ZAHNGYUXUAN commited on
Commit
ee8faeb
·
verified ·
1 Parent(s): 0974377

Add files using upload-large-folder tool

Browse files
Files changed (1) hide show
  1. config.json +76 -0
config.json ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Glm4vMoeForConditionalGeneration"
4
+ ],
5
+ "model_type": "glm4v_moe",
6
+ "text_config": {
7
+ "pad_token_id": 151329,
8
+ "vocab_size": 151552,
9
+ "eos_token_id": [
10
+ 151329,
11
+ 151336,
12
+ 151338
13
+ ],
14
+ "image_end_token_id": 151340,
15
+ "image_start_token_id": 151339,
16
+ "image_token_id": 151363,
17
+ "head_dim": 128,
18
+ "attention_bias": true,
19
+ "attention_dropout": 0.0,
20
+ "first_k_dense_replace": 1,
21
+ "hidden_act": "silu",
22
+ "hidden_size": 4096,
23
+ "initializer_range": 0.02,
24
+ "intermediate_size": 10944,
25
+ "max_position_embeddings": 65536,
26
+ "model_type": "glm4v_moe_text",
27
+ "moe_intermediate_size": 1408,
28
+ "n_group": 1,
29
+ "n_routed_experts": 128,
30
+ "n_shared_experts": 1,
31
+ "norm_topk_prob": true,
32
+ "num_attention_heads": 96,
33
+ "num_experts_per_tok": 8,
34
+ "num_hidden_layers": 46,
35
+ "num_key_value_heads": 8,
36
+ "partial_rotary_factor": 0.5,
37
+ "rms_norm_eps": 1e-05,
38
+ "torch_dtype": "bfloat16",
39
+ "rope_scaling": {
40
+ "rope_type": "default",
41
+ "mrope_section": [
42
+ 8,
43
+ 12,
44
+ 12
45
+ ]
46
+ },
47
+ "rope_theta": 10000.0,
48
+ "routed_scaling_factor": 1.0,
49
+ "topk_group": 1,
50
+ "use_cache": true,
51
+ "use_qk_norm": false
52
+ },
53
+ "torch_dtype": "bfloat16",
54
+ "transformers_version": "4.55.0.dev0",
55
+ "video_end_token_id": 151342,
56
+ "video_start_token_id": 151341,
57
+ "video_token_id": 151364,
58
+ "vision_config": {
59
+ "attention_bias": false,
60
+ "attention_dropout": 0.0,
61
+ "depth": 24,
62
+ "hidden_act": "silu",
63
+ "hidden_size": 1536,
64
+ "image_size": 336,
65
+ "in_channels": 3,
66
+ "initializer_range": 0.02,
67
+ "intermediate_size": 10944,
68
+ "model_type": "glm4v_moe",
69
+ "num_heads": 12,
70
+ "out_hidden_size": 4096,
71
+ "patch_size": 14,
72
+ "rms_norm_eps": 1e-05,
73
+ "spatial_merge_size": 2,
74
+ "temporal_patch_size": 2
75
+ }
76
+ }