File size: 2,011 Bytes
4c120d6
66b95a5
21f905d
 
 
242a3fc
 
21f905d
242a3fc
21f905d
 
ce1d661
242a3fc
 
 
 
 
 
 
16c7b3f
4c120d6
e3435fb
66c7212
e3435fb
66c7212
e3435fb
16c7b3f
 
242a3fc
e3435fb
242a3fc
 
 
 
 
 
 
e3435fb
16c7b3f
 
242a3fc
66b95a5
242a3fc
21f905d
66b95a5
66c7212
 
66b95a5
21f905d
 
242a3fc
21f905d
2b7e589
 
66c7212
16c7b3f
 
 
66b95a5
16c7b3f
66b95a5
16c7b3f
e3435fb
 
242a3fc
 
16c7b3f
 
e3435fb
242a3fc
e3435fb
16c7b3f
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
---
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
library_name: transformers
license: mit
language:
  - en
  - pt
metrics:
  - accuracy
pipeline_tag: text-generation
tags:
  - hf-inference
  - education
  - logic
  - math
  - low-resource
  - transformers
  - open-source
  - causal-lm
  - lxcorp
---

# lambda-1v-1b — Lightweight Math & Logic Reasoning Model

**lambda-1v-1b** is a compact, fine-tuned language model built on top of `TinyLlama-1.1B-Chat-v1.0`, designed for educational reasoning tasks in both Portuguese and English. It focuses on logic, number theory, and mathematics, delivering fast performance with minimal computational requirements.

---

## Model Architecture

- **Base Model**: TinyLlama-1.1B-Chat
- **Fine-Tuning Strategy**: LoRA (applied to `q_proj` and `v_proj`)
- **Quantization**: 8-bit (NF4 via `bnb_config`)
- **Dataset**: [`HuggingFaceH4/MATH`](https://huggingface.co/datasets/HuggingFaceH4/MATH) — subset: `number_theory`
- **Max Tokens per Sample**: 512
- **Batch Size**: 20 per device
- **Epochs**: 3

---

## Example Usage (Python)

```python
from transformers import AutoTokenizer, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("lxcorp/lambda-1v-1b")
tokenizer = AutoTokenizer.from_pretrained("lxcorp/lambda-1v-1b")

input_text = "Problema: Prove que 17 é um número primo."
inputs = tokenizer(input_text, return_tensors="pt")

output = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(output[0], skip_special_tokens=True))

```
---

About λχ Corp.

λχ Corp. is an indie tech corporation founded by Marius Jabami in Angola, focused on AI-driven educational tools, robotics, and lightweight software solutions. The lambdAI model is the first release in a planned series of educational LLMs optimized for reasoning, logic, and low-resource deployment.

Stay updated on the project at lxcorp.ai and huggingface.co/lxcorp.


---

Developed with care by Marius Jabami — Powered by ambition, faith, and open source.


---

---