Hugging Face's logo

zeromodels
/
glm-4-32b-base-0414

Text Generation
Keras
PyTorch
JAX
TensorFlow
English
Chinese
zeromodels
glm
glm4
glm-4
Model card Files
xet
 Community
IMvision12 commited on
Commit
079e4b0
·
0 Parent(s):

Super-squash branch 'main' using huggingface_hub

Browse files
Files changed (18) hide show
  1. .gitattributes +36 -0
  2. README.md +56 -0
  3. kf_config.json +25 -0
  4. model.weights.json +1846 -0
  5. model_00000.weights.h5 +3 -0
  6. model_00001.weights.h5 +3 -0
  7. model_00002.weights.h5 +3 -0
  8. model_00003.weights.h5 +3 -0
  9. model_00004.weights.h5 +3 -0
  10. model_00005.weights.h5 +3 -0
  11. model_00006.weights.h5 +3 -0
  12. model_00007.weights.h5 +3 -0
  13. model_00008.weights.h5 +3 -0
  14. model_00009.weights.h5 +3 -0
  15. model_00010.weights.h5 +3 -0
  16. model_00011.weights.h5 +3 -0
  17. model_00012.weights.h5 +3 -0
  18. tokenizer.json +3 -0
.gitattributes ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: text-generation
3
+ license: mit
4
+ base_model: zai-org/GLM-4-32B-Base-0414
5
+ library_name: kerasformers
6
+ language:
7
+ - en
8
+ - zh
9
+ tags:
10
+ - keras
11
+ - kerasformers
12
+ - glm
13
+ - glm4
14
+ - glm-4
15
+ - text-generation
16
+ - pytorch
17
+ - jax
18
+ - tf
19
+ ---
20
+
21
+ # Run GLM-4-0414 with Keras 3: JAX, PyTorch, or TensorFlow
22
+
23
+ [![GitHub](https://img.shields.io/badge/GitHub-KerasFormers-181717?logo=github)](https://github.com/IMvision12/KerasFormers) [![Docs](https://img.shields.io/badge/Docs-GLM-1f6feb)](https://imvision12.github.io/KerasFormers/glm4/) [![HuggingFace](https://img.shields.io/badge/HuggingFace-GLM-ffd21e?logo=huggingface&logoColor=black)](https://huggingface.co/collections/kerasformers/glm-6a82b8f9f753e8dcae3ff3f7)
24
+
25
+ # kerasformers/glm-4-32b-base-0414
26
+
27
+ Pure-**Keras 3** conversion of [`zai-org/GLM-4-32B-Base-0414`](https://huggingface.co/zai-org/GLM-4-32B-Base-0414) for [kerasformers](https://github.com/IMvision12/KerasFormers). One implementation runs unmodified on **TensorFlow / Torch / JAX**. This is a **GLM-4-0414** checkpoint served as **text -> text**; weights are stored in **bfloat16**.
28
+
29
+ For model details, license, and usage terms, see the upstream [model card](https://huggingface.co/zai-org/GLM-4-32B-Base-0414).
30
+
31
+ Paper: [ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools (arXiv:2406.12793)](https://arxiv.org/abs/2406.12793) · [HF Papers](https://huggingface.co/papers/2406.12793)
32
+
33
+ ## ✨ Quick start
34
+
35
+ ```python
36
+ import os
37
+ os.environ["KERAS_BACKEND"] = "torch" # or "jax" / "tensorflow"
38
+
39
+ from kerasformers.models.glm4 import Glm4TextGenerate, Glm4Tokenizer
40
+
41
+ model = Glm4TextGenerate.from_weights("kerasformers/glm-4-32b-base-0414")
42
+ tokenizer = Glm4Tokenizer.from_weights("kerasformers/glm-4-32b-base-0414")
43
+
44
+ messages = [{"role": "user", "content": "Name three prime numbers."}]
45
+ inputs = tokenizer(messages)
46
+ outputs = model.generate(**inputs, max_new_tokens=128)
47
+ print(tokenizer.decode(outputs[0]))
48
+ ```
49
+
50
+ Load any GLM variant the same way with `from_weights("kerasformers/<variant>")`. Browse them all in the [GLM collection](https://huggingface.co/collections/kerasformers/glm-6a82b8f9f753e8dcae3ff3f7).
51
+
52
+ ## Special Thanks
53
+
54
+ A huge thank you to the Zhipu AI / THUDM team for creating and releasing the GLM models.
55
+
56
+ License: `mit` (per the upstream model card).
kf_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "library_name": "kerasformers",
3
+ "kerasformers_version": "1.2.4",
4
+ "model_module": "kerasformers.models.glm4",
5
+ "model_class": "Glm4TextGenerate",
6
+ "variant": "glm-4-32b-base-0414",
7
+ "weights": "model.weights.json",
8
+ "schema_version": 2,
9
+ "weight_dtype": "bfloat16",
10
+ "model_type": "glm4",
11
+ "text_config": {
12
+ "vocab_size": 151552,
13
+ "embed_dim": 6144,
14
+ "num_layers": 61,
15
+ "num_heads": 48,
16
+ "num_kv_heads": 2,
17
+ "head_dim": 128,
18
+ "mlp_dim": 23040,
19
+ "partial_rotary_factor": 0.5,
20
+ "norm_eps": 1e-05,
21
+ "rope_theta": 10000.0,
22
+ "attention_bias": false,
23
+ "tie_embeddings": false
24
+ }
25
+ }
model.weights.json ADDED
@@ -0,0 +1,1846 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 65132163072.0
4
+ },
5
+ "weight_map": {
6
+ "/layers/embedding/vars": [
7
+ "model_00000.weights.h5"
8
+ ],
9
+ "/layers/glm4_decoder_layer/input_layernorm/vars": [
10
+ "model_00000.weights.h5"
11
+ ],
12
+ "/layers/glm4_decoder_layer/mlp/down/vars": [
13
+ "model_00000.weights.h5"
14
+ ],
15
+ "/layers/glm4_decoder_layer/mlp/gate_up/vars": [
16
+ "model_00000.weights.h5"
17
+ ],
18
+ "/layers/glm4_decoder_layer/post_attention_layernorm/vars": [
19
+ "model_00000.weights.h5"
20
+ ],
21
+ "/layers/glm4_decoder_layer/post_mlp_layernorm/vars": [
22
+ "model_00000.weights.h5"
23
+ ],
24
+ "/layers/glm4_decoder_layer/post_self_attn_layernorm/vars": [
25
+ "model_00000.weights.h5"
26
+ ],
27
+ "/layers/glm4_decoder_layer/attention/key/vars": [
28
+ "model_00000.weights.h5"
29
+ ],
30
+ "/layers/glm4_decoder_layer/attention/output_proj/vars": [
31
+ "model_00000.weights.h5"
32
+ ],
33
+ "/layers/glm4_decoder_layer/attention/query/vars": [
34
+ "model_00000.weights.h5"
35
+ ],
36
+ "/layers/glm4_decoder_layer/attention/value/vars": [
37
+ "model_00000.weights.h5"
38
+ ],
39
+ "/layers/glm4_decoder_layer_1/input_layernorm/vars": [
40
+ "model_00000.weights.h5"
41
+ ],
42
+ "/layers/glm4_decoder_layer_1/mlp/down/vars": [
43
+ "model_00000.weights.h5"
44
+ ],
45
+ "/layers/glm4_decoder_layer_1/mlp/gate_up/vars": [
46
+ "model_00000.weights.h5"
47
+ ],
48
+ "/layers/glm4_decoder_layer_1/post_attention_layernorm/vars": [
49
+ "model_00000.weights.h5"
50
+ ],
51
+ "/layers/glm4_decoder_layer_1/post_mlp_layernorm/vars": [
52
+ "model_00000.weights.h5"
53
+ ],
54
+ "/layers/glm4_decoder_layer_1/post_self_attn_layernorm/vars": [
55
+ "model_00000.weights.h5"
56
+ ],
57
+ "/layers/glm4_decoder_layer_1/attention/key/vars": [
58
+ "model_00000.weights.h5"
59
+ ],
60
+ "/layers/glm4_decoder_layer_1/attention/output_proj/vars": [
61
+ "model_00000.weights.h5"
62
+ ],
63
+ "/layers/glm4_decoder_layer_1/attention/query/vars": [
64
+ "model_00000.weights.h5"
65
+ ],
66
+ "/layers/glm4_decoder_layer_1/attention/value/vars": [
67
+ "model_00000.weights.h5"
68
+ ],
69
+ "/layers/glm4_decoder_layer_2/input_layernorm/vars": [
70
+ "model_00000.weights.h5"
71
+ ],
72
+ "/layers/glm4_decoder_layer_2/mlp/down/vars": [
73
+ "model_00000.weights.h5"
74
+ ],
75
+ "/layers/glm4_decoder_layer_2/mlp/gate_up/vars": [
76
+ "model_00000.weights.h5"
77
+ ],
78
+ "/layers/glm4_decoder_layer_2/post_attention_layernorm/vars": [
79
+ "model_00000.weights.h5"
80
+ ],
81
+ "/layers/glm4_decoder_layer_2/post_mlp_layernorm/vars": [
82
+ "model_00000.weights.h5"
83
+ ],
84
+ "/layers/glm4_decoder_layer_2/post_self_attn_layernorm/vars": [
85
+ "model_00000.weights.h5"
86
+ ],
87
+ "/layers/glm4_decoder_layer_2/attention/key/vars": [
88
+ "model_00000.weights.h5"
89
+ ],
90
+ "/layers/glm4_decoder_layer_2/attention/output_proj/vars": [
91
+ "model_00000.weights.h5"
92
+ ],
93
+ "/layers/glm4_decoder_layer_2/attention/query/vars": [
94
+ "model_00000.weights.h5"
95
+ ],
96
+ "/layers/glm4_decoder_layer_2/attention/value/vars": [
97
+ "model_00000.weights.h5"
98
+ ],
99
+ "/layers/glm4_decoder_layer_3/attention/key/vars": [
100
+ "model_00000.weights.h5"
101
+ ],
102
+ "/layers/glm4_decoder_layer_3/attention/query/vars": [
103
+ "model_00000.weights.h5"
104
+ ],
105
+ "/layers/glm4_decoder_layer_3/attention/value/vars": [
106
+ "model_00000.weights.h5"
107
+ ],
108
+ "/layers/glm4_decoder_layer_3/input_layernorm/vars": [
109
+ "model_00000.weights.h5"
110
+ ],
111
+ "/layers/glm4_decoder_layer_3/mlp/down/vars": [
112
+ "model_00000.weights.h5"
113
+ ],
114
+ "/layers/glm4_decoder_layer_3/mlp/gate_up/vars": [
115
+ "model_00001.weights.h5"
116
+ ],
117
+ "/layers/glm4_decoder_layer_3/post_attention_layernorm/vars": [
118
+ "model_00001.weights.h5"
119
+ ],
120
+ "/layers/glm4_decoder_layer_3/post_mlp_layernorm/vars": [
121
+ "model_00001.weights.h5"
122
+ ],
123
+ "/layers/glm4_decoder_layer_3/post_self_attn_layernorm/vars": [
124
+ "model_00001.weights.h5"
125
+ ],
126
+ "/layers/glm4_decoder_layer_3/attention/output_proj/vars": [
127
+ "model_00001.weights.h5"
128
+ ],
129
+ "/layers/glm4_decoder_layer_4/input_layernorm/vars": [
130
+ "model_00001.weights.h5"
131
+ ],
132
+ "/layers/glm4_decoder_layer_4/mlp/down/vars": [
133
+ "model_00001.weights.h5"
134
+ ],
135
+ "/layers/glm4_decoder_layer_4/mlp/gate_up/vars": [
136
+ "model_00001.weights.h5"
137
+ ],
138
+ "/layers/glm4_decoder_layer_4/post_attention_layernorm/vars": [
139
+ "model_00001.weights.h5"
140
+ ],
141
+ "/layers/glm4_decoder_layer_4/post_mlp_layernorm/vars": [
142
+ "model_00001.weights.h5"
143
+ ],
144
+ "/layers/glm4_decoder_layer_4/post_self_attn_layernorm/vars": [
145
+ "model_00001.weights.h5"
146
+ ],
147
+ "/layers/glm4_decoder_layer_4/attention/key/vars": [
148
+ "model_00001.weights.h5"
149
+ ],
150
+ "/layers/glm4_decoder_layer_4/attention/output_proj/vars": [
151
+ "model_00001.weights.h5"
152
+ ],
153
+ "/layers/glm4_decoder_layer_4/attention/query/vars": [
154
+ "model_00001.weights.h5"
155
+ ],
156
+ "/layers/glm4_decoder_layer_4/attention/value/vars": [
157
+ "model_00001.weights.h5"
158
+ ],
159
+ "/layers/glm4_decoder_layer_5/input_layernorm/vars": [
160
+ "model_00001.weights.h5"
161
+ ],
162
+ "/layers/glm4_decoder_layer_5/mlp/down/vars": [
163
+ "model_00001.weights.h5"
164
+ ],
165
+ "/layers/glm4_decoder_layer_5/mlp/gate_up/vars": [
166
+ "model_00001.weights.h5"
167
+ ],
168
+ "/layers/glm4_decoder_layer_5/post_attention_layernorm/vars": [
169
+ "model_00001.weights.h5"
170
+ ],
171
+ "/layers/glm4_decoder_layer_5/post_mlp_layernorm/vars": [
172
+ "model_00001.weights.h5"
173
+ ],
174
+ "/layers/glm4_decoder_layer_5/post_self_attn_layernorm/vars": [
175
+ "model_00001.weights.h5"
176
+ ],
177
+ "/layers/glm4_decoder_layer_5/attention/key/vars": [
178
+ "model_00001.weights.h5"
179
+ ],
180
+ "/layers/glm4_decoder_layer_5/attention/output_proj/vars": [
181
+ "model_00001.weights.h5"
182
+ ],
183
+ "/layers/glm4_decoder_layer_5/attention/query/vars": [
184
+ "model_00001.weights.h5"
185
+ ],
186
+ "/layers/glm4_decoder_layer_5/attention/value/vars": [
187
+ "model_00001.weights.h5"
188
+ ],
189
+ "/layers/glm4_decoder_layer_6/input_layernorm/vars": [
190
+ "model_00001.weights.h5"
191
+ ],
192
+ "/layers/glm4_decoder_layer_6/mlp/down/vars": [
193
+ "model_00001.weights.h5"
194
+ ],
195
+ "/layers/glm4_decoder_layer_6/mlp/gate_up/vars": [
196
+ "model_00001.weights.h5"
197
+ ],
198
+ "/layers/glm4_decoder_layer_6/post_attention_layernorm/vars": [
199
+ "model_00001.weights.h5"
200
+ ],
201
+ "/layers/glm4_decoder_layer_6/post_mlp_layernorm/vars": [
202
+ "model_00001.weights.h5"
203
+ ],
204
+ "/layers/glm4_decoder_layer_6/post_self_attn_layernorm/vars": [
205
+ "model_00001.weights.h5"
206
+ ],
207
+ "/layers/glm4_decoder_layer_6/attention/key/vars": [
208
+ "model_00001.weights.h5"
209
+ ],
210
+ "/layers/glm4_decoder_layer_6/attention/output_proj/vars": [
211
+ "model_00001.weights.h5"
212
+ ],
213
+ "/layers/glm4_decoder_layer_6/attention/query/vars": [
214
+ "model_00001.weights.h5"
215
+ ],
216
+ "/layers/glm4_decoder_layer_6/attention/value/vars": [
217
+ "model_00001.weights.h5"
218
+ ],
219
+ "/layers/glm4_decoder_layer_7/input_layernorm/vars": [
220
+ "model_00001.weights.h5"
221
+ ],
222
+ "/layers/glm4_decoder_layer_7/mlp/down/vars": [
223
+ "model_00001.weights.h5"
224
+ ],
225
+ "/layers/glm4_decoder_layer_7/mlp/gate_up/vars": [
226
+ "model_00001.weights.h5"
227
+ ],
228
+ "/layers/glm4_decoder_layer_7/post_attention_layernorm/vars": [
229
+ "model_00001.weights.h5"
230
+ ],
231
+ "/layers/glm4_decoder_layer_7/post_mlp_layernorm/vars": [
232
+ "model_00001.weights.h5"
233
+ ],
234
+ "/layers/glm4_decoder_layer_7/post_self_attn_layernorm/vars": [
235
+ "model_00001.weights.h5"
236
+ ],
237
+ "/layers/glm4_decoder_layer_7/attention/key/vars": [
238
+ "model_00001.weights.h5"
239
+ ],
240
+ "/layers/glm4_decoder_layer_7/attention/output_proj/vars": [
241
+ "model_00001.weights.h5"
242
+ ],
243
+ "/layers/glm4_decoder_layer_7/attention/query/vars": [
244
+ "model_00001.weights.h5"
245
+ ],
246
+ "/layers/glm4_decoder_layer_7/attention/value/vars": [
247
+ "model_00001.weights.h5"
248
+ ],
249
+ "/layers/glm4_decoder_layer_10/input_layernorm/vars": [
250
+ "model_00001.weights.h5"
251
+ ],
252
+ "/layers/glm4_decoder_layer_10/mlp/down/vars": [
253
+ "model_00001.weights.h5"
254
+ ],
255
+ "/layers/glm4_decoder_layer_10/mlp/gate_up/vars": [
256
+ "model_00002.weights.h5"
257
+ ],
258
+ "/layers/glm4_decoder_layer_10/post_attention_layernorm/vars": [
259
+ "model_00002.weights.h5"
260
+ ],
261
+ "/layers/glm4_decoder_layer_10/post_mlp_layernorm/vars": [
262
+ "model_00002.weights.h5"
263
+ ],
264
+ "/layers/glm4_decoder_layer_10/post_self_attn_layernorm/vars": [
265
+ "model_00002.weights.h5"
266
+ ],
267
+ "/layers/glm4_decoder_layer_10/attention/key/vars": [
268
+ "model_00002.weights.h5"
269
+ ],
270
+ "/layers/glm4_decoder_layer_10/attention/output_proj/vars": [
271
+ "model_00002.weights.h5"
272
+ ],
273
+ "/layers/glm4_decoder_layer_10/attention/query/vars": [
274
+ "model_00002.weights.h5"
275
+ ],
276
+ "/layers/glm4_decoder_layer_10/attention/value/vars": [
277
+ "model_00002.weights.h5"
278
+ ],
279
+ "/layers/glm4_decoder_layer_11/input_layernorm/vars": [
280
+ "model_00002.weights.h5"
281
+ ],
282
+ "/layers/glm4_decoder_layer_11/mlp/down/vars": [
283
+ "model_00002.weights.h5"
284
+ ],
285
+ "/layers/glm4_decoder_layer_11/mlp/gate_up/vars": [
286
+ "model_00002.weights.h5"
287
+ ],
288
+ "/layers/glm4_decoder_layer_11/post_attention_layernorm/vars": [
289
+ "model_00002.weights.h5"
290
+ ],
291
+ "/layers/glm4_decoder_layer_11/post_mlp_layernorm/vars": [
292
+ "model_00002.weights.h5"
293
+ ],
294
+ "/layers/glm4_decoder_layer_11/post_self_attn_layernorm/vars": [
295
+ "model_00002.weights.h5"
296
+ ],
297
+ "/layers/glm4_decoder_layer_11/attention/key/vars": [
298
+ "model_00002.weights.h5"
299
+ ],
300
+ "/layers/glm4_decoder_layer_11/attention/output_proj/vars": [
301
+ "model_00002.weights.h5"
302
+ ],
303
+ "/layers/glm4_decoder_layer_11/attention/query/vars": [
304
+ "model_00002.weights.h5"
305
+ ],
306
+ "/layers/glm4_decoder_layer_11/attention/value/vars": [
307
+ "model_00002.weights.h5"
308
+ ],
309
+ "/layers/glm4_decoder_layer_12/mlp/gate_up/vars": [
310
+ "model_00002.weights.h5"
311
+ ],
312
+ "/layers/glm4_decoder_layer_12/attention/key/vars": [
313
+ "model_00002.weights.h5"
314
+ ],
315
+ "/layers/glm4_decoder_layer_12/attention/output_proj/vars": [
316
+ "model_00002.weights.h5"
317
+ ],
318
+ "/layers/glm4_decoder_layer_12/attention/query/vars": [
319
+ "model_00002.weights.h5"
320
+ ],
321
+ "/layers/glm4_decoder_layer_12/attention/value/vars": [
322
+ "model_00002.weights.h5"
323
+ ],
324
+ "/layers/glm4_decoder_layer_8/input_layernorm/vars": [
325
+ "model_00002.weights.h5"
326
+ ],
327
+ "/layers/glm4_decoder_layer_8/mlp/down/vars": [
328
+ "model_00002.weights.h5"
329
+ ],
330
+ "/layers/glm4_decoder_layer_8/mlp/gate_up/vars": [
331
+ "model_00002.weights.h5"
332
+ ],
333
+ "/layers/glm4_decoder_layer_8/post_attention_layernorm/vars": [
334
+ "model_00002.weights.h5"
335
+ ],
336
+ "/layers/glm4_decoder_layer_8/post_mlp_layernorm/vars": [
337
+ "model_00002.weights.h5"
338
+ ],
339
+ "/layers/glm4_decoder_layer_8/post_self_attn_layernorm/vars": [
340
+ "model_00002.weights.h5"
341
+ ],
342
+ "/layers/glm4_decoder_layer_8/attention/key/vars": [
343
+ "model_00002.weights.h5"
344
+ ],
345
+ "/layers/glm4_decoder_layer_8/attention/output_proj/vars": [
346
+ "model_00002.weights.h5"
347
+ ],
348
+ "/layers/glm4_decoder_layer_8/attention/query/vars": [
349
+ "model_00002.weights.h5"
350
+ ],
351
+ "/layers/glm4_decoder_layer_8/attention/value/vars": [
352
+ "model_00002.weights.h5"
353
+ ],
354
+ "/layers/glm4_decoder_layer_9/input_layernorm/vars": [
355
+ "model_00002.weights.h5"
356
+ ],
357
+ "/layers/glm4_decoder_layer_9/mlp/down/vars": [
358
+ "model_00002.weights.h5"
359
+ ],
360
+ "/layers/glm4_decoder_layer_9/mlp/gate_up/vars": [
361
+ "model_00002.weights.h5"
362
+ ],
363
+ "/layers/glm4_decoder_layer_9/post_attention_layernorm/vars": [
364
+ "model_00002.weights.h5"
365
+ ],
366
+ "/layers/glm4_decoder_layer_9/post_mlp_layernorm/vars": [
367
+ "model_00002.weights.h5"
368
+ ],
369
+ "/layers/glm4_decoder_layer_9/post_self_attn_layernorm/vars": [
370
+ "model_00002.weights.h5"
371
+ ],
372
+ "/layers/glm4_decoder_layer_9/attention/key/vars": [
373
+ "model_00002.weights.h5"
374
+ ],
375
+ "/layers/glm4_decoder_layer_9/attention/output_proj/vars": [
376
+ "model_00002.weights.h5"
377
+ ],
378
+ "/layers/glm4_decoder_layer_9/attention/query/vars": [
379
+ "model_00002.weights.h5"
380
+ ],
381
+ "/layers/glm4_decoder_layer_9/attention/value/vars": [
382
+ "model_00002.weights.h5"
383
+ ],
384
+ "/layers/glm4_decoder_layer_12/input_layernorm/vars": [
385
+ "model_00002.weights.h5"
386
+ ],
387
+ "/layers/glm4_decoder_layer_12/mlp/down/vars": [
388
+ "model_00002.weights.h5"
389
+ ],
390
+ "/layers/glm4_decoder_layer_12/post_attention_layernorm/vars": [
391
+ "model_00002.weights.h5"
392
+ ],
393
+ "/layers/glm4_decoder_layer_12/post_mlp_layernorm/vars": [
394
+ "model_00002.weights.h5"
395
+ ],
396
+ "/layers/glm4_decoder_layer_12/post_self_attn_layernorm/vars": [
397
+ "model_00002.weights.h5"
398
+ ],
399
+ "/layers/glm4_decoder_layer_13/input_layernorm/vars": [
400
+ "model_00002.weights.h5"
401
+ ],
402
+ "/layers/glm4_decoder_layer_13/mlp/down/vars": [
403
+ "model_00002.weights.h5"
404
+ ],
405
+ "/layers/glm4_decoder_layer_13/mlp/gate_up/vars": [
406
+ "model_00003.weights.h5"
407
+ ],
408
+ "/layers/glm4_decoder_layer_13/post_attention_layernorm/vars": [
409
+ "model_00003.weights.h5"
410
+ ],
411
+ "/layers/glm4_decoder_layer_13/post_mlp_layernorm/vars": [
412
+ "model_00003.weights.h5"
413
+ ],
414
+ "/layers/glm4_decoder_layer_13/post_self_attn_layernorm/vars": [
415
+ "model_00003.weights.h5"
416
+ ],
417
+ "/layers/glm4_decoder_layer_13/attention/key/vars": [
418
+ "model_00003.weights.h5"
419
+ ],
420
+ "/layers/glm4_decoder_layer_13/attention/output_proj/vars": [
421
+ "model_00003.weights.h5"
422
+ ],
423
+ "/layers/glm4_decoder_layer_13/attention/query/vars": [
424
+ "model_00003.weights.h5"
425
+ ],
426
+ "/layers/glm4_decoder_layer_13/attention/value/vars": [
427
+ "model_00003.weights.h5"
428
+ ],
429
+ "/layers/glm4_decoder_layer_14/input_layernorm/vars": [
430
+ "model_00003.weights.h5"
431
+ ],
432
+ "/layers/glm4_decoder_layer_14/mlp/down/vars": [
433
+ "model_00003.weights.h5"
434
+ ],
435
+ "/layers/glm4_decoder_layer_14/mlp/gate_up/vars": [
436
+ "model_00003.weights.h5"
437
+ ],
438
+ "/layers/glm4_decoder_layer_14/post_attention_layernorm/vars": [
439
+ "model_00003.weights.h5"
440
+ ],
441
+ "/layers/glm4_decoder_layer_14/post_mlp_layernorm/vars": [
442
+ "model_00003.weights.h5"
443
+ ],
444
+ "/layers/glm4_decoder_layer_14/post_self_attn_layernorm/vars": [
445
+ "model_00003.weights.h5"
446
+ ],
447
+ "/layers/glm4_decoder_layer_14/attention/key/vars": [
448
+ "model_00003.weights.h5"
449
+ ],
450
+ "/layers/glm4_decoder_layer_14/attention/output_proj/vars": [
451
+ "model_00003.weights.h5"
452
+ ],
453
+ "/layers/glm4_decoder_layer_14/attention/query/vars": [
454
+ "model_00003.weights.h5"
455
+ ],
456
+ "/layers/glm4_decoder_layer_14/attention/value/vars": [
457
+ "model_00003.weights.h5"
458
+ ],
459
+ "/layers/glm4_decoder_layer_15/input_layernorm/vars": [
460
+ "model_00003.weights.h5"
461
+ ],
462
+ "/layers/glm4_decoder_layer_15/mlp/down/vars": [
463
+ "model_00003.weights.h5"
464
+ ],
465
+ "/layers/glm4_decoder_layer_15/mlp/gate_up/vars": [
466
+ "model_00003.weights.h5"
467
+ ],
468
+ "/layers/glm4_decoder_layer_15/post_attention_layernorm/vars": [
469
+ "model_00003.weights.h5"
470
+ ],
471
+ "/layers/glm4_decoder_layer_15/post_mlp_layernorm/vars": [
472
+ "model_00003.weights.h5"
473
+ ],
474
+ "/layers/glm4_decoder_layer_15/post_self_attn_layernorm/vars": [
475
+ "model_00003.weights.h5"
476
+ ],
477
+ "/layers/glm4_decoder_layer_15/attention/key/vars": [
478
+ "model_00003.weights.h5"
479
+ ],
480
+ "/layers/glm4_decoder_layer_15/attention/output_proj/vars": [
481
+ "model_00003.weights.h5"
482
+ ],
483
+ "/layers/glm4_decoder_layer_15/attention/query/vars": [
484
+ "model_00003.weights.h5"
485
+ ],
486
+ "/layers/glm4_decoder_layer_15/attention/value/vars": [
487
+ "model_00003.weights.h5"
488
+ ],
489
+ "/layers/glm4_decoder_layer_16/input_layernorm/vars": [
490
+ "model_00003.weights.h5"
491
+ ],
492
+ "/layers/glm4_decoder_layer_16/mlp/down/vars": [
493
+ "model_00003.weights.h5"
494
+ ],
495
+ "/layers/glm4_decoder_layer_16/mlp/gate_up/vars": [
496
+ "model_00003.weights.h5"
497
+ ],
498
+ "/layers/glm4_decoder_layer_16/post_attention_layernorm/vars": [
499
+ "model_00003.weights.h5"
500
+ ],
501
+ "/layers/glm4_decoder_layer_16/post_mlp_layernorm/vars": [
502
+ "model_00003.weights.h5"
503
+ ],
504
+ "/layers/glm4_decoder_layer_16/post_self_attn_layernorm/vars": [
505
+ "model_00003.weights.h5"
506
+ ],
507
+ "/layers/glm4_decoder_layer_16/attention/key/vars": [
508
+ "model_00003.weights.h5"
509
+ ],
510
+ "/layers/glm4_decoder_layer_16/attention/output_proj/vars": [
511
+ "model_00003.weights.h5"
512
+ ],
513
+ "/layers/glm4_decoder_layer_16/attention/query/vars": [
514
+ "model_00003.weights.h5"
515
+ ],
516
+ "/layers/glm4_decoder_layer_16/attention/value/vars": [
517
+ "model_00003.weights.h5"
518
+ ],
519
+ "/layers/glm4_decoder_layer_17/attention/key/vars": [
520
+ "model_00003.weights.h5"
521
+ ],
522
+ "/layers/glm4_decoder_layer_17/attention/output_proj/vars": [
523
+ "model_00003.weights.h5"
524
+ ],
525
+ "/layers/glm4_decoder_layer_17/attention/query/vars": [
526
+ "model_00003.weights.h5"
527
+ ],
528
+ "/layers/glm4_decoder_layer_17/attention/value/vars": [
529
+ "model_00003.weights.h5"
530
+ ],
531
+ "/layers/glm4_decoder_layer_17/input_layernorm/vars": [
532
+ "model_00003.weights.h5"
533
+ ],
534
+ "/layers/glm4_decoder_layer_17/mlp/down/vars": [
535
+ "model_00003.weights.h5"
536
+ ],
537
+ "/layers/glm4_decoder_layer_17/mlp/gate_up/vars": [
538
+ "model_00003.weights.h5"
539
+ ],
540
+ "/layers/glm4_decoder_layer_17/post_attention_layernorm/vars": [
541
+ "model_00003.weights.h5"
542
+ ],
543
+ "/layers/glm4_decoder_layer_17/post_mlp_layernorm/vars": [
544
+ "model_00003.weights.h5"
545
+ ],
546
+ "/layers/glm4_decoder_layer_17/post_self_attn_layernorm/vars": [
547
+ "model_00003.weights.h5"
548
+ ],
549
+ "/layers/glm4_decoder_layer_18/input_layernorm/vars": [
550
+ "model_00003.weights.h5"
551
+ ],
552
+ "/layers/glm4_decoder_layer_18/mlp/down/vars": [
553
+ "model_00003.weights.h5"
554
+ ],
555
+ "/layers/glm4_decoder_layer_18/mlp/gate_up/vars": [
556
+ "model_00004.weights.h5"
557
+ ],
558
+ "/layers/glm4_decoder_layer_18/post_attention_layernorm/vars": [
559
+ "model_00004.weights.h5"
560
+ ],
561
+ "/layers/glm4_decoder_layer_18/post_mlp_layernorm/vars": [
562
+ "model_00004.weights.h5"
563
+ ],
564
+ "/layers/glm4_decoder_layer_18/post_self_attn_layernorm/vars": [
565
+ "model_00004.weights.h5"
566
+ ],
567
+ "/layers/glm4_decoder_layer_18/attention/key/vars": [
568
+ "model_00004.weights.h5"
569
+ ],
570
+ "/layers/glm4_decoder_layer_18/attention/output_proj/vars": [
571
+ "model_00004.weights.h5"
572
+ ],
573
+ "/layers/glm4_decoder_layer_18/attention/query/vars": [
574
+ "model_00004.weights.h5"
575
+ ],
576
+ "/layers/glm4_decoder_layer_18/attention/value/vars": [
577
+ "model_00004.weights.h5"
578
+ ],
579
+ "/layers/glm4_decoder_layer_19/input_layernorm/vars": [
580
+ "model_00004.weights.h5"
581
+ ],
582
+ "/layers/glm4_decoder_layer_19/mlp/down/vars": [
583
+ "model_00004.weights.h5"
584
+ ],
585
+ "/layers/glm4_decoder_layer_19/mlp/gate_up/vars": [
586
+ "model_00004.weights.h5"
587
+ ],
588
+ "/layers/glm4_decoder_layer_19/post_attention_layernorm/vars": [
589
+ "model_00004.weights.h5"
590
+ ],
591
+ "/layers/glm4_decoder_layer_19/post_mlp_layernorm/vars": [
592
+ "model_00004.weights.h5"
593
+ ],
594
+ "/layers/glm4_decoder_layer_19/post_self_attn_layernorm/vars": [
595
+ "model_00004.weights.h5"
596
+ ],
597
+ "/layers/glm4_decoder_layer_19/attention/key/vars": [
598
+ "model_00004.weights.h5"
599
+ ],
600
+ "/layers/glm4_decoder_layer_19/attention/output_proj/vars": [
601
+ "model_00004.weights.h5"
602
+ ],
603
+ "/layers/glm4_decoder_layer_19/attention/query/vars": [
604
+ "model_00004.weights.h5"
605
+ ],
606
+ "/layers/glm4_decoder_layer_19/attention/value/vars": [
607
+ "model_00004.weights.h5"
608
+ ],
609
+ "/layers/glm4_decoder_layer_20/input_layernorm/vars": [
610
+ "model_00004.weights.h5"
611
+ ],
612
+ "/layers/glm4_decoder_layer_20/mlp/down/vars": [
613
+ "model_00004.weights.h5"
614
+ ],
615
+ "/layers/glm4_decoder_layer_20/mlp/gate_up/vars": [
616
+ "model_00004.weights.h5"
617
+ ],
618
+ "/layers/glm4_decoder_layer_20/post_attention_layernorm/vars": [
619
+ "model_00004.weights.h5"
620
+ ],
621
+ "/layers/glm4_decoder_layer_20/post_mlp_layernorm/vars": [
622
+ "model_00004.weights.h5"
623
+ ],
624
+ "/layers/glm4_decoder_layer_20/post_self_attn_layernorm/vars": [
625
+ "model_00004.weights.h5"
626
+ ],
627
+ "/layers/glm4_decoder_layer_20/attention/key/vars": [
628
+ "model_00004.weights.h5"
629
+ ],
630
+ "/layers/glm4_decoder_layer_20/attention/output_proj/vars": [
631
+ "model_00004.weights.h5"
632
+ ],
633
+ "/layers/glm4_decoder_layer_20/attention/query/vars": [
634
+ "model_00004.weights.h5"
635
+ ],
636
+ "/layers/glm4_decoder_layer_20/attention/value/vars": [
637
+ "model_00004.weights.h5"
638
+ ],
639
+ "/layers/glm4_decoder_layer_21/input_layernorm/vars": [
640
+ "model_00004.weights.h5"
641
+ ],
642
+ "/layers/glm4_decoder_layer_21/mlp/down/vars": [
643
+ "model_00004.weights.h5"
644
+ ],
645
+ "/layers/glm4_decoder_layer_21/mlp/gate_up/vars": [
646
+ "model_00004.weights.h5"
647
+ ],
648
+ "/layers/glm4_decoder_layer_21/post_attention_layernorm/vars": [
649
+ "model_00004.weights.h5"
650
+ ],
651
+ "/layers/glm4_decoder_layer_21/post_mlp_layernorm/vars": [
652
+ "model_00004.weights.h5"
653
+ ],
654
+ "/layers/glm4_decoder_layer_21/post_self_attn_layernorm/vars": [
655
+ "model_00004.weights.h5"
656
+ ],
657
+ "/layers/glm4_decoder_layer_21/attention/key/vars": [
658
+ "model_00004.weights.h5"
659
+ ],
660
+ "/layers/glm4_decoder_layer_21/attention/output_proj/vars": [
661
+ "model_00004.weights.h5"
662
+ ],
663
+ "/layers/glm4_decoder_layer_21/attention/query/vars": [
664
+ "model_00004.weights.h5"
665
+ ],
666
+ "/layers/glm4_decoder_layer_21/attention/value/vars": [
667
+ "model_00004.weights.h5"
668
+ ],
669
+ "/layers/glm4_decoder_layer_22/attention/key/vars": [
670
+ "model_00004.weights.h5"
671
+ ],
672
+ "/layers/glm4_decoder_layer_22/attention/query/vars": [
673
+ "model_00004.weights.h5"
674
+ ],
675
+ "/layers/glm4_decoder_layer_22/attention/value/vars": [
676
+ "model_00004.weights.h5"
677
+ ],
678
+ "/layers/glm4_decoder_layer_22/input_layernorm/vars": [
679
+ "model_00004.weights.h5"
680
+ ],
681
+ "/layers/glm4_decoder_layer_22/mlp/down/vars": [
682
+ "model_00004.weights.h5"
683
+ ],
684
+ "/layers/glm4_decoder_layer_22/mlp/gate_up/vars": [
685
+ "model_00004.weights.h5"
686
+ ],
687
+ "/layers/glm4_decoder_layer_22/post_attention_layernorm/vars": [
688
+ "model_00004.weights.h5"
689
+ ],
690
+ "/layers/glm4_decoder_layer_22/post_mlp_layernorm/vars": [
691
+ "model_00004.weights.h5"
692
+ ],
693
+ "/layers/glm4_decoder_layer_22/post_self_attn_layernorm/vars": [
694
+ "model_00004.weights.h5"
695
+ ],
696
+ "/layers/glm4_decoder_layer_22/attention/output_proj/vars": [
697
+ "model_00004.weights.h5"
698
+ ],
699
+ "/layers/glm4_decoder_layer_23/input_layernorm/vars": [
700
+ "model_00004.weights.h5"
701
+ ],
702
+ "/layers/glm4_decoder_layer_23/mlp/down/vars": [
703
+ "model_00004.weights.h5"
704
+ ],
705
+ "/layers/glm4_decoder_layer_23/mlp/gate_up/vars": [
706
+ "model_00005.weights.h5"
707
+ ],
708
+ "/layers/glm4_decoder_layer_23/post_attention_layernorm/vars": [
709
+ "model_00005.weights.h5"
710
+ ],
711
+ "/layers/glm4_decoder_layer_23/post_mlp_layernorm/vars": [
712
+ "model_00005.weights.h5"
713
+ ],
714
+ "/layers/glm4_decoder_layer_23/post_self_attn_layernorm/vars": [
715
+ "model_00005.weights.h5"
716
+ ],
717
+ "/layers/glm4_decoder_layer_23/attention/key/vars": [
718
+ "model_00005.weights.h5"
719
+ ],
720
+ "/layers/glm4_decoder_layer_23/attention/output_proj/vars": [
721
+ "model_00005.weights.h5"
722
+ ],
723
+ "/layers/glm4_decoder_layer_23/attention/query/vars": [
724
+ "model_00005.weights.h5"
725
+ ],
726
+ "/layers/glm4_decoder_layer_23/attention/value/vars": [
727
+ "model_00005.weights.h5"
728
+ ],
729
+ "/layers/glm4_decoder_layer_24/input_layernorm/vars": [
730
+ "model_00005.weights.h5"
731
+ ],
732
+ "/layers/glm4_decoder_layer_24/mlp/down/vars": [
733
+ "model_00005.weights.h5"
734
+ ],
735
+ "/layers/glm4_decoder_layer_24/mlp/gate_up/vars": [
736
+ "model_00005.weights.h5"
737
+ ],
738
+ "/layers/glm4_decoder_layer_24/post_attention_layernorm/vars": [
739
+ "model_00005.weights.h5"
740
+ ],
741
+ "/layers/glm4_decoder_layer_24/post_mlp_layernorm/vars": [
742
+ "model_00005.weights.h5"
743
+ ],
744
+ "/layers/glm4_decoder_layer_24/post_self_attn_layernorm/vars": [
745
+ "model_00005.weights.h5"
746
+ ],
747
+ "/layers/glm4_decoder_layer_24/attention/key/vars": [
748
+ "model_00005.weights.h5"
749
+ ],
750
+ "/layers/glm4_decoder_layer_24/attention/output_proj/vars": [
751
+ "model_00005.weights.h5"
752
+ ],
753
+ "/layers/glm4_decoder_layer_24/attention/query/vars": [
754
+ "model_00005.weights.h5"
755
+ ],
756
+ "/layers/glm4_decoder_layer_24/attention/value/vars": [
757
+ "model_00005.weights.h5"
758
+ ],
759
+ "/layers/glm4_decoder_layer_25/input_layernorm/vars": [
760
+ "model_00005.weights.h5"
761
+ ],
762
+ "/layers/glm4_decoder_layer_25/mlp/down/vars": [
763
+ "model_00005.weights.h5"
764
+ ],
765
+ "/layers/glm4_decoder_layer_25/mlp/gate_up/vars": [
766
+ "model_00005.weights.h5"
767
+ ],
768
+ "/layers/glm4_decoder_layer_25/post_attention_layernorm/vars": [
769
+ "model_00005.weights.h5"
770
+ ],
771
+ "/layers/glm4_decoder_layer_25/post_mlp_layernorm/vars": [
772
+ "model_00005.weights.h5"
773
+ ],
774
+ "/layers/glm4_decoder_layer_25/post_self_attn_layernorm/vars": [
775
+ "model_00005.weights.h5"
776
+ ],
777
+ "/layers/glm4_decoder_layer_25/attention/key/vars": [
778
+ "model_00005.weights.h5"
779
+ ],
780
+ "/layers/glm4_decoder_layer_25/attention/output_proj/vars": [
781
+ "model_00005.weights.h5"
782
+ ],
783
+ "/layers/glm4_decoder_layer_25/attention/query/vars": [
784
+ "model_00005.weights.h5"
785
+ ],
786
+ "/layers/glm4_decoder_layer_25/attention/value/vars": [
787
+ "model_00005.weights.h5"
788
+ ],
789
+ "/layers/glm4_decoder_layer_26/input_layernorm/vars": [
790
+ "model_00005.weights.h5"
791
+ ],
792
+ "/layers/glm4_decoder_layer_26/mlp/down/vars": [
793
+ "model_00005.weights.h5"
794
+ ],
795
+ "/layers/glm4_decoder_layer_26/mlp/gate_up/vars": [
796
+ "model_00005.weights.h5"
797
+ ],
798
+ "/layers/glm4_decoder_layer_26/post_attention_layernorm/vars": [
799
+ "model_00005.weights.h5"
800
+ ],
801
+ "/layers/glm4_decoder_layer_26/post_mlp_layernorm/vars": [
802
+ "model_00005.weights.h5"
803
+ ],
804
+ "/layers/glm4_decoder_layer_26/post_self_attn_layernorm/vars": [
805
+ "model_00005.weights.h5"
806
+ ],
807
+ "/layers/glm4_decoder_layer_26/attention/key/vars": [
808
+ "model_00005.weights.h5"
809
+ ],
810
+ "/layers/glm4_decoder_layer_26/attention/output_proj/vars": [
811
+ "model_00005.weights.h5"
812
+ ],
813
+ "/layers/glm4_decoder_layer_26/attention/query/vars": [
814
+ "model_00005.weights.h5"
815
+ ],
816
+ "/layers/glm4_decoder_layer_26/attention/value/vars": [
817
+ "model_00005.weights.h5"
818
+ ],
819
+ "/layers/glm4_decoder_layer_27/input_layernorm/vars": [
820
+ "model_00005.weights.h5"
821
+ ],
822
+ "/layers/glm4_decoder_layer_27/mlp/down/vars": [
823
+ "model_00005.weights.h5"
824
+ ],
825
+ "/layers/glm4_decoder_layer_27/mlp/gate_up/vars": [
826
+ "model_00005.weights.h5"
827
+ ],
828
+ "/layers/glm4_decoder_layer_27/post_attention_layernorm/vars": [
829
+ "model_00005.weights.h5"
830
+ ],
831
+ "/layers/glm4_decoder_layer_27/post_mlp_layernorm/vars": [
832
+ "model_00005.weights.h5"
833
+ ],
834
+ "/layers/glm4_decoder_layer_27/post_self_attn_layernorm/vars": [
835
+ "model_00005.weights.h5"
836
+ ],
837
+ "/layers/glm4_decoder_layer_27/attention/key/vars": [
838
+ "model_00005.weights.h5"
839
+ ],
840
+ "/layers/glm4_decoder_layer_27/attention/output_proj/vars": [
841
+ "model_00005.weights.h5"
842
+ ],
843
+ "/layers/glm4_decoder_layer_27/attention/query/vars": [
844
+ "model_00005.weights.h5"
845
+ ],
846
+ "/layers/glm4_decoder_layer_27/attention/value/vars": [
847
+ "model_00005.weights.h5"
848
+ ],
849
+ "/layers/glm4_decoder_layer_28/input_layernorm/vars": [
850
+ "model_00005.weights.h5"
851
+ ],
852
+ "/layers/glm4_decoder_layer_28/mlp/down/vars": [
853
+ "model_00005.weights.h5"
854
+ ],
855
+ "/layers/glm4_decoder_layer_28/mlp/gate_up/vars": [
856
+ "model_00006.weights.h5"
857
+ ],
858
+ "/layers/glm4_decoder_layer_28/post_attention_layernorm/vars": [
859
+ "model_00006.weights.h5"
860
+ ],
861
+ "/layers/glm4_decoder_layer_28/post_mlp_layernorm/vars": [
862
+ "model_00006.weights.h5"
863
+ ],
864
+ "/layers/glm4_decoder_layer_28/post_self_attn_layernorm/vars": [
865
+ "model_00006.weights.h5"
866
+ ],
867
+ "/layers/glm4_decoder_layer_28/attention/key/vars": [
868
+ "model_00006.weights.h5"
869
+ ],
870
+ "/layers/glm4_decoder_layer_28/attention/output_proj/vars": [
871
+ "model_00006.weights.h5"
872
+ ],
873
+ "/layers/glm4_decoder_layer_28/attention/query/vars": [
874
+ "model_00006.weights.h5"
875
+ ],
876
+ "/layers/glm4_decoder_layer_28/attention/value/vars": [
877
+ "model_00006.weights.h5"
878
+ ],
879
+ "/layers/glm4_decoder_layer_29/input_layernorm/vars": [
880
+ "model_00006.weights.h5"
881
+ ],
882
+ "/layers/glm4_decoder_layer_29/mlp/down/vars": [
883
+ "model_00006.weights.h5"
884
+ ],
885
+ "/layers/glm4_decoder_layer_29/mlp/gate_up/vars": [
886
+ "model_00006.weights.h5"
887
+ ],
888
+ "/layers/glm4_decoder_layer_29/post_attention_layernorm/vars": [
889
+ "model_00006.weights.h5"
890
+ ],
891
+ "/layers/glm4_decoder_layer_29/post_mlp_layernorm/vars": [
892
+ "model_00006.weights.h5"
893
+ ],
894
+ "/layers/glm4_decoder_layer_29/post_self_attn_layernorm/vars": [
895
+ "model_00006.weights.h5"
896
+ ],
897
+ "/layers/glm4_decoder_layer_29/attention/key/vars": [
898
+ "model_00006.weights.h5"
899
+ ],
900
+ "/layers/glm4_decoder_layer_29/attention/output_proj/vars": [
901
+ "model_00006.weights.h5"
902
+ ],
903
+ "/layers/glm4_decoder_layer_29/attention/query/vars": [
904
+ "model_00006.weights.h5"
905
+ ],
906
+ "/layers/glm4_decoder_layer_29/attention/value/vars": [
907
+ "model_00006.weights.h5"
908
+ ],
909
+ "/layers/glm4_decoder_layer_30/input_layernorm/vars": [
910
+ "model_00006.weights.h5"
911
+ ],
912
+ "/layers/glm4_decoder_layer_30/mlp/down/vars": [
913
+ "model_00006.weights.h5"
914
+ ],
915
+ "/layers/glm4_decoder_layer_30/mlp/gate_up/vars": [
916
+ "model_00006.weights.h5"
917
+ ],
918
+ "/layers/glm4_decoder_layer_30/post_attention_layernorm/vars": [
919
+ "model_00006.weights.h5"
920
+ ],
921
+ "/layers/glm4_decoder_layer_30/post_mlp_layernorm/vars": [
922
+ "model_00006.weights.h5"
923
+ ],
924
+ "/layers/glm4_decoder_layer_30/post_self_attn_layernorm/vars": [
925
+ "model_00006.weights.h5"
926
+ ],
927
+ "/layers/glm4_decoder_layer_30/attention/key/vars": [
928
+ "model_00006.weights.h5"
929
+ ],
930
+ "/layers/glm4_decoder_layer_30/attention/output_proj/vars": [
931
+ "model_00006.weights.h5"
932
+ ],
933
+ "/layers/glm4_decoder_layer_30/attention/query/vars": [
934
+ "model_00006.weights.h5"
935
+ ],
936
+ "/layers/glm4_decoder_layer_30/attention/value/vars": [
937
+ "model_00006.weights.h5"
938
+ ],
939
+ "/layers/glm4_decoder_layer_31/mlp/gate_up/vars": [
940
+ "model_00006.weights.h5"
941
+ ],
942
+ "/layers/glm4_decoder_layer_31/attention/key/vars": [
943
+ "model_00006.weights.h5"
944
+ ],
945
+ "/layers/glm4_decoder_layer_31/attention/output_proj/vars": [
946
+ "model_00006.weights.h5"
947
+ ],
948
+ "/layers/glm4_decoder_layer_31/attention/query/vars": [
949
+ "model_00006.weights.h5"
950
+ ],
951
+ "/layers/glm4_decoder_layer_31/attention/value/vars": [
952
+ "model_00006.weights.h5"
953
+ ],
954
+ "/layers/glm4_decoder_layer_31/input_layernorm/vars": [
955
+ "model_00006.weights.h5"
956
+ ],
957
+ "/layers/glm4_decoder_layer_31/mlp/down/vars": [
958
+ "model_00006.weights.h5"
959
+ ],
960
+ "/layers/glm4_decoder_layer_31/post_attention_layernorm/vars": [
961
+ "model_00006.weights.h5"
962
+ ],
963
+ "/layers/glm4_decoder_layer_31/post_mlp_layernorm/vars": [
964
+ "model_00006.weights.h5"
965
+ ],
966
+ "/layers/glm4_decoder_layer_31/post_self_attn_layernorm/vars": [
967
+ "model_00006.weights.h5"
968
+ ],
969
+ "/layers/glm4_decoder_layer_32/input_layernorm/vars": [
970
+ "model_00006.weights.h5"
971
+ ],
972
+ "/layers/glm4_decoder_layer_32/mlp/down/vars": [
973
+ "model_00006.weights.h5"
974
+ ],
975
+ "/layers/glm4_decoder_layer_32/mlp/gate_up/vars": [
976
+ "model_00006.weights.h5"
977
+ ],
978
+ "/layers/glm4_decoder_layer_32/post_attention_layernorm/vars": [
979
+ "model_00006.weights.h5"
980
+ ],
981
+ "/layers/glm4_decoder_layer_32/post_mlp_layernorm/vars": [
982
+ "model_00006.weights.h5"
983
+ ],
984
+ "/layers/glm4_decoder_layer_32/post_self_attn_layernorm/vars": [
985
+ "model_00006.weights.h5"
986
+ ],
987
+ "/layers/glm4_decoder_layer_32/attention/key/vars": [
988
+ "model_00006.weights.h5"
989
+ ],
990
+ "/layers/glm4_decoder_layer_32/attention/output_proj/vars": [
991
+ "model_00006.weights.h5"
992
+ ],
993
+ "/layers/glm4_decoder_layer_32/attention/query/vars": [
994
+ "model_00006.weights.h5"
995
+ ],
996
+ "/layers/glm4_decoder_layer_32/attention/value/vars": [
997
+ "model_00006.weights.h5"
998
+ ],
999
+ "/layers/glm4_decoder_layer_33/input_layernorm/vars": [
1000
+ "model_00006.weights.h5"
1001
+ ],
1002
+ "/layers/glm4_decoder_layer_33/mlp/down/vars": [
1003
+ "model_00006.weights.h5"
1004
+ ],
1005
+ "/layers/glm4_decoder_layer_33/mlp/gate_up/vars": [
1006
+ "model_00007.weights.h5"
1007
+ ],
1008
+ "/layers/glm4_decoder_layer_33/post_attention_layernorm/vars": [
1009
+ "model_00007.weights.h5"
1010
+ ],
1011
+ "/layers/glm4_decoder_layer_33/post_mlp_layernorm/vars": [
1012
+ "model_00007.weights.h5"
1013
+ ],
1014
+ "/layers/glm4_decoder_layer_33/post_self_attn_layernorm/vars": [
1015
+ "model_00007.weights.h5"
1016
+ ],
1017
+ "/layers/glm4_decoder_layer_33/attention/key/vars": [
1018
+ "model_00007.weights.h5"
1019
+ ],
1020
+ "/layers/glm4_decoder_layer_33/attention/output_proj/vars": [
1021
+ "model_00007.weights.h5"
1022
+ ],
1023
+ "/layers/glm4_decoder_layer_33/attention/query/vars": [
1024
+ "model_00007.weights.h5"
1025
+ ],
1026
+ "/layers/glm4_decoder_layer_33/attention/value/vars": [
1027
+ "model_00007.weights.h5"
1028
+ ],
1029
+ "/layers/glm4_decoder_layer_34/input_layernorm/vars": [
1030
+ "model_00007.weights.h5"
1031
+ ],
1032
+ "/layers/glm4_decoder_layer_34/mlp/down/vars": [
1033
+ "model_00007.weights.h5"
1034
+ ],
1035
+ "/layers/glm4_decoder_layer_34/mlp/gate_up/vars": [
1036
+ "model_00007.weights.h5"
1037
+ ],
1038
+ "/layers/glm4_decoder_layer_34/post_attention_layernorm/vars": [
1039
+ "model_00007.weights.h5"
1040
+ ],
1041
+ "/layers/glm4_decoder_layer_34/post_mlp_layernorm/vars": [
1042
+ "model_00007.weights.h5"
1043
+ ],
1044
+ "/layers/glm4_decoder_layer_34/post_self_attn_layernorm/vars": [
1045
+ "model_00007.weights.h5"
1046
+ ],
1047
+ "/layers/glm4_decoder_layer_34/attention/key/vars": [
1048
+ "model_00007.weights.h5"
1049
+ ],
1050
+ "/layers/glm4_decoder_layer_34/attention/output_proj/vars": [
1051
+ "model_00007.weights.h5"
1052
+ ],
1053
+ "/layers/glm4_decoder_layer_34/attention/query/vars": [
1054
+ "model_00007.weights.h5"
1055
+ ],
1056
+ "/layers/glm4_decoder_layer_34/attention/value/vars": [
1057
+ "model_00007.weights.h5"
1058
+ ],
1059
+ "/layers/glm4_decoder_layer_35/input_layernorm/vars": [
1060
+ "model_00007.weights.h5"
1061
+ ],
1062
+ "/layers/glm4_decoder_layer_35/mlp/down/vars": [
1063
+ "model_00007.weights.h5"
1064
+ ],
1065
+ "/layers/glm4_decoder_layer_35/mlp/gate_up/vars": [
1066
+ "model_00007.weights.h5"
1067
+ ],
1068
+ "/layers/glm4_decoder_layer_35/post_attention_layernorm/vars": [
1069
+ "model_00007.weights.h5"
1070
+ ],
1071
+ "/layers/glm4_decoder_layer_35/post_mlp_layernorm/vars": [
1072
+ "model_00007.weights.h5"
1073
+ ],
1074
+ "/layers/glm4_decoder_layer_35/post_self_attn_layernorm/vars": [
1075
+ "model_00007.weights.h5"
1076
+ ],
1077
+ "/layers/glm4_decoder_layer_35/attention/key/vars": [
1078
+ "model_00007.weights.h5"
1079
+ ],
1080
+ "/layers/glm4_decoder_layer_35/attention/output_proj/vars": [
1081
+ "model_00007.weights.h5"
1082
+ ],
1083
+ "/layers/glm4_decoder_layer_35/attention/query/vars": [
1084
+ "model_00007.weights.h5"
1085
+ ],
1086
+ "/layers/glm4_decoder_layer_35/attention/value/vars": [
1087
+ "model_00007.weights.h5"
1088
+ ],
1089
+ "/layers/glm4_decoder_layer_36/attention/key/vars": [
1090
+ "model_00007.weights.h5"
1091
+ ],
1092
+ "/layers/glm4_decoder_layer_36/attention/output_proj/vars": [
1093
+ "model_00007.weights.h5"
1094
+ ],
1095
+ "/layers/glm4_decoder_layer_36/attention/query/vars": [
1096
+ "model_00007.weights.h5"
1097
+ ],
1098
+ "/layers/glm4_decoder_layer_36/attention/value/vars": [
1099
+ "model_00007.weights.h5"
1100
+ ],
1101
+ "/layers/glm4_decoder_layer_36/input_layernorm/vars": [
1102
+ "model_00007.weights.h5"
1103
+ ],
1104
+ "/layers/glm4_decoder_layer_36/mlp/down/vars": [
1105
+ "model_00007.weights.h5"
1106
+ ],
1107
+ "/layers/glm4_decoder_layer_36/mlp/gate_up/vars": [
1108
+ "model_00007.weights.h5"
1109
+ ],
1110
+ "/layers/glm4_decoder_layer_36/post_attention_layernorm/vars": [
1111
+ "model_00007.weights.h5"
1112
+ ],
1113
+ "/layers/glm4_decoder_layer_36/post_mlp_layernorm/vars": [
1114
+ "model_00007.weights.h5"
1115
+ ],
1116
+ "/layers/glm4_decoder_layer_36/post_self_attn_layernorm/vars": [
1117
+ "model_00007.weights.h5"
1118
+ ],
1119
+ "/layers/glm4_decoder_layer_37/input_layernorm/vars": [
1120
+ "model_00007.weights.h5"
1121
+ ],
1122
+ "/layers/glm4_decoder_layer_37/mlp/down/vars": [
1123
+ "model_00007.weights.h5"
1124
+ ],
1125
+ "/layers/glm4_decoder_layer_37/mlp/gate_up/vars": [
1126
+ "model_00007.weights.h5"
1127
+ ],
1128
+ "/layers/glm4_decoder_layer_37/post_attention_layernorm/vars": [
1129
+ "model_00007.weights.h5"
1130
+ ],
1131
+ "/layers/glm4_decoder_layer_37/post_mlp_layernorm/vars": [
1132
+ "model_00007.weights.h5"
1133
+ ],
1134
+ "/layers/glm4_decoder_layer_37/post_self_attn_layernorm/vars": [
1135
+ "model_00007.weights.h5"
1136
+ ],
1137
+ "/layers/glm4_decoder_layer_37/attention/key/vars": [
1138
+ "model_00007.weights.h5"
1139
+ ],
1140
+ "/layers/glm4_decoder_layer_37/attention/output_proj/vars": [
1141
+ "model_00007.weights.h5"
1142
+ ],
1143
+ "/layers/glm4_decoder_layer_37/attention/query/vars": [
1144
+ "model_00007.weights.h5"
1145
+ ],
1146
+ "/layers/glm4_decoder_layer_37/attention/value/vars": [
1147
+ "model_00007.weights.h5"
1148
+ ],
1149
+ "/layers/glm4_decoder_layer_38/input_layernorm/vars": [
1150
+ "model_00007.weights.h5"
1151
+ ],
1152
+ "/layers/glm4_decoder_layer_38/mlp/down/vars": [
1153
+ "model_00007.weights.h5"
1154
+ ],
1155
+ "/layers/glm4_decoder_layer_38/mlp/gate_up/vars": [
1156
+ "model_00008.weights.h5"
1157
+ ],
1158
+ "/layers/glm4_decoder_layer_38/post_attention_layernorm/vars": [
1159
+ "model_00008.weights.h5"
1160
+ ],
1161
+ "/layers/glm4_decoder_layer_38/post_mlp_layernorm/vars": [
1162
+ "model_00008.weights.h5"
1163
+ ],
1164
+ "/layers/glm4_decoder_layer_38/post_self_attn_layernorm/vars": [
1165
+ "model_00008.weights.h5"
1166
+ ],
1167
+ "/layers/glm4_decoder_layer_38/attention/key/vars": [
1168
+ "model_00008.weights.h5"
1169
+ ],
1170
+ "/layers/glm4_decoder_layer_38/attention/output_proj/vars": [
1171
+ "model_00008.weights.h5"
1172
+ ],
1173
+ "/layers/glm4_decoder_layer_38/attention/query/vars": [
1174
+ "model_00008.weights.h5"
1175
+ ],
1176
+ "/layers/glm4_decoder_layer_38/attention/value/vars": [
1177
+ "model_00008.weights.h5"
1178
+ ],
1179
+ "/layers/glm4_decoder_layer_39/input_layernorm/vars": [
1180
+ "model_00008.weights.h5"
1181
+ ],
1182
+ "/layers/glm4_decoder_layer_39/mlp/down/vars": [
1183
+ "model_00008.weights.h5"
1184
+ ],
1185
+ "/layers/glm4_decoder_layer_39/mlp/gate_up/vars": [
1186
+ "model_00008.weights.h5"
1187
+ ],
1188
+ "/layers/glm4_decoder_layer_39/post_attention_layernorm/vars": [
1189
+ "model_00008.weights.h5"
1190
+ ],
1191
+ "/layers/glm4_decoder_layer_39/post_mlp_layernorm/vars": [
1192
+ "model_00008.weights.h5"
1193
+ ],
1194
+ "/layers/glm4_decoder_layer_39/post_self_attn_layernorm/vars": [
1195
+ "model_00008.weights.h5"
1196
+ ],
1197
+ "/layers/glm4_decoder_layer_39/attention/key/vars": [
1198
+ "model_00008.weights.h5"
1199
+ ],
1200
+ "/layers/glm4_decoder_layer_39/attention/output_proj/vars": [
1201
+ "model_00008.weights.h5"
1202
+ ],
1203
+ "/layers/glm4_decoder_layer_39/attention/query/vars": [
1204
+ "model_00008.weights.h5"
1205
+ ],
1206
+ "/layers/glm4_decoder_layer_39/attention/value/vars": [
1207
+ "model_00008.weights.h5"
1208
+ ],
1209
+ "/layers/glm4_decoder_layer_40/input_layernorm/vars": [
1210
+ "model_00008.weights.h5"
1211
+ ],
1212
+ "/layers/glm4_decoder_layer_40/mlp/down/vars": [
1213
+ "model_00008.weights.h5"
1214
+ ],
1215
+ "/layers/glm4_decoder_layer_40/mlp/gate_up/vars": [
1216
+ "model_00008.weights.h5"
1217
+ ],
1218
+ "/layers/glm4_decoder_layer_40/post_attention_layernorm/vars": [
1219
+ "model_00008.weights.h5"
1220
+ ],
1221
+ "/layers/glm4_decoder_layer_40/post_mlp_layernorm/vars": [
1222
+ "model_00008.weights.h5"
1223
+ ],
1224
+ "/layers/glm4_decoder_layer_40/post_self_attn_layernorm/vars": [
1225
+ "model_00008.weights.h5"
1226
+ ],
1227
+ "/layers/glm4_decoder_layer_40/attention/key/vars": [
1228
+ "model_00008.weights.h5"
1229
+ ],
1230
+ "/layers/glm4_decoder_layer_40/attention/output_proj/vars": [
1231
+ "model_00008.weights.h5"
1232
+ ],
1233
+ "/layers/glm4_decoder_layer_40/attention/query/vars": [
1234
+ "model_00008.weights.h5"
1235
+ ],
1236
+ "/layers/glm4_decoder_layer_40/attention/value/vars": [
1237
+ "model_00008.weights.h5"
1238
+ ],
1239
+ "/layers/glm4_decoder_layer_41/attention/key/vars": [
1240
+ "model_00008.weights.h5"
1241
+ ],
1242
+ "/layers/glm4_decoder_layer_41/attention/query/vars": [
1243
+ "model_00008.weights.h5"
1244
+ ],
1245
+ "/layers/glm4_decoder_layer_41/attention/value/vars": [
1246
+ "model_00008.weights.h5"
1247
+ ],
1248
+ "/layers/glm4_decoder_layer_41/input_layernorm/vars": [
1249
+ "model_00008.weights.h5"
1250
+ ],
1251
+ "/layers/glm4_decoder_layer_41/mlp/down/vars": [
1252
+ "model_00008.weights.h5"
1253
+ ],
1254
+ "/layers/glm4_decoder_layer_41/mlp/gate_up/vars": [
1255
+ "model_00008.weights.h5"
1256
+ ],
1257
+ "/layers/glm4_decoder_layer_41/post_attention_layernorm/vars": [
1258
+ "model_00008.weights.h5"
1259
+ ],
1260
+ "/layers/glm4_decoder_layer_41/post_mlp_layernorm/vars": [
1261
+ "model_00008.weights.h5"
1262
+ ],
1263
+ "/layers/glm4_decoder_layer_41/post_self_attn_layernorm/vars": [
1264
+ "model_00008.weights.h5"
1265
+ ],
1266
+ "/layers/glm4_decoder_layer_41/attention/output_proj/vars": [
1267
+ "model_00008.weights.h5"
1268
+ ],
1269
+ "/layers/glm4_decoder_layer_42/input_layernorm/vars": [
1270
+ "model_00008.weights.h5"
1271
+ ],
1272
+ "/layers/glm4_decoder_layer_42/mlp/down/vars": [
1273
+ "model_00008.weights.h5"
1274
+ ],
1275
+ "/layers/glm4_decoder_layer_42/mlp/gate_up/vars": [
1276
+ "model_00008.weights.h5"
1277
+ ],
1278
+ "/layers/glm4_decoder_layer_42/post_attention_layernorm/vars": [
1279
+ "model_00008.weights.h5"
1280
+ ],
1281
+ "/layers/glm4_decoder_layer_42/post_mlp_layernorm/vars": [
1282
+ "model_00008.weights.h5"
1283
+ ],
1284
+ "/layers/glm4_decoder_layer_42/post_self_attn_layernorm/vars": [
1285
+ "model_00008.weights.h5"
1286
+ ],
1287
+ "/layers/glm4_decoder_layer_42/attention/key/vars": [
1288
+ "model_00008.weights.h5"
1289
+ ],
1290
+ "/layers/glm4_decoder_layer_42/attention/output_proj/vars": [
1291
+ "model_00008.weights.h5"
1292
+ ],
1293
+ "/layers/glm4_decoder_layer_42/attention/query/vars": [
1294
+ "model_00008.weights.h5"
1295
+ ],
1296
+ "/layers/glm4_decoder_layer_42/attention/value/vars": [
1297
+ "model_00008.weights.h5"
1298
+ ],
1299
+ "/layers/glm4_decoder_layer_43/input_layernorm/vars": [
1300
+ "model_00008.weights.h5"
1301
+ ],
1302
+ "/layers/glm4_decoder_layer_43/mlp/down/vars": [
1303
+ "model_00008.weights.h5"
1304
+ ],
1305
+ "/layers/glm4_decoder_layer_43/mlp/gate_up/vars": [
1306
+ "model_00009.weights.h5"
1307
+ ],
1308
+ "/layers/glm4_decoder_layer_43/post_attention_layernorm/vars": [
1309
+ "model_00009.weights.h5"
1310
+ ],
1311
+ "/layers/glm4_decoder_layer_43/post_mlp_layernorm/vars": [
1312
+ "model_00009.weights.h5"
1313
+ ],
1314
+ "/layers/glm4_decoder_layer_43/post_self_attn_layernorm/vars": [
1315
+ "model_00009.weights.h5"
1316
+ ],
1317
+ "/layers/glm4_decoder_layer_43/attention/key/vars": [
1318
+ "model_00009.weights.h5"
1319
+ ],
1320
+ "/layers/glm4_decoder_layer_43/attention/output_proj/vars": [
1321
+ "model_00009.weights.h5"
1322
+ ],
1323
+ "/layers/glm4_decoder_layer_43/attention/query/vars": [
1324
+ "model_00009.weights.h5"
1325
+ ],
1326
+ "/layers/glm4_decoder_layer_43/attention/value/vars": [
1327
+ "model_00009.weights.h5"
1328
+ ],
1329
+ "/layers/glm4_decoder_layer_44/input_layernorm/vars": [
1330
+ "model_00009.weights.h5"
1331
+ ],
1332
+ "/layers/glm4_decoder_layer_44/mlp/down/vars": [
1333
+ "model_00009.weights.h5"
1334
+ ],
1335
+ "/layers/glm4_decoder_layer_44/mlp/gate_up/vars": [
1336
+ "model_00009.weights.h5"
1337
+ ],
1338
+ "/layers/glm4_decoder_layer_44/post_attention_layernorm/vars": [
1339
+ "model_00009.weights.h5"
1340
+ ],
1341
+ "/layers/glm4_decoder_layer_44/post_mlp_layernorm/vars": [
1342
+ "model_00009.weights.h5"
1343
+ ],
1344
+ "/layers/glm4_decoder_layer_44/post_self_attn_layernorm/vars": [
1345
+ "model_00009.weights.h5"
1346
+ ],
1347
+ "/layers/glm4_decoder_layer_44/attention/key/vars": [
1348
+ "model_00009.weights.h5"
1349
+ ],
1350
+ "/layers/glm4_decoder_layer_44/attention/output_proj/vars": [
1351
+ "model_00009.weights.h5"
1352
+ ],
1353
+ "/layers/glm4_decoder_layer_44/attention/query/vars": [
1354
+ "model_00009.weights.h5"
1355
+ ],
1356
+ "/layers/glm4_decoder_layer_44/attention/value/vars": [
1357
+ "model_00009.weights.h5"
1358
+ ],
1359
+ "/layers/glm4_decoder_layer_45/input_layernorm/vars": [
1360
+ "model_00009.weights.h5"
1361
+ ],
1362
+ "/layers/glm4_decoder_layer_45/mlp/down/vars": [
1363
+ "model_00009.weights.h5"
1364
+ ],
1365
+ "/layers/glm4_decoder_layer_45/mlp/gate_up/vars": [
1366
+ "model_00009.weights.h5"
1367
+ ],
1368
+ "/layers/glm4_decoder_layer_45/post_attention_layernorm/vars": [
1369
+ "model_00009.weights.h5"
1370
+ ],
1371
+ "/layers/glm4_decoder_layer_45/post_mlp_layernorm/vars": [
1372
+ "model_00009.weights.h5"
1373
+ ],
1374
+ "/layers/glm4_decoder_layer_45/post_self_attn_layernorm/vars": [
1375
+ "model_00009.weights.h5"
1376
+ ],
1377
+ "/layers/glm4_decoder_layer_45/attention/key/vars": [
1378
+ "model_00009.weights.h5"
1379
+ ],
1380
+ "/layers/glm4_decoder_layer_45/attention/output_proj/vars": [
1381
+ "model_00009.weights.h5"
1382
+ ],
1383
+ "/layers/glm4_decoder_layer_45/attention/query/vars": [
1384
+ "model_00009.weights.h5"
1385
+ ],
1386
+ "/layers/glm4_decoder_layer_45/attention/value/vars": [
1387
+ "model_00009.weights.h5"
1388
+ ],
1389
+ "/layers/glm4_decoder_layer_46/input_layernorm/vars": [
1390
+ "model_00009.weights.h5"
1391
+ ],
1392
+ "/layers/glm4_decoder_layer_46/mlp/down/vars": [
1393
+ "model_00009.weights.h5"
1394
+ ],
1395
+ "/layers/glm4_decoder_layer_46/mlp/gate_up/vars": [
1396
+ "model_00009.weights.h5"
1397
+ ],
1398
+ "/layers/glm4_decoder_layer_46/post_attention_layernorm/vars": [
1399
+ "model_00009.weights.h5"
1400
+ ],
1401
+ "/layers/glm4_decoder_layer_46/post_mlp_layernorm/vars": [
1402
+ "model_00009.weights.h5"
1403
+ ],
1404
+ "/layers/glm4_decoder_layer_46/post_self_attn_layernorm/vars": [
1405
+ "model_00009.weights.h5"
1406
+ ],
1407
+ "/layers/glm4_decoder_layer_46/attention/key/vars": [
1408
+ "model_00009.weights.h5"
1409
+ ],
1410
+ "/layers/glm4_decoder_layer_46/attention/output_proj/vars": [
1411
+ "model_00009.weights.h5"
1412
+ ],
1413
+ "/layers/glm4_decoder_layer_46/attention/query/vars": [
1414
+ "model_00009.weights.h5"
1415
+ ],
1416
+ "/layers/glm4_decoder_layer_46/attention/value/vars": [
1417
+ "model_00009.weights.h5"
1418
+ ],
1419
+ "/layers/glm4_decoder_layer_47/input_layernorm/vars": [
1420
+ "model_00009.weights.h5"
1421
+ ],
1422
+ "/layers/glm4_decoder_layer_47/mlp/down/vars": [
1423
+ "model_00009.weights.h5"
1424
+ ],
1425
+ "/layers/glm4_decoder_layer_47/mlp/gate_up/vars": [
1426
+ "model_00009.weights.h5"
1427
+ ],
1428
+ "/layers/glm4_decoder_layer_47/post_attention_layernorm/vars": [
1429
+ "model_00009.weights.h5"
1430
+ ],
1431
+ "/layers/glm4_decoder_layer_47/post_mlp_layernorm/vars": [
1432
+ "model_00009.weights.h5"
1433
+ ],
1434
+ "/layers/glm4_decoder_layer_47/post_self_attn_layernorm/vars": [
1435
+ "model_00009.weights.h5"
1436
+ ],
1437
+ "/layers/glm4_decoder_layer_47/attention/key/vars": [
1438
+ "model_00009.weights.h5"
1439
+ ],
1440
+ "/layers/glm4_decoder_layer_47/attention/output_proj/vars": [
1441
+ "model_00009.weights.h5"
1442
+ ],
1443
+ "/layers/glm4_decoder_layer_47/attention/query/vars": [
1444
+ "model_00009.weights.h5"
1445
+ ],
1446
+ "/layers/glm4_decoder_layer_47/attention/value/vars": [
1447
+ "model_00009.weights.h5"
1448
+ ],
1449
+ "/layers/glm4_decoder_layer_48/input_layernorm/vars": [
1450
+ "model_00009.weights.h5"
1451
+ ],
1452
+ "/layers/glm4_decoder_layer_48/mlp/down/vars": [
1453
+ "model_00009.weights.h5"
1454
+ ],
1455
+ "/layers/glm4_decoder_layer_48/mlp/gate_up/vars": [
1456
+ "model_00010.weights.h5"
1457
+ ],
1458
+ "/layers/glm4_decoder_layer_48/post_attention_layernorm/vars": [
1459
+ "model_00010.weights.h5"
1460
+ ],
1461
+ "/layers/glm4_decoder_layer_48/post_mlp_layernorm/vars": [
1462
+ "model_00010.weights.h5"
1463
+ ],
1464
+ "/layers/glm4_decoder_layer_48/post_self_attn_layernorm/vars": [
1465
+ "model_00010.weights.h5"
1466
+ ],
1467
+ "/layers/glm4_decoder_layer_48/attention/key/vars": [
1468
+ "model_00010.weights.h5"
1469
+ ],
1470
+ "/layers/glm4_decoder_layer_48/attention/output_proj/vars": [
1471
+ "model_00010.weights.h5"
1472
+ ],
1473
+ "/layers/glm4_decoder_layer_48/attention/query/vars": [
1474
+ "model_00010.weights.h5"
1475
+ ],
1476
+ "/layers/glm4_decoder_layer_48/attention/value/vars": [
1477
+ "model_00010.weights.h5"
1478
+ ],
1479
+ "/layers/glm4_decoder_layer_49/input_layernorm/vars": [
1480
+ "model_00010.weights.h5"
1481
+ ],
1482
+ "/layers/glm4_decoder_layer_49/mlp/down/vars": [
1483
+ "model_00010.weights.h5"
1484
+ ],
1485
+ "/layers/glm4_decoder_layer_49/mlp/gate_up/vars": [
1486
+ "model_00010.weights.h5"
1487
+ ],
1488
+ "/layers/glm4_decoder_layer_49/post_attention_layernorm/vars": [
1489
+ "model_00010.weights.h5"
1490
+ ],
1491
+ "/layers/glm4_decoder_layer_49/post_mlp_layernorm/vars": [
1492
+ "model_00010.weights.h5"
1493
+ ],
1494
+ "/layers/glm4_decoder_layer_49/post_self_attn_layernorm/vars": [
1495
+ "model_00010.weights.h5"
1496
+ ],
1497
+ "/layers/glm4_decoder_layer_49/attention/key/vars": [
1498
+ "model_00010.weights.h5"
1499
+ ],
1500
+ "/layers/glm4_decoder_layer_49/attention/output_proj/vars": [
1501
+ "model_00010.weights.h5"
1502
+ ],
1503
+ "/layers/glm4_decoder_layer_49/attention/query/vars": [
1504
+ "model_00010.weights.h5"
1505
+ ],
1506
+ "/layers/glm4_decoder_layer_49/attention/value/vars": [
1507
+ "model_00010.weights.h5"
1508
+ ],
1509
+ "/layers/glm4_decoder_layer_50/mlp/gate_up/vars": [
1510
+ "model_00010.weights.h5"
1511
+ ],
1512
+ "/layers/glm4_decoder_layer_50/attention/key/vars": [
1513
+ "model_00010.weights.h5"
1514
+ ],
1515
+ "/layers/glm4_decoder_layer_50/attention/output_proj/vars": [
1516
+ "model_00010.weights.h5"
1517
+ ],
1518
+ "/layers/glm4_decoder_layer_50/attention/query/vars": [
1519
+ "model_00010.weights.h5"
1520
+ ],
1521
+ "/layers/glm4_decoder_layer_50/attention/value/vars": [
1522
+ "model_00010.weights.h5"
1523
+ ],
1524
+ "/layers/glm4_decoder_layer_50/input_layernorm/vars": [
1525
+ "model_00010.weights.h5"
1526
+ ],
1527
+ "/layers/glm4_decoder_layer_50/mlp/down/vars": [
1528
+ "model_00010.weights.h5"
1529
+ ],
1530
+ "/layers/glm4_decoder_layer_50/post_attention_layernorm/vars": [
1531
+ "model_00010.weights.h5"
1532
+ ],
1533
+ "/layers/glm4_decoder_layer_50/post_mlp_layernorm/vars": [
1534
+ "model_00010.weights.h5"
1535
+ ],
1536
+ "/layers/glm4_decoder_layer_50/post_self_attn_layernorm/vars": [
1537
+ "model_00010.weights.h5"
1538
+ ],
1539
+ "/layers/glm4_decoder_layer_51/input_layernorm/vars": [
1540
+ "model_00010.weights.h5"
1541
+ ],
1542
+ "/layers/glm4_decoder_layer_51/mlp/down/vars": [
1543
+ "model_00010.weights.h5"
1544
+ ],
1545
+ "/layers/glm4_decoder_layer_51/mlp/gate_up/vars": [
1546
+ "model_00010.weights.h5"
1547
+ ],
1548
+ "/layers/glm4_decoder_layer_51/post_attention_layernorm/vars": [
1549
+ "model_00010.weights.h5"
1550
+ ],
1551
+ "/layers/glm4_decoder_layer_51/post_mlp_layernorm/vars": [
1552
+ "model_00010.weights.h5"
1553
+ ],
1554
+ "/layers/glm4_decoder_layer_51/post_self_attn_layernorm/vars": [
1555
+ "model_00010.weights.h5"
1556
+ ],
1557
+ "/layers/glm4_decoder_layer_51/attention/key/vars": [
1558
+ "model_00010.weights.h5"
1559
+ ],
1560
+ "/layers/glm4_decoder_layer_51/attention/output_proj/vars": [
1561
+ "model_00010.weights.h5"
1562
+ ],
1563
+ "/layers/glm4_decoder_layer_51/attention/query/vars": [
1564
+ "model_00010.weights.h5"
1565
+ ],
1566
+ "/layers/glm4_decoder_layer_51/attention/value/vars": [
1567
+ "model_00010.weights.h5"
1568
+ ],
1569
+ "/layers/glm4_decoder_layer_52/input_layernorm/vars": [
1570
+ "model_00010.weights.h5"
1571
+ ],
1572
+ "/layers/glm4_decoder_layer_52/mlp/down/vars": [
1573
+ "model_00010.weights.h5"
1574
+ ],
1575
+ "/layers/glm4_decoder_layer_52/mlp/gate_up/vars": [
1576
+ "model_00010.weights.h5"
1577
+ ],
1578
+ "/layers/glm4_decoder_layer_52/post_attention_layernorm/vars": [
1579
+ "model_00010.weights.h5"
1580
+ ],
1581
+ "/layers/glm4_decoder_layer_52/post_mlp_layernorm/vars": [
1582
+ "model_00010.weights.h5"
1583
+ ],
1584
+ "/layers/glm4_decoder_layer_52/post_self_attn_layernorm/vars": [
1585
+ "model_00010.weights.h5"
1586
+ ],
1587
+ "/layers/glm4_decoder_layer_52/attention/key/vars": [
1588
+ "model_00010.weights.h5"
1589
+ ],
1590
+ "/layers/glm4_decoder_layer_52/attention/output_proj/vars": [
1591
+ "model_00010.weights.h5"
1592
+ ],
1593
+ "/layers/glm4_decoder_layer_52/attention/query/vars": [
1594
+ "model_00010.weights.h5"
1595
+ ],
1596
+ "/layers/glm4_decoder_layer_52/attention/value/vars": [
1597
+ "model_00010.weights.h5"
1598
+ ],
1599
+ "/layers/glm4_decoder_layer_53/input_layernorm/vars": [
1600
+ "model_00010.weights.h5"
1601
+ ],
1602
+ "/layers/glm4_decoder_layer_53/mlp/down/vars": [
1603
+ "model_00010.weights.h5"
1604
+ ],
1605
+ "/layers/glm4_decoder_layer_53/mlp/gate_up/vars": [
1606
+ "model_00011.weights.h5"
1607
+ ],
1608
+ "/layers/glm4_decoder_layer_53/post_attention_layernorm/vars": [
1609
+ "model_00011.weights.h5"
1610
+ ],
1611
+ "/layers/glm4_decoder_layer_53/post_mlp_layernorm/vars": [
1612
+ "model_00011.weights.h5"
1613
+ ],
1614
+ "/layers/glm4_decoder_layer_53/post_self_attn_layernorm/vars": [
1615
+ "model_00011.weights.h5"
1616
+ ],
1617
+ "/layers/glm4_decoder_layer_53/attention/key/vars": [
1618
+ "model_00011.weights.h5"
1619
+ ],
1620
+ "/layers/glm4_decoder_layer_53/attention/output_proj/vars": [
1621
+ "model_00011.weights.h5"
1622
+ ],
1623
+ "/layers/glm4_decoder_layer_53/attention/query/vars": [
1624
+ "model_00011.weights.h5"
1625
+ ],
1626
+ "/layers/glm4_decoder_layer_53/attention/value/vars": [
1627
+ "model_00011.weights.h5"
1628
+ ],
1629
+ "/layers/glm4_decoder_layer_54/input_layernorm/vars": [
1630
+ "model_00011.weights.h5"
1631
+ ],
1632
+ "/layers/glm4_decoder_layer_54/mlp/down/vars": [
1633
+ "model_00011.weights.h5"
1634
+ ],
1635
+ "/layers/glm4_decoder_layer_54/mlp/gate_up/vars": [
1636
+ "model_00011.weights.h5"
1637
+ ],
1638
+ "/layers/glm4_decoder_layer_54/post_attention_layernorm/vars": [
1639
+ "model_00011.weights.h5"
1640
+ ],
1641
+ "/layers/glm4_decoder_layer_54/post_mlp_layernorm/vars": [
1642
+ "model_00011.weights.h5"
1643
+ ],
1644
+ "/layers/glm4_decoder_layer_54/post_self_attn_layernorm/vars": [
1645
+ "model_00011.weights.h5"
1646
+ ],
1647
+ "/layers/glm4_decoder_layer_54/attention/key/vars": [
1648
+ "model_00011.weights.h5"
1649
+ ],
1650
+ "/layers/glm4_decoder_layer_54/attention/output_proj/vars": [
1651
+ "model_00011.weights.h5"
1652
+ ],
1653
+ "/layers/glm4_decoder_layer_54/attention/query/vars": [
1654
+ "model_00011.weights.h5"
1655
+ ],
1656
+ "/layers/glm4_decoder_layer_54/attention/value/vars": [
1657
+ "model_00011.weights.h5"
1658
+ ],
1659
+ "/layers/glm4_decoder_layer_55/attention/key/vars": [
1660
+ "model_00011.weights.h5"
1661
+ ],
1662
+ "/layers/glm4_decoder_layer_55/attention/output_proj/vars": [
1663
+ "model_00011.weights.h5"
1664
+ ],
1665
+ "/layers/glm4_decoder_layer_55/attention/query/vars": [
1666
+ "model_00011.weights.h5"
1667
+ ],
1668
+ "/layers/glm4_decoder_layer_55/attention/value/vars": [
1669
+ "model_00011.weights.h5"
1670
+ ],
1671
+ "/layers/glm4_decoder_layer_55/input_layernorm/vars": [
1672
+ "model_00011.weights.h5"
1673
+ ],
1674
+ "/layers/glm4_decoder_layer_55/mlp/down/vars": [
1675
+ "model_00011.weights.h5"
1676
+ ],
1677
+ "/layers/glm4_decoder_layer_55/mlp/gate_up/vars": [
1678
+ "model_00011.weights.h5"
1679
+ ],
1680
+ "/layers/glm4_decoder_layer_55/post_attention_layernorm/vars": [
1681
+ "model_00011.weights.h5"
1682
+ ],
1683
+ "/layers/glm4_decoder_layer_55/post_mlp_layernorm/vars": [
1684
+ "model_00011.weights.h5"
1685
+ ],
1686
+ "/layers/glm4_decoder_layer_55/post_self_attn_layernorm/vars": [
1687
+ "model_00011.weights.h5"
1688
+ ],
1689
+ "/layers/glm4_decoder_layer_56/input_layernorm/vars": [
1690
+ "model_00011.weights.h5"
1691
+ ],
1692
+ "/layers/glm4_decoder_layer_56/mlp/down/vars": [
1693
+ "model_00011.weights.h5"
1694
+ ],
1695
+ "/layers/glm4_decoder_layer_56/mlp/gate_up/vars": [
1696
+ "model_00011.weights.h5"
1697
+ ],
1698
+ "/layers/glm4_decoder_layer_56/post_attention_layernorm/vars": [
1699
+ "model_00011.weights.h5"
1700
+ ],
1701
+ "/layers/glm4_decoder_layer_56/post_mlp_layernorm/vars": [
1702
+ "model_00011.weights.h5"
1703
+ ],
1704
+ "/layers/glm4_decoder_layer_56/post_self_attn_layernorm/vars": [
1705
+ "model_00011.weights.h5"
1706
+ ],
1707
+ "/layers/glm4_decoder_layer_56/attention/key/vars": [
1708
+ "model_00011.weights.h5"
1709
+ ],
1710
+ "/layers/glm4_decoder_layer_56/attention/output_proj/vars": [
1711
+ "model_00011.weights.h5"
1712
+ ],
1713
+ "/layers/glm4_decoder_layer_56/attention/query/vars": [
1714
+ "model_00011.weights.h5"
1715
+ ],
1716
+ "/layers/glm4_decoder_layer_56/attention/value/vars": [
1717
+ "model_00011.weights.h5"
1718
+ ],
1719
+ "/layers/glm4_decoder_layer_57/input_layernorm/vars": [
1720
+ "model_00011.weights.h5"
1721
+ ],
1722
+ "/layers/glm4_decoder_layer_57/mlp/down/vars": [
1723
+ "model_00011.weights.h5"
1724
+ ],
1725
+ "/layers/glm4_decoder_layer_57/mlp/gate_up/vars": [
1726
+ "model_00011.weights.h5"
1727
+ ],
1728
+ "/layers/glm4_decoder_layer_57/post_attention_layernorm/vars": [
1729
+ "model_00011.weights.h5"
1730
+ ],
1731
+ "/layers/glm4_decoder_layer_57/post_mlp_layernorm/vars": [
1732
+ "model_00011.weights.h5"
1733
+ ],
1734
+ "/layers/glm4_decoder_layer_57/post_self_attn_layernorm/vars": [
1735
+ "model_00011.weights.h5"
1736
+ ],
1737
+ "/layers/glm4_decoder_layer_57/attention/key/vars": [
1738
+ "model_00011.weights.h5"
1739
+ ],
1740
+ "/layers/glm4_decoder_layer_57/attention/output_proj/vars": [
1741
+ "model_00011.weights.h5"
1742
+ ],
1743
+ "/layers/glm4_decoder_layer_57/attention/query/vars": [
1744
+ "model_00011.weights.h5"
1745
+ ],
1746
+ "/layers/glm4_decoder_layer_57/attention/value/vars": [
1747
+ "model_00011.weights.h5"
1748
+ ],
1749
+ "/layers/glm4_decoder_layer_58/input_layernorm/vars": [
1750
+ "model_00011.weights.h5"
1751
+ ],
1752
+ "/layers/glm4_decoder_layer_58/mlp/down/vars": [
1753
+ "model_00011.weights.h5"
1754
+ ],
1755
+ "/layers/glm4_decoder_layer_58/mlp/gate_up/vars": [
1756
+ "model_00012.weights.h5"
1757
+ ],
1758
+ "/layers/glm4_decoder_layer_58/post_attention_layernorm/vars": [
1759
+ "model_00012.weights.h5"
1760
+ ],
1761
+ "/layers/glm4_decoder_layer_58/post_mlp_layernorm/vars": [
1762
+ "model_00012.weights.h5"
1763
+ ],
1764
+ "/layers/glm4_decoder_layer_58/post_self_attn_layernorm/vars": [
1765
+ "model_00012.weights.h5"
1766
+ ],
1767
+ "/layers/glm4_decoder_layer_58/attention/key/vars": [
1768
+ "model_00012.weights.h5"
1769
+ ],
1770
+ "/layers/glm4_decoder_layer_58/attention/output_proj/vars": [
1771
+ "model_00012.weights.h5"
1772
+ ],
1773
+ "/layers/glm4_decoder_layer_58/attention/query/vars": [
1774
+ "model_00012.weights.h5"
1775
+ ],
1776
+ "/layers/glm4_decoder_layer_58/attention/value/vars": [
1777
+ "model_00012.weights.h5"
1778
+ ],
1779
+ "/layers/glm4_decoder_layer_59/input_layernorm/vars": [
1780
+ "model_00012.weights.h5"
1781
+ ],
1782
+ "/layers/glm4_decoder_layer_59/mlp/down/vars": [
1783
+ "model_00012.weights.h5"
1784
+ ],
1785
+ "/layers/glm4_decoder_layer_59/mlp/gate_up/vars": [
1786
+ "model_00012.weights.h5"
1787
+ ],
1788
+ "/layers/glm4_decoder_layer_59/post_attention_layernorm/vars": [
1789
+ "model_00012.weights.h5"
1790
+ ],
1791
+ "/layers/glm4_decoder_layer_59/post_mlp_layernorm/vars": [
1792
+ "model_00012.weights.h5"
1793
+ ],
1794
+ "/layers/glm4_decoder_layer_59/post_self_attn_layernorm/vars": [
1795
+ "model_00012.weights.h5"
1796
+ ],
1797
+ "/layers/glm4_decoder_layer_59/attention/key/vars": [
1798
+ "model_00012.weights.h5"
1799
+ ],
1800
+ "/layers/glm4_decoder_layer_59/attention/output_proj/vars": [
1801
+ "model_00012.weights.h5"
1802
+ ],
1803
+ "/layers/glm4_decoder_layer_59/attention/query/vars": [
1804
+ "model_00012.weights.h5"
1805
+ ],
1806
+ "/layers/glm4_decoder_layer_59/attention/value/vars": [
1807
+ "model_00012.weights.h5"
1808
+ ],
1809
+ "/layers/glm4_decoder_layer_60/attention/key/vars": [
1810
+ "model_00012.weights.h5"
1811
+ ],
1812
+ "/layers/glm4_decoder_layer_60/attention/query/vars": [
1813
+ "model_00012.weights.h5"
1814
+ ],
1815
+ "/layers/glm4_decoder_layer_60/attention/value/vars": [
1816
+ "model_00012.weights.h5"
1817
+ ],
1818
+ "/layers/dense/vars": [
1819
+ "model_00012.weights.h5"
1820
+ ],
1821
+ "/layers/glm4_decoder_layer_60/input_layernorm/vars": [
1822
+ "model_00012.weights.h5"
1823
+ ],
1824
+ "/layers/glm4_decoder_layer_60/mlp/down/vars": [
1825
+ "model_00012.weights.h5"
1826
+ ],
1827
+ "/layers/glm4_decoder_layer_60/mlp/gate_up/vars": [
1828
+ "model_00012.weights.h5"
1829
+ ],
1830
+ "/layers/glm4_decoder_layer_60/post_attention_layernorm/vars": [
1831
+ "model_00012.weights.h5"
1832
+ ],
1833
+ "/layers/glm4_decoder_layer_60/post_mlp_layernorm/vars": [
1834
+ "model_00012.weights.h5"
1835
+ ],
1836
+ "/layers/glm4_decoder_layer_60/post_self_attn_layernorm/vars": [
1837
+ "model_00012.weights.h5"
1838
+ ],
1839
+ "/layers/glm4_decoder_layer_60/attention/output_proj/vars": [
1840
+ "model_00012.weights.h5"
1841
+ ],
1842
+ "/layers/glm4rms_norm/vars": [
1843
+ "model_00012.weights.h5"
1844
+ ]
1845
+ }
1846
+ }
model_00000.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b88007c220b957c52e1f33af9cf19ffe87bfcefe1f05e28b4a6b73904bdabf2
3
+ size 5251424192
model_00001.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb8fe99eb1c29e90eace79a03cc2b962e33962d4bf3c90c6934cdb187ec6abac
3
+ size 4955819968
model_00002.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f0bf24ebb35c03e6b1edab7b27aca53b1bd90610328310af00ef5977949ce71
3
+ size 5037610944
model_00003.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e70512579936f7863756f785460ff408b199b000092188d2faeb641e1710a7a4
3
+ size 5037610944
model_00004.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2b6d4bbf7f67673c0e2c1e4b9339bd951f5f25b1bd5c0c91a09321c5b8ccc23
3
+ size 5037610944
model_00005.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbd0e3e7d009d613916b6576c7f199b8ed8203287774b9cca2864fb4c00c4f06
3
+ size 5037610944
model_00006.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8fd15f8ac0fc626a13733855e7e1323db89e427b794031bced7bf181345c1683
3
+ size 5037610944
model_00007.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4972bdf23622b3b832ebd2a8f9abeacdc1251db334475d94a9641a2075916071
3
+ size 5037610944
model_00008.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d23af49396b7cd12aab3b4a605ba41813bc2a29fad1c8db6acef9c3326c3b64d
3
+ size 5037610944
model_00009.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67951b141abedcaf2991ebc66f8abb6bdc2192714bb1cd0873af9202ba4d3316
3
+ size 5037610944
model_00010.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15b0dec65e2b79d3143a5806b0ca0b28d6b570d4a468da81c43bb0280494e14f
3
+ size 5037610944
model_00011.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:020622c01cfe669d4a37aac48303b13bd337c31683f438f9adf34f06260d4907
3
+ size 5037610944
model_00012.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4bb6d7b8abb8b69cd60691c700716be394c5a34c361181e292d5a946808ee0c
3
+ size 4603387840
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76ebeac0d8bd7879ead7b43c16b44981f277e47225de2bd7de9ae1a6cc664a8c
3
+ size 19966496