Upload folder using huggingface_hub
Browse files- config.json +62 -0
- model.safetensors +3 -0
- quant_log.csv +225 -0
- quantize_config.json +21 -0
config.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_attn_implementation_autoset": true,
|
| 3 |
+
"_name_or_path": "CohereForAI/c4ai-command-r7b-12-2024",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"Cohere2ForCausalLM"
|
| 6 |
+
],
|
| 7 |
+
"attention_bias": false,
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"bos_token_id": 5,
|
| 10 |
+
"cache_implementation": "hybrid",
|
| 11 |
+
"eos_token_id": 255001,
|
| 12 |
+
"head_dim": 128,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 4096,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 14336,
|
| 17 |
+
"layer_norm_eps": 1e-05,
|
| 18 |
+
"layer_switch": 4,
|
| 19 |
+
"logit_scale": 0.25,
|
| 20 |
+
"max_position_embeddings": 8192,
|
| 21 |
+
"model_type": "cohere2",
|
| 22 |
+
"num_attention_heads": 32,
|
| 23 |
+
"num_hidden_layers": 32,
|
| 24 |
+
"num_key_value_heads": 8,
|
| 25 |
+
"order_of_interleaved_layers": "local_attn_first",
|
| 26 |
+
"pad_token_id": 0,
|
| 27 |
+
"position_embedding_type": "rope_gptj",
|
| 28 |
+
"quantization_config": {
|
| 29 |
+
"bits": 4,
|
| 30 |
+
"checkpoint_format": "gptq",
|
| 31 |
+
"desc_act": true,
|
| 32 |
+
"dynamic": null,
|
| 33 |
+
"group_size": 128,
|
| 34 |
+
"lm_head": false,
|
| 35 |
+
"meta": {
|
| 36 |
+
"damp_auto_increment": 0.0025,
|
| 37 |
+
"damp_percent": 0.01,
|
| 38 |
+
"mse": 0.0,
|
| 39 |
+
"quantizer": [
|
| 40 |
+
"gptqmodel:1.4.5"
|
| 41 |
+
],
|
| 42 |
+
"static_groups": false,
|
| 43 |
+
"true_sequential": true,
|
| 44 |
+
"uri": "https://github.com/modelcloud/gptqmodel"
|
| 45 |
+
},
|
| 46 |
+
"quant_method": "gptq",
|
| 47 |
+
"sym": true
|
| 48 |
+
},
|
| 49 |
+
"rope_scaling": null,
|
| 50 |
+
"rope_theta": 50000,
|
| 51 |
+
"rotary_pct": 1.0,
|
| 52 |
+
"sliding_window": 4096,
|
| 53 |
+
"sliding_window_pattern": 4,
|
| 54 |
+
"torch_dtype": "bfloat16",
|
| 55 |
+
"transformers_version": "4.48.0.dev0",
|
| 56 |
+
"use_cache": true,
|
| 57 |
+
"use_embedding_sharing": true,
|
| 58 |
+
"use_gated_activation": true,
|
| 59 |
+
"use_parallel_block": true,
|
| 60 |
+
"use_parallel_embedding": true,
|
| 61 |
+
"vocab_size": 256000
|
| 62 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d2b241b2a5964f95a04c80627001d19699f94432b3881654a865a4375212633b
|
| 3 |
+
size 7825635360
|
quant_log.csv
ADDED
|
@@ -0,0 +1,225 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
layer,module,loss,damp,time
|
| 2 |
+
0,self_attn.k_proj,0.05621,0.01000,1.196
|
| 3 |
+
0,self_attn.v_proj,0.00545,0.01000,0.893
|
| 4 |
+
0,self_attn.q_proj,0.09766,0.01000,0.550
|
| 5 |
+
0,self_attn.o_proj,0.00720,0.01000,1.127
|
| 6 |
+
0,mlp.up_proj,0.13284,0.01000,1.074
|
| 7 |
+
0,mlp.gate_proj,0.14490,0.01000,0.966
|
| 8 |
+
0,mlp.down_proj,0.00287,0.01000,2.250
|
| 9 |
+
1,self_attn.k_proj,0.13365,0.01000,0.510
|
| 10 |
+
1,self_attn.v_proj,0.01556,0.01000,0.506
|
| 11 |
+
1,self_attn.q_proj,0.25788,0.01000,0.553
|
| 12 |
+
1,self_attn.o_proj,0.00839,0.01000,0.997
|
| 13 |
+
1,mlp.up_proj,0.36149,0.01000,1.224
|
| 14 |
+
1,mlp.gate_proj,0.45005,0.01000,0.723
|
| 15 |
+
1,mlp.down_proj,0.20013,0.01000,2.297
|
| 16 |
+
2,self_attn.k_proj,0.24724,0.01000,0.506
|
| 17 |
+
2,self_attn.v_proj,0.03933,0.01000,0.509
|
| 18 |
+
2,self_attn.q_proj,0.47321,0.01000,0.568
|
| 19 |
+
2,self_attn.o_proj,0.01924,0.01000,0.560
|
| 20 |
+
2,mlp.up_proj,0.75024,0.01000,0.737
|
| 21 |
+
2,mlp.gate_proj,0.98029,0.01000,0.739
|
| 22 |
+
2,mlp.down_proj,0.01554,0.01000,2.684
|
| 23 |
+
3,self_attn.k_proj,0.60100,0.01000,0.859
|
| 24 |
+
3,self_attn.v_proj,0.09872,0.01000,0.512
|
| 25 |
+
3,self_attn.q_proj,0.94429,0.01000,0.570
|
| 26 |
+
3,self_attn.o_proj,0.00530,0.01000,0.568
|
| 27 |
+
3,mlp.up_proj,1.42404,0.01000,0.742
|
| 28 |
+
3,mlp.gate_proj,2.03389,0.01000,0.738
|
| 29 |
+
3,mlp.down_proj,0.04570,0.01000,2.777
|
| 30 |
+
4,self_attn.k_proj,0.37175,0.01000,0.917
|
| 31 |
+
4,self_attn.v_proj,0.09440,0.01000,0.506
|
| 32 |
+
4,self_attn.q_proj,0.85723,0.01000,0.567
|
| 33 |
+
4,self_attn.o_proj,0.02324,0.01000,0.999
|
| 34 |
+
4,mlp.up_proj,1.75096,0.01000,1.218
|
| 35 |
+
4,mlp.gate_proj,2.32645,0.01000,0.734
|
| 36 |
+
4,mlp.down_proj,0.06497,0.01000,2.299
|
| 37 |
+
5,self_attn.k_proj,0.42039,0.01000,0.508
|
| 38 |
+
5,self_attn.v_proj,0.12333,0.01000,0.516
|
| 39 |
+
5,self_attn.q_proj,0.99933,0.01000,0.566
|
| 40 |
+
5,self_attn.o_proj,0.03446,0.01000,0.561
|
| 41 |
+
5,mlp.up_proj,2.02846,0.01000,0.742
|
| 42 |
+
5,mlp.gate_proj,2.61058,0.01000,0.738
|
| 43 |
+
5,mlp.down_proj,0.09458,0.01000,2.726
|
| 44 |
+
6,self_attn.k_proj,0.43745,0.01000,0.507
|
| 45 |
+
6,self_attn.v_proj,0.12596,0.01000,0.505
|
| 46 |
+
6,self_attn.q_proj,1.12090,0.01000,0.557
|
| 47 |
+
6,self_attn.o_proj,0.05046,0.01000,0.566
|
| 48 |
+
6,mlp.up_proj,2.24570,0.01000,0.759
|
| 49 |
+
6,mlp.gate_proj,3.58803,0.01000,0.736
|
| 50 |
+
6,mlp.down_proj,0.10299,0.01000,2.669
|
| 51 |
+
7,self_attn.k_proj,0.84814,0.01000,0.927
|
| 52 |
+
7,self_attn.v_proj,0.18806,0.01000,0.505
|
| 53 |
+
7,self_attn.q_proj,1.51966,0.01000,0.567
|
| 54 |
+
7,self_attn.o_proj,0.00890,0.01000,0.559
|
| 55 |
+
7,mlp.up_proj,2.41810,0.01000,0.730
|
| 56 |
+
7,mlp.gate_proj,3.40998,0.01000,0.728
|
| 57 |
+
7,mlp.down_proj,0.12915,0.01000,2.756
|
| 58 |
+
8,self_attn.k_proj,0.47351,0.01000,0.963
|
| 59 |
+
8,self_attn.v_proj,0.13747,0.01000,0.507
|
| 60 |
+
8,self_attn.q_proj,1.14740,0.01000,0.566
|
| 61 |
+
8,self_attn.o_proj,0.05132,0.01000,1.080
|
| 62 |
+
8,mlp.up_proj,2.51129,0.01000,1.058
|
| 63 |
+
8,mlp.gate_proj,3.12458,0.01000,0.728
|
| 64 |
+
8,mlp.down_proj,0.13478,0.01000,2.295
|
| 65 |
+
9,self_attn.k_proj,0.59965,0.01000,0.504
|
| 66 |
+
9,self_attn.v_proj,0.17365,0.01000,0.504
|
| 67 |
+
9,self_attn.q_proj,1.38692,0.01000,0.562
|
| 68 |
+
9,self_attn.o_proj,0.06629,0.01000,0.562
|
| 69 |
+
9,mlp.up_proj,2.76733,0.01000,0.732
|
| 70 |
+
9,mlp.gate_proj,3.32531,0.01000,0.735
|
| 71 |
+
9,mlp.down_proj,0.16287,0.01000,2.715
|
| 72 |
+
10,self_attn.k_proj,0.52339,0.01000,0.501
|
| 73 |
+
10,self_attn.v_proj,0.19670,0.01000,0.500
|
| 74 |
+
10,self_attn.q_proj,1.25550,0.01000,0.557
|
| 75 |
+
10,self_attn.o_proj,0.07569,0.01000,0.551
|
| 76 |
+
10,mlp.up_proj,2.73953,0.01000,0.737
|
| 77 |
+
10,mlp.gate_proj,3.12449,0.01000,0.727
|
| 78 |
+
10,mlp.down_proj,0.17240,0.01000,2.656
|
| 79 |
+
11,self_attn.k_proj,1.07680,0.01000,0.681
|
| 80 |
+
11,self_attn.v_proj,0.24638,0.01000,0.504
|
| 81 |
+
11,self_attn.q_proj,2.13249,0.01000,0.555
|
| 82 |
+
11,self_attn.o_proj,0.03390,0.01000,0.559
|
| 83 |
+
11,mlp.up_proj,3.17669,0.01000,1.025
|
| 84 |
+
11,mlp.gate_proj,4.94614,0.01000,0.733
|
| 85 |
+
11,mlp.down_proj,0.16393,0.01000,2.320
|
| 86 |
+
12,self_attn.k_proj,0.55052,0.01000,0.503
|
| 87 |
+
12,self_attn.v_proj,0.14572,0.01000,0.501
|
| 88 |
+
12,self_attn.q_proj,1.24151,0.01000,0.555
|
| 89 |
+
12,self_attn.o_proj,0.06163,0.01000,0.554
|
| 90 |
+
12,mlp.up_proj,2.77750,0.01000,0.733
|
| 91 |
+
12,mlp.gate_proj,3.39289,0.01000,0.730
|
| 92 |
+
12,mlp.down_proj,0.16477,0.01000,2.622
|
| 93 |
+
13,self_attn.k_proj,0.62645,0.01000,0.891
|
| 94 |
+
13,self_attn.v_proj,0.19116,0.01000,0.503
|
| 95 |
+
13,self_attn.q_proj,1.35932,0.01000,0.564
|
| 96 |
+
13,self_attn.o_proj,0.07144,0.01000,1.141
|
| 97 |
+
13,mlp.up_proj,3.03934,0.01000,1.102
|
| 98 |
+
13,mlp.gate_proj,3.55067,0.01000,0.743
|
| 99 |
+
13,mlp.down_proj,0.18427,0.01000,2.299
|
| 100 |
+
14,self_attn.k_proj,0.58719,0.01000,0.502
|
| 101 |
+
14,self_attn.v_proj,0.21528,0.01000,0.504
|
| 102 |
+
14,self_attn.q_proj,1.32953,0.01000,0.558
|
| 103 |
+
14,self_attn.o_proj,0.10515,0.01000,0.553
|
| 104 |
+
14,mlp.up_proj,2.95010,0.01000,0.732
|
| 105 |
+
14,mlp.gate_proj,3.19144,0.01000,0.735
|
| 106 |
+
14,mlp.down_proj,0.20802,0.01000,2.640
|
| 107 |
+
15,self_attn.k_proj,1.05477,0.01000,0.504
|
| 108 |
+
15,self_attn.v_proj,0.28173,0.01000,0.507
|
| 109 |
+
15,self_attn.q_proj,2.18319,0.01000,0.563
|
| 110 |
+
15,self_attn.o_proj,0.05205,0.01000,0.560
|
| 111 |
+
15,mlp.up_proj,3.56478,0.01000,0.740
|
| 112 |
+
15,mlp.gate_proj,4.29289,0.01000,0.730
|
| 113 |
+
15,mlp.down_proj,0.22124,0.01000,2.734
|
| 114 |
+
16,self_attn.k_proj,0.63467,0.01000,0.792
|
| 115 |
+
16,self_attn.v_proj,0.25075,0.01000,0.500
|
| 116 |
+
16,self_attn.q_proj,1.50353,0.01000,0.561
|
| 117 |
+
16,self_attn.o_proj,0.10116,0.01000,1.137
|
| 118 |
+
16,mlp.up_proj,3.35414,0.01000,1.107
|
| 119 |
+
16,mlp.gate_proj,3.98696,0.01000,0.724
|
| 120 |
+
16,mlp.down_proj,0.23792,0.01000,2.295
|
| 121 |
+
17,self_attn.k_proj,0.64182,0.01000,0.733
|
| 122 |
+
17,self_attn.v_proj,0.26062,0.01000,0.502
|
| 123 |
+
17,self_attn.q_proj,1.53473,0.01000,0.566
|
| 124 |
+
17,self_attn.o_proj,0.14796,0.01000,1.152
|
| 125 |
+
17,mlp.up_proj,3.49386,0.01000,1.064
|
| 126 |
+
17,mlp.gate_proj,4.21796,0.01000,0.734
|
| 127 |
+
17,mlp.down_proj,0.24264,0.01000,2.339
|
| 128 |
+
18,self_attn.k_proj,0.68419,0.01000,0.820
|
| 129 |
+
18,self_attn.v_proj,0.28049,0.01000,0.501
|
| 130 |
+
18,self_attn.q_proj,1.66727,0.01000,0.552
|
| 131 |
+
18,self_attn.o_proj,0.11068,0.01000,0.562
|
| 132 |
+
18,mlp.up_proj,3.71261,0.01000,0.730
|
| 133 |
+
18,mlp.gate_proj,4.59650,0.01000,0.735
|
| 134 |
+
18,mlp.down_proj,0.29909,0.01000,2.722
|
| 135 |
+
19,self_attn.k_proj,1.02704,0.01000,0.948
|
| 136 |
+
19,self_attn.v_proj,0.43806,0.01000,0.502
|
| 137 |
+
19,self_attn.q_proj,2.43219,0.01000,0.553
|
| 138 |
+
19,self_attn.o_proj,0.07447,0.01000,0.555
|
| 139 |
+
19,mlp.up_proj,4.35553,0.01000,0.734
|
| 140 |
+
19,mlp.gate_proj,5.86754,0.01000,0.734
|
| 141 |
+
19,mlp.down_proj,0.36460,0.01000,2.814
|
| 142 |
+
20,self_attn.k_proj,0.66072,0.01000,0.506
|
| 143 |
+
20,self_attn.v_proj,0.30183,0.01000,0.503
|
| 144 |
+
20,self_attn.q_proj,1.68154,0.01000,0.551
|
| 145 |
+
20,self_attn.o_proj,0.20820,0.01000,0.821
|
| 146 |
+
20,mlp.up_proj,4.02871,0.01000,1.167
|
| 147 |
+
20,mlp.gate_proj,5.31456,0.01000,0.738
|
| 148 |
+
20,mlp.down_proj,0.39856,0.01000,2.304
|
| 149 |
+
21,self_attn.k_proj,0.71785,0.01000,0.502
|
| 150 |
+
21,self_attn.v_proj,0.35489,0.01000,0.500
|
| 151 |
+
21,self_attn.q_proj,1.89252,0.01000,0.554
|
| 152 |
+
21,self_attn.o_proj,0.16916,0.01000,0.555
|
| 153 |
+
21,mlp.up_proj,4.70385,0.01000,0.737
|
| 154 |
+
21,mlp.gate_proj,6.40686,0.01000,0.739
|
| 155 |
+
21,mlp.down_proj,0.48923,0.01000,2.751
|
| 156 |
+
22,self_attn.k_proj,0.65530,0.01000,0.863
|
| 157 |
+
22,self_attn.v_proj,0.37394,0.01000,0.502
|
| 158 |
+
22,self_attn.q_proj,1.91038,0.01000,0.554
|
| 159 |
+
22,self_attn.o_proj,0.19113,0.01000,1.079
|
| 160 |
+
22,mlp.up_proj,5.04891,0.01000,1.066
|
| 161 |
+
22,mlp.gate_proj,6.76659,0.01000,0.732
|
| 162 |
+
22,mlp.down_proj,0.59243,0.01000,2.354
|
| 163 |
+
23,self_attn.k_proj,1.00629,0.01000,0.512
|
| 164 |
+
23,self_attn.v_proj,0.65785,0.01000,0.505
|
| 165 |
+
23,self_attn.q_proj,2.80004,0.01000,0.565
|
| 166 |
+
23,self_attn.o_proj,0.11185,0.01000,0.561
|
| 167 |
+
23,mlp.up_proj,5.98257,0.01000,0.948
|
| 168 |
+
23,mlp.gate_proj,8.26061,0.01000,0.740
|
| 169 |
+
23,mlp.down_proj,0.71816,0.01000,2.330
|
| 170 |
+
24,self_attn.k_proj,0.75695,0.01000,0.956
|
| 171 |
+
24,self_attn.v_proj,0.56502,0.01000,0.506
|
| 172 |
+
24,self_attn.q_proj,2.20552,0.01000,0.565
|
| 173 |
+
24,self_attn.o_proj,0.20811,0.01000,1.127
|
| 174 |
+
24,mlp.up_proj,6.07553,0.01000,1.115
|
| 175 |
+
24,mlp.gate_proj,8.25385,0.01000,0.731
|
| 176 |
+
24,mlp.down_proj,0.81732,0.01000,2.311
|
| 177 |
+
25,self_attn.k_proj,0.72482,0.01000,0.511
|
| 178 |
+
25,self_attn.v_proj,0.63889,0.01000,0.505
|
| 179 |
+
25,self_attn.q_proj,2.23772,0.01000,0.563
|
| 180 |
+
25,self_attn.o_proj,0.37058,0.01000,0.557
|
| 181 |
+
25,mlp.up_proj,6.89948,0.01000,1.003
|
| 182 |
+
25,mlp.gate_proj,9.06814,0.01000,0.738
|
| 183 |
+
25,mlp.down_proj,0.92696,0.01000,2.311
|
| 184 |
+
26,self_attn.k_proj,0.72621,0.01000,0.501
|
| 185 |
+
26,self_attn.v_proj,0.67210,0.01000,0.512
|
| 186 |
+
26,self_attn.q_proj,2.30942,0.01000,0.561
|
| 187 |
+
26,self_attn.o_proj,0.34593,0.01000,0.564
|
| 188 |
+
26,mlp.up_proj,7.81135,0.01000,0.766
|
| 189 |
+
26,mlp.gate_proj,9.74435,0.01000,0.733
|
| 190 |
+
26,mlp.down_proj,1.24495,0.01000,2.766
|
| 191 |
+
27,self_attn.k_proj,0.82907,0.01000,0.940
|
| 192 |
+
27,self_attn.v_proj,1.38522,0.01000,0.507
|
| 193 |
+
27,self_attn.q_proj,2.77574,0.01000,0.563
|
| 194 |
+
27,self_attn.o_proj,0.16704,0.01000,0.553
|
| 195 |
+
27,mlp.up_proj,9.27288,0.01000,0.731
|
| 196 |
+
27,mlp.gate_proj,11.15920,0.01000,0.734
|
| 197 |
+
27,mlp.down_proj,1.72660,0.01000,2.300
|
| 198 |
+
28,self_attn.k_proj,0.65320,0.01000,0.504
|
| 199 |
+
28,self_attn.v_proj,1.41018,0.01000,0.502
|
| 200 |
+
28,self_attn.q_proj,2.39974,0.01000,0.560
|
| 201 |
+
28,self_attn.o_proj,0.32783,0.01000,0.564
|
| 202 |
+
28,mlp.up_proj,9.96800,0.01000,0.732
|
| 203 |
+
28,mlp.gate_proj,11.59946,0.01000,0.735
|
| 204 |
+
28,mlp.down_proj,2.22566,0.01000,2.705
|
| 205 |
+
29,self_attn.k_proj,0.77413,0.01000,0.762
|
| 206 |
+
29,self_attn.v_proj,1.39446,0.01000,0.505
|
| 207 |
+
29,self_attn.q_proj,2.66277,0.01000,0.555
|
| 208 |
+
29,self_attn.o_proj,0.94737,0.01000,0.994
|
| 209 |
+
29,mlp.up_proj,10.98967,0.01000,1.175
|
| 210 |
+
29,mlp.gate_proj,12.69676,0.01000,0.730
|
| 211 |
+
29,mlp.down_proj,2.79368,0.01000,2.307
|
| 212 |
+
30,self_attn.k_proj,0.71118,0.01000,0.504
|
| 213 |
+
30,self_attn.v_proj,1.45648,0.01000,0.503
|
| 214 |
+
30,self_attn.q_proj,2.53372,0.01000,0.551
|
| 215 |
+
30,self_attn.o_proj,1.33298,0.01000,0.559
|
| 216 |
+
30,mlp.up_proj,11.20086,0.01000,0.744
|
| 217 |
+
30,mlp.gate_proj,13.27178,0.01000,0.730
|
| 218 |
+
30,mlp.down_proj,4.00105,0.01000,2.530
|
| 219 |
+
31,self_attn.k_proj,1.14293,0.01000,0.881
|
| 220 |
+
31,self_attn.v_proj,1.13456,0.01000,0.502
|
| 221 |
+
31,self_attn.q_proj,2.49532,0.01000,0.562
|
| 222 |
+
31,self_attn.o_proj,0.23442,0.01000,0.565
|
| 223 |
+
31,mlp.up_proj,9.34217,0.01000,0.733
|
| 224 |
+
31,mlp.gate_proj,11.87304,0.01000,0.736
|
| 225 |
+
31,mlp.down_proj,6.81150,0.01000,2.751
|
quantize_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bits": 4,
|
| 3 |
+
"dynamic": null,
|
| 4 |
+
"group_size": 128,
|
| 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:1.4.5"
|
| 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 |
+
}
|