Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- en
|
4 |
+
tags:
|
5 |
+
- gguf
|
6 |
+
- llama.cpp
|
7 |
+
- tiny
|
8 |
+
- tinyllama
|
9 |
+
---
|
10 |
+
# TinyLlama-1.1B-Chat-v1.0
|
11 |
+
![Image](image.jpeg)
|
12 |
+
|
13 |
+
## Requisitos
|
14 |
+
|
15 |
+
Para usar este modelo, necesitas tener instalado llama.cpp en tu equipo. Puedes obtener llama.cpp desde el siguiente repositorio:
|
16 |
+
|
17 |
+
- [Repositorio de llama.cpp](https://github.com/ggerganov/llama.cpp)
|
18 |
+
|
19 |
+
Para instalar llama.cpp, sigue estos pasos:
|
20 |
+
|
21 |
+
```bash
|
22 |
+
git clone https://github.com/ggerganov/llama.cpp
|
23 |
+
cd llama.cpp
|
24 |
+
make
|
25 |
+
```
|
26 |
+
|
27 |
+
## Uso del modelo
|
28 |
+
|
29 |
+
La plantilla del modelo es la siguiente:
|
30 |
+
|
31 |
+
```plaintext
|
32 |
+
<|system|>\nAnswer user questions</s>\n<|user|>\n{prompt}</s>\n<|assistant|>
|
33 |
+
```
|
34 |
+
|
35 |
+
Puedes utilizar el modelo en llama.cpp con el siguiente comando:
|
36 |
+
|
37 |
+
```bash
|
38 |
+
./main -m ggml-model-Q8_0.gguf -p "<|system|>\nAnswer user questions</s>\n<|user|>\nHi</s>\n<|assistant|>" --log-disable
|
39 |
+
```
|
40 |
+
|
41 |
+
LM Studio config-presets
|
42 |
+
|
43 |
+
Filename:phi-3.preset.json
|
44 |
+
```json
|
45 |
+
{
|
46 |
+
"name": "Phi-3",
|
47 |
+
"inference_params": {
|
48 |
+
"input_prefix": "<|user|>\n",
|
49 |
+
"input_suffix": "</s>\n<|assistant|>\n",
|
50 |
+
"antiprompt": [
|
51 |
+
"<|user|>\n",
|
52 |
+
"</s>\n<|assistant|>\n"
|
53 |
+
],
|
54 |
+
"pre_prompt": "<|system|>\nAnswer user questions</s>",
|
55 |
+
"pre_prompt_prefix": "",
|
56 |
+
"pre_prompt_suffix": ""
|
57 |
+
},
|
58 |
+
"load_params": {
|
59 |
+
"rope_freq_scale": 0,
|
60 |
+
"rope_freq_base": 0
|
61 |
+
}
|
62 |
+
}
|
63 |
+
```
|
64 |
+
|
65 |
+
## Referencias
|
66 |
+
|
67 |
+
- [Repositorio original](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0)
|
68 |
+
- [Repositorio de llama.cpp](https://github.com/ggerganov/llama.cpp)
|