LLM EXPLORER 59,420 MODELS INDEXED

Mergkit 1 by GhaithChaabane

By GhaithChaabane · 5 downloads
HF Repo ↗ Copy URL

Mergkit 1 is an open-source language model by GhaithChaabane. Features: 8b LLM, VRAM: 22.6GB, Context: 8K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.13.

  Conversational   Endpoints compatible   Instruct   Mixtral   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/GhaithChaabane/mergkit_1 

Mergkit 1 Parameters and Internals

LLM NameMergkit 1
Repository 🤗https://huggingface.co/GhaithChaabane/mergkit_1 
Model Size8b
Required VRAM22.6 GB
Updated2026-05-24
MaintainerGhaithChaabane
Model Typemixtral
Instruction-BasedYes
Model Files  1.1 GB: 1-of-24   1.0 GB: 2-of-24   1.0 GB: 3-of-24   1.0 GB: 4-of-24   1.0 GB: 5-of-24   1.0 GB: 6-of-24   1.0 GB: 7-of-24   1.0 GB: 8-of-24   1.0 GB: 9-of-24   1.0 GB: 10-of-24   1.0 GB: 11-of-24   1.0 GB: 12-of-24   1.0 GB: 13-of-24   1.0 GB: 14-of-24   1.0 GB: 15-of-24   1.0 GB: 16-of-24   1.0 GB: 17-of-24   1.0 GB: 18-of-24   1.0 GB: 19-of-24   1.0 GB: 20-of-24   1.0 GB: 21-of-24   0.4 GB: 22-of-24   1.1 GB: 23-of-24   0.0 GB: 24-of-24
Model ArchitectureMixtralForCausalLM
Licenseapache-2.0
Context Length8192
Model Max Length8192
Transformers Version4.41.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|begin_of_text|>
Vocabulary Size128256
Torch Data Typefloat16

Best Alternatives to Mergkit 1

Best Alternatives
Context / RAM
Downloads
Likes
Lamma3merge3 15B MoE8K / 27.5 GB111
Lamma3merge2 15B MoE8K / 27.5 GB100
Llama3merge8 15B MoE8K / 27.5 GB50
Note: green Score (e.g. "73.2") means that the model is better than GhaithChaabane/mergkit_1.