1B version of StarCoderBase. LoRa finetuned on Evol-Instruct-Code-80k-v1 dataset generated using the Evol-Teacher repo. Based on Previous works of @WizardLM

Training Report: wandb
Repository: bigcode/Starcoder
Project Website: bigcode-project.org

Use

Generation

python

# pip install -q transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

checkpoint = "epinnock/wizardcoder-1b-merged"
device = "cuda" # for GPU usage or "cpu" for CPU usage

tokenizer = AutoTokenizer.from_pretrained(checkpoint)
model = AutoModelForCausalLM.from_pretrained(checkpoint).to(device)

inputs = tokenizer.encode("def print_hello_world():", return_tensors="pt").to(device)
outputs = model.generate(inputs)
print(tokenizer.decode(outputs[0]))

Fill-in-the-middle

Fill-in-the-middle uses special tokens to identify the prefix/middle/suffix part of the input and output:

python

input_text = "<fim_prefix>def print_hello_world():\n    <fim_suffix>\n    print('Hello world!')<fim_middle>"
inputs = tokenizer.encode(input_text, return_tensors="pt").to(device)
outputs = model.generate(inputs)
print(tokenizer.decode(outputs[0]))

Attribution & Other Requirements

The pretraining dataset of the model was filtered for permissive licenses only. Nevertheless, the model can generate source code verbatim from the dataset. The code's license might require attribution and/or other specific requirements that must be respected. We provide a search index that let's you search through the pretraining data to identify where generated code came from and apply the proper attribution to your code.

Limitations

The model has been trained on source code from 80+ programming languages. The predominant natural language in source code is English although other languages are also present. As such the model is capable of generating code snippets provided some context but the generated code is not guaranteed to work as intended. It can be inefficient, contain bugs or exploits. See the paper for an in-depth discussion of the model limitations.

Training

starcoder/finetune/finetune.py --model_path=bigcode/starcoderbase-1b --dataset_name=nickrosh/Evol-Instruct-Code-80k-v1 --split=train --size_valid_set 5000 --streaming --seq_length 2048 --max_steps 2400 --batch_size 6 --input_column_name=instruction --gradient_accumulation_steps 16 --learning_rate 1e-4 --lr_scheduler_type=cosine --num_warmup_steps 100 --weight_decay 0.05

Model

Architecture: StarCoder GPT-2 model with multi-query attention and Fill-in-the-Middle objective with LoRA layers merged
Fintune steps: 2400
Pretraining tokens: 1 trillion
Precision: bfloat16

Hardware

GPUs: 8 A5000
Training time: 2 days

License

The model is licensed under the BigCode OpenRAIL-M v1 license agreement. You can find the full agreement here.

epinnock/wizardcoder-1b-merged

epinnock/WizardCoder-1B-merged

Table of Contents

Model Summary

Use