear_attn.in_proj_b",
809
+ "language_model.model.layers.49.linear_attn.in_proj_ba",
810
+ "language_model.model.layers.49.linear_attn.in_proj_qkv",
811
+ "language_model.model.layers.49.linear_attn.in_proj_qkvz",
812
+ "language_model.model.layers.49.linear_attn.in_proj_z",
813
+ "language_model.model.layers.5.linear_attn.in_proj_a",
814
+ "language_model.model.layers.5.linear_attn.in_proj_b",
815
+ "language_model.model.layers.5.linear_attn.in_proj_ba",
816
+ "language_model.model.layers.5.linear_attn.in_proj_qkv",
817
+ "language_model.model.layers.5.linear_attn.in_proj_qkvz",
818
+ "language_model.model.layers.5.linear_attn.in_proj_z",
819
+ "language_model.model.layers.50.linear_attn.in_proj_a",
820
+ "language_model.model.layers.50.linear_attn.in_proj_b",
821
+ "language_model.model.layers.50.linear_attn.in_proj_ba",
822
+ "language_model.model.layers.52.linear_attn.in_proj_a",
823
+ "language_model.model.layers.52.linear_attn.in_proj_b",
824
+ "language_model.model.layers.52.linear_attn.in_proj_ba",
825
+ "language_model.model.layers.53.linear_attn.in_proj_a",
826
+ "language_model.model.layers.53.linear_attn.in_proj_b",
827
+ "language_model.model.layers.53.linear_attn.in_proj_ba",
828
+ "language_model.model.layers.54.linear_attn.in_proj_a",
829
+ "language_model.model.layers.54.linear_attn.in_proj_b",
830
+ "language_model.model.layers.54.linear_attn.in_proj_ba",
831
+ "language_model.model.layers.56.linear_attn.in_proj_a",
832
+ "language_model.model.layers.56.linear_attn.in_proj_b",
833
+ "language_model.model.layers.56.linear_attn.in_proj_ba",
834
+ "language_model.model.layers.57.linear_attn.in_proj_a",
835
+ "language_model.model.layers.57.linear_attn.in_proj_b",
836
+ "language_model.model.layers.57.linear_attn.in_proj_ba",
837
+ "language_model.model.layers.58.linear_attn.in_proj_a",
838
+ "language_model.model.layers.58.linear_attn.in_proj_b",
839
+ "language_model.model.layers.58.linear_attn.in_proj_ba",
840
+ "language_model.model.layers.58.mlp.down_proj",
841
+ "language_model.model.layers.6.linear_attn.in_proj_a",
842
+ "language_model.model.layers.6.linear_attn.in_proj_b",
843
+ "language_model.model.layers.6.linear_attn.in_proj_ba",
844
+ "language_model.model.layers.6.linear_attn.in_proj_qkv",
845
+ "language_model.model.layers.6.linear_attn.in_proj_qkvz",
846
+ "language_model.model.layers.6.linear_attn.in_proj_z",
847
+ "language_model.model.layers.60.linear_attn.in_proj_a",
848
+ "language_model.model.layers.60.linear_attn.in_proj_b",
849
+ "language_model.model.layers.60.linear_attn.in_proj_ba",
850
+ "language_model.model.layers.60.linear_attn.out_proj",
851
+ "language_model.model.layers.61.linear_attn.in_proj_a",
852
+ "language_model.model.layers.61.linear_attn.in_proj_b",
853
+ "language_model.model.layers.61.linear_attn.in_proj_ba",
854
+ "language_model.model.layers.62.linear_attn.in_proj_a",
855
+ "language_model.model.layers.62.linear_attn.in_proj_b",
856
+ "language_model.model.layers.62.linear_attn.in_proj_ba",
857
+ "language_model.model.layers.62.mlp.down_proj",
858
+ "language_model.model.layers.63.mlp.down_proj",
859
+ "language_model.model.layers.7.self_attn.k_proj",
860
+ "language_model.model.layers.7.self_attn.q_proj",
861
+ "language_model.model.layers.7.self_attn.qkv_proj",
862
+ "language_model.model.layers.7.self_attn.v_proj",
863
+ "language_model.model.layers.8.linear_attn.in_proj_a",
864
+ "language_model.model.layers.8.linear_attn.in_proj_b",
865
+ "language_model.model.layers.8.linear_attn.in_proj_ba",
866
+ "language_model.model.layers.8.linear_attn.in_proj_qkv",
867
+ "language_model.model.layers.8.linear_attn.in_proj_qkvz",
868
+ "language_model.model.layers.8.linear_attn.in_proj_z",
869
+ "language_model.model.layers.9.linear_attn.in_proj_a",
870
+ "language_model.model.layers.9.linear_attn.in_proj_b",
871
+ "language_model.model.layers.9.linear_attn.in_proj_ba",
872
+ "language_model.model.layers.9.linear_attn.in_proj_qkv",
873
+ "language_model.model.layers.9.linear_attn.in_proj_qkvz",
874
+ "language_model.model.layers.9.linear_attn.in_proj_z",
875
+ "mtp.fc",
876
+ "mtp.layers.0.mlp.down_proj",
877
+ "mtp.layers.0.mlp.gate_proj",
878
+ "mtp.layers.0.mlp.gate_up_proj",
879
+ "mtp.layers.0.mlp.up_proj",
880
+ "mtp.layers.0.self_attn.k_proj",
881
+ "mtp.layers.0.self_attn.o_proj",
882
+ "mtp.layers.0.self_attn.q_proj",
883
+ "mtp.layers.0.self_attn.qkv_proj",
884
+ "mtp.layers.0.self_attn.v_proj",
885
+ "visual.blocks.0.attn.proj",
886
+ "visual.blocks.0.attn.qkv",
887
+ "visual.blocks.0.mlp.linear_fc1",
888
+ "visual.blocks.0.mlp.linear_fc2",
889
+ "visual.blocks.1.attn.proj",
890
+ "visual.blocks.1.attn.qkv",
891
+ "visual.blocks.1.mlp.linear_fc1",
892
+ "visual.blocks.1.mlp.linear_fc2",
893
+ "visual.blocks.10.attn.proj",
894
+ "visual.blocks.10.attn.qkv",
895
+ "visual.blocks.10.mlp.linear_fc1",
896
+ "visual.blocks.10.mlp.linear_fc2",
897
+ "visual.blocks.11.attn.proj",
898
+ "visual.blocks.11.attn.qkv",
899
+ "visual.blocks.11.mlp.linear_fc1",
900
+ "visual.blocks.11.mlp.linear_fc2",
901
+ "visual.blocks.12.attn.proj",
902
+ "visual.blocks.12.attn.qkv",
903
+ "visual.blocks.12.mlp.linear_fc1",
904
+ "visual.blocks.12.mlp.linear_fc2",
905
+ "visual.blocks.13.attn.proj",
906
+ "visual.blocks.13.attn.qkv",
907
+ "visual.blocks.13.mlp.linear_fc1",
908
+ "visual.blocks.13.mlp.linear_fc2",
909
+ "visual.blocks.14.attn.proj",
910
+ "visual.blocks.14.attn.qkv",
911
+ "visual.blocks.14.mlp.linear_fc1",
912
+ "visual.blocks.14.mlp.linear_fc2",
913
+ "visual.blocks.15.attn.proj",
914
+ "visual.blocks.15.attn.qkv",
915
+ "visual.blocks.15.mlp.linear_fc1",
916
+ "visual.blocks.15.mlp.linear_fc2",
917
+ "visual.blocks.16.attn.proj",
918
+ "visual.blocks.16.attn.qkv",
919
+ "visual.blocks.16.mlp.linear_fc1",
920
+ "visual.blocks.16.mlp.linear_fc2",
921
+ "visual.blocks.17.attn.proj",
922
+ "visual.blocks.17.attn.qkv",
923
+ "visual.blocks.17.mlp.linear_fc1",
924
+ "visual.blocks.17.mlp.linear_fc2",
925
+ "visual.blocks.18.attn.proj",
926
+ "visual.blocks.18.attn.qkv",
927
+ "visual.blocks.18.mlp.linear_fc1",
928
+ "visual.blocks.18.mlp.linear_fc2",
929
+ "visual.blocks.19.attn.proj",
930
+ "visual.blocks.19.attn.qkv",
931
+ "visual.blocks.19.mlp.linear_fc1",
932
+ "visual.blocks.19.mlp.linear_fc2",
933
+ "visual.blocks.2.attn.proj",
934
+ "visual.blocks.2.attn.qkv",
935
+ "visual.blocks.2.mlp.linear_fc1",
936
+ "visual.blocks.2.mlp.linear_fc2",
937
+ "visual.blocks.20.attn.proj",
938
+ "visual.blocks.20.attn.qkv",
939
+ "visual.blocks.20.mlp.linear_fc1",
940
+ "visual.blocks.20.mlp.linear_fc2",
941
+ "visual.blocks.21.attn.proj",
942
+ "visual.blocks.21.attn.qkv",
943
+ "visual.blocks.21.mlp.linear_fc1",
944
+ "visual.blocks.21.mlp.linear_fc2",
945
+ "visual.blocks.22.attn.proj",
946
+ "visual.blocks.22.attn.qkv",
947
+ "visual.blocks.22.mlp.linear_fc1",
948
+ "visual.blocks.22.mlp.linear_fc2",
949
+ "visual.blocks.23.attn.proj",
950
+ "visual.blocks.23.attn.qkv",
951
+ "visual.blocks.23.mlp.linear_fc1",
952
+ "visual.blocks.23.mlp.linear_fc2",
953
+ "visual.blocks.24.attn.proj",
954
+ "visual.blocks.24.attn.qkv",
955
+ "visual.blocks.24.mlp.linear_fc1",
956
+ "visual.blocks.24.mlp.linear_fc2",
957
+ "visual.blocks.25.attn.proj",
958
+ "visual.blocks.25.attn.qkv",
959
+ "visual.blocks.25.mlp.linear_fc1",
960
+ "visual.blocks.25.mlp.linear_fc2",
961
+ "visual.blocks.26.attn.proj",
962
+ "visual.blocks.26.attn.qkv",
963
+ "visual.blocks.26.mlp.linear_fc1",
964
+ "visual.blocks.26.mlp.linear_fc2",
965
+ "visual.blocks.3.attn.proj",
966
+ "visual.blocks.3.attn.qkv",
967
+ "visual.blocks.3.mlp.linear_fc1",
968
+ "visual.blocks.3.mlp.linear_fc2",
969
+ "visual.blocks.4.attn.proj",
970
+ "visual.blocks.4.attn.qkv",
971
+ "visual.blocks.4.mlp.linear_fc1",
972
+ "visual.blocks.4.mlp.linear_fc2",
973
+ "visual.blocks.5.attn.proj",
974
+ "visual.blocks.5.attn.qkv",
975
+ "visual.blocks.5.mlp.linear_fc1",
976
+ "visual.blocks.5.mlp.linear_fc2",
977
+ "visual.blocks.6.attn.proj",
978
+ "visual.blocks.6.attn.qkv",
979
+ "visual.blocks.6.mlp.linear_fc1",
980
+ "visual.blocks.6.mlp.linear_fc2",
981
+ "visual.blocks.7.attn.proj",
982
+ "visual.blocks.7.attn.qkv",
983
+ "visual.blocks.7.mlp.linear_fc1",
984
+ "visual.blocks.7.mlp.linear_fc2",
985
+ "visual.blocks.8.attn.proj",
986
+ "visual.blocks.8.attn.qkv",
987
+ "visual.blocks.8.mlp.linear_fc1",
988
+ "visual.blocks.8.mlp.linear_fc2",
989
+ "visual.blocks.9.attn.proj",
990
+ "visual.blocks.9.attn.qkv",
991
+ "visual.blocks.9.mlp.linear_fc1",
992
+ "visual.blocks.9.mlp.linear_fc2",
993
+ "visual.merger.linear_fc1",
994
+ "visual.merger.linear_fc2",
995
+ "visual.pos_embed"
996
+ ],
997
+ "quantization_status": "compressed"
998
+ }
999
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 248044,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 248046,
6
+ 248044
7
+ ],
8
+ "pad_token_id": 248044,
9
+ "temperature": 1.0,
10
+ "top_k": 20,
11
+ "top_p": 0.95,
12
+ "transformers_version": "5.8.0"
13
+ }
mixed_native_manifest.json ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8375995a6b7630e8740a0c83b2f27be223af072836caa5845745bac69f0c5f05
3
+ size 5337029568
model-00002-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e23aba4be5b51566c0c5f204a13aec8a0b97f41b2832155e5c20db10e25f765c
3
+ size 5323328336
model-00003-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe64034283849f5a3e066e6fc346774623ed716da6db2aa31f334f4702a422bb
3
+ size 5317082704
model-00004-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85a6169cb9d87b34617ee8e0afb3edb42f8a691bf7af883dd1c17c80e5235146
3
+ size 5265568632
model-00005-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92fa472320176c26115a67c07687549509384cb211281d63a2db5845b982639f
3
+ size 5282606760
model-00006-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1965bf75b47e27785d1f7133c67692a7e20568962054c1e665e8378ded5fb121
3
+ size 4687075176
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "size": {
3
+ "longest_edge": 16777216,
4
+ "shortest_edge": 65536
5
+ },
6
+ "patch_size": 16,
7
+ "temporal_patch_size": 2,
8
+ "merge_size": 2,
9
+ "image_mean": [
10
+ 0.5,
11
+ 0.5,
12
+ 0.5
13
+ ],
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "processor_class": "Qwen3VLProcessor",
20
+ "image_processor_type": "Qwen2VLImageProcessorFast"
21
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f32ce20dc35f57a7f9ad1eac03525bd7d30f9df8cea6507e958279cc3657706
3
+ size 19989492
tokenizer_config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "audio_bos_token": "<|audio_start|>",
4
+ "audio_eos_token": "<|audio_end|>",
5
+ "audio_token": "<|audio_pad|>",
6
+ "backend": "tokenizers",
7
+ "bos_token": null,
8
+ "clean_up_tokenization_spaces": false,
9
+ "eos_token": "<|im_end|>",
10
+ "errors": "replace",
11
+ "image_token": "<|image_pad|>",
12
+ "is_local": true,
13
+ "local_files_only": false,
14
+ "model_max_length": 262144,
15
+ "model_specific_special_tokens": {
16
+ "audio_bos_token": "<|audio_start|>",
17
+ "audio_eos_token": "<|audio_end|>",
18
+ "audio_token": "<|audio_pad|>",
19
+ "image_token": "<|image_pad|>",
20
+ "video_token": "<|video_pad|>",
21
+ "vision_bos_token": "<|vision_start|>",
22
+ "vision_eos_token": "<|vision_end|>"
23
+ },
24
+ "pad_token": "<|endoftext|>",
25
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
26
+ "split_special_tokens": false,
27
+ "tokenizer_class": "Qwen2Tokenizer",
28
+ "unk_token": null,
29
+ "video_token": "<|video_pad|>",
30
+ "vision_bos_token": "<|vision_start|>",
31
+ "vision_eos_token": "<|vision_end|>"
32
+ }
video_preprocessor_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "size": {
3
+ "longest_edge": 25165824,
4
+ "shortest_edge": 4096
5
+ },
6
+ "patch_size": 16,
7
+ "temporal_patch_size": 2,
8
+ "merge_size": 2,
9
+ "image_mean": [
10
+ 0.5,
11
+ 0.5,
12
+ 0.5
13
+ ],
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "processor_class": "Qwen3VLProcessor",
20
+ "video_processor_type": "Qwen3VLVideoProcessor"
21
+ }