CoPaw-Flash-9B-DataAnalyst-LoRA
Agentic Data Analyst that autonomously explores, analyzes, and visualizes your datasets.
Data Analyst Demo
What It Does
This model functions as an autonomous data analyst :
📂 Loads and explores datasets (CSV, Excel, JSON)
🔍 Performs statistical analysis and data profiling
📊 Creates visualizations (matplotlib, seaborn, plotly)
🐍 Writes and executes Python analysis scripts
📝 Generates summary reports and insights
🔄 Iterates through multi-step analysis workflows
🎯 Completes 90% of tasks autonomously (no human intervention)
Model Details
Property Value Base Model agentscope-ai/CoPaw-Flash-9B (Qwen3.5-9B architecture) Task Type Data Analysis Agent LoRA Rank 64 LoRA Alpha 128 Precision bfloat16 PEFT Version 0.18.1
Performance Benchmark
Tested on
29 real Kaggle datasets with
Data Analyst framework (max_turns=50, context=128K):
Metric Qwen3.5-9B Base DataAnalyst-LoRA Improvement Avg iterations 1.2 26.0 21.7x Python files 0 100+ ∞ Charts generated 0 290+ ∞ Total tokens ~5K 18.5M 3700x Natural completion rate *0% 89.7% +89.7pp Hit turn limit N/A 10.3% - Usable output 0/29 (0%) 26/29 (90%) +90pp User intervention Required every step Autonomous Autonomous
*Natural completion = Model autonomously outputs final summary report within 50 turns
Performance Benchmark: Base Model vs DataAnalyst-LoRA
Key Findings
Base Model (Qwen3.5-9B):
❌ Understands tool call format but cannot execute autonomously
❌ Stops after 1-2 iterations
❌ Requires continuous user "continue" prompts
❌ Produces zero analysis output
❌ Not usable for real data analysis tasks
CoPaw-Flash-9B-DataAnalyst-LoRA:
✅ Fully autonomous execution (26 iterations average)
✅ Generates complete analysis pipelines
✅ Creates visualizations and reports
✅ 90% success rate on real-world datasets
✅ Production-ready for data analysis workflows
Conclusion: LoRA training is essential, not optional . Base model lacks autonomous data analyst capabilities despite understanding the tool calling format. This LoRA transforms the base model into a production-ready AI data analyst that can handle real-world datasets independently.
Quick Start
Step 1: Deploy with vLLM
1 export HF_TOKEN = your_huggingface_token
2
3 CUDA_VISIBLE_DEVICES = 0,1 vllm serve agentscope-ai/QwenPaw-Flash-9B \
4 --enable-lora \
5 --lora-modules agent-lora = jason1966/CoPaw-Flash-9B-DataAnalyst-LoRA \
6 --max-lora-rank 64 \
7 --tensor-parallel-size 2 \
8 --gpu-memory-utilization 0.85 \
9 --max-model-len 131072 \
10 --gdn-prefill-backend triton \
11 --trust-remote-code \
12 --reasoning-parser qwen3 \
13 --enable-auto-tool-choice \
14 --tool-call-parser qwen3_xml \
15 --port 8000
Step 2: Setup Data Analyst Framework
1 git clone https://github.com/IIIIQIIII/data-analyst.git
2 cd data-analyst
3 bun install
Configure .env:
1 CLAUDE_CODE_USE_OPENAI = 1
2 OPENAI_BASE_URL = http://localhost:8000/v1
3 OPENAI_API_KEY = unused
4 OPENAI_MODEL = agent-lora
Step 3: Start Analyzing
Then input your analysis task:
Analyze sales_2024.csv and identify trendsThe model will autonomously load data, perform analysis, create visualizations, and generate reports—all without requiring manual "continue" prompts.
vLLM Parameters
Parameter Description --enable-loraEnable LoRA adapter support --lora-modules agent-lora=...Load DataAnalyst-LoRA adapter --max-lora-rank 64LoRA rank (must match adapter) --reasoning-parser qwen3Enable reasoning process visibility --enable-auto-tool-choiceAutomatic tool selection --tool-call-parser qwen3_xmlParse XML-format tool calls --gdn-prefill-backend tritonOptimize prefill with Triton
Hardware Requirements
Configuration VRAM Required Dual GPU (bf16, TP=2) ~11GB per GPU Single GPU (bf16) ~22GB 8-bit quantized ~12GB 4-bit quantized ~6GB
Tested: 2x NVIDIA H200, vLLM 0.19.1, CUDA 13.0, Python 3.12
Troubleshooting
Issue Solution FlashInfer errors Add --gdn-prefill-backend triton Out of memory Reduce --max-model-len or --gpu-memory-utilization Connection refused Check netstat -tlnp | grep 8000
Acknowledgments
CoPaw-Flash-9B — Base model by AgentScope AIBrev.dev — GPU cloud infrastructure by NVIDIALocoreMind — Research and development
License
Apache 2.0
