README.md

---
license: apache-2.0
tags:
- code generation
metrics:
- code_eval
pipeline_tag: text-generation
inference: true
widget:
- text: 'def print_hello_world():'
  example_title: Hello world
  group: Python
model-index:
- name: StarCoder
  results:
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (Python)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.191
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (JavaScript)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.184
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (Java)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.166
      verified: false
datasets:
- bigcode/starcoderdata
---

# Model Card for DeciCoder 1B

DeciCoder 1B is a 1 billion parameter decoder-only code completion model
trained on the Python, Java, and Javascript subsets of [Starcoder Training Dataset](https://huggingface.co/datasets/bigcode/starcoderdata).
The model uses Grouped Query Attention and has a context window of 2048
tokens. It was trained using a Fill-in-the-Middle training objective. The model's
architecture was generated by Deci's proprietary Neural Architecture
Search-based technology, AutoNAC.

## Model Details

- **Developed by:** Deci 
- **Model type:** DeciCoder is an auto-regressive language model based on the transformer decoder architecture, using Grouped Query Attention.
- **Language(s):** Python, Java, JavaScript
- **License:** Model checkpoints are licensed under the [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)

## Model Architecture

| Parameters | Layers | Heads  | Sequence Length  | GQA num_key_value_heads  | Hidden Size  |
|:----------|:----------|:----------|:----------|:----------|:----------|
| 1.1B    | 20    | 32    | 2048   | 4  | 2048 |  |


- **Decoder layer:** Grouped Query Attention [Ainslie et al., 2023](https://arxiv.org/abs/2305.13245)
- **Position Embeddings:** Rotary Position Embeddings [Su et al., 2021](https://arxiv.org/abs/2104.09864)

## Uses

The model is intended to do single/multiline code completion from a
context window of up to 2048k tokens. It is *not* an instruction model
and commands like \"Write a function that computes the absolute value of
an integer,\" won't yield the desired results. A more effective approach
is to frame instructions in the style of source code comments (e.g. \#
this function calculates the absolute value of an integer) or to present
a function signature and docstring, enabling the model to complete the
function's body.

### How to Use

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

checkpoint = "Deci/DeciCoder-1b"
device = "cuda" # for GPU usage or "cpu" for CPU usage

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

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

### Attribution

DeciCoder was trained on StarCoder Training Dataset, filtered for
Python, Java, and Javascript code. For additional information, please
refer to [https://huggingface.co/datasets/bigcode/starcoderdata](https://huggingface.co/datasets/bigcode/starcoderdata).

### Limitations

The model has undergone training with source code from Python, Java, and
JavaScript. While the primary language in the source is English, it does
contain other languages. Therefore, the model can produce code snippets
given some context. However, there\'s no assurance that the resulting
code will function as expected. It might be suboptimal, contain bugs, or
even exploits.

## Training Details

### Training Data

DeciCoder was trained on the Python, Java, and Javascript subsets of [Starcoder Training Dataset](https://huggingface.co/datasets/bigcode/starcoderdata)


### Training Procedure 

- **Warm-Up Steps**: 9000
- **Total Training Steps**: 284k
- **Total Tokenes**: 446B
- **Global Batch Size**: 768
- **Optimizer**: AdamW
- **Optimizer Parameters**: beta1=0.9, beta2=0.95
- **Weight Decay**: 0.1
- **Learning Rate**: 4e-4
- **Learning Rate Schedule**: cosine

## Evaluation

Below are DeciCoder's pass@1 on MultiPL HumanEval scores

| Python | JavaScript | Java  |
|:----------|:----------|:----------|
| 19.1%    | 18.4%    | 16.6%    |


### Runtime Benchmarks

|Inference Tool/Hardware | A10G (tokens/sec) | A100 (tokens/sec)  |
|:----------|:----------|:----------|
| HF Inference Endpoints  | 1,364.2 | 3,244.4    |
| Infery LLM | 3,889.3   |  11,676.8  |

## Documentation

- [Notebook](https://colab.research.google.com/drive/1JCxvBsWCZKHfIcHSMVf7GZCs3ClMQPjs)
- Blog post: [Introducing DeciCoder: The New Gold Standard in Efficient and Accurate Code Generation](https://deci.ai/blog/decicoder-efficient-and-accurate-code-generation-llm/)
- Questions:Feel free to contact us via our [Discord Community!](https://discord.com/invite/p9ecgRhDR8/)

## How to Cite

Please cite this model using this format.

```bibtex
@misc{DeciFoundationModels,
title = {DeciCoder},
author = {DeciAI Research Team},
year = {2023}
url={[https://huggingface.co/deci/decicoder-1b](https://huggingface.co/deci/decicoder-1b)},
}
```