|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
base_model: |
|
|
- prithivMLmods/Qwen3-4B-ft-bf16 |
|
|
pipeline_tag: text-generation |
|
|
library_name: transformers |
|
|
tags: |
|
|
- text-generation-inference |
|
|
- code |
|
|
- RL |
|
|
- moe |
|
|
datasets: |
|
|
- livecodebench/code_generation_lite |
|
|
- PrimeIntellect/verifiable-coding-problems |
|
|
- likaixin/TACO-verified |
|
|
- open-r1/codeforces-cots |
|
|
--- |
|
|
|
|
|
 |
|
|
|
|
|
# **Blitzar-Coder-4B-F.1** |
|
|
|
|
|
> **Blitzar-Coder-4B-F.1** is a high-efficiency, multi-language coding model fine-tuned on **Qwen3-4B** using **larger coding traces datasets** spanning **10+ programming languages** including Python, Java, C#, C++, C, Go, JavaScript, TypeScript, Rust, and more. This model delivers exceptional code generation, debugging, and reasoning capabilities—making it an ideal tool for developers seeking advanced programming assistance under constrained compute. |
|
|
|
|
|
> \[!note] |
|
|
> GGUF: [https://huggingface.co/prithivMLmods/Blitzar-Coder-4B-F.1-GGUF](https://huggingface.co/prithivMLmods/Blitzar-Coder-4B-F.1-GGUF) |
|
|
|
|
|
--- |
|
|
|
|
|
## **Key Features** |
|
|
|
|
|
1. **Multi-Language Code Mastery** |
|
|
Fine-tuned on **extensive coding traces datasets** covering **10+ programming languages** (Python, Java, C#, C++, C, Go, JavaScript, TypeScript, Rust, Swift, Kotlin, and more), enabling cross-language development and translation. |
|
|
|
|
|
2. **Advanced Code Generation & Reasoning** |
|
|
Supports complex algorithm synthesis, code optimization, debugging workflows, and architectural design patterns across multiple paradigms—from systems programming to web development. |
|
|
|
|
|
3. **Cross-Language Development Support** |
|
|
Seamlessly handles polyglot codebases, API integrations, and framework-specific implementations while maintaining language-specific best practices and idioms. |
|
|
|
|
|
4. **Intelligent Code Analysis** |
|
|
Performs code reviews, identifies performance bottlenecks, suggests refactoring opportunities, and provides detailed explanations for complex programming concepts. |
|
|
|
|
|
5. **Structured Output for Development** |
|
|
Generates clean code documentation, API specifications, configuration files, and technical documentation in various formats including **Markdown**, **JSON**, **YAML**, and inline comments. |
|
|
|
|
|
6. **Optimized 4B Footprint for Developer Workflows** |
|
|
Balanced for performance and efficiency, deployable on **developer workstations**, **CI/CD pipelines**, and **edge development environments** without compromising code quality. |
|
|
|
|
|
--- |
|
|
|
|
|
## **Quickstart with Transformers** |
|
|
|
|
|
```python |
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer |
|
|
|
|
|
model_name = "prithivMLmods/Blitzar-Coder-4B-F.1" |
|
|
|
|
|
model = AutoModelForCausalLM.from_pretrained( |
|
|
model_name, |
|
|
torch_dtype="auto", |
|
|
device_map="auto" |
|
|
) |
|
|
tokenizer = AutoTokenizer.from_pretrained(model_name) |
|
|
|
|
|
prompt = "Create a REST API endpoint in Python using FastAPI that handles file uploads with validation and returns processing status." |
|
|
|
|
|
messages = [ |
|
|
{"role": "system", "content": "You are an expert programming assistant skilled in multiple languages and development practices."}, |
|
|
{"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=512 |
|
|
) |
|
|
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) |
|
|
``` |
|
|
|
|
|
--- |
|
|
|
|
|
## **Intended Use** |
|
|
|
|
|
* Multi-language code generation and debugging assistance |
|
|
* Cross-platform development and code translation between languages |
|
|
* Code review, optimization, and refactoring suggestions |
|
|
* Technical documentation and API specification generation |
|
|
* Developer productivity tools and IDE integrations |
|
|
* Educational coding tutorials and programming concept explanations |
|
|
|
|
|
--- |
|
|
|
|
|
## **Limitations** |
|
|
|
|
|
* Optimized for coding tasks—may underperform on general conversation |
|
|
* Context limitations may affect analysis of very large codebases |
|
|
* Focused on programming domains—creative writing capabilities are limited |
|
|
* Best suited for technical development workflows rather than casual chat |
|
|
|
|
|
--- |
|
|
|
|
|
## **References** |
|
|
|
|
|
1. [Qwen2.5 Technical Report (2024)](https://arxiv.org/pdf/2412.15115) |
|
|
2. [YaRN: Efficient Context Window Extension of Large Language Models](https://arxiv.org/pdf/2309.00071) |
