Schnuckade commited on
Commit
202075c
·
verified ·
1 Parent(s): 2f7114e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +45 -13
README.md CHANGED
@@ -1,21 +1,53 @@
1
- ---
2
- base_model: unsloth/Qwen3.5-4B
3
- tags:
4
- - text-generation-inference
5
- - transformers
6
- - unsloth
7
- - qwen3_5
8
  license: apache-2.0
9
  language:
 
10
  - en
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  ---
12
 
13
- # Uploaded finetuned model
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14
 
15
- - **Developed by:** JallyAI
16
- - **License:** apache-2.0
17
- - **Finetuned from model :** unsloth/Qwen3.5-4B
18
 
19
- This qwen3_5 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
  license: apache-2.0
2
  language:
3
+ - de
4
  - en
5
+ base_model:
6
+ - Qwen/Qwen3.5-4B
7
+ pipeline_tag: text-generation
8
+ library_name: adapter-transformers
9
+ tags:
10
+ - efficient
11
+ - qwen
12
+ - qwen3.5
13
+ - gguf
14
+ - ollama
15
+ - instruction-finetuning
16
+ - nomi
17
+ - lazyloopstudio
18
+ - unsloth
19
+ - nomi2.0
20
  ---
21
 
22
+ <p align="center">
23
+ <img src="https://cdn-uploads.huggingface.co/production/uploads/6921fa6332f7fb129563d495/jemMkfi73Fck611ID18ts.png" width="128">
24
+ </p>
25
+
26
+ # Nomi 2.0
27
+
28
+ ## Introduction
29
+
30
+ Interducing Nomi 2.0, an 4B LLM based on Qwen 3.5 4B, that thinks efficient while keeping great performance.
31
+
32
+ In this training, we aimed to improve Nomi's reasoning. It's base, Qwen 3.5 4B was great, but overthought most of the requests und looped a lot.
33
+ Nomi 2.0 fixes that. Even at more complicated prompts it only thinks for about 10 sec on consumer hardware (~500 Reasoning tokens)
34
+
35
+
36
+ ## 🌟 Key Features & Improvements
37
+
38
+ * **Architecture:** Qwen-3.5-4B (runs on 8 GB VRAM GPUs like the RTX 4060).
39
+ * **Multilingual Support:** Can understand and generate text in German and English.
40
+ * **Efficiency:** Runs with 60+ Tokens/s at Q4 with an Context Window of almost 100k tokens.
41
+
42
+ ## 🧠 Training Details
43
+
44
+ * **Base Model:** `Qwen/Qwen3.5-4B`
45
+ * **Fine-tuning:** SFT (Supervised Fine-Tuning).
46
+ * **Training Tool:** **Unsloth** (for 4-bit optimized training).
47
+
48
+ ---
49
 
 
 
 
50
 
51
+ ## ⚠️ Limitations
52
 
53
+ As a 4B parameter model, Nomi-2.0 is not a replacement for Claude or larger models when it comes to deep world knowledge or complex mathematical reasoning. It is a specialized tool for speed, local privacy, and high-quality document structure.