ed001/datascience-coder-6.7b

The Data Science Coder

Data Science coder is a group of fine tuned models designed to help with coding for data science applications. It comes in 2 variants: 1.3b and 6.7b. Models are fine tuned from DeepSeek Coder instruct versions. Fine tuning was performed on the ed001/ds-coder-instruct-v1 dataset which is constructed by filtering publicly available datasets on HuggingFace.

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

def build_instruction_prompt(instruction):
    return '''
    You are the Data Science Coder, a helpful AI assistant created by a man named Ed.
    You help people with data science coding and you answer questions about data science in a helpful manner.
    ### Instruction:
    {}
    ### Response:
    '''.format(instruction.strip()).lstrip()

tokenizer = AutoTokenizer.from_pretrained("ed001/datascience-coder-6.7b", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("ed001/datascience-coder-6.7b", trust_remote_code=True).cuda()
pipe = pipeline(task="text-generation", model=model, tokenizer=tokenizer, max_length=1024, top_p=0.95)
result = pipe(build_instruction_prompt("Perform EDA on the Iris dataset"))
print(result[0]['generated_text'])

Training Details

lorar: 16 loraalpha: 8 loradropout: 0.05 targetmodules: q, k, v, o, gateproj, downproj, upproj, lmhead weightdecay: 0 optmizer: pagedadamw32bit lr: 1e-4 lrscheduler: cosine maxseqlen: 4096 batchsize: 4 maxgradnorm: 0.5 warmupratio: 0.05 num_epochs: 1

The model was trained on the python susbet of the ds-coder-instruct dataset.

Samples

<img src="https://cdn-uploads.huggingface.co/production/uploads/62618f3e6dae705b2567fb13/0H8lj26xLOfLuCD0yVmER.png" width="90%"/>

<img src="https://cdn-uploads.huggingface.co/production/uploads/62618f3e6dae705b2567fb13/8W62qr1cPSLsq6lLfLCib.png" width="90%"/>

<img src="https://cdn-uploads.huggingface.co/production/uploads/62618f3e6dae705b2567fb13/XNLclcr4KQqtPseGg2Gzn.png" width="90%"/>

Contact

GitHub: Ea0011

Open LLM Leaderboard Evaluation Results

Detailed results can be found here