LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Replit V1 CodeInstruct 3B Fp16 by teknium

 »  All LLMs  »  teknium  »  Replit V1 CodeInstruct 3B Fp16   URL Share it on

  169   
  Autotrain compatible   Code   Custom code Dataset:bigcode/the-stack-dedu... Dataset:sahil2801/codealpaca-2... Dataset:teknium/gpteacher-code...   Endpoints compatible   Fp16   Instruct   Mpt   Pytorch   Quantized   Region:us   Self-instruct
Model Card on HF 🤗: https://huggingface.co/teknium/Replit-v1-CodeInstruct-3B-fp16 

Replit V1 CodeInstruct 3B Fp16 Benchmarks

LLME Score: 0.10752
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Replit V1 CodeInstruct 3B Fp16 (teknium/Replit-v1-CodeInstruct-3B-fp16)

Replit V1 CodeInstruct 3B Fp16 Parameters and Internals

Model Type 
code, instruct, self instruct
Use Cases 
Areas:
code instruction, code generation
Additional Notes 
Issues with device="auto" in model arguments, requires trust_remote_code=True.
Supported Languages 
Markdown (proficient), Java (proficient), JavaScript (proficient), Python (proficient), TypeScript (proficient), PHP (proficient), SQL (proficient), JSX (proficient), reStructuredText (proficient), Rust (proficient), C (proficient), CSS (proficient), Go (proficient), C++ (proficient), HTML (proficient), Vue (proficient), Ruby (proficient), Jupyter Notebook (proficient), R (proficient), Shell (proficient)
Training Details 
Data Sources:
bigcode/the-stack-dedup, sahil2801/CodeAlpaca-20k, teknium/GPTeacher-CodeInstruct
Data Volume:
~25,000 code instruction/response pairs
Methodology:
fine-tuning using CodeAlpaca & GPTeacher datasets to add instruct capabilities
Training Time:
1 hour
Hardware Used:
2x a100 80gb
Input Output 
Input Format:
"### Instruction: ### Input: ### Response:" or "### Instruction: ### Response:"
Output Format:
custom prompt-based
Performance Tips:
Sampler settings: max_new_tokens=128, do_sample=True, use_cache=True, temperature=0.2, top_p=0.9, eos_token_id=self.tokenizer.eos_token_id. Tokenizer decode arguments: skip_special_tokens=True, clean_up_tokenization_space=False
LLM NameReplit V1 CodeInstruct 3B Fp16
Repository 🤗https://huggingface.co/teknium/Replit-v1-CodeInstruct-3B-fp16 
Base Model(s)  Replit V1 CodeInstruct 3B   teknium/Replit-v1-CodeInstruct-3B
Model Size3b
Required VRAM5.2 GB
Updated2025-02-22
Maintainerteknium
Model Typempt
Model Files  5.2 GB   0.0 GB
Supported Languagescode
Quantization Typefp16
Model ArchitectureMPTForCausalLM
Licensecc-by-sa-4.0
Model Max Length512
Transformers Version4.29.2
Tokenizer ClassReplitLMTokenizer
Padding Token<|pad|>
Vocabulary Size32769
Torch Data Typefloat16

Best Alternatives to Replit V1 CodeInstruct 3B Fp16

Best Alternatives
Context / RAM
Downloads
Likes
Replit Code V1.5 3B0K / 6.6 GB90716289
Code Millenials 3B0K / 5.2 GB1321
Sea Lion 3B0K / 6.4 GB51317
Mpt 3B 8K Instruct0K / 6.9 GB43
Replit Code V1 3B0K / 10.4 GB994725
Glaive Function Calling V10K / 10.4 GB11268
...aive Function Calling V2 Small0K / 10.4 GB8914
Replit V2 CodeInstruct 3B GGML0K / 1.5 GB6433
Evol Replit V10K / 10.4 GB138
Replit V2 CodeInstruct 3B0K / 10.4 GB4672
Note: green Score (e.g. "73.2") means that the model is better than teknium/Replit-v1-CodeInstruct-3B-fp16.

Rank the Replit V1 CodeInstruct 3B Fp16 Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation  
Factuality and Completeness of Knowledge  
Censorship and Alignment  
Data Analysis and Insight Generation  
Text Generation  
Text Summarization and Feature Extraction  
Code Generation  
Multi-Language Support and Translation  

What open-source LLMs or SLMs are you in search of? 43508 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227
  Support LLM Explorer