metadata

license: apache-2.0
pipeline_tag: text-generation
tags:
  - text-generation
  - causal-lm
  - instruction-tuned
  - serverless
library_name: transformers
inference: true
language:
  - en
base_model: automatedstockminingorg/expert-on-investment-valuation-mypricermodel
datasets:
  - automatedstockminingorg/investment-valuation-chunks

Expert on Investment Valuation Model

Introduction

This model is fine-tuned on data specifically curated for investment valuation, helping users with insights and explanations on various valuation techniques, including the discounted cash flow (DCF) model and comparable company analysis.

Designed for generating text that follows instructions and role-playing in a financial advisory setting.
Supports long-context processing to handle in-depth questions.
Multilingual support available in English.

This repo contains the instruction-tuned version of the model:

Type: Causal Language Model (instruction-tuned)
Language: English
Model Architecture: Transformers

For more details, please refer to our documentation.

Requirements

To ensure compatibility, use the latest version of transformers.

Quickstart

Here is a code snippet to show how to load the tokenizer and model and generate responses.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "automatedstockminingorg/14b-stockanalyst-14b-stockanalyst"

model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

prompt = "Explain the discounted cash flow (DCF) model in investment valuation."
messages = [
    {"role": "system", "content": "You are an expert in investment valuation."},
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=300
)
generated_ids = [
    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]

response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)