PerryCheng614 commited on
Commit
7813cf9
·
verified ·
1 Parent(s): 2e58a1d

Upload folder using huggingface_hub

Browse files
Files changed (4) hide show
  1. config.json +62 -0
  2. model.safetensors +3 -0
  3. quant_log.csv +113 -0
  4. quantize_config.json +21 -0
config.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_attn_implementation_autoset": true,
3
+ "_name_or_path": "meta-llama/Llama-3.2-1B-Instruct",
4
+ "architectures": [
5
+ "LlamaForCausalLM"
6
+ ],
7
+ "attention_bias": false,
8
+ "attention_dropout": 0.0,
9
+ "bos_token_id": 128000,
10
+ "eos_token_id": [
11
+ 128001,
12
+ 128008,
13
+ 128009
14
+ ],
15
+ "head_dim": 64,
16
+ "hidden_act": "silu",
17
+ "hidden_size": 2048,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 8192,
20
+ "max_position_embeddings": 131072,
21
+ "mlp_bias": false,
22
+ "model_type": "llama",
23
+ "num_attention_heads": 32,
24
+ "num_hidden_layers": 16,
25
+ "num_key_value_heads": 8,
26
+ "pretraining_tp": 1,
27
+ "quantization_config": {
28
+ "bits": 4,
29
+ "checkpoint_format": "gptq",
30
+ "desc_act": true,
31
+ "dynamic": null,
32
+ "group_size": 32,
33
+ "lm_head": false,
34
+ "meta": {
35
+ "damp_auto_increment": 0.0025,
36
+ "damp_percent": 0.01,
37
+ "mse": 0.0,
38
+ "quantizer": [
39
+ "gptqmodel:1.4.6-dev"
40
+ ],
41
+ "static_groups": false,
42
+ "true_sequential": true,
43
+ "uri": "https://github.com/modelcloud/gptqmodel"
44
+ },
45
+ "quant_method": "gptq",
46
+ "sym": true
47
+ },
48
+ "rms_norm_eps": 1e-05,
49
+ "rope_scaling": {
50
+ "factor": 32.0,
51
+ "high_freq_factor": 4.0,
52
+ "low_freq_factor": 1.0,
53
+ "original_max_position_embeddings": 8192,
54
+ "rope_type": "llama3"
55
+ },
56
+ "rope_theta": 500000.0,
57
+ "tie_word_embeddings": true,
58
+ "torch_dtype": "bfloat16",
59
+ "transformers_version": "4.47.1",
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62814e6a3861364aeb6c9f0409990c2a72817f8cd8fd72c498b0ea637523673f
3
+ size 1614733472
quant_log.csv ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ layer,module,loss,damp,time
2
+ 0,self_attn.k_proj,0.02168,0.01000,0.732
3
+ 0,self_attn.v_proj,0.00053,0.01000,0.632
4
+ 0,self_attn.q_proj,0.04424,0.01000,0.614
5
+ 0,self_attn.o_proj,0.00005,0.01000,0.607
6
+ 0,mlp.up_proj,0.03671,0.01000,0.636
7
+ 0,mlp.gate_proj,0.04622,0.01000,0.776
8
+ 0,mlp.down_proj,0.00027,0.01000,2.563
9
+ 1,self_attn.k_proj,0.03438,0.01000,0.654
10
+ 1,self_attn.v_proj,0.00195,0.01000,0.643
11
+ 1,self_attn.q_proj,0.06312,0.01000,0.658
12
+ 1,self_attn.o_proj,0.00014,0.01000,0.665
13
+ 1,mlp.up_proj,0.05251,0.01000,0.660
14
+ 1,mlp.gate_proj,0.07196,0.01000,0.793
15
+ 1,mlp.down_proj,0.13592,0.01000,2.671
16
+ 2,self_attn.k_proj,0.06812,0.01000,0.635
17
+ 2,self_attn.v_proj,0.00459,0.01000,0.633
18
+ 2,self_attn.q_proj,0.13188,0.01000,0.648
19
+ 2,self_attn.o_proj,0.00021,0.01000,0.653
20
+ 2,mlp.up_proj,0.06672,0.01000,0.667
21
+ 2,mlp.gate_proj,0.10655,0.01000,0.795
22
+ 2,mlp.down_proj,0.00084,0.01000,2.664
23
+ 3,self_attn.k_proj,0.04493,0.01000,0.652
24
+ 3,self_attn.v_proj,0.00586,0.01000,0.620
25
+ 3,self_attn.q_proj,0.09710,0.01000,0.623
26
+ 3,self_attn.o_proj,0.00041,0.01000,0.633
27
+ 3,mlp.up_proj,0.08134,0.01000,0.654
28
+ 3,mlp.gate_proj,0.16285,0.01000,0.647
29
+ 3,mlp.down_proj,0.00128,0.01000,2.671
30
+ 4,self_attn.k_proj,0.04807,0.01000,0.604
31
+ 4,self_attn.v_proj,0.00549,0.01000,0.591
32
+ 4,self_attn.q_proj,0.09806,0.01000,0.611
33
+ 4,self_attn.o_proj,0.00059,0.01000,0.626
34
+ 4,mlp.up_proj,0.08266,0.01000,0.643
35
+ 4,mlp.gate_proj,0.17954,0.01000,0.627
36
+ 4,mlp.down_proj,0.00149,0.01000,2.526
37
+ 5,self_attn.k_proj,0.07132,0.01000,0.603
38
+ 5,self_attn.v_proj,0.00474,0.01000,0.623
39
+ 5,self_attn.q_proj,0.12509,0.01000,0.610
40
+ 5,self_attn.o_proj,0.00061,0.01000,0.634
41
+ 5,mlp.up_proj,0.09010,0.01000,0.655
42
+ 5,mlp.gate_proj,0.16730,0.01000,0.623
43
+ 5,mlp.down_proj,0.00180,0.01000,2.580
44
+ 6,self_attn.k_proj,0.05819,0.01000,0.615
45
+ 6,self_attn.v_proj,0.00612,0.01000,0.594
46
+ 6,self_attn.q_proj,0.09133,0.01000,0.603
47
+ 6,self_attn.o_proj,0.00093,0.01000,0.615
48
+ 6,mlp.up_proj,0.09097,0.01000,0.661
49
+ 6,mlp.gate_proj,0.16461,0.01000,0.752
50
+ 6,mlp.down_proj,0.00181,0.01000,2.506
51
+ 7,self_attn.k_proj,0.05784,0.01000,0.603
52
+ 7,self_attn.v_proj,0.00698,0.01000,0.590
53
+ 7,self_attn.q_proj,0.10795,0.01000,0.609
54
+ 7,self_attn.o_proj,0.00090,0.01000,0.631
55
+ 7,mlp.up_proj,0.09403,0.01000,0.653
56
+ 7,mlp.gate_proj,0.15263,0.01000,0.650
57
+ 7,mlp.down_proj,0.00196,0.01000,2.499
58
+ 8,self_attn.k_proj,0.06733,0.01000,0.591
59
+ 8,self_attn.v_proj,0.00663,0.01000,0.594
60
+ 8,self_attn.q_proj,0.10911,0.01000,0.611
61
+ 8,self_attn.o_proj,0.00114,0.01000,0.620
62
+ 8,mlp.up_proj,0.10367,0.01000,0.642
63
+ 8,mlp.gate_proj,0.16448,0.01000,0.628
64
+ 8,mlp.down_proj,0.00263,0.01000,2.473
65
+ 9,self_attn.k_proj,0.05555,0.01000,0.597
66
+ 9,self_attn.v_proj,0.00782,0.01000,0.598
67
+ 9,self_attn.q_proj,0.13325,0.01000,0.625
68
+ 9,self_attn.o_proj,0.00155,0.01000,0.632
69
+ 9,mlp.up_proj,0.11070,0.01000,0.666
70
+ 9,mlp.gate_proj,0.18121,0.01000,0.649
71
+ 9,mlp.down_proj,0.00325,0.01000,2.563
72
+ 10,self_attn.k_proj,0.06845,0.01000,0.592
73
+ 10,self_attn.v_proj,0.00844,0.01000,0.592
74
+ 10,self_attn.q_proj,0.13737,0.01000,0.607
75
+ 10,self_attn.o_proj,0.00105,0.01000,0.600
76
+ 10,mlp.up_proj,0.13314,0.01000,0.625
77
+ 10,mlp.gate_proj,0.21042,0.01000,0.613
78
+ 10,mlp.down_proj,0.00412,0.01000,2.506
79
+ 11,self_attn.k_proj,0.08247,0.01000,0.590
80
+ 11,self_attn.v_proj,0.00849,0.01000,0.594
81
+ 11,self_attn.q_proj,0.13637,0.01000,0.599
82
+ 11,self_attn.o_proj,0.00090,0.01000,0.601
83
+ 11,mlp.up_proj,0.14766,0.01000,0.640
84
+ 11,mlp.gate_proj,0.22769,0.01000,0.632
85
+ 11,mlp.down_proj,0.00453,0.01000,2.479
86
+ 12,self_attn.k_proj,0.08029,0.01000,0.581
87
+ 12,self_attn.v_proj,0.00880,0.01000,0.579
88
+ 12,self_attn.q_proj,0.13529,0.01000,0.603
89
+ 12,self_attn.o_proj,0.00088,0.01000,0.585
90
+ 12,mlp.up_proj,0.15437,0.01000,0.612
91
+ 12,mlp.gate_proj,0.22473,0.01000,0.610
92
+ 12,mlp.down_proj,0.00506,0.01000,2.522
93
+ 13,self_attn.k_proj,0.07901,0.01000,0.610
94
+ 13,self_attn.v_proj,0.01435,0.01000,0.578
95
+ 13,self_attn.q_proj,0.15220,0.01000,0.602
96
+ 13,self_attn.o_proj,0.00121,0.01000,0.613
97
+ 13,mlp.up_proj,0.18113,0.01000,0.642
98
+ 13,mlp.gate_proj,0.24019,0.01000,0.611
99
+ 13,mlp.down_proj,0.00736,0.01000,2.476
100
+ 14,self_attn.k_proj,0.08390,0.01000,0.579
101
+ 14,self_attn.v_proj,0.03109,0.01000,0.576
102
+ 14,self_attn.q_proj,0.16015,0.01000,0.601
103
+ 14,self_attn.o_proj,0.00296,0.01000,0.600
104
+ 14,mlp.up_proj,0.20902,0.01000,0.663
105
+ 14,mlp.gate_proj,0.29858,0.01000,0.608
106
+ 14,mlp.down_proj,0.00978,0.01000,2.525
107
+ 15,self_attn.k_proj,0.07619,0.01000,0.582
108
+ 15,self_attn.v_proj,0.03300,0.01000,0.577
109
+ 15,self_attn.q_proj,0.14672,0.01000,0.598
110
+ 15,self_attn.o_proj,0.00982,0.01000,0.612
111
+ 15,mlp.up_proj,0.25285,0.01000,0.633
112
+ 15,mlp.gate_proj,0.33463,0.01000,0.617
113
+ 15,mlp.down_proj,0.02326,0.01000,2.467
quantize_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 4,
3
+ "dynamic": null,
4
+ "group_size": 32,
5
+ "desc_act": true,
6
+ "sym": true,
7
+ "lm_head": false,
8
+ "quant_method": "gptq",
9
+ "checkpoint_format": "gptq",
10
+ "meta": {
11
+ "quantizer": [
12
+ "gptqmodel:1.4.6-dev"
13
+ ],
14
+ "uri": "https://github.com/modelcloud/gptqmodel",
15
+ "damp_percent": 0.01,
16
+ "damp_auto_increment": 0.0025,
17
+ "static_groups": false,
18
+ "true_sequential": true,
19
+ "mse": 0.0
20
+ }
21
+ }