LLM EXPLORER 59,962 MODELS INDEXED

Mallam 1.1B 20K Instructions by mesolitica

By mesolitica · 67 downloads
HF Repo ↗ Copy URL

Mallam 1.1B 20K Instructions is an open-source language model by mesolitica. Features: 1b LLM, VRAM: 2.2GB, Context: 32K, Instruction-Based, LLM Explorer Score: 0.1.

  Conversational   Endpoints compatible   Instruct   Mistral   Ms   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/mesolitica/mallam-1.1b-20k-instructions 

Mallam 1.1B 20K Instructions Parameters and Internals

Model Type 
text generation
Use Cases 
Areas:
Research, Quick demonstration
Primary Use Cases:
Fine-tuning for text generation
Limitations:
Minimal moderation mechanisms
Additional Notes 
Demonstrates ease of fine-tuning Mistral-based models.
Supported Languages 
Malay (fluent)
Training Details 
Data Sources:
https://huggingface.co/collections/mesolitica/malaysian-synthetic-dataset-656c2673fe7fe0b1e9e25fe2
Data Volume:
Malaysian instructions dataset
Methodology:
Finetuning on Mistral Instruct chat template
Context Length:
20480
Input Output 
Input Format:
Mistral chat template
Accepted Modalities:
text
Output Format:
Generated text response
LLM NameMallam 1.1B 20K Instructions
Repository 🤗https://huggingface.co/mesolitica/mallam-1.1b-20k-instructions 
Model Size1b
Required VRAM2.2 GB
Updated2026-08-08
Maintainermesolitica
Model Typemistral
Instruction-BasedYes
Model Files  2.2 GB
Supported Languagesms
Model ArchitectureMistralForCausalLM
Context Length32768
Model Max Length32768
Transformers Version4.35.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<pad>
Vocabulary Size32000
Torch Data Typebfloat16