Update README.md
Browse files
README.md
CHANGED
@@ -38,7 +38,44 @@ The code to replicate the experiments is available on [GitHub](https://github.co
|
|
38 |
|
39 |
## Model Inference
|
40 |
|
41 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
|
43 |
# License
|
44 |
This model is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
|
|
|
38 |
|
39 |
## Model Inference
|
40 |
|
41 |
+
To utilize the LlamaLens model for inference, follow these steps:
|
42 |
+
|
43 |
+
1. **Install the Required Libraries**:
|
44 |
+
|
45 |
+
Ensure you have the necessary libraries installed. You can do this using pip:
|
46 |
+
|
47 |
+
```bash
|
48 |
+
pip install transformers torch
|
49 |
+
```
|
50 |
+
2. **Load the Model and Tokenizer:**:
|
51 |
+
Use the transformers library to load the LlamaLens model and its tokenizer:
|
52 |
+
|
53 |
+
```python
|
54 |
+
from transformers import AutoTokenizer, AutoModelForCausalLM
|
55 |
+
|
56 |
+
model_name = "QCRI/LlamaLens"
|
57 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
58 |
+
model = AutoModelForCausalLM.from_pretrained(model_name)
|
59 |
+
```
|
60 |
+
3. **Prepare the Input:**:
|
61 |
+
Tokenize your input text:
|
62 |
+
```python
|
63 |
+
input_text = "Your input text here"
|
64 |
+
inputs = tokenizer(input_text, return_tensors="pt")
|
65 |
+
```
|
66 |
+
4. **Generate the Output:**:
|
67 |
+
Generate a response using the model:
|
68 |
+
```python
|
69 |
+
output = model.generate(**inputs)
|
70 |
+
response = tokenizer.decode(output[0], skip_special_tokens=True)
|
71 |
+
print(response)
|
72 |
+
```
|
73 |
+
|
74 |
+
## Paper
|
75 |
+
For an in-depth understanding, refer to our paper: [**LlamaLens: Specialized Multilingual LLM for Analyzing News and Social Media Content**](https://arxiv.org/pdf/2410.15308).
|
76 |
+
|
77 |
+
|
78 |
+
|
79 |
|
80 |
# License
|
81 |
This model is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
|