tgetsov commited on
Commit
65ccee3
Β·
verified Β·
1 Parent(s): 1197702

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +53 -4
README.md CHANGED
@@ -1,10 +1,59 @@
1
  ---
2
- title: README
3
- emoji: 🐨
4
  colorFrom: blue
5
- colorTo: pink
6
  sdk: static
7
  pinned: false
8
  ---
9
 
10
- Edit this `README.md` markdown file to author your organization card.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ title: MainStack
3
+ emoji: πŸ› οΈ
4
  colorFrom: blue
5
+ colorTo: indigo
6
  sdk: static
7
  pinned: false
8
  ---
9
 
10
+ # MainStack
11
+
12
+ **Specializing in ServiceNow Agentic Delivery.**
13
+
14
+ MainStack builds AI systems and open models for the full ServiceNow delivery
15
+ lifecycle β€” from business analysis and solution design through user stories,
16
+ implementation, testing, and validation.
17
+
18
+ ## πŸš€ Featured release
19
+
20
+ ### [marvy-14B](https://huggingface.co/MainStack/marvy-14B) β€” the first open fine-tuned model for full-cycle ServiceNow delivery
21
+
22
+ `marvy-14B` is a fine-tune of [Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct)
23
+ (Apache-2.0) specialized for end-to-end ServiceNow delivery work. It was trained
24
+ on a curated, fully anonymized corpus of real engagement artifacts spanning the
25
+ delivery chain practiced by our agents:
26
+
27
+ > business-analyst β†’ solutions-architect β†’ technical-consultant β†’ validation-architect
28
+
29
+ It produces implementation-grade first drafts of:
30
+
31
+ | | |
32
+ |---|---|
33
+ | πŸ“‹ Business analyses | πŸ“ Solution Design Documents (SDD) |
34
+ | βœ… Requirements & acceptance criteria | πŸ“ User stories |
35
+ | πŸ‘₯ Stakeholder maps | πŸ”§ Implementation plans (tables, plugins, sys_ids) |
36
+ | πŸ—‚οΈ Systems inventories | πŸ§ͺ Test cases |
37
+ | πŸ” Validation & gap analysis | πŸ”— End-to-end delivery chains (story β†’ build β†’ test) |
38
+
39
+ **Available formats:**
40
+
41
+ - [`MainStack/marvy-14B`](https://huggingface.co/MainStack/marvy-14B) β€” merged FP16 (Transformers / vLLM / MLX)
42
+ - [`MainStack/marvy-14B-lora`](https://huggingface.co/MainStack/marvy-14B-lora) β€” LoRA adapter (~175 MB)
43
+ - [`MainStack/marvy-14B-GGUF`](https://huggingface.co/MainStack/marvy-14B-GGUF) β€” Q4_K_M & Q8_0 quants for Ollama / llama.cpp / LM Studio
44
+
45
+ ```bash
46
+ ollama run hf.co/MainStack/marvy-14B-GGUF:Q4_K_M
47
+ ```
48
+
49
+ ## πŸ”’ Built responsibly
50
+
51
+ Our training corpus is rigorously anonymized: customer/partner names aliased,
52
+ emails/hostnames/IPs scrubbed, credential files excluded, and a leakage scanner
53
+ that asserts **zero** residual PII in training content.
54
+
55
+ ## 🌐 About
56
+
57
+ MainStack β€” [mainstack.co.uk](https://www.mainstack.co.uk/) Β· [LinkedIn](https://linkedin.com/company/mainstack-it)
58
+
59
+ *Released under Apache-2.0. Built with Qwen.*