Stable Code 3B GPTQ by TechxGenus

 ยป  All LLMs  ยป  TechxGenus  ยป  Stable Code 3B GPTQ   URL Share it on

  Arxiv:1910.02054   Arxiv:2104.09864   Arxiv:2204.06745   Arxiv:2305.06161   Arxiv:2307.09288   Arxiv:2309.12284   Arxiv:2310.10631   4-bit   Autotrain compatible   Code   Dataset:bigcode/commitpackft   Dataset:bigcode/starcoderdata Dataset:bigcode/the-stack-gith... Dataset:eleutherai/proof-pile-...   Dataset:meta-math/metamathqa Dataset:tiiuae/falcon-refinedw...   En   Endpoints compatible   Gptq   Model-index   Quantized   Region:us   Safetensors   Stablelm

Stable Code 3B GPTQ Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Stable Code 3B GPTQ (TechxGenus/stable-code-3b-GPTQ)

Stable Code 3B GPTQ Parameters and Internals

Model Type 
text generation, code
Use Cases 
Areas:
research, commercial applications
Limitations:
Potentially unreliable, unsafe, or undesirable behaviors without further fine-tuning.
Considerations:
Evaluate and fine-tune for safe performance in downstream applications.
Additional Notes 
Demo example code snippets to get started with the model are provided.
Supported Languages 
English (fluent), Code (multi-language)
Training Details 
Data Sources:
tiiuae/falcon-refinedweb, bigcode/the-stack-github-issues, bigcode/commitpackft, bigcode/starcoderdata, meta-math/MetaMathQA
Data Volume:
1.3 trillion tokens
Methodology:
decoder-only transformer, similar to LLaMA with modifications
Context Length:
16384
Hardware Used:
256 NVIDIA A100 40GB GPUs (AWS P4d instances)
Model Architecture:
decoder-only transformer architecture with rotary position embeddings and SwiGLU with flash-attention.
Input Output 
Accepted Modalities:
text, code
LLM NameStable Code 3B GPTQ
Repository ๐Ÿค—https://huggingface.co/TechxGenus/stable-code-3b-GPTQ 
Model Size3b
Required VRAM1.8 GB
Updated2024-12-23
MaintainerTechxGenus
Model Typestablelm
Model Files  1.8 GB
Supported Languagesen
GPTQ QuantizationYes
Quantization Typegptq
Model ArchitectureStableLmForCausalLM
Licenseother
Context Length16384
Model Max Length16384
Transformers Version4.39.3
Tokenizer ClassGPTNeoXTokenizer
Vocabulary Size50304
Torch Data Typefloat16

Best Alternatives to Stable Code 3B GPTQ

Best Alternatives
Context / RAM
Downloads
Likes
Stable Code Instruct 3B 4bit16K / 1.8 GB335
Stable Code 3B 4bit16K / 1.8 GB221
Stable Code 3B16K / 5.6 GB8139635
Stable Cypher Instruct 3B16K / 1.7 GB82922
Stable Code Instruct 3B16K / 5.6 GB1869167
...t 3B Mix Spider Bird 200 Steps16K / 5.6 GB110
...struct 3B Spider 1500 Steps Q416K / 1.7 GB150
... Instruct 3B Spider 3500 Steps16K / 11.2 GB100
Stable Code Instruct 3B AWQ16K / 1.8 GB170
Stablelm Zephyr 3B4K / 5.6 GB10011249
Note: green Score (e.g. "73.2") means that the model is better than TechxGenus/stable-code-3b-GPTQ.

Rank the Stable Code 3B GPTQ Capabilities

๐Ÿ†˜ 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! ๐ŸŒŸ

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? 40123 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241217