Hugging Face's logo

EdgeCompress01
/
Qwen3-4B-GPTQ-4bit

Safetensors
qwen3
4-bit precision
gptq
Model card Files
xet
 Community
Abdohaaland commited on
Commit
2d696fe
·
verified ·
1 Parent(s): 87a9b64

Upload GPTQ quantized model

Browse files
Files changed (9) hide show
  1. .gitattributes +1 -0
  2. chat_template.jinja +89 -0
  3. config.json +117 -0
  4. generation_config.json +12 -0
  5. model.safetensors +3 -0
  6. quant_log.csv +253 -0
  7. quantize_config.json +46 -0
  8. tokenizer.json +3 -0
  9. tokenizer_config.json +29 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ {%- if enable_thinking is defined and enable_thinking is false %}
87
+ {{- '<think>\n\n</think>\n\n' }}
88
+ {%- endif %}
89
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3ForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2560,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 9728,
15
+ "layer_types": [
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention"
52
+ ],
53
+ "max_position_embeddings": 40960,
54
+ "max_window_layers": 36,
55
+ "model_type": "qwen3",
56
+ "num_attention_heads": 32,
57
+ "num_hidden_layers": 36,
58
+ "num_key_value_heads": 8,
59
+ "pad_token_id": null,
60
+ "quantization_config": {
61
+ "bits": 4,
62
+ "checkpoint_format": "gptq",
63
+ "desc_act": false,
64
+ "format": "gptq",
65
+ "group_size": 128,
66
+ "lm_head": false,
67
+ "meta": {
68
+ "act_group_aware": true,
69
+ "auto_forward_data_parallel": true,
70
+ "damp_auto_increment": 0.01,
71
+ "damp_percent": 0.05,
72
+ "failsafe": {
73
+ "smooth": {
74
+ "group_size_threshold": 128,
75
+ "k": 2.75,
76
+ "type": "mad"
77
+ },
78
+ "strategy": "rtn",
79
+ "threshold": "0.5%"
80
+ },
81
+ "gc_mode": "interval",
82
+ "gptaq": null,
83
+ "hessian": {
84
+ "chunk_bytes": null,
85
+ "chunk_size": null,
86
+ "staging_dtype": "float32"
87
+ },
88
+ "mock_quantization": false,
89
+ "mse": 0.0,
90
+ "offload_to_disk": true,
91
+ "offload_to_disk_path": "./gptqmodel_offload/wvhbwycm-bttdmogw/",
92
+ "pack_impl": "cpu",
93
+ "quantizer": [
94
+ "gptqmodel:5.7.0"
95
+ ],
96
+ "static_groups": false,
97
+ "true_sequential": true,
98
+ "uri": "https://github.com/modelcloud/gptqmodel",
99
+ "vram_strategy": "exclusive",
100
+ "wait_for_submodule_finalizers": false
101
+ },
102
+ "pack_dtype": "int32",
103
+ "quant_method": "gptq",
104
+ "sym": true
105
+ },
106
+ "rms_norm_eps": 1e-06,
107
+ "rope_parameters": {
108
+ "rope_theta": 1000000,
109
+ "rope_type": "default"
110
+ },
111
+ "sliding_window": null,
112
+ "tie_word_embeddings": true,
113
+ "transformers_version": "5.2.0",
114
+ "use_cache": true,
115
+ "use_sliding_window": false,
116
+ "vocab_size": 151936
117
+ }
generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "temperature": 0.6,
9
+ "top_k": 20,
10
+ "top_p": 0.95,
11
+ "transformers_version": "5.2.0"
12
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:422d4597f1cc666a60c602a9ce88525be70623bbfe08f5b6683fe937b4ca463f
3
+ size 2669888992
quant_log.csv ADDED
@@ -0,0 +1,253 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ layer,module,loss,samples,damp,time
2
+ 0,self_attn.k_proj,0.0000005446,0.05000,4.435
3
+ 0,self_attn.v_proj,0.0000004809,0.05000,4.490
4
+ 0,self_attn.q_proj,0.0000019953,0.05000,4.522
5
+ 0,self_attn.o_proj,0.0000013446,0.05000,1.474
6
+ 0,mlp.gate_proj,0.0001399555,0.05000,2.281
7
+ 0,mlp.up_proj,0.0001233887,0.05000,2.308
8
+ 0,mlp.down_proj,0.0000115761,0.05000,3.801
9
+ 1,self_attn.k_proj,0.0000008938,0.05000,4.646
10
+ 1,self_attn.v_proj,0.0000009459,0.05000,4.716
11
+ 1,self_attn.q_proj,0.0000033180,0.05000,4.747
12
+ 1,self_attn.o_proj,0.0000019172,0.05000,1.496
13
+ 1,mlp.gate_proj,0.0041653160,0.05000,2.307
14
+ 1,mlp.up_proj,0.0018882145,0.05000,2.328
15
+ 1,mlp.down_proj,0.0000226382,0.05000,3.953
16
+ 2,self_attn.q_proj,0.0000094249,0.05000,4.620
17
+ 2,self_attn.v_proj,0.0000027178,0.05000,4.675
18
+ 2,self_attn.k_proj,0.0000026439,0.05000,4.706
19
+ 2,self_attn.o_proj,0.0000024441,0.05000,1.528
20
+ 2,mlp.up_proj,0.0056557384,0.05000,2.318
21
+ 2,mlp.gate_proj,0.0073925028,0.05000,2.362
22
+ 2,mlp.down_proj,0.0000110274,0.05000,3.862
23
+ 3,self_attn.v_proj,0.0000051766,0.05000,4.551
24
+ 3,self_attn.k_proj,0.0000047641,0.05000,4.605
25
+ 3,self_attn.q_proj,0.0000187628,0.05000,4.655
26
+ 3,self_attn.o_proj,0.0000038860,0.05000,1.532
27
+ 3,mlp.gate_proj,0.0075710094,0.05000,2.311
28
+ 3,mlp.up_proj,0.0039994645,0.05000,2.335
29
+ 3,mlp.down_proj,0.0000351612,0.05000,3.938
30
+ 4,self_attn.v_proj,0.0000105205,0.05000,4.585
31
+ 4,self_attn.q_proj,0.0000371834,0.05000,4.616
32
+ 4,self_attn.k_proj,0.0000098234,0.05000,4.657
33
+ 4,self_attn.o_proj,0.0000067732,0.05000,1.475
34
+ 4,mlp.gate_proj,0.0071237201,0.05000,2.310
35
+ 4,mlp.up_proj,0.0029747248,0.05000,2.344
36
+ 4,mlp.down_proj,0.0000574324,0.05000,3.899
37
+ 5,self_attn.q_proj,0.0000413305,0.05000,4.582
38
+ 5,self_attn.k_proj,0.0000102927,0.05000,4.682
39
+ 5,self_attn.v_proj,0.0000116420,0.05000,4.687
40
+ 5,self_attn.o_proj,0.0000104877,0.05000,1.517
41
+ 5,mlp.gate_proj,0.0025908125,0.05000,2.189
42
+ 5,mlp.up_proj,0.0013660692,0.05000,2.231
43
+ 5,mlp.down_proj,0.0000756003,0.05000,3.986
44
+ 6,self_attn.v_proj,0.0000247758,0.05000,4.602
45
+ 6,self_attn.k_proj,0.0000211257,0.05000,4.647
46
+ 6,self_attn.q_proj,0.0000905059,0.05000,4.674
47
+ 6,self_attn.o_proj,0.0000300698,0.05000,1.613
48
+ 6,mlp.gate_proj,0.0030917147,0.05000,2.328
49
+ 6,mlp.up_proj,0.0018886445,0.05000,2.331
50
+ 6,mlp.down_proj,0.0332592290,0.05000,3.771
51
+ 7,self_attn.k_proj,0.0000501618,0.05000,4.714
52
+ 7,self_attn.v_proj,0.0000571124,0.05000,4.742
53
+ 7,self_attn.q_proj,0.0001914269,0.05000,4.786
54
+ 7,self_attn.o_proj,0.0000288241,0.05000,1.529
55
+ 7,mlp.gate_proj,0.0034643194,0.05000,2.267
56
+ 7,mlp.up_proj,0.0022614451,0.05000,2.286
57
+ 7,mlp.down_proj,0.0001239395,0.05000,3.985
58
+ 8,self_attn.v_proj,0.0000967413,0.05000,4.581
59
+ 8,self_attn.q_proj,0.0003134763,0.05000,4.654
60
+ 8,self_attn.k_proj,0.0000789401,0.05000,4.675
61
+ 8,self_attn.o_proj,0.0000430230,0.05000,1.496
62
+ 8,mlp.gate_proj,0.0026727254,0.05000,2.208
63
+ 8,mlp.up_proj,0.0022354008,0.05000,2.211
64
+ 8,mlp.down_proj,0.0002008462,0.05000,3.878
65
+ 9,self_attn.v_proj,0.0001148402,0.05000,4.588
66
+ 9,self_attn.q_proj,0.0003638978,0.05000,4.640
67
+ 9,self_attn.k_proj,0.0000975615,0.05000,4.660
68
+ 9,self_attn.o_proj,0.0000505605,0.05000,1.508
69
+ 9,mlp.up_proj,0.0028585258,0.05000,2.225
70
+ 9,mlp.gate_proj,0.0043162701,0.05000,2.256
71
+ 9,mlp.down_proj,0.0002047321,0.05000,3.899
72
+ 10,self_attn.k_proj,0.0001473746,0.05000,4.548
73
+ 10,self_attn.q_proj,0.0005854961,0.05000,4.614
74
+ 10,self_attn.v_proj,0.0001894117,0.05000,4.653
75
+ 10,self_attn.o_proj,0.0000811983,0.05000,1.509
76
+ 10,mlp.gate_proj,0.0033544317,0.05000,2.200
77
+ 10,mlp.up_proj,0.0023989295,0.05000,2.222
78
+ 10,mlp.down_proj,0.0001772248,0.05000,4.003
79
+ 11,self_attn.v_proj,0.0000827167,0.05000,4.690
80
+ 11,self_attn.q_proj,0.0002662091,0.05000,4.735
81
+ 11,self_attn.k_proj,0.0000691783,0.05000,4.782
82
+ 11,self_attn.o_proj,0.0000499789,0.05000,1.552
83
+ 11,mlp.gate_proj,0.0026805456,0.05000,2.212
84
+ 11,mlp.up_proj,0.0021580690,0.05000,2.233
85
+ 11,mlp.down_proj,0.0001650185,0.05000,3.973
86
+ 12,self_attn.q_proj,0.0003080287,0.05000,4.437
87
+ 12,self_attn.v_proj,0.0000938578,0.05000,4.512
88
+ 12,self_attn.k_proj,0.0000781946,0.05000,4.542
89
+ 12,self_attn.o_proj,0.0000578502,0.05000,1.546
90
+ 12,mlp.gate_proj,0.0023073073,0.05000,2.179
91
+ 12,mlp.up_proj,0.0020454499,0.05000,2.207
92
+ 12,mlp.down_proj,0.0001695974,0.05000,3.923
93
+ 13,self_attn.q_proj,0.0002273831,0.05000,4.594
94
+ 13,self_attn.v_proj,0.0000638472,0.05000,4.635
95
+ 13,self_attn.k_proj,0.0000578739,0.05000,4.654
96
+ 13,self_attn.o_proj,0.0000429728,0.05000,1.494
97
+ 13,mlp.up_proj,0.0020749440,0.05000,2.240
98
+ 13,mlp.gate_proj,0.0021318999,0.05000,2.263
99
+ 13,mlp.down_proj,0.0001738223,0.05000,3.795
100
+ 14,self_attn.q_proj,0.0004071593,0.05000,4.546
101
+ 14,self_attn.v_proj,0.0001184261,0.05000,4.615
102
+ 14,self_attn.k_proj,0.0001007172,0.05000,4.637
103
+ 14,self_attn.o_proj,0.0000832822,0.05000,1.471
104
+ 14,mlp.gate_proj,0.0021145169,0.05000,2.254
105
+ 14,mlp.up_proj,0.0020794049,0.05000,2.288
106
+ 14,mlp.down_proj,0.0001682187,0.05000,3.931
107
+ 15,self_attn.q_proj,0.0003949634,0.05000,4.506
108
+ 15,self_attn.k_proj,0.0000972706,0.05000,4.567
109
+ 15,self_attn.v_proj,0.0001075819,0.05000,4.595
110
+ 15,self_attn.o_proj,0.0000702698,0.05000,1.511
111
+ 15,mlp.up_proj,0.0019951703,0.05000,2.234
112
+ 15,mlp.gate_proj,0.0019534257,0.05000,2.254
113
+ 15,mlp.down_proj,0.0001479562,0.05000,3.895
114
+ 16,self_attn.k_proj,0.0001644773,0.05000,4.625
115
+ 16,self_attn.v_proj,0.0002094128,0.05000,4.660
116
+ 16,self_attn.q_proj,0.0006918110,0.05000,4.681
117
+ 16,self_attn.o_proj,0.0000710968,0.05000,1.467
118
+ 16,mlp.up_proj,0.0021764796,0.05000,2.290
119
+ 16,mlp.gate_proj,0.0022640927,0.05000,2.310
120
+ 16,mlp.down_proj,0.0033198516,0.05000,3.882
121
+ 17,self_attn.k_proj,0.0001204673,0.05000,4.521
122
+ 17,self_attn.q_proj,0.0005616263,0.05000,4.611
123
+ 17,self_attn.v_proj,0.0001526302,0.05000,4.643
124
+ 17,self_attn.o_proj,0.0000714830,0.05000,1.504
125
+ 17,mlp.gate_proj,0.0018528427,0.05000,2.291
126
+ 17,mlp.up_proj,0.0018205950,0.05000,2.312
127
+ 17,mlp.down_proj,0.0001379722,0.05000,3.952
128
+ 18,self_attn.q_proj,0.0006088453,0.05000,4.630
129
+ 18,self_attn.v_proj,0.0001791300,0.05000,4.672
130
+ 18,self_attn.k_proj,0.0001419126,0.05000,4.703
131
+ 18,self_attn.o_proj,0.0000788589,0.05000,1.453
132
+ 18,mlp.up_proj,0.0019597744,0.05000,2.320
133
+ 18,mlp.gate_proj,0.0019572557,0.05000,2.347
134
+ 18,mlp.down_proj,0.0001711005,0.05000,3.881
135
+ 19,self_attn.k_proj,0.0002610412,0.05000,4.595
136
+ 19,self_attn.v_proj,0.0003096798,0.05000,4.659
137
+ 19,self_attn.q_proj,0.0011777417,0.05000,4.678
138
+ 19,self_attn.o_proj,0.0001044770,0.05000,1.502
139
+ 19,mlp.gate_proj,0.0020530066,0.05000,2.263
140
+ 19,mlp.up_proj,0.0020742333,0.05000,2.289
141
+ 19,mlp.down_proj,0.0001970563,0.05000,3.836
142
+ 20,self_attn.q_proj,0.0010637725,0.05000,4.590
143
+ 20,self_attn.v_proj,0.0002682226,0.05000,4.641
144
+ 20,self_attn.k_proj,0.0002198448,0.05000,4.642
145
+ 20,self_attn.o_proj,0.0000939247,0.05000,1.528
146
+ 20,mlp.up_proj,0.0022529307,0.05000,2.304
147
+ 20,mlp.gate_proj,0.0021745552,0.05000,2.330
148
+ 20,mlp.down_proj,0.0002165000,0.05000,3.838
149
+ 21,self_attn.q_proj,0.0013895572,0.05000,4.544
150
+ 21,self_attn.v_proj,0.0003581413,0.05000,4.568
151
+ 21,self_attn.k_proj,0.0002983564,0.05000,4.609
152
+ 21,self_attn.o_proj,0.0001169767,0.05000,1.541
153
+ 21,mlp.gate_proj,0.0025367997,0.05000,2.320
154
+ 21,mlp.up_proj,0.0026352093,0.05000,2.351
155
+ 21,mlp.down_proj,0.0002726330,0.05000,3.887
156
+ 22,self_attn.k_proj,0.0005701520,0.05000,4.537
157
+ 22,self_attn.q_proj,0.0026601848,0.05000,4.591
158
+ 22,self_attn.v_proj,0.0007194130,0.05000,4.608
159
+ 22,self_attn.o_proj,0.0002020238,0.05000,1.531
160
+ 22,mlp.gate_proj,0.0028571501,0.05000,2.316
161
+ 22,mlp.up_proj,0.0028531155,0.05000,2.345
162
+ 22,mlp.down_proj,0.0004290918,0.05000,3.869
163
+ 23,self_attn.v_proj,0.0006810734,0.05000,4.633
164
+ 23,self_attn.q_proj,0.0026127356,0.05000,4.709
165
+ 23,self_attn.k_proj,0.0005518920,0.05000,4.724
166
+ 23,self_attn.o_proj,0.0002539472,0.05000,1.481
167
+ 23,mlp.up_proj,0.0033200211,0.05000,2.287
168
+ 23,mlp.gate_proj,0.0035143472,0.05000,2.324
169
+ 23,mlp.down_proj,0.0006081266,0.05000,3.900
170
+ 24,self_attn.q_proj,0.0042166189,0.05000,4.470
171
+ 24,self_attn.k_proj,0.0009021544,0.05000,4.528
172
+ 24,self_attn.v_proj,0.0012101352,0.05000,4.542
173
+ 24,self_attn.o_proj,0.0002619612,0.05000,1.513
174
+ 24,mlp.gate_proj,0.0039473643,0.05000,2.217
175
+ 24,mlp.up_proj,0.0036371494,0.05000,2.262
176
+ 24,mlp.down_proj,0.0007358780,0.05000,3.948
177
+ 25,self_attn.v_proj,0.0008588059,0.05000,4.613
178
+ 25,self_attn.k_proj,0.0007072059,0.05000,4.680
179
+ 25,self_attn.q_proj,0.0030050377,0.05000,4.714
180
+ 25,self_attn.o_proj,0.0001392516,0.05000,1.505
181
+ 25,mlp.up_proj,0.0041476647,0.05000,2.311
182
+ 25,mlp.gate_proj,0.0045728512,0.05000,2.333
183
+ 25,mlp.down_proj,0.0008995740,0.05000,3.832
184
+ 26,self_attn.v_proj,0.0013400384,0.05000,4.516
185
+ 26,self_attn.q_proj,0.0048303925,0.05000,4.567
186
+ 26,self_attn.k_proj,0.0010677429,0.05000,4.585
187
+ 26,self_attn.o_proj,0.0001389864,0.05000,1.591
188
+ 26,mlp.up_proj,0.0051213020,0.05000,2.350
189
+ 26,mlp.gate_proj,0.0054934720,0.05000,2.364
190
+ 26,mlp.down_proj,0.0010003413,0.05000,3.923
191
+ 27,self_attn.q_proj,0.0059571299,0.05000,4.440
192
+ 27,self_attn.k_proj,0.0012676684,0.05000,4.500
193
+ 27,self_attn.v_proj,0.0017379407,0.05000,4.525
194
+ 27,self_attn.o_proj,0.0001889476,0.05000,1.517
195
+ 27,mlp.up_proj,0.0058125995,0.05000,2.315
196
+ 27,mlp.gate_proj,0.0060353899,0.05000,2.331
197
+ 27,mlp.down_proj,0.0013123413,0.05000,3.845
198
+ 28,self_attn.k_proj,0.0014861768,0.05000,4.549
199
+ 28,self_attn.v_proj,0.0018484022,0.05000,4.578
200
+ 28,self_attn.q_proj,0.0063890707,0.05000,4.628
201
+ 28,self_attn.o_proj,0.0003140973,0.05000,1.510
202
+ 28,mlp.gate_proj,0.0067501478,0.05000,2.178
203
+ 28,mlp.up_proj,0.0067735156,0.05000,2.204
204
+ 28,mlp.down_proj,0.0018620787,0.05000,3.949
205
+ 29,self_attn.v_proj,0.0043428479,0.05000,4.569
206
+ 29,self_attn.k_proj,0.0031365861,0.05000,4.624
207
+ 29,self_attn.q_proj,0.0145508105,0.05000,4.633
208
+ 29,self_attn.o_proj,0.0002052081,0.05000,1.491
209
+ 29,mlp.up_proj,0.0082491106,0.05000,2.178
210
+ 29,mlp.gate_proj,0.0078799528,0.05000,2.198
211
+ 29,mlp.down_proj,0.0022954082,0.05000,4.057
212
+ 30,self_attn.k_proj,0.0040841100,0.05000,4.601
213
+ 30,self_attn.v_proj,0.0054587066,0.05000,4.635
214
+ 30,self_attn.q_proj,0.0169272007,0.05000,4.642
215
+ 30,self_attn.o_proj,0.0005446325,0.05000,1.495
216
+ 30,mlp.gate_proj,0.0084495943,0.05000,2.342
217
+ 30,mlp.up_proj,0.0091387503,0.05000,2.392
218
+ 30,mlp.down_proj,0.0031828321,0.05000,4.041
219
+ 31,self_attn.k_proj,0.0053448655,0.05000,4.500
220
+ 31,self_attn.v_proj,0.0078707677,0.05000,4.542
221
+ 31,self_attn.q_proj,0.0216492466,0.05000,4.579
222
+ 31,self_attn.o_proj,0.0005467853,0.05000,1.504
223
+ 31,mlp.gate_proj,0.0085940339,0.05000,2.143
224
+ 31,mlp.up_proj,0.0097583888,0.05000,2.184
225
+ 31,mlp.down_proj,0.0042225099,0.05000,3.889
226
+ 32,self_attn.q_proj,0.0314087043,0.05000,4.484
227
+ 32,self_attn.k_proj,0.0073322198,0.05000,4.576
228
+ 32,self_attn.v_proj,0.0116100621,0.05000,4.598
229
+ 32,self_attn.o_proj,0.0007367414,0.05000,1.519
230
+ 32,mlp.up_proj,0.0105576177,0.05000,2.267
231
+ 32,mlp.gate_proj,0.0090539314,0.05000,2.292
232
+ 32,mlp.down_proj,0.0053806565,0.05000,3.901
233
+ 33,self_attn.v_proj,0.0256433210,0.05000,4.733
234
+ 33,self_attn.q_proj,0.0648465085,0.05000,4.771
235
+ 33,self_attn.k_proj,0.0127558404,0.05000,4.786
236
+ 33,self_attn.o_proj,0.0010363033,0.05000,1.516
237
+ 33,mlp.gate_proj,0.0097379385,0.05000,2.270
238
+ 33,mlp.up_proj,0.0115533778,0.05000,2.308
239
+ 33,mlp.down_proj,0.0071556654,0.05000,3.931
240
+ 34,self_attn.q_proj,0.0531877879,0.05000,4.851
241
+ 34,self_attn.k_proj,0.0115072968,0.05000,4.920
242
+ 34,self_attn.v_proj,0.0209449343,0.05000,4.921
243
+ 34,self_attn.o_proj,0.0028885654,0.05000,1.596
244
+ 34,mlp.gate_proj,0.0124351588,0.05000,2.418
245
+ 34,mlp.up_proj,0.0137425665,0.05000,2.451
246
+ 34,mlp.down_proj,0.0108865213,0.05000,4.057
247
+ 35,self_attn.q_proj,0.0295646570,0.05000,4.796
248
+ 35,self_attn.v_proj,0.0100270835,0.05000,4.859
249
+ 35,self_attn.k_proj,0.0073113912,0.05000,4.915
250
+ 35,self_attn.o_proj,0.0038638669,0.05000,1.477
251
+ 35,mlp.gate_proj,0.0334204830,0.05000,2.426
252
+ 35,mlp.up_proj,0.0434612004,0.05000,2.442
253
+ 35,mlp.down_proj,0.0230910449,0.05000,3.926
quantize_config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 4,
3
+ "group_size": 128,
4
+ "desc_act": false,
5
+ "lm_head": false,
6
+ "quant_method": "gptq",
7
+ "checkpoint_format": "gptq",
8
+ "pack_dtype": "int32",
9
+ "meta": {
10
+ "quantizer": [
11
+ "gptqmodel:5.7.0"
12
+ ],
13
+ "uri": "https://github.com/modelcloud/gptqmodel",
14
+ "damp_percent": 0.05,
15
+ "damp_auto_increment": 0.01,
16
+ "static_groups": false,
17
+ "true_sequential": true,
18
+ "mse": 0.0,
19
+ "gptaq": null,
20
+ "act_group_aware": true,
21
+ "failsafe": {
22
+ "strategy": "rtn",
23
+ "threshold": "0.5%",
24
+ "smooth": {
25
+ "type": "mad",
26
+ "group_size_threshold": 128,
27
+ "k": 2.75
28
+ }
29
+ },
30
+ "offload_to_disk": true,
31
+ "offload_to_disk_path": "./gptqmodel_offload/wvhbwycm-bttdmogw/",
32
+ "pack_impl": "cpu",
33
+ "mock_quantization": false,
34
+ "gc_mode": "interval",
35
+ "wait_for_submodule_finalizers": false,
36
+ "auto_forward_data_parallel": true,
37
+ "hessian": {
38
+ "chunk_size": null,
39
+ "chunk_bytes": null,
40
+ "staging_dtype": "float32"
41
+ },
42
+ "vram_strategy": "exclusive"
43
+ },
44
+ "sym": true,
45
+ "format": "gptq"
46
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6adb90dd2c38b281238c0f606f444f129358c61dc015dd772cf77654df557cee
3
+ size 11422748
tokenizer_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 131072,
25
+ "pad_token": "<|endoftext|>",
26
+ "split_special_tokens": false,
27
+ "tokenizer_class": "Qwen2Tokenizer",
28
+ "unk_token": null
29
+ }