Hugging Face's logo

tiiuae
/
Falcon3-10B-Instruct-GGUF

Text Generation
Transformers
GGUF
falcon3
Inference Endpoints
conversational
Model card Files Community
2
Falcon3-10B-Instruct-GGUF
File size: 4,152 Bytes 
c78228d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
 

---
language:
- en
- fr
- es
- pt
base_model:
- tiiuae/Falcon3-10B-Instruct
pipeline_tag: text-generation
library_name: transformers
tags:
- falcon3

---

<div align="center">
    <img src="https://huggingface.co/datasets/tiiuae/documentation-images/resolve/main/general/falco3-logo.png" alt="drawing" width="500"/>
</div>

# Falcon3-10B-Instruct-GGUF


**Falcon3** family of Open Foundation Models is a set of pretrained and instruct LLMs ranging from 1B to 10B parameters.

**Falcon3-10B-Instruct** achieves state-of-the-art results (at release's time) on reasoning, language understanding, instruction following, code and mathematics tasks.
Falcon3-10B-Instruct supports 4 languages (English, French, Spanish, Portuguese) and a context length of up to 32K.

This repository contains the GGUFs instruction-tuned 10B Falcon3 model.

## Model Details
- Architecture
  - Transformer-based causal decoder-only architecture
  - 40 decoder blocks
  - Grouped Query Attention (GQA) for faster inference: 12 query heads and 4 key-value heads
  - Wider head dimension: 256
  - High RoPE value to support long context understanding: 1000042
  - Uses SwiGLu and RMSNorm
  - 32K context length
  - 131K vocab size
- Depth up-scaled from **Falcon3-7B-Base** with 2 Teratokens of datasets comprising of web, code, STEM, high quality and mutlilingual data using 1024 H100 GPU chips
- Posttrained on 1.2 million samples of STEM, conversational, code, safety and function call data
- Supports EN, FR, ES, PT
- Developed by [Technology Innovation Institute](https://www.tii.ae)
- License: TII Falcon-LLM License 2.0
- Model Release Date: December 2024
- Quantization: q2_K, q3_K_M, q4_0, q4_K_M, q5_0, q5_K_M, q6_K, q8_0

  
## Getting started

### 1. Download GGUF models from hugging face

First, download the model from Hugging Face. You can use the `huggingface_hub` library or download it manually:

```bash
pip  install  huggingface_hub
huggingface-cli  download  {model_name}
```

This will download the model to your current directory. Make sure to replace {model_name} with the actual username and model name from your Hugging Face repository.

## 2. Install llama.cpp

You have several options for installing llama.cpp:

**1. Build from source:**

This gives you the most flexibility and control. Follow the instructions in the llama.cpp repository to build from source:

```bash

git  clone  https://github.com/ggerganov/llama.cpp
cd llama.cpp
cmake -B build
cmake --build build --config Release
```

For more information about how to build llama.cpp from source please refere to llama.cpp documentation on how to build from source: **[llama.cpp build from source](https://github.com/ggerganov/llama.cpp/blob/master/docs/build.md)**.

**2. Download pre-built binaries:**

If you prefer a quicker setup, you can download pre-built binaries for your operating system. Check the llama.cpp repository for available binaries.

**3. Use Docker:**

For a more contained environment, you can use the official llama.cpp Docker image. Refer to the llama.cpp documentation for instructions on how to use the Docker image.

For detailed instructions and more information, please check the llama.cpp documentation on docker: **[llama.cpp docker](https://github.com/ggerganov/llama.cpp/blob/master/docs/docker.mdg)**.

### 3. Start playing with your model

Run simple text completion
```bash
llama-cli -m {path-to-gguf-model}  -p  "I believe the meaning of life is"  -n  128
```

Run in conversation mode
```bash
llama-cli -m {path-to-gguf-model}  -p  "You are a helpful assistant"  -cnv  -co
```

## Useful links
- View our [release blogpost](https://huggingface.co/blog/falcon3).
- Feel free to join [our discord server](https://discord.gg/fwXpMyGc) if you have any questions or to interact with our researchers and developers.
  
## Technical Report

Coming soon....

## Citation
If the Falcon3 family of models were helpful to your work, feel free to give us a cite.
 
```
@misc{Falcon3,
    title = {The Falcon 3 Family of Open Models},
    url = {https://huggingface.co/blog/falcon3},
    author = {Falcon-LLM Team},
    month = {December},
    year = {2024}
}
```