PerryCheng614 commited on
Commit
946ecea
1 Parent(s): b902540

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 sha256:751aa2c38ca552ae90ac9c5968b7ce3ace0e08784a1d1056863b8424ca014299
3
+ size 1614733472
quant_log.csv ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ layer,module,loss,damp,time
2
+ 0,self_attn.k_proj,0.20627,0.01000,0.986
3
+ 0,self_attn.v_proj,0.00507,0.01000,0.607
4
+ 0,self_attn.q_proj,0.42227,0.01000,0.610
5
+ 0,self_attn.o_proj,0.00047,0.01000,0.603
6
+ 0,mlp.up_proj,0.35409,0.01000,0.636
7
+ 0,mlp.gate_proj,0.44565,0.01000,0.749
8
+ 0,mlp.down_proj,0.00251,0.01000,2.774
9
+ 1,self_attn.k_proj,0.33142,0.01000,0.600
10
+ 1,self_attn.v_proj,0.01882,0.01000,0.615
11
+ 1,self_attn.q_proj,0.61006,0.01000,0.611
12
+ 1,self_attn.o_proj,0.00132,0.01000,0.608
13
+ 1,mlp.up_proj,0.50808,0.01000,0.643
14
+ 1,mlp.gate_proj,0.69618,0.01000,0.790
15
+ 1,mlp.down_proj,1.31996,0.01000,2.535
16
+ 2,self_attn.k_proj,0.65871,0.01000,0.607
17
+ 2,self_attn.v_proj,0.04459,0.01000,0.610
18
+ 2,self_attn.q_proj,1.27351,0.01000,0.619
19
+ 2,self_attn.o_proj,0.00206,0.01000,0.596
20
+ 2,mlp.up_proj,0.65205,0.01000,0.631
21
+ 2,mlp.gate_proj,1.04195,0.01000,0.784
22
+ 2,mlp.down_proj,0.00808,0.01000,2.504
23
+ 3,self_attn.k_proj,0.43784,0.01000,0.592
24
+ 3,self_attn.v_proj,0.05698,0.01000,0.586
25
+ 3,self_attn.q_proj,0.94526,0.01000,0.610
26
+ 3,self_attn.o_proj,0.00391,0.01000,0.644
27
+ 3,mlp.up_proj,0.79450,0.01000,0.621
28
+ 3,mlp.gate_proj,1.58985,0.01000,0.798
29
+ 3,mlp.down_proj,0.01225,0.01000,2.532
30
+ 4,self_attn.k_proj,0.46703,0.01000,0.583
31
+ 4,self_attn.v_proj,0.05344,0.01000,0.588
32
+ 4,self_attn.q_proj,0.95165,0.01000,0.609
33
+ 4,self_attn.o_proj,0.00573,0.01000,0.604
34
+ 4,mlp.up_proj,0.80078,0.01000,0.630
35
+ 4,mlp.gate_proj,1.73697,0.01000,0.794
36
+ 4,mlp.down_proj,0.01413,0.01000,2.501
37
+ 5,self_attn.k_proj,0.69090,0.01000,0.596
38
+ 5,self_attn.v_proj,0.04588,0.01000,0.589
39
+ 5,self_attn.q_proj,1.21408,0.01000,0.664
40
+ 5,self_attn.o_proj,0.00586,0.01000,0.612
41
+ 5,mlp.up_proj,0.86928,0.01000,0.637
42
+ 5,mlp.gate_proj,1.61338,0.01000,0.800
43
+ 5,mlp.down_proj,0.01706,0.01000,2.504
44
+ 6,self_attn.k_proj,0.56137,0.01000,0.592
45
+ 6,self_attn.v_proj,0.05914,0.01000,0.583
46
+ 6,self_attn.q_proj,0.88346,0.01000,0.589
47
+ 6,self_attn.o_proj,0.00905,0.01000,0.614
48
+ 6,mlp.up_proj,0.88001,0.01000,0.630
49
+ 6,mlp.gate_proj,1.59125,0.01000,0.793
50
+ 6,mlp.down_proj,0.01736,0.01000,2.431
51
+ 7,self_attn.k_proj,0.56204,0.01000,0.584
52
+ 7,self_attn.v_proj,0.06787,0.01000,0.578
53
+ 7,self_attn.q_proj,1.04871,0.01000,0.627
54
+ 7,self_attn.o_proj,0.00866,0.01000,0.598
55
+ 7,mlp.up_proj,0.91172,0.01000,0.630
56
+ 7,mlp.gate_proj,1.47971,0.01000,0.801
57
+ 7,mlp.down_proj,0.01878,0.01000,2.453
58
+ 8,self_attn.k_proj,0.65364,0.01000,0.587
59
+ 8,self_attn.v_proj,0.06459,0.01000,0.599
60
+ 8,self_attn.q_proj,1.06362,0.01000,0.589
61
+ 8,self_attn.o_proj,0.01113,0.01000,0.585
62
+ 8,mlp.up_proj,1.00755,0.01000,0.611
63
+ 8,mlp.gate_proj,1.59544,0.01000,0.796
64
+ 8,mlp.down_proj,0.02531,0.01000,2.416
65
+ 9,self_attn.k_proj,0.54082,0.01000,0.584
66
+ 9,self_attn.v_proj,0.07637,0.01000,0.580
67
+ 9,self_attn.q_proj,1.30084,0.01000,0.591
68
+ 9,self_attn.o_proj,0.01507,0.01000,0.603
69
+ 9,mlp.up_proj,1.07703,0.01000,0.622
70
+ 9,mlp.gate_proj,1.76000,0.01000,0.796
71
+ 9,mlp.down_proj,0.03086,0.01000,2.426
72
+ 10,self_attn.k_proj,0.66062,0.01000,0.590
73
+ 10,self_attn.v_proj,0.08113,0.01000,0.583
74
+ 10,self_attn.q_proj,1.32715,0.01000,0.583
75
+ 10,self_attn.o_proj,0.01019,0.01000,0.594
76
+ 10,mlp.up_proj,1.28599,0.01000,0.622
77
+ 10,mlp.gate_proj,2.02958,0.01000,0.791
78
+ 10,mlp.down_proj,0.03905,0.01000,2.438
79
+ 11,self_attn.k_proj,0.79215,0.01000,0.583
80
+ 11,self_attn.v_proj,0.08159,0.01000,0.590
81
+ 11,self_attn.q_proj,1.31279,0.01000,0.571
82
+ 11,self_attn.o_proj,0.00862,0.01000,0.625
83
+ 11,mlp.up_proj,1.42253,0.01000,0.641
84
+ 11,mlp.gate_proj,2.19144,0.01000,0.801
85
+ 11,mlp.down_proj,0.04266,0.01000,2.499
86
+ 12,self_attn.k_proj,0.76721,0.01000,0.585
87
+ 12,self_attn.v_proj,0.08425,0.01000,0.580
88
+ 12,self_attn.q_proj,1.29475,0.01000,0.582
89
+ 12,self_attn.o_proj,0.00837,0.01000,0.586
90
+ 12,mlp.up_proj,1.47983,0.01000,0.626
91
+ 12,mlp.gate_proj,2.15212,0.01000,0.804
92
+ 12,mlp.down_proj,0.04750,0.01000,2.422
93
+ 13,self_attn.k_proj,0.73892,0.01000,0.589
94
+ 13,self_attn.v_proj,0.13723,0.01000,0.574
95
+ 13,self_attn.q_proj,1.45466,0.01000,0.589
96
+ 13,self_attn.o_proj,0.01131,0.01000,0.578
97
+ 13,mlp.up_proj,1.73117,0.01000,0.623
98
+ 13,mlp.gate_proj,2.29325,0.01000,0.791
99
+ 13,mlp.down_proj,0.06902,0.01000,2.415
100
+ 14,self_attn.k_proj,0.79778,0.01000,0.588
101
+ 14,self_attn.v_proj,0.29438,0.01000,0.582
102
+ 14,self_attn.q_proj,1.52455,0.01000,0.580
103
+ 14,self_attn.o_proj,0.02765,0.01000,0.586
104
+ 14,mlp.up_proj,1.98990,0.01000,0.606
105
+ 14,mlp.gate_proj,2.84552,0.01000,0.621
106
+ 14,mlp.down_proj,0.09091,0.01000,2.440
107
+ 15,self_attn.k_proj,0.71842,0.01000,0.607
108
+ 15,self_attn.v_proj,0.31329,0.01000,0.577
109
+ 15,self_attn.q_proj,1.39225,0.01000,0.585
110
+ 15,self_attn.o_proj,0.09505,0.01000,0.585
111
+ 15,mlp.up_proj,2.41008,0.01000,0.636
112
+ 15,mlp.gate_proj,3.18340,0.01000,0.802
113
+ 15,mlp.down_proj,0.21390,0.01000,2.454
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
+ }