LLM EXPLORER 60,783 MODELS INDEXED

LaMini Cerebras 1.3B by MBZUAI

By MBZUAI · 90 downloads
HF Repo ↗ Copy URL

LaMini Cerebras 1.3B is an open-source language model by MBZUAI. Features: 1.3b LLM, VRAM: 5.4GB, License: cc-by-nc-4.0, LLM Explorer Score: 0.06.

  Arxiv:2304.14402   En   Endpoints compatible   Gpt2   Pytorch   Region:us
Model Card on HF 🤗: https://huggingface.co/MBZUAI/LaMini-Cerebras-1.3B 

LaMini Cerebras 1.3B Parameters and Internals

Model Type 
text generation
Use Cases 
Areas:
responding to human instructions
Primary Use Cases:
natural language instruction following
Additional Notes 
The model is a fine-tuned version of Cerebras-GPT-1.3B for improved performance on instruction tasks.
Supported Languages 
en (fluent)
Training Details 
Data Sources:
LaMini-instruction dataset
Data Volume:
2.58M samples
Model Architecture:
Cerebras-GPT
Input Output 
Input Format:
Instruction text
Accepted Modalities:
text
Output Format:
Formatted text response
Performance Tips:
Use the same wrapper text for improved performance
LLM NameLaMini Cerebras 1.3B
Repository 🤗https://huggingface.co/MBZUAI/LaMini-Cerebras-1.3B 
Model Size1.3b
Required VRAM5.4 GB
Updated2026-07-29
MaintainerMBZUAI
Model Typegpt2
Model Files  5.4 GB
Supported Languagesen
Model ArchitectureGPT2LMHeadModel
Licensecc-by-nc-4.0
Model Max Length512
Transformers Version4.28.0.dev0
Tokenizer ClassGPT2Tokenizer
Beginning of Sentence Token<|endoftext|>
End of Sentence Token<|endoftext|>
Unk Token<|endoftext|>
Vocabulary Size50258
Torch Data Typefloat32
Activation Functiongelu
Errorsreplace

Best Alternatives to LaMini Cerebras 1.3B

Best Alternatives
Context / RAM
Downloads
Likes
Novi Nano Base0K / 0 GB1671
Cerebras 1.3b Quantized0K / 1.4 GB660
GPT Sw3 1.3B Instruct0K / 5.5 GB10683
Quokka 1.3B0K / 2.7 GB750
1.3B0K / 2.7 GB802
GPT Sw3 1.3B0K / 5.5 GB426
MGPT 1.3B Kazakh0K / 5.8 GB218
Note: green Score (e.g. "73.2") means that the model is better than MBZUAI/LaMini-Cerebras-1.3B.