gemma-2-9b-linkscout-v1

gemma-2-9b-linkscout-v1

This model is a fine-tuned version of unsloth/gemma-2-9b-it-bnb-4bit for fact-checking and misinformation detection.

Model Description

• Base Model: unsloth/gemma-2-9b-it-bnb-4bit
• Fine-tuned for: Fact-checking and bias detection
• Training Method: LoRA/QLoRA with Unsloth
• Merged: Yes (adapter merged with base model)

Usage

from unsloth import FastLanguageModel

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name="Adi-Evolve/gemma-2-9b-linkscout-v1",
    max_seq_length=2048,
    dtype=None,
    load_in_4bit=True,
)

FastLanguageModel.for_inference(model)

# Generate analysis
prompt = """<start_of_turn>user
Analyze this article for misinformation: [ARTICLE TEXT]
<end_of_turn>
<start_of_turn>model
"""

inputs = tokenizer([prompt], return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training Details

• Merged from LoRA adapter
• Original base: unsloth/gemma-2-9b-it-bnb-4bit

Limitations

This model should be used as a tool to assist in fact-checking, not as a sole source of truth.