Hugging Face's logo

keras
/
qwen2.5_coder_instruct_1.5b

Text Generation
KerasHub
Model card Files
xet
 Community
prasadsachin commited on
Commit
b27a2b3
·
verified ·
1 Parent(s): a44da76

Upload folder using huggingface_hub

Browse files
Files changed (9) hide show
  1. README.md +26 -0
  2. assets/tokenizer/merges.txt +0 -0
  3. assets/tokenizer/vocabulary.json +0 -0
  4. config.json +30 -0
  5. metadata.json +9 -0
  6. model.weights.h5 +3 -0
  7. preprocessor.json +65 -0
  8. task.json +103 -0
  9. tokenizer.json +44 -0
README.md ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: keras-hub
3
+ pipeline_tag: text-generation
4
+ ---
5
+ This is a [`Qwen` model](https://keras.io/api/keras_hub/models/qwen) uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
6
+ This model is related to a `CausalLM` task.
7
+
8
+ Model config:
9
+ * **name:** qwen_backbone
10
+ * **trainable:** True
11
+ * **dtype:** {'module': 'keras', 'class_name': 'DTypePolicy', 'config': {'name': 'float32'}, 'registered_name': None}
12
+ * **vocabulary_size:** 151936
13
+ * **num_layers:** 28
14
+ * **num_query_heads:** 12
15
+ * **hidden_dim:** 1536
16
+ * **intermediate_dim:** 8960
17
+ * **rope_max_wavelength:** 1000000.0
18
+ * **rope_scaling_factor:** 1.0
19
+ * **num_key_value_heads:** 2
20
+ * **layer_norm_epsilon:** 1e-06
21
+ * **dropout:** 0
22
+ * **tie_word_embeddings:** True
23
+ * **use_sliding_window_attention:** False
24
+ * **sliding_window_size:** 32768
25
+
26
+ This model card has been generated automatically and should be completed by the model author. See [Model Cards documentation](https://huggingface.co/docs/hub/model-cards) for more information.
assets/tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
assets/tokenizer/vocabulary.json ADDED
The diff for this file is too large to render. See raw diff
 
config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "module": "keras_hub.src.models.qwen.qwen_backbone",
3
+ "class_name": "QwenBackbone",
4
+ "config": {
5
+ "name": "qwen_backbone",
6
+ "trainable": true,
7
+ "dtype": {
8
+ "module": "keras",
9
+ "class_name": "DTypePolicy",
10
+ "config": {
11
+ "name": "float32"
12
+ },
13
+ "registered_name": null
14
+ },
15
+ "vocabulary_size": 151936,
16
+ "num_layers": 28,
17
+ "num_query_heads": 12,
18
+ "hidden_dim": 1536,
19
+ "intermediate_dim": 8960,
20
+ "rope_max_wavelength": 1000000.0,
21
+ "rope_scaling_factor": 1.0,
22
+ "num_key_value_heads": 2,
23
+ "layer_norm_epsilon": 1e-06,
24
+ "dropout": 0,
25
+ "tie_word_embeddings": true,
26
+ "use_sliding_window_attention": false,
27
+ "sliding_window_size": 32768
28
+ },
29
+ "registered_name": "keras_hub>QwenBackbone"
30
+ }
metadata.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "keras_version": "3.12.0",
3
+ "keras_hub_version": "0.26.0.dev0",
4
+ "parameter_count": 1543714304,
5
+ "date_saved": "2026-01-28@06:44:32",
6
+ "tasks": [
7
+ "CausalLM"
8
+ ]
9
+ }
model.weights.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e30e5af3ef287e8db168b1a751171a4fa3c35912e170cfd08e260d4b7f72d815
3
+ size 6175843232
preprocessor.json ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "module": "keras_hub.src.models.qwen.qwen_causal_lm_preprocessor",
3
+ "class_name": "QwenCausalLMPreprocessor",
4
+ "config": {
5
+ "name": "qwen_causal_lm_preprocessor_2",
6
+ "trainable": true,
7
+ "dtype": {
8
+ "module": "keras",
9
+ "class_name": "DTypePolicy",
10
+ "config": {
11
+ "name": "float32"
12
+ },
13
+ "registered_name": null
14
+ },
15
+ "tokenizer": {
16
+ "module": "keras_hub.src.models.qwen.qwen_tokenizer",
17
+ "class_name": "QwenTokenizer",
18
+ "config": {
19
+ "name": "qwen_tokenizer",
20
+ "trainable": true,
21
+ "dtype": {
22
+ "module": "keras",
23
+ "class_name": "DTypePolicy",
24
+ "config": {
25
+ "name": "int32"
26
+ },
27
+ "registered_name": null
28
+ },
29
+ "config_file": "tokenizer.json",
30
+ "sequence_length": null,
31
+ "add_prefix_space": false,
32
+ "unsplittable_tokens": [
33
+ "<tool_call>",
34
+ "<|vision_end|>",
35
+ "<|quad_end|>",
36
+ "<|video_pad|>",
37
+ "<|repo_name|>",
38
+ "</tool_call>",
39
+ "<|file_sep|>",
40
+ "<|fim_prefix|>",
41
+ "<|box_end|>",
42
+ "<|object_ref_start|>",
43
+ "<|vision_start|>",
44
+ "<|box_start|>",
45
+ "<|fim_suffix|>",
46
+ "<|image_pad|>",
47
+ "<|im_start|>",
48
+ "<|quad_start|>",
49
+ "<|vision_pad|>",
50
+ "<|fim_pad|>",
51
+ "<|fim_middle|>",
52
+ "<|im_end|>",
53
+ "<|object_ref_end|>",
54
+ "<|endoftext|>"
55
+ ]
56
+ },
57
+ "registered_name": "keras_hub>QwenTokenizer"
58
+ },
59
+ "config_file": "preprocessor.json",
60
+ "sequence_length": 1024,
61
+ "add_start_token": true,
62
+ "add_end_token": true
63
+ },
64
+ "registered_name": "keras_hub>QwenCausalLMPreprocessor"
65
+ }
task.json ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "module": "keras_hub.src.models.qwen.qwen_causal_lm",
3
+ "class_name": "QwenCausalLM",
4
+ "config": {
5
+ "backbone": {
6
+ "module": "keras_hub.src.models.qwen.qwen_backbone",
7
+ "class_name": "QwenBackbone",
8
+ "config": {
9
+ "name": "qwen_backbone",
10
+ "trainable": true,
11
+ "dtype": {
12
+ "module": "keras",
13
+ "class_name": "DTypePolicy",
14
+ "config": {
15
+ "name": "float32"
16
+ },
17
+ "registered_name": null
18
+ },
19
+ "vocabulary_size": 151936,
20
+ "num_layers": 28,
21
+ "num_query_heads": 12,
22
+ "hidden_dim": 1536,
23
+ "intermediate_dim": 8960,
24
+ "rope_max_wavelength": 1000000.0,
25
+ "rope_scaling_factor": 1.0,
26
+ "num_key_value_heads": 2,
27
+ "layer_norm_epsilon": 1e-06,
28
+ "dropout": 0,
29
+ "tie_word_embeddings": true,
30
+ "use_sliding_window_attention": false,
31
+ "sliding_window_size": 32768
32
+ },
33
+ "registered_name": "keras_hub>QwenBackbone"
34
+ },
35
+ "preprocessor": {
36
+ "module": "keras_hub.src.models.qwen.qwen_causal_lm_preprocessor",
37
+ "class_name": "QwenCausalLMPreprocessor",
38
+ "config": {
39
+ "name": "qwen_causal_lm_preprocessor_2",
40
+ "trainable": true,
41
+ "dtype": {
42
+ "module": "keras",
43
+ "class_name": "DTypePolicy",
44
+ "config": {
45
+ "name": "float32"
46
+ },
47
+ "registered_name": null
48
+ },
49
+ "tokenizer": {
50
+ "module": "keras_hub.src.models.qwen.qwen_tokenizer",
51
+ "class_name": "QwenTokenizer",
52
+ "config": {
53
+ "name": "qwen_tokenizer",
54
+ "trainable": true,
55
+ "dtype": {
56
+ "module": "keras",
57
+ "class_name": "DTypePolicy",
58
+ "config": {
59
+ "name": "int32"
60
+ },
61
+ "registered_name": null
62
+ },
63
+ "config_file": "tokenizer.json",
64
+ "sequence_length": null,
65
+ "add_prefix_space": false,
66
+ "unsplittable_tokens": [
67
+ "<tool_call>",
68
+ "<|vision_end|>",
69
+ "<|quad_end|>",
70
+ "<|video_pad|>",
71
+ "<|repo_name|>",
72
+ "</tool_call>",
73
+ "<|file_sep|>",
74
+ "<|fim_prefix|>",
75
+ "<|box_end|>",
76
+ "<|object_ref_start|>",
77
+ "<|vision_start|>",
78
+ "<|box_start|>",
79
+ "<|fim_suffix|>",
80
+ "<|image_pad|>",
81
+ "<|im_start|>",
82
+ "<|quad_start|>",
83
+ "<|vision_pad|>",
84
+ "<|fim_pad|>",
85
+ "<|fim_middle|>",
86
+ "<|im_end|>",
87
+ "<|object_ref_end|>",
88
+ "<|endoftext|>"
89
+ ]
90
+ },
91
+ "registered_name": "keras_hub>QwenTokenizer"
92
+ },
93
+ "config_file": "preprocessor.json",
94
+ "sequence_length": 1024,
95
+ "add_start_token": true,
96
+ "add_end_token": true
97
+ },
98
+ "registered_name": "keras_hub>QwenCausalLMPreprocessor"
99
+ },
100
+ "name": "qwen_causal_lm"
101
+ },
102
+ "registered_name": "keras_hub>QwenCausalLM"
103
+ }
tokenizer.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "module": "keras_hub.src.models.qwen.qwen_tokenizer",
3
+ "class_name": "QwenTokenizer",
4
+ "config": {
5
+ "name": "qwen_tokenizer",
6
+ "trainable": true,
7
+ "dtype": {
8
+ "module": "keras",
9
+ "class_name": "DTypePolicy",
10
+ "config": {
11
+ "name": "int32"
12
+ },
13
+ "registered_name": null
14
+ },
15
+ "config_file": "tokenizer.json",
16
+ "sequence_length": null,
17
+ "add_prefix_space": false,
18
+ "unsplittable_tokens": [
19
+ "<tool_call>",
20
+ "<|vision_end|>",
21
+ "<|quad_end|>",
22
+ "<|video_pad|>",
23
+ "<|repo_name|>",
24
+ "</tool_call>",
25
+ "<|file_sep|>",
26
+ "<|fim_prefix|>",
27
+ "<|box_end|>",
28
+ "<|object_ref_start|>",
29
+ "<|vision_start|>",
30
+ "<|box_start|>",
31
+ "<|fim_suffix|>",
32
+ "<|image_pad|>",
33
+ "<|im_start|>",
34
+ "<|quad_start|>",
35
+ "<|vision_pad|>",
36
+ "<|fim_pad|>",
37
+ "<|fim_middle|>",
38
+ "<|im_end|>",
39
+ "<|object_ref_end|>",
40
+ "<|endoftext|>"
41
+ ]
42
+ },
43
+ "registered_name": "keras_hub>QwenTokenizer"
44
+ }