PerryCheng614 commited on
Commit
de20b03
·
verified ·
1 Parent(s): cc6406f

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:cdb9356789500a7b33446a9807f3f82dacbca2177630d79ae86c4e8d9e447ee3
3
+ size 1614733472
quant_log.csv ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ layer,module,loss,damp,time
2
+ 0,self_attn.k_proj,0.03175,0.01000,1.017
3
+ 0,self_attn.v_proj,0.00079,0.01000,0.900
4
+ 0,self_attn.q_proj,0.06525,0.01000,0.697
5
+ 0,self_attn.o_proj,0.00007,0.01000,0.725
6
+ 0,mlp.up_proj,0.05552,0.01000,0.760
7
+ 0,mlp.gate_proj,0.06979,0.01000,1.191
8
+ 0,mlp.down_proj,0.00038,0.01000,2.984
9
+ 1,self_attn.k_proj,0.05085,0.01000,0.757
10
+ 1,self_attn.v_proj,0.00290,0.01000,0.688
11
+ 1,self_attn.q_proj,0.09413,0.01000,0.707
12
+ 1,self_attn.o_proj,0.00013,0.01000,0.738
13
+ 1,mlp.up_proj,0.07888,0.01000,0.762
14
+ 1,mlp.gate_proj,0.10826,0.01000,0.880
15
+ 1,mlp.down_proj,0.28590,0.01000,2.819
16
+ 2,self_attn.k_proj,0.09612,0.01000,0.699
17
+ 2,self_attn.v_proj,0.00669,0.01000,0.639
18
+ 2,self_attn.q_proj,0.18759,0.01000,0.643
19
+ 2,self_attn.o_proj,0.00022,0.01000,0.677
20
+ 2,mlp.up_proj,0.09868,0.01000,0.738
21
+ 2,mlp.gate_proj,0.15738,0.01000,0.793
22
+ 2,mlp.down_proj,0.00116,0.01000,2.830
23
+ 3,self_attn.k_proj,0.06407,0.01000,0.683
24
+ 3,self_attn.v_proj,0.00853,0.01000,0.627
25
+ 3,self_attn.q_proj,0.13942,0.01000,0.636
26
+ 3,self_attn.o_proj,0.00042,0.01000,0.680
27
+ 3,mlp.up_proj,0.12063,0.01000,0.704
28
+ 3,mlp.gate_proj,0.24089,0.01000,0.821
29
+ 3,mlp.down_proj,0.00180,0.01000,2.678
30
+ 4,self_attn.k_proj,0.06845,0.01000,0.657
31
+ 4,self_attn.v_proj,0.00802,0.01000,0.598
32
+ 4,self_attn.q_proj,0.14062,0.01000,0.603
33
+ 4,self_attn.o_proj,0.00080,0.01000,0.657
34
+ 4,mlp.up_proj,0.12564,0.01000,0.720
35
+ 4,mlp.gate_proj,0.27190,0.01000,1.176
36
+ 4,mlp.down_proj,0.00229,0.01000,2.712
37
+ 5,self_attn.k_proj,0.10515,0.01000,0.656
38
+ 5,self_attn.v_proj,0.00722,0.01000,0.618
39
+ 5,self_attn.q_proj,0.18627,0.01000,0.614
40
+ 5,self_attn.o_proj,0.00093,0.01000,0.654
41
+ 5,mlp.up_proj,0.13925,0.01000,0.695
42
+ 5,mlp.gate_proj,0.25759,0.01000,0.810
43
+ 5,mlp.down_proj,0.00276,0.01000,2.709
44
+ 6,self_attn.k_proj,0.08841,0.01000,0.672
45
+ 6,self_attn.v_proj,0.00952,0.01000,0.597
46
+ 6,self_attn.q_proj,0.14054,0.01000,0.631
47
+ 6,self_attn.o_proj,0.00132,0.01000,0.653
48
+ 6,mlp.up_proj,0.14352,0.01000,0.732
49
+ 6,mlp.gate_proj,0.25903,0.01000,0.809
50
+ 6,mlp.down_proj,0.00290,0.01000,2.758
51
+ 7,self_attn.k_proj,0.08830,0.01000,0.631
52
+ 7,self_attn.v_proj,0.01086,0.01000,0.601
53
+ 7,self_attn.q_proj,0.16545,0.01000,0.613
54
+ 7,self_attn.o_proj,0.00131,0.01000,0.648
55
+ 7,mlp.up_proj,0.15137,0.01000,0.718
56
+ 7,mlp.gate_proj,0.24540,0.01000,0.781
57
+ 7,mlp.down_proj,0.00321,0.01000,2.713
58
+ 8,self_attn.k_proj,0.10563,0.01000,0.666
59
+ 8,self_attn.v_proj,0.01057,0.01000,0.594
60
+ 8,self_attn.q_proj,0.17346,0.01000,0.611
61
+ 8,self_attn.o_proj,0.00192,0.01000,0.670
62
+ 8,mlp.up_proj,0.17157,0.01000,0.714
63
+ 8,mlp.gate_proj,0.27059,0.01000,0.806
64
+ 8,mlp.down_proj,0.00437,0.01000,2.705
65
+ 9,self_attn.k_proj,0.08775,0.01000,0.661
66
+ 9,self_attn.v_proj,0.01261,0.01000,0.618
67
+ 9,self_attn.q_proj,0.21340,0.01000,0.628
68
+ 9,self_attn.o_proj,0.00241,0.01000,0.637
69
+ 9,mlp.up_proj,0.18442,0.01000,0.690
70
+ 9,mlp.gate_proj,0.30010,0.01000,0.839
71
+ 9,mlp.down_proj,0.00531,0.01000,2.706
72
+ 10,self_attn.k_proj,0.10957,0.01000,0.676
73
+ 10,self_attn.v_proj,0.01352,0.01000,0.586
74
+ 10,self_attn.q_proj,0.22256,0.01000,0.613
75
+ 10,self_attn.o_proj,0.00179,0.01000,0.652
76
+ 10,mlp.up_proj,0.22136,0.01000,0.715
77
+ 10,mlp.gate_proj,0.34884,0.01000,0.811
78
+ 10,mlp.down_proj,0.00678,0.01000,2.739
79
+ 11,self_attn.k_proj,0.12984,0.01000,0.639
80
+ 11,self_attn.v_proj,0.01349,0.01000,0.609
81
+ 11,self_attn.q_proj,0.21828,0.01000,0.636
82
+ 11,self_attn.o_proj,0.00152,0.01000,0.631
83
+ 11,mlp.up_proj,0.24861,0.01000,0.693
84
+ 11,mlp.gate_proj,0.38283,0.01000,0.805
85
+ 11,mlp.down_proj,0.00769,0.01000,2.708
86
+ 12,self_attn.k_proj,0.12568,0.01000,0.689
87
+ 12,self_attn.v_proj,0.01422,0.01000,0.635
88
+ 12,self_attn.q_proj,0.21687,0.01000,0.637
89
+ 12,self_attn.o_proj,0.00137,0.01000,0.677
90
+ 12,mlp.up_proj,0.26780,0.01000,0.713
91
+ 12,mlp.gate_proj,0.38886,0.01000,0.841
92
+ 12,mlp.down_proj,0.00909,0.01000,2.751
93
+ 13,self_attn.k_proj,0.12808,0.01000,0.667
94
+ 13,self_attn.v_proj,0.02392,0.01000,0.630
95
+ 13,self_attn.q_proj,0.25501,0.01000,0.639
96
+ 13,self_attn.o_proj,0.00206,0.01000,0.669
97
+ 13,mlp.up_proj,0.31638,0.01000,0.709
98
+ 13,mlp.gate_proj,0.41893,0.01000,0.787
99
+ 13,mlp.down_proj,0.01304,0.01000,2.773
100
+ 14,self_attn.k_proj,0.13626,0.01000,0.663
101
+ 14,self_attn.v_proj,0.05109,0.01000,0.637
102
+ 14,self_attn.q_proj,0.26423,0.01000,0.630
103
+ 14,self_attn.o_proj,0.00437,0.01000,0.641
104
+ 14,mlp.up_proj,0.36103,0.01000,0.723
105
+ 14,mlp.gate_proj,0.51521,0.01000,0.780
106
+ 14,mlp.down_proj,0.01679,0.01000,2.741
107
+ 15,self_attn.k_proj,0.12817,0.01000,0.651
108
+ 15,self_attn.v_proj,0.05531,0.01000,0.617
109
+ 15,self_attn.q_proj,0.24657,0.01000,0.597
110
+ 15,self_attn.o_proj,0.01504,0.01000,0.672
111
+ 15,mlp.up_proj,0.43401,0.01000,0.715
112
+ 15,mlp.gate_proj,0.57454,0.01000,0.765
113
+ 15,mlp.down_proj,0.03761,0.01000,2.710
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
+ }