Crystalcareai
commited on
Commit
•
264696f
1
Parent(s):
ee0a4d8
Update README.md
Browse files
README.md
CHANGED
@@ -63,7 +63,39 @@ Llama-3-SEC has been trained using the llama3 chat template, which allows for ef
|
|
63 |
|
64 |
To run inference with the Llama-3-SEC model using the llama3 chat template, use the following code:
|
65 |
|
66 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
67 |
|
68 |
## Limitations and Future Work
|
69 |
|
@@ -71,7 +103,7 @@ This release represents the initial checkpoint of the Llama-3-SEC model, trained
|
|
71 |
|
72 |
## Usage
|
73 |
|
74 |
-
|
75 |
|
76 |
## Citation
|
77 |
|
|
|
63 |
|
64 |
To run inference with the Llama-3-SEC model using the llama3 chat template, use the following code:
|
65 |
|
66 |
+
```python
|
67 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
68 |
+
device = "cuda"
|
69 |
+
|
70 |
+
model = AutoModelForCausalLM.from_pretrained(
|
71 |
+
"arcee-ai/Llama-3-SEC",
|
72 |
+
torch_dtype="auto",
|
73 |
+
device_map="auto"
|
74 |
+
)
|
75 |
+
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2-72B-Instruct")
|
76 |
+
|
77 |
+
prompt = "What are the key regulatory considerations for a company planning to conduct an initial public offering (IPO) in the United States?"
|
78 |
+
messages = [
|
79 |
+
{"role": "system", "content": "You are an expert financial assistant - specializing in governance and regulatory domains."},
|
80 |
+
{"role": "user", "content": prompt}
|
81 |
+
]
|
82 |
+
text = tokenizer.apply_chat_template(
|
83 |
+
messages,
|
84 |
+
tokenize=False,
|
85 |
+
add_generation_prompt=True
|
86 |
+
)
|
87 |
+
model_inputs = tokenizer([text], return_tensors="pt").to(device)
|
88 |
+
|
89 |
+
generated_ids = model.generate(
|
90 |
+
model_inputs.input_ids,
|
91 |
+
max_new_tokens=512
|
92 |
+
)
|
93 |
+
generated_ids = [
|
94 |
+
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
|
95 |
+
]
|
96 |
+
|
97 |
+
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
98 |
+
```
|
99 |
|
100 |
## Limitations and Future Work
|
101 |
|
|
|
103 |
|
104 |
## Usage
|
105 |
|
106 |
+
The model is available for both commercial and non-commercial use under the Llama-3 license. We encourage users to explore the model's capabilities and provide feedback to help us continuously improve its performance and usability. For more information - please see our detailed blog on Llama-3-SEC.
|
107 |
|
108 |
## Citation
|
109 |
|