fte-benchmark-gemma

FTE Benchmark Pricing Model

Fine-tuned Gemma 2-2B model for FTE (Full-Time Equivalent) benchmark pricing analysis.

Model Description

This model has been fine-tuned on salary benchmarking data to provide detailed reasoning about FTE pricing across different roles, geographies, and experience levels.

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

# Load model
base_model = AutoModelForCausalLM.from_pretrained("google/gemma-2-2b-it")
model = PeftModel.from_pretrained(base_model, "HexAdvisory/fte-benchmark-gemma")
tokenizer = AutoTokenizer.from_pretrained("HexAdvisory/fte-benchmark-gemma")

# Create prompt
prompt = """<start_of_turn>user
What is the FTE benchmark pricing for a Senior Software Engineer with 5-8 years experience in San Francisco, United States?

Input: {"role": "Senior Software Engineer", "geography": "United States", "city": "San Francisco", "experience_level": "5-8 years"}
Action: analyze_benchmark_pricing<end_of_turn>
<start_of_turn>model
"""

# Generate
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=300)
print(tokenizer.decode(outputs[0]))

Training

• Base Model: google/gemma-2-2b-it
• Method: LoRA fine-tuning
• Dataset: 202 custom FTE pricing examples
• Training: 1 epoch

Capabilities

The model can:

• Analyze FTE benchmark pricing
• Provide multi-step reasoning
• Consider geography, experience level, and role
• Output structured recommendations

License

Same as base Gemma 2 model