yasserrmd commited on
Commit
26aed62
·
verified ·
1 Parent(s): 58b225e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +64 -21
README.md CHANGED
@@ -8,38 +8,81 @@ tags:
8
  - merge
9
 
10
  ---
11
- # merged-medical-reasoning
12
 
13
- This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
14
 
15
- ## Merge Details
16
- ### Merge Method
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
17
 
18
- This model was merged using the [SLERP](https://en.wikipedia.org/wiki/Slerp) merge method.
19
 
20
- ### Models Merged
21
 
22
- The following models were included in the merge:
23
- * [ertghiu256/qwen3-4b-code-reasoning](https://huggingface.co/ertghiu256/qwen3-4b-code-reasoning)
24
- * [Menlo/Jan-nano](https://huggingface.co/Menlo/Jan-nano)
25
 
26
- ### Configuration
 
 
 
 
 
27
 
28
- The following YAML configuration was used to produce this model:
29
 
30
- ```yaml
31
 
32
- merge_method: slerp
33
- base_model: Menlo/Jan-nano
34
 
35
- models:
36
- - model: Menlo/Jan-nano
37
- - model: ertghiu256/qwen3-4b-code-reasoning
38
 
39
- parameters:
40
- t: 0.4 # 70% base (MedScholar), 30% Nemotron reasoning
41
 
42
- dtype: float16
43
- tokenizer_source: Menlo/Jan-nano
44
 
45
  ```
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
  - merge
9
 
10
  ---
 
11
 
 
12
 
13
+ # 🧠 AgenticCoder‑4B
14
+
15
+ **AgenticCoder‑4B** is a compact 4B parameter language model designed for autonomous agent workflows and intelligent code reasoning. It merges the planning and tool-use strengths of `Jan-nano` with the coding and logic capabilities of `Qwen3‑4B‑Code‑Reasoning`, creating a balanced model ideal for real-world assistant scenarios, research agents, and smart development tools.
16
+
17
+ ---
18
+
19
+ ## ✨ Key Features
20
+
21
+ - 🔁 **Agentic Planning & MCP Alignment**
22
+ Trained on datasets and architectures optimized for multi-step reasoning, task decomposition, and memory–contextual workflows.
23
+
24
+ - 💻 **Code Understanding & Reasoning**
25
+ Strong capabilities in Python code generation, script explanation, optimization, and multi-turn task development.
26
+
27
+ - 🧰 **Tool Use Simulation**
28
+ Handles realistic tool interaction prompts such as CSV analysis, OCR, and file parsing in code.
29
+
30
+ - 📦 **Compact & Efficient (4B)**
31
+ Lightweight enough for cost-efficient deployment, edge device integration, and fine-tuning.
32
+
33
+ ---
34
+
35
+ ## 🛠️ Merge Details
36
+
37
+ - **Merge Method:** SLERP (`t = 0.4`)
38
+ - **Base Model:** [`Menlo/Jan-nano`](https://huggingface.co/Menlo/Jan-nano)
39
+ - **Merged With:** [`ertghiu256/qwen3-4b-code-reasoning`](https://huggingface.co/ertghiu256/qwen3-4b-code-reasoning)
40
+ - **Precision:** `float16`
41
+ - **Tokenizer Source:** `Menlo/Jan-nano`
42
+
43
 
 
44
 
 
45
 
46
+ ---
47
+
48
+ ## 📎 Example Use Cases
49
 
50
+ ```text
51
+ ✅ "Design a 3-week beginner Python curriculum including AI tools."
52
+ ✅ "Write a Python function to recursively scan JSON for a key, without using recursion."
53
+ ✅ "Read a folder of images and extract text using OCR, save to files."
54
+ ✅ "Summarize trends in a sales CSV and visualize monthly performance."
55
+ ````
56
 
57
+ ---
58
 
59
+ ## 📁 License & Use
60
 
61
+ This model is provided for research and development use under the terms of the base models’ respective licenses. Please ensure compliance before commercial usage.
 
62
 
63
+ ---
 
 
64
 
65
+ ## 🧬 Citation
 
66
 
67
+ If you use this model, consider citing it as:
 
68
 
69
  ```
70
+ @misc{agenticcoder4b2025,
71
+ title={AgenticCoder-4B: A Compact Agent + Code Reasoning Model},
72
+ author={Yasser, M.},
73
+ year={2025},
74
+ url={https://huggingface.co/your-username/AgenticCoder-4B}
75
+ }
76
+ ```
77
+
78
+ ---
79
+
80
+ ## 🤝 Acknowledgements
81
+
82
+ * [Menlo/Jan-nano](https://huggingface.co/Menlo/Jan-nano) by Menlo Systems
83
+ * [Qwen3‑4B‑Code‑Reasoning](https://huggingface.co/ertghiu256/qwen3-4b-code-reasoning) by ertghiu256
84
+ * MergeKit, SLERP, Hugging Face
85
+
86
+ ---
87
+
88
+