Files changed (1) hide show
  1. Qwen_Qwen3-4B.json +67 -0
Qwen_Qwen3-4B.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bomFormat": "CycloneDX",
3
+ "specVersion": "1.6",
4
+ "serialNumber": "urn:uuid:f24b7da0-02aa-40b9-97d8-8c76756a806e",
5
+ "version": 1,
6
+ "metadata": {
7
+ "timestamp": "2025-06-05T09:37:34.037173+00:00",
8
+ "component": {
9
+ "type": "machine-learning-model",
10
+ "bom-ref": "Qwen/Qwen3-4B-d504bcdf-8093-598c-937c-206e9c62b7df",
11
+ "name": "Qwen/Qwen3-4B",
12
+ "externalReferences": [
13
+ {
14
+ "url": "https://huggingface.co/Qwen/Qwen3-4B",
15
+ "type": "documentation"
16
+ }
17
+ ],
18
+ "modelCard": {
19
+ "modelParameters": {
20
+ "task": "text-generation",
21
+ "architectureFamily": "qwen3",
22
+ "modelArchitecture": "Qwen3ForCausalLM"
23
+ },
24
+ "properties": [
25
+ {
26
+ "name": "library_name",
27
+ "value": "transformers"
28
+ },
29
+ {
30
+ "name": "base_model",
31
+ "value": "Qwen/Qwen3-4B-Base"
32
+ }
33
+ ]
34
+ },
35
+ "authors": [
36
+ {
37
+ "name": "Qwen"
38
+ }
39
+ ],
40
+ "licenses": [
41
+ {
42
+ "license": {
43
+ "id": "Apache-2.0",
44
+ "url": "https://spdx.org/licenses/Apache-2.0.html"
45
+ }
46
+ }
47
+ ],
48
+ "description": "**Qwen3-4B** has the following features:- Type: Causal Language Models- Training Stage: Pretraining & Post-training- Number of Parameters: 4.0B- Number of Paramaters (Non-Embedding): 3.6B- Number of Layers: 36- Number of Attention Heads (GQA): 32 for Q and 8 for KV- Context Length: 32,768 natively and [131,072 tokens with YaRN](#processing-long-texts).For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).> [!TIP]> If you encounter significant endless repetitions, please refer to the [Best Practices](#best-practices) section for optimal sampling parameters, and set the ``presence_penalty`` to 1.5.",
49
+ "tags": [
50
+ "transformers",
51
+ "safetensors",
52
+ "qwen3",
53
+ "text-generation",
54
+ "conversational",
55
+ "arxiv:2309.00071",
56
+ "arxiv:2505.09388",
57
+ "base_model:Qwen/Qwen3-4B-Base",
58
+ "base_model:finetune:Qwen/Qwen3-4B-Base",
59
+ "license:apache-2.0",
60
+ "autotrain_compatible",
61
+ "text-generation-inference",
62
+ "endpoints_compatible",
63
+ "region:us"
64
+ ]
65
+ }
66
+ }
67
+ }