Astronomy hypothesis generation with Falcon-7B

It was fine-tuned on several thousand astronomy abstracts collected on Arxiv.

Model Details

from transformers import AutoModelForCausalLM, AutoTokenizer
import transformers
import torch

online_model = AutoModelForCausalLM.from_pretrained("universeTBD/falcon-7b-abstracts-tiny", torch_dtype=torch.bfloat16,
                                                    device_map="auto", trust_remote_code=True)

tokenizer = AutoTokenizer.from_pretrained("tiiuae/falcon-7b")
pipeline = transformers.pipeline(
    "text-generation",
    model=online_model,
    tokenizer=tokenizer,
    torch_dtype=torch.bfloat16,
    trust_remote_code=True,
    device_map="auto",
)

sequences = pipeline(
   "### Instruction: Generate a scientific hypothesis about astronomy in the style of an Arxiv paper.\n ### Hypothesis:",
    max_length=500,
    do_sample=True,
    top_k=10,
    num_return_sequences=1,
    eos_token_id=tokenizer.eos_token_id,
)

def format_output(output):
    output = output.replace("\n", " ")  # Replace newline characters with spaces
    output = output.replace("\\n", " ")
    parts = output.split("###")  # Split string at '###'
    
    # Get and clean instruction part
    instruction = parts[1].strip() 
    
    # Get and clean hypothesis part
    hypothesis = parts[2].strip()  
    
    # Format the output
    formatted_output = f"{instruction}\n\n{hypothesis}"
    
    return formatted_output

format_output(sequences[0]['generated_text'])

Example generation:

Using 3D positions and K magnitudes of stars from the Gaia DR2 for which we have spectroscopic information from the RAVE database, we derive distances to the stellar populations in different parts of the bulge of the Milky Way. We find that the metal-rich (blue) stars in the inner part of the bulge have a disk component, while the metal-poor (red) stars in the inner part of the bulge do not have a discernible disk component and are dominated by halo components. Spectral parameters indicate that the red stars are enhanced in nitrogen and the blue stars are enhanced in iron, suggesting that the red stars may have a faster rotation curve than the blue stars. These morpho-chemical properties are similar to those of the classical thick disk populations. However, the inner part of the bulge stars with metallicity about -1.0 <[Fe/H] < -0.5 do not have a discernible disk component and are also found in the halo component. Stars with metallicity about -2.5 <[Fe/H] < -1.0 in the inner part of the bulge also have a faint halo component and are enhanced in nitrogen. We suggest that the metal-rich blue stars in the inner part of the bulge came from a disk formed in situ and the red stars in the inner part of the bulge came from two different disk-to-halo transition zones which may be associated with the late low-density and late high-density spiral arms, respectively.

Model Description

  • Developed by: [More Information Needed]
  • Shared by [optional]: [More Information Needed]
  • Model type: [More Information Needed]
  • Language(s) (NLP): [More Information Needed]
  • License: [More Information Needed]
  • Finetuned from model [optional]: [More Information Needed]

Model Sources [optional]

  • Repository: [More Information Needed]
  • Paper [optional]: [More Information Needed]
  • Demo [optional]: [More Information Needed]

Uses

Direct Use

[More Information Needed]

Downstream Use [optional]

[More Information Needed]

Out-of-Scope Use

[More Information Needed]

Bias, Risks, and Limitations

[More Information Needed]

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

Training Details

Training Data

[More Information Needed]

Training Procedure

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

  • Training regime: [More Information Needed]

Speeds, Sizes, Times [optional]

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

[More Information Needed]

Factors

[More Information Needed]

Metrics

[More Information Needed]

Results

[More Information Needed]

Summary

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

  • Hardware Type: [More Information Needed]
  • Hours used: [More Information Needed]
  • Cloud Provider: [More Information Needed]
  • Compute Region: [More Information Needed]
  • Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

[More Information Needed]

Downloads last month
15
Inference Examples
Inference API (serverless) does not yet support model repos that contain custom code.