Update README.md
Browse files
README.md
CHANGED
@@ -1,9 +1,45 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
-
|
9 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language: en
|
3 |
+
tags:
|
4 |
+
- text-generation
|
5 |
+
- ollama
|
6 |
+
- aya
|
7 |
+
- llm
|
8 |
+
- conversational
|
9 |
+
pipeline_tag: text-generation
|
10 |
+
library_name: transformers
|
11 |
+
inference: true
|
12 |
+
---
|
13 |
+
|
14 |
+
# Aya-8B
|
15 |
+
|
16 |
+
## Model Description
|
17 |
+
|
18 |
+
This is the Aya-8B model, originally designed for Ollama and converted to be compatible with Hugging Face. Aya is an open-source language model known for its conversational abilities and text generation capabilities.
|
19 |
+
|
20 |
+
## Usage
|
21 |
+
|
22 |
+
```python
|
23 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
24 |
+
|
25 |
+
tokenizer = AutoTokenizer.from_pretrained("Danna8/aya-8b")
|
26 |
+
model = AutoModelForCausalLM.from_pretrained("Danna8/aya-8b")
|
27 |
+
|
28 |
+
inputs = tokenizer("Hello, how are you today?", return_tensors="pt")
|
29 |
+
outputs = model.generate(inputs["input_ids"], max_length=100)
|
30 |
+
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
31 |
+
print(response)
|
32 |
+
```
|
33 |
+
|
34 |
+
## Model Details
|
35 |
+
|
36 |
+
- **Model Type:** Transformer-based language model
|
37 |
+
- **Size:** 8 billion parameters
|
38 |
+
|
39 |
+
|
40 |
+
## Limitations and Biases
|
41 |
+
|
42 |
+
Like all language models, Aya-8B may reproduce biases present in its training data. Users should be aware of these limitations when deploying the model.
|
43 |
+
|
44 |
+
## License
|
45 |
+
|