99 lines
3.2 KiB
Markdown
99 lines
3.2 KiB
Markdown
|
---
|
|||
|
|
license: apache-2.0
|
|||
|
|
language:
|
|||
|
|
- en
|
|||
|
|
base_model:
|
|||
|
|
- Qwen/Qwen3-0.6B
|
|||
|
|
tags:
|
|||
|
|
- event-data
|
|||
|
|
- political-science
|
|||
|
|
- computational-social-science
|
|||
|
|
---
|
|||
|
|
|
|||
|
|
|
|||
|
|
|
|||
|
|
# Example usage with vLLM
|
|||
|
|
|
|||
|
|
## Load the model and tokenizer
|
|||
|
|
|
|||
|
|
```
|
|||
|
|
from vllm import LLM, SamplingParams
|
|||
|
|
from transformers import AutoTokenizer
|
|||
|
|
|
|||
|
|
model = LLM(model="ahalt/event-attribute-extractor",
|
|||
|
|
enable_prefix_caching=True,
|
|||
|
|
max_model_len=8000,
|
|||
|
|
gpu_memory_utilization=0.80)
|
|||
|
|
|
|||
|
|
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-0.6B")
|
|||
|
|
|
|||
|
|
|
|||
|
|
sampling_params = SamplingParams(
|
|||
|
|
temperature=0.5, # Greedy decoding breaks Qwen
|
|||
|
|
top_p=0.8, # Qwen3 non-thinking recommendation
|
|||
|
|
top_k=20, # Qwen3 recommendation
|
|||
|
|
presence_penalty=1.5, # Recommended for quantized models
|
|||
|
|
min_p=0.0,
|
|||
|
|
#guided_decoding=guided_decoding_params, # Optionally, set a JSON schema for contrained decoding
|
|||
|
|
max_tokens=1024,
|
|||
|
|
)
|
|||
|
|
```
|
|||
|
|
|
|||
|
|
|
|||
|
|
## Prompt setup
|
|||
|
|
|
|||
|
|
```
|
|||
|
|
system_content_short = """Extract political events as JSON.
|
|||
|
|
|
|||
|
|
OUTPUT FORMAT:
|
|||
|
|
[
|
|||
|
|
{
|
|||
|
|
"event_type": "EVENT_TYPE",
|
|||
|
|
"anchor_quote": "quote from text",
|
|||
|
|
"actor": "who performed action OR N/A",
|
|||
|
|
"recipient": "who was targeted OR N/A",
|
|||
|
|
"date": "when occurred OR N/A",
|
|||
|
|
"location": "where occurred OR N/A"
|
|||
|
|
}
|
|||
|
|
]
|
|||
|
|
|
|||
|
|
Return valid JSON only. Empty array [] if no events."""
|
|||
|
|
|
|||
|
|
|
|||
|
|
def make_prompt(doc, event_type, tokenizer):
|
|||
|
|
messages = [
|
|||
|
|
{"role": "system", "content": system_content_short},
|
|||
|
|
{"role": "user", "content": f"## Document: {doc}\n\n## Event Type: {event_type}"},
|
|||
|
|
]
|
|||
|
|
prompt = tokenizer.apply_chat_template(
|
|||
|
|
messages,
|
|||
|
|
tokenize=False,
|
|||
|
|
add_generation_prompt=True,
|
|||
|
|
enable_thinking=False
|
|||
|
|
)
|
|||
|
|
return prompt
|
|||
|
|
```
|
|||
|
|
|
|||
|
|
|
|||
|
|
## Example usage
|
|||
|
|
|
|||
|
|
```
|
|||
|
|
text = """KYIV, Ukraine (AP) — Ukraine’s anti-corruption agencies said they had uncovered a major graft scheme involving inflated military procurement contracts, just two days after Ukraine’s parliament voted to restore the agencies’ independence.
|
|||
|
|
|
|||
|
|
In a joint statement published Saturday on social media, the National Anti-Corruption Bureau (NABU) and the Specialized Anti-Corruption Prosecutor’s Office (SAPO) said the suspects had taken bribes in a scheme that used state funds to buy drones and other military equipment at inflated prices.
|
|||
|
|
|
|||
|
|
“The essence of the scheme was to conclude state contracts with supplier companies at deliberately inflated prices,” the statement said, adding that offenders had received kickbacks of up to 30% of the contracts’ value.
|
|||
|
|
|
|||
|
|
event_type = "Investigate, charge, or prosecute"
|
|||
|
|
|
|||
|
|
prompt = make_prompt(text, event_type, tokenizer)
|
|||
|
|
output = model.generate(prompt, sampling_params=sampling_params)
|
|||
|
|
response = output[0].outputs[0].text.strip()
|
|||
|
|
|
|||
|
|
[{"event_type": "Investigate, charge, or prosecute",
|
|||
|
|
"anchor_quote": "Ukraine\u2019s anti-corruption agencies said they had uncovered a major graft scheme involving inflated military procurement contracts",
|
|||
|
|
"actor": "National Anti-Corruption Bureau (NABU); Specialized Anti-Corruption Prosecutor\u2019s Office (SAPO)",
|
|||
|
|
"recipient": "suspects involved in the scheme",
|
|||
|
|
"date": "Saturday",
|
|||
|
|
"location": "Ukraine"}
|
|||
|
|
```
|