Hugging Face's logo

freestuff2875
/
GLM-Image-bnb-4bit

Text-to-Image
Diffusers
Safetensors
Chinese
English
GlmImagePipeline
4-bit precision
bitsandbytes
Model card Files
xet
 Community
GLM-Image-bnb-4bit
File size: 2,560 Bytes
cbdc95b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
{
  "_class_name": "GlmImageTransformer2DModel",
  "_diffusers_version": "0.37.1",
  "_name_or_path": "zai-org/GLM-Image",
  "attention_head_dim": 128,
  "condition_dim": 256,
  "in_channels": 16,
  "num_attention_heads": 32,
  "num_layers": 30,
  "out_channels": 16,
  "patch_size": 2,
  "prior_vq_quantizer_codebook_size": 16384,
  "quantization_config": {
    "_load_in_4bit": true,
    "_load_in_8bit": false,
    "bnb_4bit_compute_dtype": "bfloat16",
    "bnb_4bit_quant_storage": "uint8",
    "bnb_4bit_quant_type": "nf4",
    "bnb_4bit_use_double_quant": true,
    "llm_int8_enable_fp32_cpu_offload": false,
    "llm_int8_has_fp16_weight": false,
    "llm_int8_skip_modules": [
      "time_condition_embed",
      "image_projector",
      "proj_out",
      "prior_projector",
      "glyph_projector",
      "norm_out",
      "transformer_blocks.0.norm1.linear",
      "transformer_blocks.23.norm1.linear",
      "transformer_blocks.24.norm1.linear",
      "transformer_blocks.25.norm1.linear",
      "transformer_blocks.0.attn1.to_k",
      "transformer_blocks.0.attn1.to_v",
      "transformer_blocks.0.ff.net.0.proj",
      "transformer_blocks.1.attn1.to_k",
      "transformer_blocks.1.attn1.to_q",
      "transformer_blocks.1.attn1.to_v",
      "transformer_blocks.1.ff.net.0.proj",
      "transformer_blocks.10.attn1.to_v",
      "transformer_blocks.2.attn1.to_k",
      "transformer_blocks.2.attn1.to_q",
      "transformer_blocks.2.attn1.to_v",
      "transformer_blocks.3.attn1.to_k",
      "transformer_blocks.3.attn1.to_q",
      "transformer_blocks.3.attn1.to_v",
      "transformer_blocks.3.ff.net.0.proj",
      "transformer_blocks.4.attn1.to_k",
      "transformer_blocks.4.attn1.to_q",
      "transformer_blocks.4.attn1.to_v",
      "transformer_blocks.5.attn1.to_v",
      "transformer_blocks.6.attn1.to_v",
      "transformer_blocks.7.attn1.to_v",
      "transformer_blocks.8.attn1.to_v",
      "transformer_blocks.9.attn1.to_v",
      "transformer_blocks.21.norm1.linear",
      "transformer_blocks.22.norm1.linear",
      "transformer_blocks.25.ff.net.2",
      "transformer_blocks.26.attn1.to_v",
      "transformer_blocks.26.ff.net.2",
      "transformer_blocks.26.norm1.linear",
      "transformer_blocks.27.attn1.to_v",
      "transformer_blocks.27.norm1.linear",
      "transformer_blocks.28.attn1.to_v",
      "transformer_blocks.29.attn1.to_v"
    ],
    "llm_int8_threshold": 6.0,
    "load_in_4bit": true,
    "load_in_8bit": false,
    "quant_method": "bitsandbytes"
  },
  "text_embed_dim": 1472,
  "time_embed_dim": 512
}