LLM EXPLORER 59,511 MODELS INDEXED

Upstage Llama 30B Instruct 2048 AWQ by TheBloke

By TheBloke · 3 downloads
HF Repo ↗ Copy URL

Upstage Llama 30B Instruct 2048 AWQ is an open-source language model by TheBloke. Features: 30b LLM, VRAM: 17.6GB, Context: 2K, License: other, Quantized, Instruction-Based, LLM Explorer Score: 0.09.

  4-bit   Awq Base model:quantized:upstage/l... Base model:upstage/llama-30b-i...   Dataset:gair/lima Dataset:metaeval/scienceqa tex...   Dataset:open-orca/openorca   Dataset:openbookqa   Dataset:sciq   En   Instruct   Instruction   Llama   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Upstage
Model Card on HF 🤗: https://huggingface.co/TheBloke/upstage-llama-30b-instruct-2048-AWQ 

Upstage Llama 30B Instruct 2048 AWQ Parameters and Internals

Model Type 
text-generation
Use Cases 
Areas:
research, commercial applications
Supported Languages 
en (English - proficient)
Training Details 
Data Sources:
openbookqa, sciq, Open-Orca/OpenOrca, metaeval/ScienceQA_text_only, GAIR/lima
Methodology:
fine-tuned from LLaMA
Context Length:
10000
Hardware Used:
A100x8 * 1
Model Architecture:
Transformer-based
Responsible Ai Considerations 
Mitigation Strategies:
no ethical issues as benchmark test set or training set were not included in training
Input Output 
Input Format:
### System: {System} ### User: {User} ### Assistant: {Assistant}
Accepted Modalities:
text
Output Format:
text
LLM NameUpstage Llama 30B Instruct 2048 AWQ
Repository 🤗https://huggingface.co/TheBloke/upstage-llama-30b-instruct-2048-AWQ 
Model NameLlama 30B Instruct 2048
Model Creatorupstage
Base Model(s)  Llama 30B Instruct 2048   upstage/llama-30b-instruct-2048
Model Size30b
Required VRAM17.6 GB
Updated2026-07-28
MaintainerTheBloke
Model Typellama
Instruction-BasedYes
Model Files  10.0 GB: 1-of-2   7.6 GB: 2-of-2
Supported Languagesen
AWQ QuantizationYes
Quantization Typeawq
Model ArchitectureLlamaForCausalLM
Licenseother
Context Length2048
Model Max Length2048
Transformers Version4.30.2
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size32000
Torch Data Typebfloat16

Best Alternatives to Upstage Llama 30B Instruct 2048 AWQ

Best Alternatives
Context / RAM
Downloads
Likes
WizardLM 30B Uncensored AWQ2K / 17.6 GB74
Tildeopen 30B Mu Instruct64K / 60.9 GB416
...ct 2048 Open Platypus 100steps2K / 65.2 GB880
Llama 30B Instruct 20482K / 65.2 GB374103
...azarus Instruct PL Lora Unload2K / 65.2 GB850
Llama 30B Instruct2K / 65.2 GB33324
...B 2048 Instruct PL Lora Unload2K / 65.2 GB971
...lama 30B Instruct 2048 PL Lora2K / 65.2 GB890
...e Llama 30B Instruct 2048 GPTQ2K / 16.9 GB2722
Open Instruct Human Mix 30B2K / 130.5 GB581
Note: green Score (e.g. "73.2") means that the model is better than TheBloke/upstage-llama-30b-instruct-2048-AWQ.