nightmedia commited on
Commit
12ac7fa
·
verified ·
1 Parent(s): c297af6

Add files using upload-large-folder tool

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
README.md ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ language:
4
+ - en
5
+ - zh
6
+ base_model: YOYO-AI/Qwen3-30B-A3B-YOYO-V4
7
+ pipeline_tag: text-generation
8
+ tags:
9
+ - merge
10
+ - mlx
11
+ library_name: mlx
12
+ ---
13
+
14
+ # Qwen3-30B-A3B-YOYO-V4-qx86x-mlx
15
+
16
+ This model [Qwen3-30B-A3B-YOYO-V4-qx86x-mlx](https://huggingface.co/Qwen3-30B-A3B-YOYO-V4-qx86x-mlx) was
17
+ converted to MLX format from [YOYO-AI/Qwen3-30B-A3B-YOYO-V4](https://huggingface.co/YOYO-AI/Qwen3-30B-A3B-YOYO-V4)
18
+ using mlx-lm version **0.28.3**.
19
+
20
+ ## Use with mlx
21
+
22
+ ```bash
23
+ pip install mlx-lm
24
+ ```
25
+
26
+ ```python
27
+ from mlx_lm import load, generate
28
+
29
+ model, tokenizer = load("Qwen3-30B-A3B-YOYO-V4-qx86x-mlx")
30
+
31
+ prompt = "hello"
32
+
33
+ if tokenizer.chat_template is not None:
34
+ messages = [{"role": "user", "content": prompt}]
35
+ prompt = tokenizer.apply_chat_template(
36
+ messages, add_generation_prompt=True
37
+ )
38
+
39
+ response = generate(model, tokenizer, prompt=prompt, verbose=True)
40
+ ```
added_tokens.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</think>": 151668,
3
+ "</tool_call>": 151658,
4
+ "</tool_response>": 151666,
5
+ "<think>": 151667,
6
+ "<tool_call>": 151657,
7
+ "<tool_response>": 151665,
8
+ "<|box_end|>": 151649,
9
+ "<|box_start|>": 151648,
10
+ "<|endoftext|>": 151643,
11
+ "<|file_sep|>": 151664,
12
+ "<|fim_middle|>": 151660,
13
+ "<|fim_pad|>": 151662,
14
+ "<|fim_prefix|>": 151659,
15
+ "<|fim_suffix|>": 151661,
16
+ "<|im_end|>": 151645,
17
+ "<|im_start|>": 151644,
18
+ "<|image_pad|>": 151655,
19
+ "<|object_ref_end|>": 151647,
20
+ "<|object_ref_start|>": 151646,
21
+ "<|quad_end|>": 151651,
22
+ "<|quad_start|>": 151650,
23
+ "<|repo_name|>": 151663,
24
+ "<|video_pad|>": 151656,
25
+ "<|vision_end|>": 151653,
26
+ "<|vision_pad|>": 151654,
27
+ "<|vision_start|>": 151652
28
+ }
chat_template.jinja ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# 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>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\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" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if message.content is string %}
27
+ {%- set content = message.content %}
28
+ {%- else %}
29
+ {%- set content = '' %}
30
+ {%- endif %}
31
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
32
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
33
+ {%- elif message.role == "assistant" %}
34
+ {%- set reasoning_content = '' %}
35
+ {%- if message.reasoning_content is string %}
36
+ {%- set reasoning_content = message.reasoning_content %}
37
+ {%- else %}
38
+ {%- if '</think>' in content %}
39
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
40
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
41
+ {%- endif %}
42
+ {%- endif %}
43
+ {%- if loop.index0 > ns.last_query_index %}
44
+ {%- if loop.last or (not loop.last and reasoning_content) %}
45
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
46
+ {%- else %}
47
+ {{- '<|im_start|>' + message.role + '\n' + content }}
48
+ {%- endif %}
49
+ {%- else %}
50
+ {{- '<|im_start|>' + message.role + '\n' + content }}
51
+ {%- endif %}
52
+ {%- if message.tool_calls %}
53
+ {%- for tool_call in message.tool_calls %}
54
+ {%- if (loop.first and content) or (not loop.first) %}
55
+ {{- '\n' }}
56
+ {%- endif %}
57
+ {%- if tool_call.function %}
58
+ {%- set tool_call = tool_call.function %}
59
+ {%- endif %}
60
+ {{- '<tool_call>\n{"name": "' }}
61
+ {{- tool_call.name }}
62
+ {{- '", "arguments": ' }}
63
+ {%- if tool_call.arguments is string %}
64
+ {{- tool_call.arguments }}
65
+ {%- else %}
66
+ {{- tool_call.arguments | tojson }}
67
+ {%- endif %}
68
+ {{- '}\n</tool_call>' }}
69
+ {%- endfor %}
70
+ {%- endif %}
71
+ {{- '<|im_end|>\n' }}
72
+ {%- elif message.role == "tool" %}
73
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
74
+ {{- '<|im_start|>user' }}
75
+ {%- endif %}
76
+ {{- '\n<tool_response>\n' }}
77
+ {{- content }}
78
+ {{- '\n</tool_response>' }}
79
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
80
+ {{- '<|im_end|>\n' }}
81
+ {%- endif %}
82
+ {%- endif %}
83
+ {%- endfor %}
84
+ {%- if add_generation_prompt %}
85
+ {{- '<|im_start|>assistant\n' }}
86
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,3136 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3MoeForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "decoder_sparse_step": 1,
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 6144,
15
+ "max_position_embeddings": 262144,
16
+ "max_window_layers": 48,
17
+ "mlp_only_layers": [],
18
+ "model_type": "qwen3_moe",
19
+ "moe_intermediate_size": 768,
20
+ "norm_topk_prob": true,
21
+ "num_attention_heads": 32,
22
+ "num_experts": 128,
23
+ "num_experts_per_tok": 8,
24
+ "num_hidden_layers": 48,
25
+ "num_key_value_heads": 4,
26
+ "output_router_logits": false,
27
+ "quantization": {
28
+ "group_size": 64,
29
+ "bits": 8,
30
+ "mode": "affine",
31
+ "model.embed_tokens": {
32
+ "group_size": 64,
33
+ "bits": 8
34
+ },
35
+ "model.layers.0.self_attn.q_proj": {
36
+ "group_size": 64,
37
+ "bits": 8
38
+ },
39
+ "model.layers.0.self_attn.k_proj": {
40
+ "group_size": 64,
41
+ "bits": 8
42
+ },
43
+ "model.layers.0.self_attn.v_proj": {
44
+ "group_size": 64,
45
+ "bits": 8
46
+ },
47
+ "model.layers.0.self_attn.o_proj": {
48
+ "group_size": 64,
49
+ "bits": 8
50
+ },
51
+ "model.layers.0.mlp.gate": {
52
+ "group_size": 64,
53
+ "bits": 8
54
+ },
55
+ "model.layers.0.mlp.switch_mlp.gate_proj": {
56
+ "group_size": 64,
57
+ "bits": 8
58
+ },
59
+ "model.layers.0.mlp.switch_mlp.up_proj": {
60
+ "group_size": 64,
61
+ "bits": 8
62
+ },
63
+ "model.layers.0.mlp.switch_mlp.down_proj": {
64
+ "group_size": 64,
65
+ "bits": 8
66
+ },
67
+ "model.layers.1.self_attn.q_proj": {
68
+ "group_size": 64,
69
+ "bits": 6
70
+ },
71
+ "model.layers.1.self_attn.k_proj": {
72
+ "group_size": 64,
73
+ "bits": 6
74
+ },
75
+ "model.layers.1.self_attn.v_proj": {
76
+ "group_size": 64,
77
+ "bits": 8
78
+ },
79
+ "model.layers.1.self_attn.o_proj": {
80
+ "group_size": 64,
81
+ "bits": 6
82
+ },
83
+ "model.layers.1.mlp.gate": {
84
+ "group_size": 64,
85
+ "bits": 6
86
+ },
87
+ "model.layers.1.mlp.switch_mlp.gate_proj": {
88
+ "group_size": 64,
89
+ "bits": 6
90
+ },
91
+ "model.layers.1.mlp.switch_mlp.up_proj": {
92
+ "group_size": 64,
93
+ "bits": 6
94
+ },
95
+ "model.layers.1.mlp.switch_mlp.down_proj": {
96
+ "group_size": 64,
97
+ "bits": 8
98
+ },
99
+ "model.layers.2.self_attn.q_proj": {
100
+ "group_size": 64,
101
+ "bits": 6
102
+ },
103
+ "model.layers.2.self_attn.k_proj": {
104
+ "group_size": 64,
105
+ "bits": 6
106
+ },
107
+ "model.layers.2.self_attn.v_proj": {
108
+ "group_size": 64,
109
+ "bits": 8
110
+ },
111
+ "model.layers.2.self_attn.o_proj": {
112
+ "group_size": 64,
113
+ "bits": 6
114
+ },
115
+ "model.layers.2.mlp.gate": {
116
+ "group_size": 64,
117
+ "bits": 6
118
+ },
119
+ "model.layers.2.mlp.switch_mlp.gate_proj": {
120
+ "group_size": 64,
121
+ "bits": 6
122
+ },
123
+ "model.layers.2.mlp.switch_mlp.up_proj": {
124
+ "group_size": 64,
125
+ "bits": 6
126
+ },
127
+ "model.layers.2.mlp.switch_mlp.down_proj": {
128
+ "group_size": 64,
129
+ "bits": 8
130
+ },
131
+ "model.layers.3.self_attn.q_proj": {
132
+ "group_size": 64,
133
+ "bits": 6
134
+ },
135
+ "model.layers.3.self_attn.k_proj": {
136
+ "group_size": 64,
137
+ "bits": 6
138
+ },
139
+ "model.layers.3.self_attn.v_proj": {
140
+ "group_size": 64,
141
+ "bits": 8
142
+ },
143
+ "model.layers.3.self_attn.o_proj": {
144
+ "group_size": 64,
145
+ "bits": 6
146
+ },
147
+ "model.layers.3.mlp.gate": {
148
+ "group_size": 64,
149
+ "bits": 6
150
+ },
151
+ "model.layers.3.mlp.switch_mlp.gate_proj": {
152
+ "group_size": 64,
153
+ "bits": 6
154
+ },
155
+ "model.layers.3.mlp.switch_mlp.up_proj": {
156
+ "group_size": 64,
157
+ "bits": 6
158
+ },
159
+ "model.layers.3.mlp.switch_mlp.down_proj": {
160
+ "group_size": 64,
161
+ "bits": 8
162
+ },
163
+ "model.layers.4.self_attn.q_proj": {
164
+ "group_size": 64,
165
+ "bits": 6
166
+ },
167
+ "model.layers.4.self_attn.k_proj": {
168
+ "group_size": 64,
169
+ "bits": 6
170
+ },
171
+ "model.layers.4.self_attn.v_proj": {
172
+ "group_size": 64,
173
+ "bits": 8
174
+ },
175
+ "model.layers.4.self_attn.o_proj": {
176
+ "group_size": 64,
177
+ "bits": 6
178
+ },
179
+ "model.layers.4.mlp.gate": {
180
+ "group_size": 64,
181
+ "bits": 6
182
+ },
183
+ "model.layers.4.mlp.switch_mlp.gate_proj": {
184
+ "group_size": 64,
185
+ "bits": 6
186
+ },
187
+ "model.layers.4.mlp.switch_mlp.up_proj": {
188
+ "group_size": 64,
189
+ "bits": 6
190
+ },
191
+ "model.layers.4.mlp.switch_mlp.down_proj": {
192
+ "group_size": 64,
193
+ "bits": 8
194
+ },
195
+ "model.layers.5.self_attn.q_proj": {
196
+ "group_size": 64,
197
+ "bits": 6
198
+ },
199
+ "model.layers.5.self_attn.k_proj": {
200
+ "group_size": 64,
201
+ "bits": 6
202
+ },
203
+ "model.layers.5.self_attn.v_proj": {
204
+ "group_size": 64,
205
+ "bits": 8
206
+ },
207
+ "model.layers.5.self_attn.o_proj": {
208
+ "group_size": 64,
209
+ "bits": 6
210
+ },
211
+ "model.layers.5.mlp.gate": {
212
+ "group_size": 64,
213
+ "bits": 6
214
+ },
215
+ "model.layers.5.mlp.switch_mlp.gate_proj": {
216
+ "group_size": 64,
217
+ "bits": 6
218
+ },
219
+ "model.layers.5.mlp.switch_mlp.up_proj": {
220
+ "group_size": 64,
221
+ "bits": 6
222
+ },
223
+ "model.layers.5.mlp.switch_mlp.down_proj": {
224
+ "group_size": 64,
225
+ "bits": 8
226
+ },
227
+ "model.layers.6.self_attn.q_proj": {
228
+ "group_size": 64,
229
+ "bits": 6
230
+ },
231
+ "model.layers.6.self_attn.k_proj": {
232
+ "group_size": 64,
233
+ "bits": 6
234
+ },
235
+ "model.layers.6.self_attn.v_proj": {
236
+ "group_size": 64,
237
+ "bits": 8
238
+ },
239
+ "model.layers.6.self_attn.o_proj": {
240
+ "group_size": 64,
241
+ "bits": 6
242
+ },
243
+ "model.layers.6.mlp.gate": {
244
+ "group_size": 64,
245
+ "bits": 6
246
+ },
247
+ "model.layers.6.mlp.switch_mlp.gate_proj": {
248
+ "group_size": 64,
249
+ "bits": 6
250
+ },
251
+ "model.layers.6.mlp.switch_mlp.up_proj": {
252
+ "group_size": 64,
253
+ "bits": 6
254
+ },
255
+ "model.layers.6.mlp.switch_mlp.down_proj": {
256
+ "group_size": 64,
257
+ "bits": 6
258
+ },
259
+ "model.layers.7.self_attn.q_proj": {
260
+ "group_size": 64,
261
+ "bits": 6
262
+ },
263
+ "model.layers.7.self_attn.k_proj": {
264
+ "group_size": 64,
265
+ "bits": 6
266
+ },
267
+ "model.layers.7.self_attn.v_proj": {
268
+ "group_size": 64,
269
+ "bits": 8
270
+ },
271
+ "model.layers.7.self_attn.o_proj": {
272
+ "group_size": 64,
273
+ "bits": 6
274
+ },
275
+ "model.layers.7.mlp.gate": {
276
+ "group_size": 64,
277
+ "bits": 6
278
+ },
279
+ "model.layers.7.mlp.switch_mlp.gate_proj": {
280
+ "group_size": 64,
281
+ "bits": 6
282
+ },
283
+ "model.layers.7.mlp.switch_mlp.up_proj": {
284
+ "group_size": 64,
285
+ "bits": 6
286
+ },
287
+ "model.layers.7.mlp.switch_mlp.down_proj": {
288
+ "group_size": 64,
289
+ "bits": 6
290
+ },
291
+ "model.layers.8.self_attn.q_proj": {
292
+ "group_size": 64,
293
+ "bits": 6
294
+ },
295
+ "model.layers.8.self_attn.k_proj": {
296
+ "group_size": 64,
297
+ "bits": 6
298
+ },
299
+ "model.layers.8.self_attn.v_proj": {
300
+ "group_size": 64,
301
+ "bits": 8
302
+ },
303
+ "model.layers.8.self_attn.o_proj": {
304
+ "group_size": 64,
305
+ "bits": 6
306
+ },
307
+ "model.layers.8.mlp.gate": {
308
+ "group_size": 64,
309
+ "bits": 6
310
+ },
311
+ "model.layers.8.mlp.switch_mlp.gate_proj": {
312
+ "group_size": 64,
313
+ "bits": 6
314
+ },
315
+ "model.layers.8.mlp.switch_mlp.up_proj": {
316
+ "group_size": 64,
317
+ "bits": 6
318
+ },
319
+ "model.layers.8.mlp.switch_mlp.down_proj": {
320
+ "group_size": 64,
321
+ "bits": 8
322
+ },
323
+ "model.layers.9.self_attn.q_proj": {
324
+ "group_size": 64,
325
+ "bits": 6
326
+ },
327
+ "model.layers.9.self_attn.k_proj": {
328
+ "group_size": 64,
329
+ "bits": 6
330
+ },
331
+ "model.layers.9.self_attn.v_proj": {
332
+ "group_size": 64,
333
+ "bits": 8
334
+ },
335
+ "model.layers.9.self_attn.o_proj": {
336
+ "group_size": 64,
337
+ "bits": 6
338
+ },
339
+ "model.layers.9.mlp.gate": {
340
+ "group_size": 64,
341
+ "bits": 6
342
+ },
343
+ "model.layers.9.mlp.switch_mlp.gate_proj": {
344
+ "group_size": 64,
345
+ "bits": 6
346
+ },
347
+ "model.layers.9.mlp.switch_mlp.up_proj": {
348
+ "group_size": 64,
349
+ "bits": 6
350
+ },
351
+ "model.layers.9.mlp.switch_mlp.down_proj": {
352
+ "group_size": 64,
353
+ "bits": 6
354
+ },
355
+ "model.layers.10.self_attn.q_proj": {
356
+ "group_size": 64,
357
+ "bits": 6
358
+ },
359
+ "model.layers.10.self_attn.k_proj": {
360
+ "group_size": 64,
361
+ "bits": 6
362
+ },
363
+ "model.layers.10.self_attn.v_proj": {
364
+ "group_size": 64,
365
+ "bits": 8
366
+ },
367
+ "model.layers.10.self_attn.o_proj": {
368
+ "group_size": 64,
369
+ "bits": 6
370
+ },
371
+ "model.layers.10.mlp.gate": {
372
+ "group_size": 64,
373
+ "bits": 6
374
+ },
375
+ "model.layers.10.mlp.switch_mlp.gate_proj": {
376
+ "group_size": 64,
377
+ "bits": 6
378
+ },
379
+ "model.layers.10.mlp.switch_mlp.up_proj": {
380
+ "group_size": 64,
381
+ "bits": 6
382
+ },
383
+ "model.layers.10.mlp.switch_mlp.down_proj": {
384
+ "group_size": 64,
385
+ "bits": 6
386
+ },
387
+ "model.layers.11.self_attn.q_proj": {
388
+ "group_size": 64,
389
+ "bits": 6
390
+ },
391
+ "model.layers.11.self_attn.k_proj": {
392
+ "group_size": 64,
393
+ "bits": 6
394
+ },
395
+ "model.layers.11.self_attn.v_proj": {
396
+ "group_size": 64,
397
+ "bits": 8
398
+ },
399
+ "model.layers.11.self_attn.o_proj": {
400
+ "group_size": 64,
401
+ "bits": 6
402
+ },
403
+ "model.layers.11.mlp.gate": {
404
+ "group_size": 64,
405
+ "bits": 6
406
+ },
407
+ "model.layers.11.mlp.switch_mlp.gate_proj": {
408
+ "group_size": 64,
409
+ "bits": 6
410
+ },
411
+ "model.layers.11.mlp.switch_mlp.up_proj": {
412
+ "group_size": 64,
413
+ "bits": 6
414
+ },
415
+ "model.layers.11.mlp.switch_mlp.down_proj": {
416
+ "group_size": 64,
417
+ "bits": 8
418
+ },
419
+ "model.layers.12.self_attn.q_proj": {
420
+ "group_size": 64,
421
+ "bits": 6
422
+ },
423
+ "model.layers.12.self_attn.k_proj": {
424
+ "group_size": 64,
425
+ "bits": 6
426
+ },
427
+ "model.layers.12.self_attn.v_proj": {
428
+ "group_size": 64,
429
+ "bits": 8
430
+ },
431
+ "model.layers.12.self_attn.o_proj": {
432
+ "group_size": 64,
433
+ "bits": 6
434
+ },
435
+ "model.layers.12.mlp.gate": {
436
+ "group_size": 64,
437
+ "bits": 6
438
+ },
439
+ "model.layers.12.mlp.switch_mlp.gate_proj": {
440
+ "group_size": 64,
441
+ "bits": 6
442
+ },
443
+ "model.layers.12.mlp.switch_mlp.up_proj": {
444
+ "group_size": 64,
445
+ "bits": 6
446
+ },
447
+ "model.layers.12.mlp.switch_mlp.down_proj": {
448
+ "group_size": 64,
449
+ "bits": 6
450
+ },
451
+ "model.layers.13.self_attn.q_proj": {
452
+ "group_size": 64,
453
+ "bits": 6
454
+ },
455
+ "model.layers.13.self_attn.k_proj": {
456
+ "group_size": 64,
457
+ "bits": 6
458
+ },
459
+ "model.layers.13.self_attn.v_proj": {
460
+ "group_size": 64,
461
+ "bits": 8
462
+ },
463
+ "model.layers.13.self_attn.o_proj": {
464
+ "group_size": 64,
465
+ "bits": 6
466
+ },
467
+ "model.layers.13.mlp.gate": {
468
+ "group_size": 64,
469
+ "bits": 6
470
+ },
471
+ "model.layers.13.mlp.switch_mlp.gate_proj": {
472
+ "group_size": 64,
473
+ "bits": 6
474
+ },
475
+ "model.layers.13.mlp.switch_mlp.up_proj": {
476
+ "group_size": 64,
477
+ "bits": 6
478
+ },
479
+ "model.layers.13.mlp.switch_mlp.down_proj": {
480
+ "group_size": 64,
481
+ "bits": 6
482
+ },
483
+ "model.layers.14.self_attn.q_proj": {
484
+ "group_size": 64,
485
+ "bits": 6
486
+ },
487
+ "model.layers.14.self_attn.k_proj": {
488
+ "group_size": 64,
489
+ "bits": 6
490
+ },
491
+ "model.layers.14.self_attn.v_proj": {
492
+ "group_size": 64,
493
+ "bits": 8
494
+ },
495
+ "model.layers.14.self_attn.o_proj": {
496
+ "group_size": 64,
497
+ "bits": 6
498
+ },
499
+ "model.layers.14.mlp.gate": {
500
+ "group_size": 64,
501
+ "bits": 6
502
+ },
503
+ "model.layers.14.mlp.switch_mlp.gate_proj": {
504
+ "group_size": 64,
505
+ "bits": 6
506
+ },
507
+ "model.layers.14.mlp.switch_mlp.up_proj": {
508
+ "group_size": 64,
509
+ "bits": 6
510
+ },
511
+ "model.layers.14.mlp.switch_mlp.down_proj": {
512
+ "group_size": 64,
513
+ "bits": 8
514
+ },
515
+ "model.layers.15.self_attn.q_proj": {
516
+ "group_size": 64,
517
+ "bits": 6
518
+ },
519
+ "model.layers.15.self_attn.k_proj": {
520
+ "group_size": 64,
521
+ "bits": 6
522
+ },
523
+ "model.layers.15.self_attn.v_proj": {
524
+ "group_size": 64,
525
+ "bits": 8
526
+ },
527
+ "model.layers.15.self_attn.o_proj": {
528
+ "group_size": 64,
529
+ "bits": 6
530
+ },
531
+ "model.layers.15.mlp.gate": {
532
+ "group_size": 64,
533
+ "bits": 6
534
+ },
535
+ "model.layers.15.mlp.switch_mlp.gate_proj": {
536
+ "group_size": 64,
537
+ "bits": 6
538
+ },
539
+ "model.layers.15.mlp.switch_mlp.up_proj": {
540
+ "group_size": 64,
541
+ "bits": 6
542
+ },
543
+ "model.layers.15.mlp.switch_mlp.down_proj": {
544
+ "group_size": 64,
545
+ "bits": 6
546
+ },
547
+ "model.layers.16.self_attn.q_proj": {
548
+ "group_size": 64,
549
+ "bits": 6
550
+ },
551
+ "model.layers.16.self_attn.k_proj": {
552
+ "group_size": 64,
553
+ "bits": 6
554
+ },
555
+ "model.layers.16.self_attn.v_proj": {
556
+ "group_size": 64,
557
+ "bits": 8
558
+ },
559
+ "model.layers.16.self_attn.o_proj": {
560
+ "group_size": 64,
561
+ "bits": 6
562
+ },
563
+ "model.layers.16.mlp.gate": {
564
+ "group_size": 64,
565
+ "bits": 6
566
+ },
567
+ "model.layers.16.mlp.switch_mlp.gate_proj": {
568
+ "group_size": 64,
569
+ "bits": 6
570
+ },
571
+ "model.layers.16.mlp.switch_mlp.up_proj": {
572
+ "group_size": 64,
573
+ "bits": 6
574
+ },
575
+ "model.layers.16.mlp.switch_mlp.down_proj": {
576
+ "group_size": 64,
577
+ "bits": 6
578
+ },
579
+ "model.layers.17.self_attn.q_proj": {
580
+ "group_size": 64,
581
+ "bits": 6
582
+ },
583
+ "model.layers.17.self_attn.k_proj": {
584
+ "group_size": 64,
585
+ "bits": 6
586
+ },
587
+ "model.layers.17.self_attn.v_proj": {
588
+ "group_size": 64,
589
+ "bits": 8
590
+ },
591
+ "model.layers.17.self_attn.o_proj": {
592
+ "group_size": 64,
593
+ "bits": 6
594
+ },
595
+ "model.layers.17.mlp.gate": {
596
+ "group_size": 64,
597
+ "bits": 6
598
+ },
599
+ "model.layers.17.mlp.switch_mlp.gate_proj": {
600
+ "group_size": 64,
601
+ "bits": 6
602
+ },
603
+ "model.layers.17.mlp.switch_mlp.up_proj": {
604
+ "group_size": 64,
605
+ "bits": 6
606
+ },
607
+ "model.layers.17.mlp.switch_mlp.down_proj": {
608
+ "group_size": 64,
609
+ "bits": 8
610
+ },
611
+ "model.layers.18.self_attn.q_proj": {
612
+ "group_size": 64,
613
+ "bits": 6
614
+ },
615
+ "model.layers.18.self_attn.k_proj": {
616
+ "group_size": 64,
617
+ "bits": 6
618
+ },
619
+ "model.layers.18.self_attn.v_proj": {
620
+ "group_size": 64,
621
+ "bits": 8
622
+ },
623
+ "model.layers.18.self_attn.o_proj": {
624
+ "group_size": 64,
625
+ "bits": 6
626
+ },
627
+ "model.layers.18.mlp.gate": {
628
+ "group_size": 64,
629
+ "bits": 6
630
+ },
631
+ "model.layers.18.mlp.switch_mlp.gate_proj": {
632
+ "group_size": 64,
633
+ "bits": 6
634
+ },
635
+ "model.layers.18.mlp.switch_mlp.up_proj": {
636
+ "group_size": 64,
637
+ "bits": 6
638
+ },
639
+ "model.layers.18.mlp.switch_mlp.down_proj": {
640
+ "group_size": 64,
641
+ "bits": 6
642
+ },
643
+ "model.layers.19.self_attn.q_proj": {
644
+ "group_size": 64,
645
+ "bits": 6
646
+ },
647
+ "model.layers.19.self_attn.k_proj": {
648
+ "group_size": 64,
649
+ "bits": 6
650
+ },
651
+ "model.layers.19.self_attn.v_proj": {
652
+ "group_size": 64,
653
+ "bits": 8
654
+ },
655
+ "model.layers.19.self_attn.o_proj": {
656
+ "group_size": 64,
657
+ "bits": 6
658
+ },
659
+ "model.layers.19.mlp.gate": {
660
+ "group_size": 64,
661
+ "bits": 6
662
+ },
663
+ "model.layers.19.mlp.switch_mlp.gate_proj": {
664
+ "group_size": 64,
665
+ "bits": 6
666
+ },
667
+ "model.layers.19.mlp.switch_mlp.up_proj": {
668
+ "group_size": 64,
669
+ "bits": 6
670
+ },
671
+ "model.layers.19.mlp.switch_mlp.down_proj": {
672
+ "group_size": 64,
673
+ "bits": 6
674
+ },
675
+ "model.layers.20.self_attn.q_proj": {
676
+ "group_size": 64,
677
+ "bits": 6
678
+ },
679
+ "model.layers.20.self_attn.k_proj": {
680
+ "group_size": 64,
681
+ "bits": 6
682
+ },
683
+ "model.layers.20.self_attn.v_proj": {
684
+ "group_size": 64,
685
+ "bits": 8
686
+ },
687
+ "model.layers.20.self_attn.o_proj": {
688
+ "group_size": 64,
689
+ "bits": 6
690
+ },
691
+ "model.layers.20.mlp.gate": {
692
+ "group_size": 64,
693
+ "bits": 6
694
+ },
695
+ "model.layers.20.mlp.switch_mlp.gate_proj": {
696
+ "group_size": 64,
697
+ "bits": 6
698
+ },
699
+ "model.layers.20.mlp.switch_mlp.up_proj": {
700
+ "group_size": 64,
701
+ "bits": 6
702
+ },
703
+ "model.layers.20.mlp.switch_mlp.down_proj": {
704
+ "group_size": 64,
705
+ "bits": 8
706
+ },
707
+ "model.layers.21.self_attn.q_proj": {
708
+ "group_size": 64,
709
+ "bits": 6
710
+ },
711
+ "model.layers.21.self_attn.k_proj": {
712
+ "group_size": 64,
713
+ "bits": 6
714
+ },
715
+ "model.layers.21.self_attn.v_proj": {
716
+ "group_size": 64,
717
+ "bits": 8
718
+ },
719
+ "model.layers.21.self_attn.o_proj": {
720
+ "group_size": 64,
721
+ "bits": 6
722
+ },
723
+ "model.layers.21.mlp.gate": {
724
+ "group_size": 64,
725
+ "bits": 6
726
+ },
727
+ "model.layers.21.mlp.switch_mlp.gate_proj": {
728
+ "group_size": 64,
729
+ "bits": 6
730
+ },
731
+ "model.layers.21.mlp.switch_mlp.up_proj": {
732
+ "group_size": 64,
733
+ "bits": 6
734
+ },
735
+ "model.layers.21.mlp.switch_mlp.down_proj": {
736
+ "group_size": 64,
737
+ "bits": 6
738
+ },
739
+ "model.layers.22.self_attn.q_proj": {
740
+ "group_size": 64,
741
+ "bits": 6
742
+ },
743
+ "model.layers.22.self_attn.k_proj": {
744
+ "group_size": 64,
745
+ "bits": 6
746
+ },
747
+ "model.layers.22.self_attn.v_proj": {
748
+ "group_size": 64,
749
+ "bits": 8
750
+ },
751
+ "model.layers.22.self_attn.o_proj": {
752
+ "group_size": 64,
753
+ "bits": 6
754
+ },
755
+ "model.layers.22.mlp.gate": {
756
+ "group_size": 64,
757
+ "bits": 6
758
+ },
759
+ "model.layers.22.mlp.switch_mlp.gate_proj": {
760
+ "group_size": 64,
761
+ "bits": 6
762
+ },
763
+ "model.layers.22.mlp.switch_mlp.up_proj": {
764
+ "group_size": 64,
765
+ "bits": 6
766
+ },
767
+ "model.layers.22.mlp.switch_mlp.down_proj": {
768
+ "group_size": 64,
769
+ "bits": 6
770
+ },
771
+ "model.layers.23.self_attn.q_proj": {
772
+ "group_size": 64,
773
+ "bits": 6
774
+ },
775
+ "model.layers.23.self_attn.k_proj": {
776
+ "group_size": 64,
777
+ "bits": 6
778
+ },
779
+ "model.layers.23.self_attn.v_proj": {
780
+ "group_size": 64,
781
+ "bits": 8
782
+ },
783
+ "model.layers.23.self_attn.o_proj": {
784
+ "group_size": 64,
785
+ "bits": 6
786
+ },
787
+ "model.layers.23.mlp.gate": {
788
+ "group_size": 64,
789
+ "bits": 6
790
+ },
791
+ "model.layers.23.mlp.switch_mlp.gate_proj": {
792
+ "group_size": 64,
793
+ "bits": 6
794
+ },
795
+ "model.layers.23.mlp.switch_mlp.up_proj": {
796
+ "group_size": 64,
797
+ "bits": 6
798
+ },
799
+ "model.layers.23.mlp.switch_mlp.down_proj": {
800
+ "group_size": 64,
801
+ "bits": 8
802
+ },
803
+ "model.layers.24.self_attn.q_proj": {
804
+ "group_size": 64,
805
+ "bits": 6
806
+ },
807
+ "model.layers.24.self_attn.k_proj": {
808
+ "group_size": 64,
809
+ "bits": 6
810
+ },
811
+ "model.layers.24.self_attn.v_proj": {
812
+ "group_size": 64,
813
+ "bits": 8
814
+ },
815
+ "model.layers.24.self_attn.o_proj": {
816
+ "group_size": 64,
817
+ "bits": 6
818
+ },
819
+ "model.layers.24.mlp.gate": {
820
+ "group_size": 64,
821
+ "bits": 6
822
+ },
823
+ "model.layers.24.mlp.switch_mlp.gate_proj": {
824
+ "group_size": 64,
825
+ "bits": 6
826
+ },
827
+ "model.layers.24.mlp.switch_mlp.up_proj": {
828
+ "group_size": 64,
829
+ "bits": 6
830
+ },
831
+ "model.layers.24.mlp.switch_mlp.down_proj": {
832
+ "group_size": 64,
833
+ "bits": 6
834
+ },
835
+ "model.layers.25.self_attn.q_proj": {
836
+ "group_size": 64,
837
+ "bits": 6
838
+ },
839
+ "model.layers.25.self_attn.k_proj": {
840
+ "group_size": 64,
841
+ "bits": 6
842
+ },
843
+ "model.layers.25.self_attn.v_proj": {
844
+ "group_size": 64,
845
+ "bits": 8
846
+ },
847
+ "model.layers.25.self_attn.o_proj": {
848
+ "group_size": 64,
849
+ "bits": 6
850
+ },
851
+ "model.layers.25.mlp.gate": {
852
+ "group_size": 64,
853
+ "bits": 6
854
+ },
855
+ "model.layers.25.mlp.switch_mlp.gate_proj": {
856
+ "group_size": 64,
857
+ "bits": 6
858
+ },
859
+ "model.layers.25.mlp.switch_mlp.up_proj": {
860
+ "group_size": 64,
861
+ "bits": 6
862
+ },
863
+ "model.layers.25.mlp.switch_mlp.down_proj": {
864
+ "group_size": 64,
865
+ "bits": 6
866
+ },
867
+ "model.layers.26.self_attn.q_proj": {
868
+ "group_size": 64,
869
+ "bits": 6
870
+ },
871
+ "model.layers.26.self_attn.k_proj": {
872
+ "group_size": 64,
873
+ "bits": 6
874
+ },
875
+ "model.layers.26.self_attn.v_proj": {
876
+ "group_size": 64,
877
+ "bits": 8
878
+ },
879
+ "model.layers.26.self_attn.o_proj": {
880
+ "group_size": 64,
881
+ "bits": 6
882
+ },
883
+ "model.layers.26.mlp.gate": {
884
+ "group_size": 64,
885
+ "bits": 6
886
+ },
887
+ "model.layers.26.mlp.switch_mlp.gate_proj": {
888
+ "group_size": 64,
889
+ "bits": 6
890
+ },
891
+ "model.layers.26.mlp.switch_mlp.up_proj": {
892
+ "group_size": 64,
893
+ "bits": 6
894
+ },
895
+ "model.layers.26.mlp.switch_mlp.down_proj": {
896
+ "group_size": 64,
897
+ "bits": 8
898
+ },
899
+ "model.layers.27.self_attn.q_proj": {
900
+ "group_size": 64,
901
+ "bits": 6
902
+ },
903
+ "model.layers.27.self_attn.k_proj": {
904
+ "group_size": 64,
905
+ "bits": 6
906
+ },
907
+ "model.layers.27.self_attn.v_proj": {
908
+ "group_size": 64,
909
+ "bits": 8
910
+ },
911
+ "model.layers.27.self_attn.o_proj": {
912
+ "group_size": 64,
913
+ "bits": 6
914
+ },
915
+ "model.layers.27.mlp.gate": {
916
+ "group_size": 64,
917
+ "bits": 6
918
+ },
919
+ "model.layers.27.mlp.switch_mlp.gate_proj": {
920
+ "group_size": 64,
921
+ "bits": 6
922
+ },
923
+ "model.layers.27.mlp.switch_mlp.up_proj": {
924
+ "group_size": 64,
925
+ "bits": 6
926
+ },
927
+ "model.layers.27.mlp.switch_mlp.down_proj": {
928
+ "group_size": 64,
929
+ "bits": 6
930
+ },
931
+ "model.layers.28.self_attn.q_proj": {
932
+ "group_size": 64,
933
+ "bits": 6
934
+ },
935
+ "model.layers.28.self_attn.k_proj": {
936
+ "group_size": 64,
937
+ "bits": 6
938
+ },
939
+ "model.layers.28.self_attn.v_proj": {
940
+ "group_size": 64,
941
+ "bits": 8
942
+ },
943
+ "model.layers.28.self_attn.o_proj": {
944
+ "group_size": 64,
945
+ "bits": 6
946
+ },
947
+ "model.layers.28.mlp.gate": {
948
+ "group_size": 64,
949
+ "bits": 6
950
+ },
951
+ "model.layers.28.mlp.switch_mlp.gate_proj": {
952
+ "group_size": 64,
953
+ "bits": 6
954
+ },
955
+ "model.layers.28.mlp.switch_mlp.up_proj": {
956
+ "group_size": 64,
957
+ "bits": 6
958
+ },
959
+ "model.layers.28.mlp.switch_mlp.down_proj": {
960
+ "group_size": 64,
961
+ "bits": 6
962
+ },
963
+ "model.layers.29.self_attn.q_proj": {
964
+ "group_size": 64,
965
+ "bits": 6
966
+ },
967
+ "model.layers.29.self_attn.k_proj": {
968
+ "group_size": 64,
969
+ "bits": 6
970
+ },
971
+ "model.layers.29.self_attn.v_proj": {
972
+ "group_size": 64,
973
+ "bits": 8
974
+ },
975
+ "model.layers.29.self_attn.o_proj": {
976
+ "group_size": 64,
977
+ "bits": 6
978
+ },
979
+ "model.layers.29.mlp.gate": {
980
+ "group_size": 64,
981
+ "bits": 6
982
+ },
983
+ "model.layers.29.mlp.switch_mlp.gate_proj": {
984
+ "group_size": 64,
985
+ "bits": 6
986
+ },
987
+ "model.layers.29.mlp.switch_mlp.up_proj": {
988
+ "group_size": 64,
989
+ "bits": 6
990
+ },
991
+ "model.layers.29.mlp.switch_mlp.down_proj": {
992
+ "group_size": 64,
993
+ "bits": 8
994
+ },
995
+ "model.layers.30.self_attn.q_proj": {
996
+ "group_size": 64,
997
+ "bits": 6
998
+ },
999
+ "model.layers.30.self_attn.k_proj": {
1000
+ "group_size": 64,
1001
+ "bits": 6
1002
+ },
1003
+ "model.layers.30.self_attn.v_proj": {
1004
+ "group_size": 64,
1005
+ "bits": 8
1006
+ },
1007
+ "model.layers.30.self_attn.o_proj": {
1008
+ "group_size": 64,
1009
+ "bits": 6
1010
+ },
1011
+ "model.layers.30.mlp.gate": {
1012
+ "group_size": 64,
1013
+ "bits": 6
1014
+ },
1015
+ "model.layers.30.mlp.switch_mlp.gate_proj": {
1016
+ "group_size": 64,
1017
+ "bits": 6
1018
+ },
1019
+ "model.layers.30.mlp.switch_mlp.up_proj": {
1020
+ "group_size": 64,
1021
+ "bits": 6
1022
+ },
1023
+ "model.layers.30.mlp.switch_mlp.down_proj": {
1024
+ "group_size": 64,
1025
+ "bits": 6
1026
+ },
1027
+ "model.layers.31.self_attn.q_proj": {
1028
+ "group_size": 64,
1029
+ "bits": 6
1030
+ },
1031
+ "model.layers.31.self_attn.k_proj": {
1032
+ "group_size": 64,
1033
+ "bits": 6
1034
+ },
1035
+ "model.layers.31.self_attn.v_proj": {
1036
+ "group_size": 64,
1037
+ "bits": 8
1038
+ },
1039
+ "model.layers.31.self_attn.o_proj": {
1040
+ "group_size": 64,
1041
+ "bits": 6
1042
+ },
1043
+ "model.layers.31.mlp.gate": {
1044
+ "group_size": 64,
1045
+ "bits": 6
1046
+ },
1047
+ "model.layers.31.mlp.switch_mlp.gate_proj": {
1048
+ "group_size": 64,
1049
+ "bits": 6
1050
+ },
1051
+ "model.layers.31.mlp.switch_mlp.up_proj": {
1052
+ "group_size": 64,
1053
+ "bits": 6
1054
+ },
1055
+ "model.layers.31.mlp.switch_mlp.down_proj": {
1056
+ "group_size": 64,
1057
+ "bits": 6
1058
+ },
1059
+ "model.layers.32.self_attn.q_proj": {
1060
+ "group_size": 64,
1061
+ "bits": 6
1062
+ },
1063
+ "model.layers.32.self_attn.k_proj": {
1064
+ "group_size": 64,
1065
+ "bits": 6
1066
+ },
1067
+ "model.layers.32.self_attn.v_proj": {
1068
+ "group_size": 64,
1069
+ "bits": 8
1070
+ },
1071
+ "model.layers.32.self_attn.o_proj": {
1072
+ "group_size": 64,
1073
+ "bits": 6
1074
+ },
1075
+ "model.layers.32.mlp.gate": {
1076
+ "group_size": 64,
1077
+ "bits": 6
1078
+ },
1079
+ "model.layers.32.mlp.switch_mlp.gate_proj": {
1080
+ "group_size": 64,
1081
+ "bits": 6
1082
+ },
1083
+ "model.layers.32.mlp.switch_mlp.up_proj": {
1084
+ "group_size": 64,
1085
+ "bits": 6
1086
+ },
1087
+ "model.layers.32.mlp.switch_mlp.down_proj": {
1088
+ "group_size": 64,
1089
+ "bits": 8
1090
+ },
1091
+ "model.layers.33.self_attn.q_proj": {
1092
+ "group_size": 64,
1093
+ "bits": 6
1094
+ },
1095
+ "model.layers.33.self_attn.k_proj": {
1096
+ "group_size": 64,
1097
+ "bits": 6
1098
+ },
1099
+ "model.layers.33.self_attn.v_proj": {
1100
+ "group_size": 64,
1101
+ "bits": 8
1102
+ },
1103
+ "model.layers.33.self_attn.o_proj": {
1104
+ "group_size": 64,
1105
+ "bits": 6
1106
+ },
1107
+ "model.layers.33.mlp.gate": {
1108
+ "group_size": 64,
1109
+ "bits": 6
1110
+ },
1111
+ "model.layers.33.mlp.switch_mlp.gate_proj": {
1112
+ "group_size": 64,
1113
+ "bits": 6
1114
+ },
1115
+ "model.layers.33.mlp.switch_mlp.up_proj": {
1116
+ "group_size": 64,
1117
+ "bits": 6
1118
+ },
1119
+ "model.layers.33.mlp.switch_mlp.down_proj": {
1120
+ "group_size": 64,
1121
+ "bits": 6
1122
+ },
1123
+ "model.layers.34.self_attn.q_proj": {
1124
+ "group_size": 64,
1125
+ "bits": 6
1126
+ },
1127
+ "model.layers.34.self_attn.k_proj": {
1128
+ "group_size": 64,
1129
+ "bits": 6
1130
+ },
1131
+ "model.layers.34.self_attn.v_proj": {
1132
+ "group_size": 64,
1133
+ "bits": 8
1134
+ },
1135
+ "model.layers.34.self_attn.o_proj": {
1136
+ "group_size": 64,
1137
+ "bits": 6
1138
+ },
1139
+ "model.layers.34.mlp.gate": {
1140
+ "group_size": 64,
1141
+ "bits": 6
1142
+ },
1143
+ "model.layers.34.mlp.switch_mlp.gate_proj": {
1144
+ "group_size": 64,
1145
+ "bits": 6
1146
+ },
1147
+ "model.layers.34.mlp.switch_mlp.up_proj": {
1148
+ "group_size": 64,
1149
+ "bits": 6
1150
+ },
1151
+ "model.layers.34.mlp.switch_mlp.down_proj": {
1152
+ "group_size": 64,
1153
+ "bits": 6
1154
+ },
1155
+ "model.layers.35.self_attn.q_proj": {
1156
+ "group_size": 64,
1157
+ "bits": 6
1158
+ },
1159
+ "model.layers.35.self_attn.k_proj": {
1160
+ "group_size": 64,
1161
+ "bits": 6
1162
+ },
1163
+ "model.layers.35.self_attn.v_proj": {
1164
+ "group_size": 64,
1165
+ "bits": 8
1166
+ },
1167
+ "model.layers.35.self_attn.o_proj": {
1168
+ "group_size": 64,
1169
+ "bits": 6
1170
+ },
1171
+ "model.layers.35.mlp.gate": {
1172
+ "group_size": 64,
1173
+ "bits": 6
1174
+ },
1175
+ "model.layers.35.mlp.switch_mlp.gate_proj": {
1176
+ "group_size": 64,
1177
+ "bits": 6
1178
+ },
1179
+ "model.layers.35.mlp.switch_mlp.up_proj": {
1180
+ "group_size": 64,
1181
+ "bits": 6
1182
+ },
1183
+ "model.layers.35.mlp.switch_mlp.down_proj": {
1184
+ "group_size": 64,
1185
+ "bits": 8
1186
+ },
1187
+ "model.layers.36.self_attn.q_proj": {
1188
+ "group_size": 64,
1189
+ "bits": 6
1190
+ },
1191
+ "model.layers.36.self_attn.k_proj": {
1192
+ "group_size": 64,
1193
+ "bits": 6
1194
+ },
1195
+ "model.layers.36.self_attn.v_proj": {
1196
+ "group_size": 64,
1197
+ "bits": 8
1198
+ },
1199
+ "model.layers.36.self_attn.o_proj": {
1200
+ "group_size": 64,
1201
+ "bits": 6
1202
+ },
1203
+ "model.layers.36.mlp.gate": {
1204
+ "group_size": 64,
1205
+ "bits": 6
1206
+ },
1207
+ "model.layers.36.mlp.switch_mlp.gate_proj": {
1208
+ "group_size": 64,
1209
+ "bits": 6
1210
+ },
1211
+ "model.layers.36.mlp.switch_mlp.up_proj": {
1212
+ "group_size": 64,
1213
+ "bits": 6
1214
+ },
1215
+ "model.layers.36.mlp.switch_mlp.down_proj": {
1216
+ "group_size": 64,
1217
+ "bits": 6
1218
+ },
1219
+ "model.layers.37.self_attn.q_proj": {
1220
+ "group_size": 64,
1221
+ "bits": 6
1222
+ },
1223
+ "model.layers.37.self_attn.k_proj": {
1224
+ "group_size": 64,
1225
+ "bits": 6
1226
+ },
1227
+ "model.layers.37.self_attn.v_proj": {
1228
+ "group_size": 64,
1229
+ "bits": 8
1230
+ },
1231
+ "model.layers.37.self_attn.o_proj": {
1232
+ "group_size": 64,
1233
+ "bits": 6
1234
+ },
1235
+ "model.layers.37.mlp.gate": {
1236
+ "group_size": 64,
1237
+ "bits": 6
1238
+ },
1239
+ "model.layers.37.mlp.switch_mlp.gate_proj": {
1240
+ "group_size": 64,
1241
+ "bits": 6
1242
+ },
1243
+ "model.layers.37.mlp.switch_mlp.up_proj": {
1244
+ "group_size": 64,
1245
+ "bits": 6
1246
+ },
1247
+ "model.layers.37.mlp.switch_mlp.down_proj": {
1248
+ "group_size": 64,
1249
+ "bits": 6
1250
+ },
1251
+ "model.layers.38.self_attn.q_proj": {
1252
+ "group_size": 64,
1253
+ "bits": 6
1254
+ },
1255
+ "model.layers.38.self_attn.k_proj": {
1256
+ "group_size": 64,
1257
+ "bits": 6
1258
+ },
1259
+ "model.layers.38.self_attn.v_proj": {
1260
+ "group_size": 64,
1261
+ "bits": 8
1262
+ },
1263
+ "model.layers.38.self_attn.o_proj": {
1264
+ "group_size": 64,
1265
+ "bits": 6
1266
+ },
1267
+ "model.layers.38.mlp.gate": {
1268
+ "group_size": 64,
1269
+ "bits": 6
1270
+ },
1271
+ "model.layers.38.mlp.switch_mlp.gate_proj": {
1272
+ "group_size": 64,
1273
+ "bits": 6
1274
+ },
1275
+ "model.layers.38.mlp.switch_mlp.up_proj": {
1276
+ "group_size": 64,
1277
+ "bits": 6
1278
+ },
1279
+ "model.layers.38.mlp.switch_mlp.down_proj": {
1280
+ "group_size": 64,
1281
+ "bits": 8
1282
+ },
1283
+ "model.layers.39.self_attn.q_proj": {
1284
+ "group_size": 64,
1285
+ "bits": 6
1286
+ },
1287
+ "model.layers.39.self_attn.k_proj": {
1288
+ "group_size": 64,
1289
+ "bits": 6
1290
+ },
1291
+ "model.layers.39.self_attn.v_proj": {
1292
+ "group_size": 64,
1293
+ "bits": 8
1294
+ },
1295
+ "model.layers.39.self_attn.o_proj": {
1296
+ "group_size": 64,
1297
+ "bits": 6
1298
+ },
1299
+ "model.layers.39.mlp.gate": {
1300
+ "group_size": 64,
1301
+ "bits": 6
1302
+ },
1303
+ "model.layers.39.mlp.switch_mlp.gate_proj": {
1304
+ "group_size": 64,
1305
+ "bits": 6
1306
+ },
1307
+ "model.layers.39.mlp.switch_mlp.up_proj": {
1308
+ "group_size": 64,
1309
+ "bits": 6
1310
+ },
1311
+ "model.layers.39.mlp.switch_mlp.down_proj": {
1312
+ "group_size": 64,
1313
+ "bits": 6
1314
+ },
1315
+ "model.layers.40.self_attn.q_proj": {
1316
+ "group_size": 64,
1317
+ "bits": 6
1318
+ },
1319
+ "model.layers.40.self_attn.k_proj": {
1320
+ "group_size": 64,
1321
+ "bits": 6
1322
+ },
1323
+ "model.layers.40.self_attn.v_proj": {
1324
+ "group_size": 64,
1325
+ "bits": 8
1326
+ },
1327
+ "model.layers.40.self_attn.o_proj": {
1328
+ "group_size": 64,
1329
+ "bits": 6
1330
+ },
1331
+ "model.layers.40.mlp.gate": {
1332
+ "group_size": 64,
1333
+ "bits": 6
1334
+ },
1335
+ "model.layers.40.mlp.switch_mlp.gate_proj": {
1336
+ "group_size": 64,
1337
+ "bits": 6
1338
+ },
1339
+ "model.layers.40.mlp.switch_mlp.up_proj": {
1340
+ "group_size": 64,
1341
+ "bits": 6
1342
+ },
1343
+ "model.layers.40.mlp.switch_mlp.down_proj": {
1344
+ "group_size": 64,
1345
+ "bits": 6
1346
+ },
1347
+ "model.layers.41.self_attn.q_proj": {
1348
+ "group_size": 64,
1349
+ "bits": 6
1350
+ },
1351
+ "model.layers.41.self_attn.k_proj": {
1352
+ "group_size": 64,
1353
+ "bits": 6
1354
+ },
1355
+ "model.layers.41.self_attn.v_proj": {
1356
+ "group_size": 64,
1357
+ "bits": 8
1358
+ },
1359
+ "model.layers.41.self_attn.o_proj": {
1360
+ "group_size": 64,
1361
+ "bits": 6
1362
+ },
1363
+ "model.layers.41.mlp.gate": {
1364
+ "group_size": 64,
1365
+ "bits": 6
1366
+ },
1367
+ "model.layers.41.mlp.switch_mlp.gate_proj": {
1368
+ "group_size": 64,
1369
+ "bits": 6
1370
+ },
1371
+ "model.layers.41.mlp.switch_mlp.up_proj": {
1372
+ "group_size": 64,
1373
+ "bits": 6
1374
+ },
1375
+ "model.layers.41.mlp.switch_mlp.down_proj": {
1376
+ "group_size": 64,
1377
+ "bits": 8
1378
+ },
1379
+ "model.layers.42.self_attn.q_proj": {
1380
+ "group_size": 64,
1381
+ "bits": 6
1382
+ },
1383
+ "model.layers.42.self_attn.k_proj": {
1384
+ "group_size": 64,
1385
+ "bits": 6
1386
+ },
1387
+ "model.layers.42.self_attn.v_proj": {
1388
+ "group_size": 64,
1389
+ "bits": 8
1390
+ },
1391
+ "model.layers.42.self_attn.o_proj": {
1392
+ "group_size": 64,
1393
+ "bits": 6
1394
+ },
1395
+ "model.layers.42.mlp.gate": {
1396
+ "group_size": 64,
1397
+ "bits": 6
1398
+ },
1399
+ "model.layers.42.mlp.switch_mlp.gate_proj": {
1400
+ "group_size": 64,
1401
+ "bits": 6
1402
+ },
1403
+ "model.layers.42.mlp.switch_mlp.up_proj": {
1404
+ "group_size": 64,
1405
+ "bits": 6
1406
+ },
1407
+ "model.layers.42.mlp.switch_mlp.down_proj": {
1408
+ "group_size": 64,
1409
+ "bits": 8
1410
+ },
1411
+ "model.layers.43.self_attn.q_proj": {
1412
+ "group_size": 64,
1413
+ "bits": 6
1414
+ },
1415
+ "model.layers.43.self_attn.k_proj": {
1416
+ "group_size": 64,
1417
+ "bits": 6
1418
+ },
1419
+ "model.layers.43.self_attn.v_proj": {
1420
+ "group_size": 64,
1421
+ "bits": 8
1422
+ },
1423
+ "model.layers.43.self_attn.o_proj": {
1424
+ "group_size": 64,
1425
+ "bits": 6
1426
+ },
1427
+ "model.layers.43.mlp.gate": {
1428
+ "group_size": 64,
1429
+ "bits": 6
1430
+ },
1431
+ "model.layers.43.mlp.switch_mlp.gate_proj": {
1432
+ "group_size": 64,
1433
+ "bits": 6
1434
+ },
1435
+ "model.layers.43.mlp.switch_mlp.up_proj": {
1436
+ "group_size": 64,
1437
+ "bits": 6
1438
+ },
1439
+ "model.layers.43.mlp.switch_mlp.down_proj": {
1440
+ "group_size": 64,
1441
+ "bits": 8
1442
+ },
1443
+ "model.layers.44.self_attn.q_proj": {
1444
+ "group_size": 64,
1445
+ "bits": 6
1446
+ },
1447
+ "model.layers.44.self_attn.k_proj": {
1448
+ "group_size": 64,
1449
+ "bits": 6
1450
+ },
1451
+ "model.layers.44.self_attn.v_proj": {
1452
+ "group_size": 64,
1453
+ "bits": 8
1454
+ },
1455
+ "model.layers.44.self_attn.o_proj": {
1456
+ "group_size": 64,
1457
+ "bits": 6
1458
+ },
1459
+ "model.layers.44.mlp.gate": {
1460
+ "group_size": 64,
1461
+ "bits": 6
1462
+ },
1463
+ "model.layers.44.mlp.switch_mlp.gate_proj": {
1464
+ "group_size": 64,
1465
+ "bits": 6
1466
+ },
1467
+ "model.layers.44.mlp.switch_mlp.up_proj": {
1468
+ "group_size": 64,
1469
+ "bits": 6
1470
+ },
1471
+ "model.layers.44.mlp.switch_mlp.down_proj": {
1472
+ "group_size": 64,
1473
+ "bits": 8
1474
+ },
1475
+ "model.layers.45.self_attn.q_proj": {
1476
+ "group_size": 64,
1477
+ "bits": 6
1478
+ },
1479
+ "model.layers.45.self_attn.k_proj": {
1480
+ "group_size": 64,
1481
+ "bits": 6
1482
+ },
1483
+ "model.layers.45.self_attn.v_proj": {
1484
+ "group_size": 64,
1485
+ "bits": 8
1486
+ },
1487
+ "model.layers.45.self_attn.o_proj": {
1488
+ "group_size": 64,
1489
+ "bits": 6
1490
+ },
1491
+ "model.layers.45.mlp.gate": {
1492
+ "group_size": 64,
1493
+ "bits": 6
1494
+ },
1495
+ "model.layers.45.mlp.switch_mlp.gate_proj": {
1496
+ "group_size": 64,
1497
+ "bits": 6
1498
+ },
1499
+ "model.layers.45.mlp.switch_mlp.up_proj": {
1500
+ "group_size": 64,
1501
+ "bits": 6
1502
+ },
1503
+ "model.layers.45.mlp.switch_mlp.down_proj": {
1504
+ "group_size": 64,
1505
+ "bits": 8
1506
+ },
1507
+ "model.layers.46.self_attn.q_proj": {
1508
+ "group_size": 64,
1509
+ "bits": 6
1510
+ },
1511
+ "model.layers.46.self_attn.k_proj": {
1512
+ "group_size": 64,
1513
+ "bits": 6
1514
+ },
1515
+ "model.layers.46.self_attn.v_proj": {
1516
+ "group_size": 64,
1517
+ "bits": 8
1518
+ },
1519
+ "model.layers.46.self_attn.o_proj": {
1520
+ "group_size": 64,
1521
+ "bits": 6
1522
+ },
1523
+ "model.layers.46.mlp.gate": {
1524
+ "group_size": 64,
1525
+ "bits": 6
1526
+ },
1527
+ "model.layers.46.mlp.switch_mlp.gate_proj": {
1528
+ "group_size": 64,
1529
+ "bits": 6
1530
+ },
1531
+ "model.layers.46.mlp.switch_mlp.up_proj": {
1532
+ "group_size": 64,
1533
+ "bits": 6
1534
+ },
1535
+ "model.layers.46.mlp.switch_mlp.down_proj": {
1536
+ "group_size": 64,
1537
+ "bits": 8
1538
+ },
1539
+ "model.layers.47.self_attn.q_proj": {
1540
+ "group_size": 64,
1541
+ "bits": 6
1542
+ },
1543
+ "model.layers.47.self_attn.k_proj": {
1544
+ "group_size": 64,
1545
+ "bits": 6
1546
+ },
1547
+ "model.layers.47.self_attn.v_proj": {
1548
+ "group_size": 64,
1549
+ "bits": 8
1550
+ },
1551
+ "model.layers.47.self_attn.o_proj": {
1552
+ "group_size": 64,
1553
+ "bits": 6
1554
+ },
1555
+ "model.layers.47.mlp.gate": {
1556
+ "group_size": 64,
1557
+ "bits": 6
1558
+ },
1559
+ "model.layers.47.mlp.switch_mlp.gate_proj": {
1560
+ "group_size": 64,
1561
+ "bits": 6
1562
+ },
1563
+ "model.layers.47.mlp.switch_mlp.up_proj": {
1564
+ "group_size": 64,
1565
+ "bits": 6
1566
+ },
1567
+ "model.layers.47.mlp.switch_mlp.down_proj": {
1568
+ "group_size": 64,
1569
+ "bits": 8
1570
+ },
1571
+ "lm_head": {
1572
+ "group_size": 64,
1573
+ "bits": 8
1574
+ }
1575
+ },
1576
+ "quantization_config": {
1577
+ "group_size": 64,
1578
+ "bits": 8,
1579
+ "mode": "affine",
1580
+ "model.embed_tokens": {
1581
+ "group_size": 64,
1582
+ "bits": 8
1583
+ },
1584
+ "model.layers.0.self_attn.q_proj": {
1585
+ "group_size": 64,
1586
+ "bits": 8
1587
+ },
1588
+ "model.layers.0.self_attn.k_proj": {
1589
+ "group_size": 64,
1590
+ "bits": 8
1591
+ },
1592
+ "model.layers.0.self_attn.v_proj": {
1593
+ "group_size": 64,
1594
+ "bits": 8
1595
+ },
1596
+ "model.layers.0.self_attn.o_proj": {
1597
+ "group_size": 64,
1598
+ "bits": 8
1599
+ },
1600
+ "model.layers.0.mlp.gate": {
1601
+ "group_size": 64,
1602
+ "bits": 8
1603
+ },
1604
+ "model.layers.0.mlp.switch_mlp.gate_proj": {
1605
+ "group_size": 64,
1606
+ "bits": 8
1607
+ },
1608
+ "model.layers.0.mlp.switch_mlp.up_proj": {
1609
+ "group_size": 64,
1610
+ "bits": 8
1611
+ },
1612
+ "model.layers.0.mlp.switch_mlp.down_proj": {
1613
+ "group_size": 64,
1614
+ "bits": 8
1615
+ },
1616
+ "model.layers.1.self_attn.q_proj": {
1617
+ "group_size": 64,
1618
+ "bits": 6
1619
+ },
1620
+ "model.layers.1.self_attn.k_proj": {
1621
+ "group_size": 64,
1622
+ "bits": 6
1623
+ },
1624
+ "model.layers.1.self_attn.v_proj": {
1625
+ "group_size": 64,
1626
+ "bits": 8
1627
+ },
1628
+ "model.layers.1.self_attn.o_proj": {
1629
+ "group_size": 64,
1630
+ "bits": 6
1631
+ },
1632
+ "model.layers.1.mlp.gate": {
1633
+ "group_size": 64,
1634
+ "bits": 6
1635
+ },
1636
+ "model.layers.1.mlp.switch_mlp.gate_proj": {
1637
+ "group_size": 64,
1638
+ "bits": 6
1639
+ },
1640
+ "model.layers.1.mlp.switch_mlp.up_proj": {
1641
+ "group_size": 64,
1642
+ "bits": 6
1643
+ },
1644
+ "model.layers.1.mlp.switch_mlp.down_proj": {
1645
+ "group_size": 64,
1646
+ "bits": 8
1647
+ },
1648
+ "model.layers.2.self_attn.q_proj": {
1649
+ "group_size": 64,
1650
+ "bits": 6
1651
+ },
1652
+ "model.layers.2.self_attn.k_proj": {
1653
+ "group_size": 64,
1654
+ "bits": 6
1655
+ },
1656
+ "model.layers.2.self_attn.v_proj": {
1657
+ "group_size": 64,
1658
+ "bits": 8
1659
+ },
1660
+ "model.layers.2.self_attn.o_proj": {
1661
+ "group_size": 64,
1662
+ "bits": 6
1663
+ },
1664
+ "model.layers.2.mlp.gate": {
1665
+ "group_size": 64,
1666
+ "bits": 6
1667
+ },
1668
+ "model.layers.2.mlp.switch_mlp.gate_proj": {
1669
+ "group_size": 64,
1670
+ "bits": 6
1671
+ },
1672
+ "model.layers.2.mlp.switch_mlp.up_proj": {
1673
+ "group_size": 64,
1674
+ "bits": 6
1675
+ },
1676
+ "model.layers.2.mlp.switch_mlp.down_proj": {
1677
+ "group_size": 64,
1678
+ "bits": 8
1679
+ },
1680
+ "model.layers.3.self_attn.q_proj": {
1681
+ "group_size": 64,
1682
+ "bits": 6
1683
+ },
1684
+ "model.layers.3.self_attn.k_proj": {
1685
+ "group_size": 64,
1686
+ "bits": 6
1687
+ },
1688
+ "model.layers.3.self_attn.v_proj": {
1689
+ "group_size": 64,
1690
+ "bits": 8
1691
+ },
1692
+ "model.layers.3.self_attn.o_proj": {
1693
+ "group_size": 64,
1694
+ "bits": 6
1695
+ },
1696
+ "model.layers.3.mlp.gate": {
1697
+ "group_size": 64,
1698
+ "bits": 6
1699
+ },
1700
+ "model.layers.3.mlp.switch_mlp.gate_proj": {
1701
+ "group_size": 64,
1702
+ "bits": 6
1703
+ },
1704
+ "model.layers.3.mlp.switch_mlp.up_proj": {
1705
+ "group_size": 64,
1706
+ "bits": 6
1707
+ },
1708
+ "model.layers.3.mlp.switch_mlp.down_proj": {
1709
+ "group_size": 64,
1710
+ "bits": 8
1711
+ },
1712
+ "model.layers.4.self_attn.q_proj": {
1713
+ "group_size": 64,
1714
+ "bits": 6
1715
+ },
1716
+ "model.layers.4.self_attn.k_proj": {
1717
+ "group_size": 64,
1718
+ "bits": 6
1719
+ },
1720
+ "model.layers.4.self_attn.v_proj": {
1721
+ "group_size": 64,
1722
+ "bits": 8
1723
+ },
1724
+ "model.layers.4.self_attn.o_proj": {
1725
+ "group_size": 64,
1726
+ "bits": 6
1727
+ },
1728
+ "model.layers.4.mlp.gate": {
1729
+ "group_size": 64,
1730
+ "bits": 6
1731
+ },
1732
+ "model.layers.4.mlp.switch_mlp.gate_proj": {
1733
+ "group_size": 64,
1734
+ "bits": 6
1735
+ },
1736
+ "model.layers.4.mlp.switch_mlp.up_proj": {
1737
+ "group_size": 64,
1738
+ "bits": 6
1739
+ },
1740
+ "model.layers.4.mlp.switch_mlp.down_proj": {
1741
+ "group_size": 64,
1742
+ "bits": 8
1743
+ },
1744
+ "model.layers.5.self_attn.q_proj": {
1745
+ "group_size": 64,
1746
+ "bits": 6
1747
+ },
1748
+ "model.layers.5.self_attn.k_proj": {
1749
+ "group_size": 64,
1750
+ "bits": 6
1751
+ },
1752
+ "model.layers.5.self_attn.v_proj": {
1753
+ "group_size": 64,
1754
+ "bits": 8
1755
+ },
1756
+ "model.layers.5.self_attn.o_proj": {
1757
+ "group_size": 64,
1758
+ "bits": 6
1759
+ },
1760
+ "model.layers.5.mlp.gate": {
1761
+ "group_size": 64,
1762
+ "bits": 6
1763
+ },
1764
+ "model.layers.5.mlp.switch_mlp.gate_proj": {
1765
+ "group_size": 64,
1766
+ "bits": 6
1767
+ },
1768
+ "model.layers.5.mlp.switch_mlp.up_proj": {
1769
+ "group_size": 64,
1770
+ "bits": 6
1771
+ },
1772
+ "model.layers.5.mlp.switch_mlp.down_proj": {
1773
+ "group_size": 64,
1774
+ "bits": 8
1775
+ },
1776
+ "model.layers.6.self_attn.q_proj": {
1777
+ "group_size": 64,
1778
+ "bits": 6
1779
+ },
1780
+ "model.layers.6.self_attn.k_proj": {
1781
+ "group_size": 64,
1782
+ "bits": 6
1783
+ },
1784
+ "model.layers.6.self_attn.v_proj": {
1785
+ "group_size": 64,
1786
+ "bits": 8
1787
+ },
1788
+ "model.layers.6.self_attn.o_proj": {
1789
+ "group_size": 64,
1790
+ "bits": 6
1791
+ },
1792
+ "model.layers.6.mlp.gate": {
1793
+ "group_size": 64,
1794
+ "bits": 6
1795
+ },
1796
+ "model.layers.6.mlp.switch_mlp.gate_proj": {
1797
+ "group_size": 64,
1798
+ "bits": 6
1799
+ },
1800
+ "model.layers.6.mlp.switch_mlp.up_proj": {
1801
+ "group_size": 64,
1802
+ "bits": 6
1803
+ },
1804
+ "model.layers.6.mlp.switch_mlp.down_proj": {
1805
+ "group_size": 64,
1806
+ "bits": 6
1807
+ },
1808
+ "model.layers.7.self_attn.q_proj": {
1809
+ "group_size": 64,
1810
+ "bits": 6
1811
+ },
1812
+ "model.layers.7.self_attn.k_proj": {
1813
+ "group_size": 64,
1814
+ "bits": 6
1815
+ },
1816
+ "model.layers.7.self_attn.v_proj": {
1817
+ "group_size": 64,
1818
+ "bits": 8
1819
+ },
1820
+ "model.layers.7.self_attn.o_proj": {
1821
+ "group_size": 64,
1822
+ "bits": 6
1823
+ },
1824
+ "model.layers.7.mlp.gate": {
1825
+ "group_size": 64,
1826
+ "bits": 6
1827
+ },
1828
+ "model.layers.7.mlp.switch_mlp.gate_proj": {
1829
+ "group_size": 64,
1830
+ "bits": 6
1831
+ },
1832
+ "model.layers.7.mlp.switch_mlp.up_proj": {
1833
+ "group_size": 64,
1834
+ "bits": 6
1835
+ },
1836
+ "model.layers.7.mlp.switch_mlp.down_proj": {
1837
+ "group_size": 64,
1838
+ "bits": 6
1839
+ },
1840
+ "model.layers.8.self_attn.q_proj": {
1841
+ "group_size": 64,
1842
+ "bits": 6
1843
+ },
1844
+ "model.layers.8.self_attn.k_proj": {
1845
+ "group_size": 64,
1846
+ "bits": 6
1847
+ },
1848
+ "model.layers.8.self_attn.v_proj": {
1849
+ "group_size": 64,
1850
+ "bits": 8
1851
+ },
1852
+ "model.layers.8.self_attn.o_proj": {
1853
+ "group_size": 64,
1854
+ "bits": 6
1855
+ },
1856
+ "model.layers.8.mlp.gate": {
1857
+ "group_size": 64,
1858
+ "bits": 6
1859
+ },
1860
+ "model.layers.8.mlp.switch_mlp.gate_proj": {
1861
+ "group_size": 64,
1862
+ "bits": 6
1863
+ },
1864
+ "model.layers.8.mlp.switch_mlp.up_proj": {
1865
+ "group_size": 64,
1866
+ "bits": 6
1867
+ },
1868
+ "model.layers.8.mlp.switch_mlp.down_proj": {
1869
+ "group_size": 64,
1870
+ "bits": 8
1871
+ },
1872
+ "model.layers.9.self_attn.q_proj": {
1873
+ "group_size": 64,
1874
+ "bits": 6
1875
+ },
1876
+ "model.layers.9.self_attn.k_proj": {
1877
+ "group_size": 64,
1878
+ "bits": 6
1879
+ },
1880
+ "model.layers.9.self_attn.v_proj": {
1881
+ "group_size": 64,
1882
+ "bits": 8
1883
+ },
1884
+ "model.layers.9.self_attn.o_proj": {
1885
+ "group_size": 64,
1886
+ "bits": 6
1887
+ },
1888
+ "model.layers.9.mlp.gate": {
1889
+ "group_size": 64,
1890
+ "bits": 6
1891
+ },
1892
+ "model.layers.9.mlp.switch_mlp.gate_proj": {
1893
+ "group_size": 64,
1894
+ "bits": 6
1895
+ },
1896
+ "model.layers.9.mlp.switch_mlp.up_proj": {
1897
+ "group_size": 64,
1898
+ "bits": 6
1899
+ },
1900
+ "model.layers.9.mlp.switch_mlp.down_proj": {
1901
+ "group_size": 64,
1902
+ "bits": 6
1903
+ },
1904
+ "model.layers.10.self_attn.q_proj": {
1905
+ "group_size": 64,
1906
+ "bits": 6
1907
+ },
1908
+ "model.layers.10.self_attn.k_proj": {
1909
+ "group_size": 64,
1910
+ "bits": 6
1911
+ },
1912
+ "model.layers.10.self_attn.v_proj": {
1913
+ "group_size": 64,
1914
+ "bits": 8
1915
+ },
1916
+ "model.layers.10.self_attn.o_proj": {
1917
+ "group_size": 64,
1918
+ "bits": 6
1919
+ },
1920
+ "model.layers.10.mlp.gate": {
1921
+ "group_size": 64,
1922
+ "bits": 6
1923
+ },
1924
+ "model.layers.10.mlp.switch_mlp.gate_proj": {
1925
+ "group_size": 64,
1926
+ "bits": 6
1927
+ },
1928
+ "model.layers.10.mlp.switch_mlp.up_proj": {
1929
+ "group_size": 64,
1930
+ "bits": 6
1931
+ },
1932
+ "model.layers.10.mlp.switch_mlp.down_proj": {
1933
+ "group_size": 64,
1934
+ "bits": 6
1935
+ },
1936
+ "model.layers.11.self_attn.q_proj": {
1937
+ "group_size": 64,
1938
+ "bits": 6
1939
+ },
1940
+ "model.layers.11.self_attn.k_proj": {
1941
+ "group_size": 64,
1942
+ "bits": 6
1943
+ },
1944
+ "model.layers.11.self_attn.v_proj": {
1945
+ "group_size": 64,
1946
+ "bits": 8
1947
+ },
1948
+ "model.layers.11.self_attn.o_proj": {
1949
+ "group_size": 64,
1950
+ "bits": 6
1951
+ },
1952
+ "model.layers.11.mlp.gate": {
1953
+ "group_size": 64,
1954
+ "bits": 6
1955
+ },
1956
+ "model.layers.11.mlp.switch_mlp.gate_proj": {
1957
+ "group_size": 64,
1958
+ "bits": 6
1959
+ },
1960
+ "model.layers.11.mlp.switch_mlp.up_proj": {
1961
+ "group_size": 64,
1962
+ "bits": 6
1963
+ },
1964
+ "model.layers.11.mlp.switch_mlp.down_proj": {
1965
+ "group_size": 64,
1966
+ "bits": 8
1967
+ },
1968
+ "model.layers.12.self_attn.q_proj": {
1969
+ "group_size": 64,
1970
+ "bits": 6
1971
+ },
1972
+ "model.layers.12.self_attn.k_proj": {
1973
+ "group_size": 64,
1974
+ "bits": 6
1975
+ },
1976
+ "model.layers.12.self_attn.v_proj": {
1977
+ "group_size": 64,
1978
+ "bits": 8
1979
+ },
1980
+ "model.layers.12.self_attn.o_proj": {
1981
+ "group_size": 64,
1982
+ "bits": 6
1983
+ },
1984
+ "model.layers.12.mlp.gate": {
1985
+ "group_size": 64,
1986
+ "bits": 6
1987
+ },
1988
+ "model.layers.12.mlp.switch_mlp.gate_proj": {
1989
+ "group_size": 64,
1990
+ "bits": 6
1991
+ },
1992
+ "model.layers.12.mlp.switch_mlp.up_proj": {
1993
+ "group_size": 64,
1994
+ "bits": 6
1995
+ },
1996
+ "model.layers.12.mlp.switch_mlp.down_proj": {
1997
+ "group_size": 64,
1998
+ "bits": 6
1999
+ },
2000
+ "model.layers.13.self_attn.q_proj": {
2001
+ "group_size": 64,
2002
+ "bits": 6
2003
+ },
2004
+ "model.layers.13.self_attn.k_proj": {
2005
+ "group_size": 64,
2006
+ "bits": 6
2007
+ },
2008
+ "model.layers.13.self_attn.v_proj": {
2009
+ "group_size": 64,
2010
+ "bits": 8
2011
+ },
2012
+ "model.layers.13.self_attn.o_proj": {
2013
+ "group_size": 64,
2014
+ "bits": 6
2015
+ },
2016
+ "model.layers.13.mlp.gate": {
2017
+ "group_size": 64,
2018
+ "bits": 6
2019
+ },
2020
+ "model.layers.13.mlp.switch_mlp.gate_proj": {
2021
+ "group_size": 64,
2022
+ "bits": 6
2023
+ },
2024
+ "model.layers.13.mlp.switch_mlp.up_proj": {
2025
+ "group_size": 64,
2026
+ "bits": 6
2027
+ },
2028
+ "model.layers.13.mlp.switch_mlp.down_proj": {
2029
+ "group_size": 64,
2030
+ "bits": 6
2031
+ },
2032
+ "model.layers.14.self_attn.q_proj": {
2033
+ "group_size": 64,
2034
+ "bits": 6
2035
+ },
2036
+ "model.layers.14.self_attn.k_proj": {
2037
+ "group_size": 64,
2038
+ "bits": 6
2039
+ },
2040
+ "model.layers.14.self_attn.v_proj": {
2041
+ "group_size": 64,
2042
+ "bits": 8
2043
+ },
2044
+ "model.layers.14.self_attn.o_proj": {
2045
+ "group_size": 64,
2046
+ "bits": 6
2047
+ },
2048
+ "model.layers.14.mlp.gate": {
2049
+ "group_size": 64,
2050
+ "bits": 6
2051
+ },
2052
+ "model.layers.14.mlp.switch_mlp.gate_proj": {
2053
+ "group_size": 64,
2054
+ "bits": 6
2055
+ },
2056
+ "model.layers.14.mlp.switch_mlp.up_proj": {
2057
+ "group_size": 64,
2058
+ "bits": 6
2059
+ },
2060
+ "model.layers.14.mlp.switch_mlp.down_proj": {
2061
+ "group_size": 64,
2062
+ "bits": 8
2063
+ },
2064
+ "model.layers.15.self_attn.q_proj": {
2065
+ "group_size": 64,
2066
+ "bits": 6
2067
+ },
2068
+ "model.layers.15.self_attn.k_proj": {
2069
+ "group_size": 64,
2070
+ "bits": 6
2071
+ },
2072
+ "model.layers.15.self_attn.v_proj": {
2073
+ "group_size": 64,
2074
+ "bits": 8
2075
+ },
2076
+ "model.layers.15.self_attn.o_proj": {
2077
+ "group_size": 64,
2078
+ "bits": 6
2079
+ },
2080
+ "model.layers.15.mlp.gate": {
2081
+ "group_size": 64,
2082
+ "bits": 6
2083
+ },
2084
+ "model.layers.15.mlp.switch_mlp.gate_proj": {
2085
+ "group_size": 64,
2086
+ "bits": 6
2087
+ },
2088
+ "model.layers.15.mlp.switch_mlp.up_proj": {
2089
+ "group_size": 64,
2090
+ "bits": 6
2091
+ },
2092
+ "model.layers.15.mlp.switch_mlp.down_proj": {
2093
+ "group_size": 64,
2094
+ "bits": 6
2095
+ },
2096
+ "model.layers.16.self_attn.q_proj": {
2097
+ "group_size": 64,
2098
+ "bits": 6
2099
+ },
2100
+ "model.layers.16.self_attn.k_proj": {
2101
+ "group_size": 64,
2102
+ "bits": 6
2103
+ },
2104
+ "model.layers.16.self_attn.v_proj": {
2105
+ "group_size": 64,
2106
+ "bits": 8
2107
+ },
2108
+ "model.layers.16.self_attn.o_proj": {
2109
+ "group_size": 64,
2110
+ "bits": 6
2111
+ },
2112
+ "model.layers.16.mlp.gate": {
2113
+ "group_size": 64,
2114
+ "bits": 6
2115
+ },
2116
+ "model.layers.16.mlp.switch_mlp.gate_proj": {
2117
+ "group_size": 64,
2118
+ "bits": 6
2119
+ },
2120
+ "model.layers.16.mlp.switch_mlp.up_proj": {
2121
+ "group_size": 64,
2122
+ "bits": 6
2123
+ },
2124
+ "model.layers.16.mlp.switch_mlp.down_proj": {
2125
+ "group_size": 64,
2126
+ "bits": 6
2127
+ },
2128
+ "model.layers.17.self_attn.q_proj": {
2129
+ "group_size": 64,
2130
+ "bits": 6
2131
+ },
2132
+ "model.layers.17.self_attn.k_proj": {
2133
+ "group_size": 64,
2134
+ "bits": 6
2135
+ },
2136
+ "model.layers.17.self_attn.v_proj": {
2137
+ "group_size": 64,
2138
+ "bits": 8
2139
+ },
2140
+ "model.layers.17.self_attn.o_proj": {
2141
+ "group_size": 64,
2142
+ "bits": 6
2143
+ },
2144
+ "model.layers.17.mlp.gate": {
2145
+ "group_size": 64,
2146
+ "bits": 6
2147
+ },
2148
+ "model.layers.17.mlp.switch_mlp.gate_proj": {
2149
+ "group_size": 64,
2150
+ "bits": 6
2151
+ },
2152
+ "model.layers.17.mlp.switch_mlp.up_proj": {
2153
+ "group_size": 64,
2154
+ "bits": 6
2155
+ },
2156
+ "model.layers.17.mlp.switch_mlp.down_proj": {
2157
+ "group_size": 64,
2158
+ "bits": 8
2159
+ },
2160
+ "model.layers.18.self_attn.q_proj": {
2161
+ "group_size": 64,
2162
+ "bits": 6
2163
+ },
2164
+ "model.layers.18.self_attn.k_proj": {
2165
+ "group_size": 64,
2166
+ "bits": 6
2167
+ },
2168
+ "model.layers.18.self_attn.v_proj": {
2169
+ "group_size": 64,
2170
+ "bits": 8
2171
+ },
2172
+ "model.layers.18.self_attn.o_proj": {
2173
+ "group_size": 64,
2174
+ "bits": 6
2175
+ },
2176
+ "model.layers.18.mlp.gate": {
2177
+ "group_size": 64,
2178
+ "bits": 6
2179
+ },
2180
+ "model.layers.18.mlp.switch_mlp.gate_proj": {
2181
+ "group_size": 64,
2182
+ "bits": 6
2183
+ },
2184
+ "model.layers.18.mlp.switch_mlp.up_proj": {
2185
+ "group_size": 64,
2186
+ "bits": 6
2187
+ },
2188
+ "model.layers.18.mlp.switch_mlp.down_proj": {
2189
+ "group_size": 64,
2190
+ "bits": 6
2191
+ },
2192
+ "model.layers.19.self_attn.q_proj": {
2193
+ "group_size": 64,
2194
+ "bits": 6
2195
+ },
2196
+ "model.layers.19.self_attn.k_proj": {
2197
+ "group_size": 64,
2198
+ "bits": 6
2199
+ },
2200
+ "model.layers.19.self_attn.v_proj": {
2201
+ "group_size": 64,
2202
+ "bits": 8
2203
+ },
2204
+ "model.layers.19.self_attn.o_proj": {
2205
+ "group_size": 64,
2206
+ "bits": 6
2207
+ },
2208
+ "model.layers.19.mlp.gate": {
2209
+ "group_size": 64,
2210
+ "bits": 6
2211
+ },
2212
+ "model.layers.19.mlp.switch_mlp.gate_proj": {
2213
+ "group_size": 64,
2214
+ "bits": 6
2215
+ },
2216
+ "model.layers.19.mlp.switch_mlp.up_proj": {
2217
+ "group_size": 64,
2218
+ "bits": 6
2219
+ },
2220
+ "model.layers.19.mlp.switch_mlp.down_proj": {
2221
+ "group_size": 64,
2222
+ "bits": 6
2223
+ },
2224
+ "model.layers.20.self_attn.q_proj": {
2225
+ "group_size": 64,
2226
+ "bits": 6
2227
+ },
2228
+ "model.layers.20.self_attn.k_proj": {
2229
+ "group_size": 64,
2230
+ "bits": 6
2231
+ },
2232
+ "model.layers.20.self_attn.v_proj": {
2233
+ "group_size": 64,
2234
+ "bits": 8
2235
+ },
2236
+ "model.layers.20.self_attn.o_proj": {
2237
+ "group_size": 64,
2238
+ "bits": 6
2239
+ },
2240
+ "model.layers.20.mlp.gate": {
2241
+ "group_size": 64,
2242
+ "bits": 6
2243
+ },
2244
+ "model.layers.20.mlp.switch_mlp.gate_proj": {
2245
+ "group_size": 64,
2246
+ "bits": 6
2247
+ },
2248
+ "model.layers.20.mlp.switch_mlp.up_proj": {
2249
+ "group_size": 64,
2250
+ "bits": 6
2251
+ },
2252
+ "model.layers.20.mlp.switch_mlp.down_proj": {
2253
+ "group_size": 64,
2254
+ "bits": 8
2255
+ },
2256
+ "model.layers.21.self_attn.q_proj": {
2257
+ "group_size": 64,
2258
+ "bits": 6
2259
+ },
2260
+ "model.layers.21.self_attn.k_proj": {
2261
+ "group_size": 64,
2262
+ "bits": 6
2263
+ },
2264
+ "model.layers.21.self_attn.v_proj": {
2265
+ "group_size": 64,
2266
+ "bits": 8
2267
+ },
2268
+ "model.layers.21.self_attn.o_proj": {
2269
+ "group_size": 64,
2270
+ "bits": 6
2271
+ },
2272
+ "model.layers.21.mlp.gate": {
2273
+ "group_size": 64,
2274
+ "bits": 6
2275
+ },
2276
+ "model.layers.21.mlp.switch_mlp.gate_proj": {
2277
+ "group_size": 64,
2278
+ "bits": 6
2279
+ },
2280
+ "model.layers.21.mlp.switch_mlp.up_proj": {
2281
+ "group_size": 64,
2282
+ "bits": 6
2283
+ },
2284
+ "model.layers.21.mlp.switch_mlp.down_proj": {
2285
+ "group_size": 64,
2286
+ "bits": 6
2287
+ },
2288
+ "model.layers.22.self_attn.q_proj": {
2289
+ "group_size": 64,
2290
+ "bits": 6
2291
+ },
2292
+ "model.layers.22.self_attn.k_proj": {
2293
+ "group_size": 64,
2294
+ "bits": 6
2295
+ },
2296
+ "model.layers.22.self_attn.v_proj": {
2297
+ "group_size": 64,
2298
+ "bits": 8
2299
+ },
2300
+ "model.layers.22.self_attn.o_proj": {
2301
+ "group_size": 64,
2302
+ "bits": 6
2303
+ },
2304
+ "model.layers.22.mlp.gate": {
2305
+ "group_size": 64,
2306
+ "bits": 6
2307
+ },
2308
+ "model.layers.22.mlp.switch_mlp.gate_proj": {
2309
+ "group_size": 64,
2310
+ "bits": 6
2311
+ },
2312
+ "model.layers.22.mlp.switch_mlp.up_proj": {
2313
+ "group_size": 64,
2314
+ "bits": 6
2315
+ },
2316
+ "model.layers.22.mlp.switch_mlp.down_proj": {
2317
+ "group_size": 64,
2318
+ "bits": 6
2319
+ },
2320
+ "model.layers.23.self_attn.q_proj": {
2321
+ "group_size": 64,
2322
+ "bits": 6
2323
+ },
2324
+ "model.layers.23.self_attn.k_proj": {
2325
+ "group_size": 64,
2326
+ "bits": 6
2327
+ },
2328
+ "model.layers.23.self_attn.v_proj": {
2329
+ "group_size": 64,
2330
+ "bits": 8
2331
+ },
2332
+ "model.layers.23.self_attn.o_proj": {
2333
+ "group_size": 64,
2334
+ "bits": 6
2335
+ },
2336
+ "model.layers.23.mlp.gate": {
2337
+ "group_size": 64,
2338
+ "bits": 6
2339
+ },
2340
+ "model.layers.23.mlp.switch_mlp.gate_proj": {
2341
+ "group_size": 64,
2342
+ "bits": 6
2343
+ },
2344
+ "model.layers.23.mlp.switch_mlp.up_proj": {
2345
+ "group_size": 64,
2346
+ "bits": 6
2347
+ },
2348
+ "model.layers.23.mlp.switch_mlp.down_proj": {
2349
+ "group_size": 64,
2350
+ "bits": 8
2351
+ },
2352
+ "model.layers.24.self_attn.q_proj": {
2353
+ "group_size": 64,
2354
+ "bits": 6
2355
+ },
2356
+ "model.layers.24.self_attn.k_proj": {
2357
+ "group_size": 64,
2358
+ "bits": 6
2359
+ },
2360
+ "model.layers.24.self_attn.v_proj": {
2361
+ "group_size": 64,
2362
+ "bits": 8
2363
+ },
2364
+ "model.layers.24.self_attn.o_proj": {
2365
+ "group_size": 64,
2366
+ "bits": 6
2367
+ },
2368
+ "model.layers.24.mlp.gate": {
2369
+ "group_size": 64,
2370
+ "bits": 6
2371
+ },
2372
+ "model.layers.24.mlp.switch_mlp.gate_proj": {
2373
+ "group_size": 64,
2374
+ "bits": 6
2375
+ },
2376
+ "model.layers.24.mlp.switch_mlp.up_proj": {
2377
+ "group_size": 64,
2378
+ "bits": 6
2379
+ },
2380
+ "model.layers.24.mlp.switch_mlp.down_proj": {
2381
+ "group_size": 64,
2382
+ "bits": 6
2383
+ },
2384
+ "model.layers.25.self_attn.q_proj": {
2385
+ "group_size": 64,
2386
+ "bits": 6
2387
+ },
2388
+ "model.layers.25.self_attn.k_proj": {
2389
+ "group_size": 64,
2390
+ "bits": 6
2391
+ },
2392
+ "model.layers.25.self_attn.v_proj": {
2393
+ "group_size": 64,
2394
+ "bits": 8
2395
+ },
2396
+ "model.layers.25.self_attn.o_proj": {
2397
+ "group_size": 64,
2398
+ "bits": 6
2399
+ },
2400
+ "model.layers.25.mlp.gate": {
2401
+ "group_size": 64,
2402
+ "bits": 6
2403
+ },
2404
+ "model.layers.25.mlp.switch_mlp.gate_proj": {
2405
+ "group_size": 64,
2406
+ "bits": 6
2407
+ },
2408
+ "model.layers.25.mlp.switch_mlp.up_proj": {
2409
+ "group_size": 64,
2410
+ "bits": 6
2411
+ },
2412
+ "model.layers.25.mlp.switch_mlp.down_proj": {
2413
+ "group_size": 64,
2414
+ "bits": 6
2415
+ },
2416
+ "model.layers.26.self_attn.q_proj": {
2417
+ "group_size": 64,
2418
+ "bits": 6
2419
+ },
2420
+ "model.layers.26.self_attn.k_proj": {
2421
+ "group_size": 64,
2422
+ "bits": 6
2423
+ },
2424
+ "model.layers.26.self_attn.v_proj": {
2425
+ "group_size": 64,
2426
+ "bits": 8
2427
+ },
2428
+ "model.layers.26.self_attn.o_proj": {
2429
+ "group_size": 64,
2430
+ "bits": 6
2431
+ },
2432
+ "model.layers.26.mlp.gate": {
2433
+ "group_size": 64,
2434
+ "bits": 6
2435
+ },
2436
+ "model.layers.26.mlp.switch_mlp.gate_proj": {
2437
+ "group_size": 64,
2438
+ "bits": 6
2439
+ },
2440
+ "model.layers.26.mlp.switch_mlp.up_proj": {
2441
+ "group_size": 64,
2442
+ "bits": 6
2443
+ },
2444
+ "model.layers.26.mlp.switch_mlp.down_proj": {
2445
+ "group_size": 64,
2446
+ "bits": 8
2447
+ },
2448
+ "model.layers.27.self_attn.q_proj": {
2449
+ "group_size": 64,
2450
+ "bits": 6
2451
+ },
2452
+ "model.layers.27.self_attn.k_proj": {
2453
+ "group_size": 64,
2454
+ "bits": 6
2455
+ },
2456
+ "model.layers.27.self_attn.v_proj": {
2457
+ "group_size": 64,
2458
+ "bits": 8
2459
+ },
2460
+ "model.layers.27.self_attn.o_proj": {
2461
+ "group_size": 64,
2462
+ "bits": 6
2463
+ },
2464
+ "model.layers.27.mlp.gate": {
2465
+ "group_size": 64,
2466
+ "bits": 6
2467
+ },
2468
+ "model.layers.27.mlp.switch_mlp.gate_proj": {
2469
+ "group_size": 64,
2470
+ "bits": 6
2471
+ },
2472
+ "model.layers.27.mlp.switch_mlp.up_proj": {
2473
+ "group_size": 64,
2474
+ "bits": 6
2475
+ },
2476
+ "model.layers.27.mlp.switch_mlp.down_proj": {
2477
+ "group_size": 64,
2478
+ "bits": 6
2479
+ },
2480
+ "model.layers.28.self_attn.q_proj": {
2481
+ "group_size": 64,
2482
+ "bits": 6
2483
+ },
2484
+ "model.layers.28.self_attn.k_proj": {
2485
+ "group_size": 64,
2486
+ "bits": 6
2487
+ },
2488
+ "model.layers.28.self_attn.v_proj": {
2489
+ "group_size": 64,
2490
+ "bits": 8
2491
+ },
2492
+ "model.layers.28.self_attn.o_proj": {
2493
+ "group_size": 64,
2494
+ "bits": 6
2495
+ },
2496
+ "model.layers.28.mlp.gate": {
2497
+ "group_size": 64,
2498
+ "bits": 6
2499
+ },
2500
+ "model.layers.28.mlp.switch_mlp.gate_proj": {
2501
+ "group_size": 64,
2502
+ "bits": 6
2503
+ },
2504
+ "model.layers.28.mlp.switch_mlp.up_proj": {
2505
+ "group_size": 64,
2506
+ "bits": 6
2507
+ },
2508
+ "model.layers.28.mlp.switch_mlp.down_proj": {
2509
+ "group_size": 64,
2510
+ "bits": 6
2511
+ },
2512
+ "model.layers.29.self_attn.q_proj": {
2513
+ "group_size": 64,
2514
+ "bits": 6
2515
+ },
2516
+ "model.layers.29.self_attn.k_proj": {
2517
+ "group_size": 64,
2518
+ "bits": 6
2519
+ },
2520
+ "model.layers.29.self_attn.v_proj": {
2521
+ "group_size": 64,
2522
+ "bits": 8
2523
+ },
2524
+ "model.layers.29.self_attn.o_proj": {
2525
+ "group_size": 64,
2526
+ "bits": 6
2527
+ },
2528
+ "model.layers.29.mlp.gate": {
2529
+ "group_size": 64,
2530
+ "bits": 6
2531
+ },
2532
+ "model.layers.29.mlp.switch_mlp.gate_proj": {
2533
+ "group_size": 64,
2534
+ "bits": 6
2535
+ },
2536
+ "model.layers.29.mlp.switch_mlp.up_proj": {
2537
+ "group_size": 64,
2538
+ "bits": 6
2539
+ },
2540
+ "model.layers.29.mlp.switch_mlp.down_proj": {
2541
+ "group_size": 64,
2542
+ "bits": 8
2543
+ },
2544
+ "model.layers.30.self_attn.q_proj": {
2545
+ "group_size": 64,
2546
+ "bits": 6
2547
+ },
2548
+ "model.layers.30.self_attn.k_proj": {
2549
+ "group_size": 64,
2550
+ "bits": 6
2551
+ },
2552
+ "model.layers.30.self_attn.v_proj": {
2553
+ "group_size": 64,
2554
+ "bits": 8
2555
+ },
2556
+ "model.layers.30.self_attn.o_proj": {
2557
+ "group_size": 64,
2558
+ "bits": 6
2559
+ },
2560
+ "model.layers.30.mlp.gate": {
2561
+ "group_size": 64,
2562
+ "bits": 6
2563
+ },
2564
+ "model.layers.30.mlp.switch_mlp.gate_proj": {
2565
+ "group_size": 64,
2566
+ "bits": 6
2567
+ },
2568
+ "model.layers.30.mlp.switch_mlp.up_proj": {
2569
+ "group_size": 64,
2570
+ "bits": 6
2571
+ },
2572
+ "model.layers.30.mlp.switch_mlp.down_proj": {
2573
+ "group_size": 64,
2574
+ "bits": 6
2575
+ },
2576
+ "model.layers.31.self_attn.q_proj": {
2577
+ "group_size": 64,
2578
+ "bits": 6
2579
+ },
2580
+ "model.layers.31.self_attn.k_proj": {
2581
+ "group_size": 64,
2582
+ "bits": 6
2583
+ },
2584
+ "model.layers.31.self_attn.v_proj": {
2585
+ "group_size": 64,
2586
+ "bits": 8
2587
+ },
2588
+ "model.layers.31.self_attn.o_proj": {
2589
+ "group_size": 64,
2590
+ "bits": 6
2591
+ },
2592
+ "model.layers.31.mlp.gate": {
2593
+ "group_size": 64,
2594
+ "bits": 6
2595
+ },
2596
+ "model.layers.31.mlp.switch_mlp.gate_proj": {
2597
+ "group_size": 64,
2598
+ "bits": 6
2599
+ },
2600
+ "model.layers.31.mlp.switch_mlp.up_proj": {
2601
+ "group_size": 64,
2602
+ "bits": 6
2603
+ },
2604
+ "model.layers.31.mlp.switch_mlp.down_proj": {
2605
+ "group_size": 64,
2606
+ "bits": 6
2607
+ },
2608
+ "model.layers.32.self_attn.q_proj": {
2609
+ "group_size": 64,
2610
+ "bits": 6
2611
+ },
2612
+ "model.layers.32.self_attn.k_proj": {
2613
+ "group_size": 64,
2614
+ "bits": 6
2615
+ },
2616
+ "model.layers.32.self_attn.v_proj": {
2617
+ "group_size": 64,
2618
+ "bits": 8
2619
+ },
2620
+ "model.layers.32.self_attn.o_proj": {
2621
+ "group_size": 64,
2622
+ "bits": 6
2623
+ },
2624
+ "model.layers.32.mlp.gate": {
2625
+ "group_size": 64,
2626
+ "bits": 6
2627
+ },
2628
+ "model.layers.32.mlp.switch_mlp.gate_proj": {
2629
+ "group_size": 64,
2630
+ "bits": 6
2631
+ },
2632
+ "model.layers.32.mlp.switch_mlp.up_proj": {
2633
+ "group_size": 64,
2634
+ "bits": 6
2635
+ },
2636
+ "model.layers.32.mlp.switch_mlp.down_proj": {
2637
+ "group_size": 64,
2638
+ "bits": 8
2639
+ },
2640
+ "model.layers.33.self_attn.q_proj": {
2641
+ "group_size": 64,
2642
+ "bits": 6
2643
+ },
2644
+ "model.layers.33.self_attn.k_proj": {
2645
+ "group_size": 64,
2646
+ "bits": 6
2647
+ },
2648
+ "model.layers.33.self_attn.v_proj": {
2649
+ "group_size": 64,
2650
+ "bits": 8
2651
+ },
2652
+ "model.layers.33.self_attn.o_proj": {
2653
+ "group_size": 64,
2654
+ "bits": 6
2655
+ },
2656
+ "model.layers.33.mlp.gate": {
2657
+ "group_size": 64,
2658
+ "bits": 6
2659
+ },
2660
+ "model.layers.33.mlp.switch_mlp.gate_proj": {
2661
+ "group_size": 64,
2662
+ "bits": 6
2663
+ },
2664
+ "model.layers.33.mlp.switch_mlp.up_proj": {
2665
+ "group_size": 64,
2666
+ "bits": 6
2667
+ },
2668
+ "model.layers.33.mlp.switch_mlp.down_proj": {
2669
+ "group_size": 64,
2670
+ "bits": 6
2671
+ },
2672
+ "model.layers.34.self_attn.q_proj": {
2673
+ "group_size": 64,
2674
+ "bits": 6
2675
+ },
2676
+ "model.layers.34.self_attn.k_proj": {
2677
+ "group_size": 64,
2678
+ "bits": 6
2679
+ },
2680
+ "model.layers.34.self_attn.v_proj": {
2681
+ "group_size": 64,
2682
+ "bits": 8
2683
+ },
2684
+ "model.layers.34.self_attn.o_proj": {
2685
+ "group_size": 64,
2686
+ "bits": 6
2687
+ },
2688
+ "model.layers.34.mlp.gate": {
2689
+ "group_size": 64,
2690
+ "bits": 6
2691
+ },
2692
+ "model.layers.34.mlp.switch_mlp.gate_proj": {
2693
+ "group_size": 64,
2694
+ "bits": 6
2695
+ },
2696
+ "model.layers.34.mlp.switch_mlp.up_proj": {
2697
+ "group_size": 64,
2698
+ "bits": 6
2699
+ },
2700
+ "model.layers.34.mlp.switch_mlp.down_proj": {
2701
+ "group_size": 64,
2702
+ "bits": 6
2703
+ },
2704
+ "model.layers.35.self_attn.q_proj": {
2705
+ "group_size": 64,
2706
+ "bits": 6
2707
+ },
2708
+ "model.layers.35.self_attn.k_proj": {
2709
+ "group_size": 64,
2710
+ "bits": 6
2711
+ },
2712
+ "model.layers.35.self_attn.v_proj": {
2713
+ "group_size": 64,
2714
+ "bits": 8
2715
+ },
2716
+ "model.layers.35.self_attn.o_proj": {
2717
+ "group_size": 64,
2718
+ "bits": 6
2719
+ },
2720
+ "model.layers.35.mlp.gate": {
2721
+ "group_size": 64,
2722
+ "bits": 6
2723
+ },
2724
+ "model.layers.35.mlp.switch_mlp.gate_proj": {
2725
+ "group_size": 64,
2726
+ "bits": 6
2727
+ },
2728
+ "model.layers.35.mlp.switch_mlp.up_proj": {
2729
+ "group_size": 64,
2730
+ "bits": 6
2731
+ },
2732
+ "model.layers.35.mlp.switch_mlp.down_proj": {
2733
+ "group_size": 64,
2734
+ "bits": 8
2735
+ },
2736
+ "model.layers.36.self_attn.q_proj": {
2737
+ "group_size": 64,
2738
+ "bits": 6
2739
+ },
2740
+ "model.layers.36.self_attn.k_proj": {
2741
+ "group_size": 64,
2742
+ "bits": 6
2743
+ },
2744
+ "model.layers.36.self_attn.v_proj": {
2745
+ "group_size": 64,
2746
+ "bits": 8
2747
+ },
2748
+ "model.layers.36.self_attn.o_proj": {
2749
+ "group_size": 64,
2750
+ "bits": 6
2751
+ },
2752
+ "model.layers.36.mlp.gate": {
2753
+ "group_size": 64,
2754
+ "bits": 6
2755
+ },
2756
+ "model.layers.36.mlp.switch_mlp.gate_proj": {
2757
+ "group_size": 64,
2758
+ "bits": 6
2759
+ },
2760
+ "model.layers.36.mlp.switch_mlp.up_proj": {
2761
+ "group_size": 64,
2762
+ "bits": 6
2763
+ },
2764
+ "model.layers.36.mlp.switch_mlp.down_proj": {
2765
+ "group_size": 64,
2766
+ "bits": 6
2767
+ },
2768
+ "model.layers.37.self_attn.q_proj": {
2769
+ "group_size": 64,
2770
+ "bits": 6
2771
+ },
2772
+ "model.layers.37.self_attn.k_proj": {
2773
+ "group_size": 64,
2774
+ "bits": 6
2775
+ },
2776
+ "model.layers.37.self_attn.v_proj": {
2777
+ "group_size": 64,
2778
+ "bits": 8
2779
+ },
2780
+ "model.layers.37.self_attn.o_proj": {
2781
+ "group_size": 64,
2782
+ "bits": 6
2783
+ },
2784
+ "model.layers.37.mlp.gate": {
2785
+ "group_size": 64,
2786
+ "bits": 6
2787
+ },
2788
+ "model.layers.37.mlp.switch_mlp.gate_proj": {
2789
+ "group_size": 64,
2790
+ "bits": 6
2791
+ },
2792
+ "model.layers.37.mlp.switch_mlp.up_proj": {
2793
+ "group_size": 64,
2794
+ "bits": 6
2795
+ },
2796
+ "model.layers.37.mlp.switch_mlp.down_proj": {
2797
+ "group_size": 64,
2798
+ "bits": 6
2799
+ },
2800
+ "model.layers.38.self_attn.q_proj": {
2801
+ "group_size": 64,
2802
+ "bits": 6
2803
+ },
2804
+ "model.layers.38.self_attn.k_proj": {
2805
+ "group_size": 64,
2806
+ "bits": 6
2807
+ },
2808
+ "model.layers.38.self_attn.v_proj": {
2809
+ "group_size": 64,
2810
+ "bits": 8
2811
+ },
2812
+ "model.layers.38.self_attn.o_proj": {
2813
+ "group_size": 64,
2814
+ "bits": 6
2815
+ },
2816
+ "model.layers.38.mlp.gate": {
2817
+ "group_size": 64,
2818
+ "bits": 6
2819
+ },
2820
+ "model.layers.38.mlp.switch_mlp.gate_proj": {
2821
+ "group_size": 64,
2822
+ "bits": 6
2823
+ },
2824
+ "model.layers.38.mlp.switch_mlp.up_proj": {
2825
+ "group_size": 64,
2826
+ "bits": 6
2827
+ },
2828
+ "model.layers.38.mlp.switch_mlp.down_proj": {
2829
+ "group_size": 64,
2830
+ "bits": 8
2831
+ },
2832
+ "model.layers.39.self_attn.q_proj": {
2833
+ "group_size": 64,
2834
+ "bits": 6
2835
+ },
2836
+ "model.layers.39.self_attn.k_proj": {
2837
+ "group_size": 64,
2838
+ "bits": 6
2839
+ },
2840
+ "model.layers.39.self_attn.v_proj": {
2841
+ "group_size": 64,
2842
+ "bits": 8
2843
+ },
2844
+ "model.layers.39.self_attn.o_proj": {
2845
+ "group_size": 64,
2846
+ "bits": 6
2847
+ },
2848
+ "model.layers.39.mlp.gate": {
2849
+ "group_size": 64,
2850
+ "bits": 6
2851
+ },
2852
+ "model.layers.39.mlp.switch_mlp.gate_proj": {
2853
+ "group_size": 64,
2854
+ "bits": 6
2855
+ },
2856
+ "model.layers.39.mlp.switch_mlp.up_proj": {
2857
+ "group_size": 64,
2858
+ "bits": 6
2859
+ },
2860
+ "model.layers.39.mlp.switch_mlp.down_proj": {
2861
+ "group_size": 64,
2862
+ "bits": 6
2863
+ },
2864
+ "model.layers.40.self_attn.q_proj": {
2865
+ "group_size": 64,
2866
+ "bits": 6
2867
+ },
2868
+ "model.layers.40.self_attn.k_proj": {
2869
+ "group_size": 64,
2870
+ "bits": 6
2871
+ },
2872
+ "model.layers.40.self_attn.v_proj": {
2873
+ "group_size": 64,
2874
+ "bits": 8
2875
+ },
2876
+ "model.layers.40.self_attn.o_proj": {
2877
+ "group_size": 64,
2878
+ "bits": 6
2879
+ },
2880
+ "model.layers.40.mlp.gate": {
2881
+ "group_size": 64,
2882
+ "bits": 6
2883
+ },
2884
+ "model.layers.40.mlp.switch_mlp.gate_proj": {
2885
+ "group_size": 64,
2886
+ "bits": 6
2887
+ },
2888
+ "model.layers.40.mlp.switch_mlp.up_proj": {
2889
+ "group_size": 64,
2890
+ "bits": 6
2891
+ },
2892
+ "model.layers.40.mlp.switch_mlp.down_proj": {
2893
+ "group_size": 64,
2894
+ "bits": 6
2895
+ },
2896
+ "model.layers.41.self_attn.q_proj": {
2897
+ "group_size": 64,
2898
+ "bits": 6
2899
+ },
2900
+ "model.layers.41.self_attn.k_proj": {
2901
+ "group_size": 64,
2902
+ "bits": 6
2903
+ },
2904
+ "model.layers.41.self_attn.v_proj": {
2905
+ "group_size": 64,
2906
+ "bits": 8
2907
+ },
2908
+ "model.layers.41.self_attn.o_proj": {
2909
+ "group_size": 64,
2910
+ "bits": 6
2911
+ },
2912
+ "model.layers.41.mlp.gate": {
2913
+ "group_size": 64,
2914
+ "bits": 6
2915
+ },
2916
+ "model.layers.41.mlp.switch_mlp.gate_proj": {
2917
+ "group_size": 64,
2918
+ "bits": 6
2919
+ },
2920
+ "model.layers.41.mlp.switch_mlp.up_proj": {
2921
+ "group_size": 64,
2922
+ "bits": 6
2923
+ },
2924
+ "model.layers.41.mlp.switch_mlp.down_proj": {
2925
+ "group_size": 64,
2926
+ "bits": 8
2927
+ },
2928
+ "model.layers.42.self_attn.q_proj": {
2929
+ "group_size": 64,
2930
+ "bits": 6
2931
+ },
2932
+ "model.layers.42.self_attn.k_proj": {
2933
+ "group_size": 64,
2934
+ "bits": 6
2935
+ },
2936
+ "model.layers.42.self_attn.v_proj": {
2937
+ "group_size": 64,
2938
+ "bits": 8
2939
+ },
2940
+ "model.layers.42.self_attn.o_proj": {
2941
+ "group_size": 64,
2942
+ "bits": 6
2943
+ },
2944
+ "model.layers.42.mlp.gate": {
2945
+ "group_size": 64,
2946
+ "bits": 6
2947
+ },
2948
+ "model.layers.42.mlp.switch_mlp.gate_proj": {
2949
+ "group_size": 64,
2950
+ "bits": 6
2951
+ },
2952
+ "model.layers.42.mlp.switch_mlp.up_proj": {
2953
+ "group_size": 64,
2954
+ "bits": 6
2955
+ },
2956
+ "model.layers.42.mlp.switch_mlp.down_proj": {
2957
+ "group_size": 64,
2958
+ "bits": 8
2959
+ },
2960
+ "model.layers.43.self_attn.q_proj": {
2961
+ "group_size": 64,
2962
+ "bits": 6
2963
+ },
2964
+ "model.layers.43.self_attn.k_proj": {
2965
+ "group_size": 64,
2966
+ "bits": 6
2967
+ },
2968
+ "model.layers.43.self_attn.v_proj": {
2969
+ "group_size": 64,
2970
+ "bits": 8
2971
+ },
2972
+ "model.layers.43.self_attn.o_proj": {
2973
+ "group_size": 64,
2974
+ "bits": 6
2975
+ },
2976
+ "model.layers.43.mlp.gate": {
2977
+ "group_size": 64,
2978
+ "bits": 6
2979
+ },
2980
+ "model.layers.43.mlp.switch_mlp.gate_proj": {
2981
+ "group_size": 64,
2982
+ "bits": 6
2983
+ },
2984
+ "model.layers.43.mlp.switch_mlp.up_proj": {
2985
+ "group_size": 64,
2986
+ "bits": 6
2987
+ },
2988
+ "model.layers.43.mlp.switch_mlp.down_proj": {
2989
+ "group_size": 64,
2990
+ "bits": 8
2991
+ },
2992
+ "model.layers.44.self_attn.q_proj": {
2993
+ "group_size": 64,
2994
+ "bits": 6
2995
+ },
2996
+ "model.layers.44.self_attn.k_proj": {
2997
+ "group_size": 64,
2998
+ "bits": 6
2999
+ },
3000
+ "model.layers.44.self_attn.v_proj": {
3001
+ "group_size": 64,
3002
+ "bits": 8
3003
+ },
3004
+ "model.layers.44.self_attn.o_proj": {
3005
+ "group_size": 64,
3006
+ "bits": 6
3007
+ },
3008
+ "model.layers.44.mlp.gate": {
3009
+ "group_size": 64,
3010
+ "bits": 6
3011
+ },
3012
+ "model.layers.44.mlp.switch_mlp.gate_proj": {
3013
+ "group_size": 64,
3014
+ "bits": 6
3015
+ },
3016
+ "model.layers.44.mlp.switch_mlp.up_proj": {
3017
+ "group_size": 64,
3018
+ "bits": 6
3019
+ },
3020
+ "model.layers.44.mlp.switch_mlp.down_proj": {
3021
+ "group_size": 64,
3022
+ "bits": 8
3023
+ },
3024
+ "model.layers.45.self_attn.q_proj": {
3025
+ "group_size": 64,
3026
+ "bits": 6
3027
+ },
3028
+ "model.layers.45.self_attn.k_proj": {
3029
+ "group_size": 64,
3030
+ "bits": 6
3031
+ },
3032
+ "model.layers.45.self_attn.v_proj": {
3033
+ "group_size": 64,
3034
+ "bits": 8
3035
+ },
3036
+ "model.layers.45.self_attn.o_proj": {
3037
+ "group_size": 64,
3038
+ "bits": 6
3039
+ },
3040
+ "model.layers.45.mlp.gate": {
3041
+ "group_size": 64,
3042
+ "bits": 6
3043
+ },
3044
+ "model.layers.45.mlp.switch_mlp.gate_proj": {
3045
+ "group_size": 64,
3046
+ "bits": 6
3047
+ },
3048
+ "model.layers.45.mlp.switch_mlp.up_proj": {
3049
+ "group_size": 64,
3050
+ "bits": 6
3051
+ },
3052
+ "model.layers.45.mlp.switch_mlp.down_proj": {
3053
+ "group_size": 64,
3054
+ "bits": 8
3055
+ },
3056
+ "model.layers.46.self_attn.q_proj": {
3057
+ "group_size": 64,
3058
+ "bits": 6
3059
+ },
3060
+ "model.layers.46.self_attn.k_proj": {
3061
+ "group_size": 64,
3062
+ "bits": 6
3063
+ },
3064
+ "model.layers.46.self_attn.v_proj": {
3065
+ "group_size": 64,
3066
+ "bits": 8
3067
+ },
3068
+ "model.layers.46.self_attn.o_proj": {
3069
+ "group_size": 64,
3070
+ "bits": 6
3071
+ },
3072
+ "model.layers.46.mlp.gate": {
3073
+ "group_size": 64,
3074
+ "bits": 6
3075
+ },
3076
+ "model.layers.46.mlp.switch_mlp.gate_proj": {
3077
+ "group_size": 64,
3078
+ "bits": 6
3079
+ },
3080
+ "model.layers.46.mlp.switch_mlp.up_proj": {
3081
+ "group_size": 64,
3082
+ "bits": 6
3083
+ },
3084
+ "model.layers.46.mlp.switch_mlp.down_proj": {
3085
+ "group_size": 64,
3086
+ "bits": 8
3087
+ },
3088
+ "model.layers.47.self_attn.q_proj": {
3089
+ "group_size": 64,
3090
+ "bits": 6
3091
+ },
3092
+ "model.layers.47.self_attn.k_proj": {
3093
+ "group_size": 64,
3094
+ "bits": 6
3095
+ },
3096
+ "model.layers.47.self_attn.v_proj": {
3097
+ "group_size": 64,
3098
+ "bits": 8
3099
+ },
3100
+ "model.layers.47.self_attn.o_proj": {
3101
+ "group_size": 64,
3102
+ "bits": 6
3103
+ },
3104
+ "model.layers.47.mlp.gate": {
3105
+ "group_size": 64,
3106
+ "bits": 6
3107
+ },
3108
+ "model.layers.47.mlp.switch_mlp.gate_proj": {
3109
+ "group_size": 64,
3110
+ "bits": 6
3111
+ },
3112
+ "model.layers.47.mlp.switch_mlp.up_proj": {
3113
+ "group_size": 64,
3114
+ "bits": 6
3115
+ },
3116
+ "model.layers.47.mlp.switch_mlp.down_proj": {
3117
+ "group_size": 64,
3118
+ "bits": 8
3119
+ },
3120
+ "lm_head": {
3121
+ "group_size": 64,
3122
+ "bits": 8
3123
+ }
3124
+ },
3125
+ "rms_norm_eps": 1e-06,
3126
+ "rope_scaling": null,
3127
+ "rope_theta": 10000000,
3128
+ "router_aux_loss_coef": 0.001,
3129
+ "sliding_window": null,
3130
+ "tie_word_embeddings": false,
3131
+ "torch_dtype": "bfloat16",
3132
+ "transformers_version": "4.51.0",
3133
+ "use_cache": true,
3134
+ "use_sliding_window": false,
3135
+ "vocab_size": 151936
3136
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "temperature": 0.7,
10
+ "top_k": 20,
11
+ "top_p": 0.8,
12
+ "transformers_version": "4.51.0"
13
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2f8dd4bf4387de6147deb6baf102c665788653f9b7373b532c6101b834da3df
3
+ size 5362946044
model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68a6083de8d8e07002a020ae3e7c427733eada6c1c47c278f3f32de6ec7219f8
3
+ size 5367551178
model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:065705612cc7addf61f26670e477d9927ba39c8a2ffc0298f3aba979ee7a9b6f
3
+ size 5229139424
model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0b128d7a4c194d344118d7b7dc49e4833e6d25bb65597cf748cf6d2eb2311a5
3
+ size 5266887880
model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:933e61664cbc8f23c693082d2c188af756043f156e569c4d1fcc474fe75b8128
3
+ size 5062544271
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
3
+ size 11422654
tokenizer_config.json ADDED
@@ -0,0 +1,239 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<tool_response>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": false
188
+ },
189
+ "151666": {
190
+ "content": "</tool_response>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": false
196
+ },
197
+ "151667": {
198
+ "content": "<think>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": false
204
+ },
205
+ "151668": {
206
+ "content": "</think>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": false
212
+ }
213
+ },
214
+ "additional_special_tokens": [
215
+ "<|im_start|>",
216
+ "<|im_end|>",
217
+ "<|object_ref_start|>",
218
+ "<|object_ref_end|>",
219
+ "<|box_start|>",
220
+ "<|box_end|>",
221
+ "<|quad_start|>",
222
+ "<|quad_end|>",
223
+ "<|vision_start|>",
224
+ "<|vision_end|>",
225
+ "<|vision_pad|>",
226
+ "<|image_pad|>",
227
+ "<|video_pad|>"
228
+ ],
229
+ "bos_token": null,
230
+ "clean_up_tokenization_spaces": false,
231
+ "eos_token": "<|im_end|>",
232
+ "errors": "replace",
233
+ "extra_special_tokens": {},
234
+ "model_max_length": 262144,
235
+ "pad_token": "<|endoftext|>",
236
+ "split_special_tokens": false,
237
+ "tokenizer_class": "Qwen2Tokenizer",
238
+ "unk_token": null
239
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff