rishiraj commited on
Commit
da536ba
·
verified ·
1 Parent(s): 77bfe94

Smolify: Intelligence Distilled.

Browse files
Files changed (1) hide show
  1. README.md +57 -12
README.md CHANGED
@@ -1,21 +1,66 @@
1
  ---
2
- base_model: unsloth/gemma-3-270m-it
3
- tags:
4
- - text-generation-inference
5
- - transformers
6
- - unsloth
7
- - gemma3_text
8
  license: apache-2.0
9
  language:
10
  - en
 
 
 
 
 
 
 
 
 
 
 
11
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
 
13
- # Uploaded finetuned model
 
 
 
 
 
 
 
 
14
 
15
- - **Developed by:** smolify
16
- - **License:** apache-2.0
17
- - **Finetuned from model :** unsloth/gemma-3-270m-it
 
 
 
 
 
18
 
19
- This gemma3_text model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
 
 
20
 
21
- [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
 
1
  ---
 
 
 
 
 
 
2
  license: apache-2.0
3
  language:
4
  - en
5
+ tags:
6
+ - text-generation-inference
7
+ - transformers
8
+ - smolify
9
+ - dslm
10
+ pipeline_tag: text-generation
11
+ inference:
12
+ parameters:
13
+ temperature: 1
14
+ top_p: 0.95
15
+ top_k: 64
16
  ---
17
+ # 🤏 smolified-contract-intelligence-engine
18
+
19
+ > **Intelligence, Distilled.**
20
+
21
+ This is a **Domain Specific Language Model (DSLM)** generated by the **Smolify Foundry**.
22
+
23
+ It has been synthetically distilled from SOTA reasoning engines into a high-efficiency architecture, optimized for deployment on edge hardware (CPU/NPU) or low-VRAM environments.
24
+
25
+ ## 📦 Asset Details
26
+ - **Origin:** Smolify Foundry (Job ID: `05086435`)
27
+ - **Architecture:** DSLM-Micro (270M Parameter Class)
28
+ - **Training Method:** Proprietary Neural Distillation
29
+ - **Optimization:** 4-bit Quantized / FP16 Mixed
30
+ - **Dataset:** [Link to Dataset](https://huggingface.co/datasets/smolify/smolified-contract-intelligence-engine)
31
+
32
+ ## 🚀 Usage (Inference)
33
+ This model is compatible with standard inference backends like vLLM.
34
+
35
+ ```python
36
+ # Example: Running your Sovereign Model
37
+ from transformers import AutoModelForCausalLM, AutoTokenizer
38
+
39
+ model_id = "smolify/smolified-contract-intelligence-engine"
40
+ tokenizer = AutoTokenizer.from_pretrained(model_id)
41
+ model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
42
 
43
+ messages = [
44
+ {'role': 'system', 'content': '''You are a privacy-first contract intelligence system (LexSovereign). Extract structured legal entities and obligations from unstructured agreements. Identify contracting parties, agreement dates, monetary values, obligations, and governing law clauses. Operate with enterprise-grade precision and support multilingual legal text.'''},
45
+ {'role': 'user', 'content': '''Software License Agreement made on July 1, 2024, by and between ByteWorks Innovations Ltd. ("Licensor"), based in Dublin, Ireland, and Alpha-Tech Solutions AG ("Licensee"), headquartered in Zurich, Switzerland. The Licensor grants Licensee a non-exclusive, perpetual license for 'Quantum ERP' software for a one-time fee of EUR 1,200,000. Maintenance and support services will be provided for 3 years at EUR 50,000 per year. Licensee must ensure all users comply with usage restrictions. Irish law governs this agreement. Any disputes under Irish law shall be heard in Dublin courts.'''}
46
+ ]
47
+ text = tokenizer.apply_chat_template(
48
+ messages,
49
+ tokenize = False,
50
+ add_generation_prompt = True,
51
+ ).removeprefix('<bos>')
52
 
53
+ from transformers import TextStreamer
54
+ _ = model.generate(
55
+ **tokenizer(text, return_tensors = "pt").to("cuda"),
56
+ max_new_tokens = 1000,
57
+ temperature = 1, top_p = 0.95, top_k = 64,
58
+ streamer = TextStreamer(tokenizer, skip_prompt = True),
59
+ )
60
+ ```
61
 
62
+ ## ⚖️ License & Ownership
63
+ This model weights are a sovereign asset owned by **smolify**.
64
+ Generated via [Smolify.ai](https://smolify.ai).
65
 
66
+ [<img src="https://smolify.ai/smolify.gif" width="100"/>](https://smolify.ai)