PerryCheng614 commited on
Commit
7819792
·
verified ·
1 Parent(s): 3528b2e

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": "/home/azureuser/.cache/huggingface/hub/models--meta-llama--Llama-3.2-1B-Instruct/snapshots/9213176726f574b556790deb65791e0c5aa438b6",
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:1.5.1-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 sha256:85c070d6fc7afa65c11681d789532f1ea79328ad47b0efd3c880af854999cdab
3
+ size 1614733472
quant_log.csv ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ layer,module,loss,damp,time
2
+ 0,self_attn.k_proj,0.03242,0.01000,0.906
3
+ 0,self_attn.v_proj,0.00080,0.01000,0.680
4
+ 0,self_attn.q_proj,0.06659,0.01000,0.697
5
+ 0,self_attn.o_proj,0.00007,0.01000,0.746
6
+ 0,mlp.up_proj,0.05609,0.01000,0.751
7
+ 0,mlp.gate_proj,0.07055,0.01000,0.817
8
+ 0,mlp.down_proj,0.00038,0.01000,3.103
9
+ 1,self_attn.k_proj,0.05143,0.01000,0.746
10
+ 1,self_attn.v_proj,0.00292,0.01000,0.728
11
+ 1,self_attn.q_proj,0.09487,0.01000,0.708
12
+ 1,self_attn.o_proj,0.00013,0.01000,0.733
13
+ 1,mlp.up_proj,0.07948,0.01000,0.788
14
+ 1,mlp.gate_proj,0.10903,0.01000,0.859
15
+ 1,mlp.down_proj,0.28001,0.01000,2.977
16
+ 2,self_attn.k_proj,0.09702,0.01000,0.750
17
+ 2,self_attn.v_proj,0.00672,0.01000,0.677
18
+ 2,self_attn.q_proj,0.18856,0.01000,0.696
19
+ 2,self_attn.o_proj,0.00022,0.01000,0.725
20
+ 2,mlp.up_proj,0.09984,0.01000,0.781
21
+ 2,mlp.gate_proj,0.15930,0.01000,0.803
22
+ 2,mlp.down_proj,0.00119,0.01000,2.998
23
+ 3,self_attn.k_proj,0.06446,0.01000,0.734
24
+ 3,self_attn.v_proj,0.00860,0.01000,0.685
25
+ 3,self_attn.q_proj,0.14038,0.01000,0.688
26
+ 3,self_attn.o_proj,0.00043,0.01000,0.744
27
+ 3,mlp.up_proj,0.12107,0.01000,0.794
28
+ 3,mlp.gate_proj,0.24166,0.01000,0.799
29
+ 3,mlp.down_proj,0.00183,0.01000,3.019
30
+ 4,self_attn.k_proj,0.06840,0.01000,0.755
31
+ 4,self_attn.v_proj,0.00803,0.01000,0.702
32
+ 4,self_attn.q_proj,0.14097,0.01000,0.707
33
+ 4,self_attn.o_proj,0.00079,0.01000,0.731
34
+ 4,mlp.up_proj,0.12585,0.01000,0.765
35
+ 4,mlp.gate_proj,0.27268,0.01000,0.796
36
+ 4,mlp.down_proj,0.00232,0.01000,2.969
37
+ 5,self_attn.k_proj,0.10500,0.01000,0.766
38
+ 5,self_attn.v_proj,0.00724,0.01000,0.684
39
+ 5,self_attn.q_proj,0.18579,0.01000,0.700
40
+ 5,self_attn.o_proj,0.00092,0.01000,0.754
41
+ 5,mlp.up_proj,0.13968,0.01000,0.765
42
+ 5,mlp.gate_proj,0.25845,0.01000,0.796
43
+ 5,mlp.down_proj,0.00277,0.01000,3.046
44
+ 6,self_attn.k_proj,0.08859,0.01000,0.727
45
+ 6,self_attn.v_proj,0.00953,0.01000,0.689
46
+ 6,self_attn.q_proj,0.14065,0.01000,0.699
47
+ 6,self_attn.o_proj,0.00129,0.01000,0.735
48
+ 6,mlp.up_proj,0.14354,0.01000,0.791
49
+ 6,mlp.gate_proj,0.25938,0.01000,0.785
50
+ 6,mlp.down_proj,0.00291,0.01000,3.035
51
+ 7,self_attn.k_proj,0.08833,0.01000,0.741
52
+ 7,self_attn.v_proj,0.01088,0.01000,0.672
53
+ 7,self_attn.q_proj,0.16550,0.01000,0.681
54
+ 7,self_attn.o_proj,0.00129,0.01000,0.746
55
+ 7,mlp.up_proj,0.15128,0.01000,0.779
56
+ 7,mlp.gate_proj,0.24538,0.01000,0.803
57
+ 7,mlp.down_proj,0.00320,0.01000,3.057
58
+ 8,self_attn.k_proj,0.10553,0.01000,0.749
59
+ 8,self_attn.v_proj,0.01056,0.01000,0.678
60
+ 8,self_attn.q_proj,0.17305,0.01000,0.717
61
+ 8,self_attn.o_proj,0.00191,0.01000,0.732
62
+ 8,mlp.up_proj,0.17159,0.01000,0.759
63
+ 8,mlp.gate_proj,0.27065,0.01000,0.804
64
+ 8,mlp.down_proj,0.00440,0.01000,3.075
65
+ 9,self_attn.k_proj,0.08821,0.01000,0.751
66
+ 9,self_attn.v_proj,0.01267,0.01000,0.671
67
+ 9,self_attn.q_proj,0.21381,0.01000,0.702
68
+ 9,self_attn.o_proj,0.00240,0.01000,0.738
69
+ 9,mlp.up_proj,0.18510,0.01000,0.789
70
+ 9,mlp.gate_proj,0.30127,0.01000,0.789
71
+ 9,mlp.down_proj,0.00538,0.01000,3.119
72
+ 10,self_attn.k_proj,0.11026,0.01000,0.753
73
+ 10,self_attn.v_proj,0.01362,0.01000,0.693
74
+ 10,self_attn.q_proj,0.22377,0.01000,0.718
75
+ 10,self_attn.o_proj,0.00176,0.01000,0.748
76
+ 10,mlp.up_proj,0.22299,0.01000,0.774
77
+ 10,mlp.gate_proj,0.35132,0.01000,0.820
78
+ 10,mlp.down_proj,0.00690,0.01000,3.018
79
+ 11,self_attn.k_proj,0.13077,0.01000,0.753
80
+ 11,self_attn.v_proj,0.01364,0.01000,0.718
81
+ 11,self_attn.q_proj,0.21940,0.01000,0.718
82
+ 11,self_attn.o_proj,0.00149,0.01000,0.758
83
+ 11,mlp.up_proj,0.25063,0.01000,0.773
84
+ 11,mlp.gate_proj,0.38616,0.01000,0.799
85
+ 11,mlp.down_proj,0.00782,0.01000,2.983
86
+ 12,self_attn.k_proj,0.12650,0.01000,0.742
87
+ 12,self_attn.v_proj,0.01431,0.01000,0.666
88
+ 12,self_attn.q_proj,0.21842,0.01000,0.733
89
+ 12,self_attn.o_proj,0.00136,0.01000,0.730
90
+ 12,mlp.up_proj,0.27024,0.01000,0.783
91
+ 12,mlp.gate_proj,0.39223,0.01000,0.795
92
+ 12,mlp.down_proj,0.00925,0.01000,2.858
93
+ 13,self_attn.k_proj,0.12914,0.01000,0.687
94
+ 13,self_attn.v_proj,0.02417,0.01000,0.633
95
+ 13,self_attn.q_proj,0.25690,0.01000,0.634
96
+ 13,self_attn.o_proj,0.00204,0.01000,0.717
97
+ 13,mlp.up_proj,0.31966,0.01000,0.770
98
+ 13,mlp.gate_proj,0.42314,0.01000,0.836
99
+ 13,mlp.down_proj,0.01332,0.01000,2.869
100
+ 14,self_attn.k_proj,0.13651,0.01000,0.691
101
+ 14,self_attn.v_proj,0.05146,0.01000,0.625
102
+ 14,self_attn.q_proj,0.26610,0.01000,0.670
103
+ 14,self_attn.o_proj,0.00420,0.01000,0.708
104
+ 14,mlp.up_proj,0.36466,0.01000,0.754
105
+ 14,mlp.gate_proj,0.52044,0.01000,0.870
106
+ 14,mlp.down_proj,0.01716,0.01000,2.821
107
+ 15,self_attn.k_proj,0.12916,0.01000,0.698
108
+ 15,self_attn.v_proj,0.05582,0.01000,0.641
109
+ 15,self_attn.q_proj,0.24866,0.01000,0.646
110
+ 15,self_attn.o_proj,0.01490,0.01000,0.705
111
+ 15,mlp.up_proj,0.43859,0.01000,0.785
112
+ 15,mlp.gate_proj,0.58076,0.01000,0.905
113
+ 15,mlp.down_proj,0.03817,0.01000,2.872
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:1.5.1-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
+ }