| --- |
| license: apache-2.0 |
| tags: |
| - finance |
| - tax |
| - banking |
| - investment |
| - indian-finance |
| - gst |
| - conversational |
| --- |
| |
| # π¦ Finxan |
|
|
| Finxan is an AI assistant specialized in Indian finance β GST, tax compliance, B2B invoicing, banking, and investments. |
|
|
| ## π¬ What Finxan can do |
|
|
| - π§Ύ Calculate GST and verify tax compliance |
| - π¦ Create and manage B2B invoices |
| - π Answer investment and banking queries |
| - β
Check financial document compliance |
| - π€ Answer questions about itself |
|
|
| ## π Quick Start |
| ```python |
| from peft import PeftModel |
| from transformers import AutoModelForCausalLM, AutoTokenizer |
| import torch |
| |
| base = AutoModelForCausalLM.from_pretrained( |
| "mistralai/Mistral-7B-v0.1", |
| torch_dtype=torch.float16, |
| device_map="auto" |
| ) |
| model = PeftModel.from_pretrained(base, "webdpro/finxan") |
| tokenizer = AutoTokenizer.from_pretrained("webdpro/finxan") |
| |
| def ask_finxan(instruction, input_text=""): |
| prompt = f"### Instruction:\n{instruction}\n\n### Input:\n{input_text}\n\n### Response:\n" |
| inputs = tokenizer(prompt, return_tensors="pt").to(model.device) |
| outputs = model.generate( |
| **inputs, |
| max_new_tokens=80, |
| temperature=0.3, |
| do_sample=True, |
| repetition_penalty=1.3, |
| pad_token_id=tokenizer.eos_token_id |
| ) |
| full = tokenizer.decode(outputs[0], skip_special_tokens=True) |
| return full.split("### Response:")[-1].strip() |
| |
| # Try it |
| print(ask_finxan("What is your name?")) |
| print(ask_finxan("Calculate GST", "Invoice Rs 50000, GST 18%")) |
| print(ask_finxan("Who built you?")) |
| ``` |
|
|
| ## π Example Outputs |
|
|
| | Question | Answer | |
| |----------|--------| |
| | What is your name? | I am Finxan, your AI finance assistant | |
| | Calculate GST for Rs 50000 at 18% | GST: Rs 9000. Total: Rs 59000 | |
| | Create a B2B invoice | Invoice created with GST applied | |
| | Check GST compliance | GST valid. Compliant. | |
|
|