Update README.md
Browse files
README.md
CHANGED
@@ -11,14 +11,17 @@ language:
|
|
11 |
<img src="https://i.imgur.com/39sQd3h.png" alt="Merak Orca" width="300" height="300"/>
|
12 |
</p>
|
13 |
|
14 |
-
**Merak-7B-v3-Mini-Orca** is Ichsan2895's [Merak-7B-v3](https://huggingface.co/Ichsan2895/Merak-7B-v3) fine-tuned
|
|
|
15 |
|
16 |
|
17 |
## Usage
|
18 |
This model fit on 16GB VRAM GPU (Google Collab T4 wil do), by using BitsandBytes it can run on 6GB VRAM GPU.
|
19 |
|
20 |
-
Quantized versions is available:
|
|
|
21 |
GPTQ: https://huggingface.co/asyafiqe/Merak-7B-v3-Mini-Orca-Indo-GPTQ
|
|
|
22 |
GGML/GGUF: I will try to make this version once GGUF merge is stable.
|
23 |
|
24 |
Start chatting with Merak Mini Orca using the following code snippet:
|
@@ -40,7 +43,9 @@ print(tokenizer.decode(output[0], skip_special_tokens=True))
|
|
40 |
```
|
41 |
|
42 |
### Prompt format
|
43 |
-
You can use [Vicuna 1.1](https://github.com/oobabooga/text-generation-webui/blob/main/instruction-templates/Vicuna-v1.1.yaml)
|
|
|
|
|
44 |
```
|
45 |
SYSTEM: Anda adalah asisten AI. Anda akan diberi tugas. Anda harus menghasilkan jawaban yang rinci dan panjang.
|
46 |
USER: <prompt> (without the <>)
|
|
|
11 |
<img src="https://i.imgur.com/39sQd3h.png" alt="Merak Orca" width="300" height="300"/>
|
12 |
</p>
|
13 |
|
14 |
+
**Merak-7B-v3-Mini-Orca** is Ichsan2895's [Merak-7B-v3](https://huggingface.co/Ichsan2895/Merak-7B-v3) fine-tuned
|
15 |
+
on Bahasa Indonesia translated psmathur's [orca_mini_v1_dataset](https://huggingface.co/datasets/psmathur/orca_mini_v1_dataset).
|
16 |
|
17 |
|
18 |
## Usage
|
19 |
This model fit on 16GB VRAM GPU (Google Collab T4 wil do), by using BitsandBytes it can run on 6GB VRAM GPU.
|
20 |
|
21 |
+
**Quantized** versions is available:
|
22 |
+
|
23 |
GPTQ: https://huggingface.co/asyafiqe/Merak-7B-v3-Mini-Orca-Indo-GPTQ
|
24 |
+
|
25 |
GGML/GGUF: I will try to make this version once GGUF merge is stable.
|
26 |
|
27 |
Start chatting with Merak Mini Orca using the following code snippet:
|
|
|
43 |
```
|
44 |
|
45 |
### Prompt format
|
46 |
+
You can use [Vicuna 1.1](https://github.com/oobabooga/text-generation-webui/blob/main/instruction-templates/Vicuna-v1.1.yaml)
|
47 |
+
format for Ooobabooga's text generation webui.
|
48 |
+
|
49 |
```
|
50 |
SYSTEM: Anda adalah asisten AI. Anda akan diberi tugas. Anda harus menghasilkan jawaban yang rinci dan panjang.
|
51 |
USER: <prompt> (without the <>)
|