LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  Agents   Benchmarks   LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Swallow 70B AWQ by TheBloke

 »  All LLMs  »  TheBloke  »  Swallow 70B AWQ   URL Share it on

  5   

Swallow 70B AWQ is an open-source language model by TheBloke. Features: 70b LLM, VRAM: 37GB, Context: 4K, License: llama2, Quantized, LLM Explorer Score: 0.11.

  4-bit   Awq Base model:quantized:tokyotech... Base model:tokyotech-llm/swall...   En   Ja   Llama   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/TheBloke/Swallow-70B-AWQ 

Swallow 70B AWQ Benchmarks

LLME Score: 0.10966
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Swallow 70B AWQ (TheBloke/Swallow-70B-AWQ)
🌟 Advertise your project 🚀

Swallow 70B AWQ Parameters and Internals

Model Type 
text-generation
Use Cases 
Areas:
Research, Text generation
Additional Notes 
Swallow model versions range from 7B to 70B, focusing mainly on Japanese-English bilingual tasks. The model employs advanced quantization techniques for efficient inference.
Supported Languages 
en (English), ja (Japanese)
Training Details 
Data Sources:
Japanese Wikipedia, RefinedWeb, Swallow Corpus, The Pile
Methodology:
Continual pre-training and instruction tuning using datasets like Anthropic HH-RLHF and Databricks Dolly 15-k.
Context Length:
4096
Model Architecture:
Llama-2 architecture based, with a Japanese vocabulary extended tokenizer.
Input Output 
Input Format:
{prompt}
Accepted Modalities:
text
Output Format:
Generated text
Performance Tips:
Efficient text generation with Japanese input due to optimized tokenizer. For fast inference, GPU usage is recommended.
LLM NameSwallow 70B AWQ
Repository 🤗https://huggingface.co/TheBloke/Swallow-70B-AWQ 
Model NameSwallow 70B
Model Creatortokyotech-llm
Base Model(s)  Swallow 70B Hf   tokyotech-llm/Swallow-70b-hf
Model Size70b
Required VRAM37 GB
Updated2026-04-09
MaintainerTheBloke
Model Typellama
Model Files  9.9 GB: 1-of-4   10.0 GB: 2-of-4   9.9 GB: 3-of-4   7.2 GB: 4-of-4
Supported Languagesen ja
AWQ QuantizationYes
Quantization Typeawq
Model ArchitectureLlamaForCausalLM
Licensellama2
Context Length4096
Model Max Length4096
Transformers Version4.37.0.dev0
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size43176
Torch Data Typefloat16

Best Alternatives to Swallow 70B AWQ

Best Alternatives
Context / RAM
Downloads
Likes
...0B Instruct Gradient 1048K AWQ1024K / 39.9 GB31
...70B Instruct Gradient 262K AWQ256K / 39.9 GB60
Llama 3.3 70B Instruct AWQ128K / 39.9 GB41904041
Llama 3.3 70B Instruct AWQ128K / 39.9 GB3474586
...lama 3.3 70B Instruct AWQ INT4128K / 39.9 GB10239430
... SauerkrautLM 70B Instruct AWQ128K / 39.9 GB145
MultiVerse 70B AWQ32K / 41.3 GB892
Opus V1.2 70B AWQ32K / 36.7 GB71
QuartetAnemoi 70B T0.0001 AWQ31K / 36.7 GB51
Senku 70B AWQ 4bit GEMM31K / 36.7 GB11
Note: green Score (e.g. "73.2") means that the model is better than TheBloke/Swallow-70B-AWQ.

Rank the Swallow 70B AWQ Capabilities

🆘 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! 🌟

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? 52721 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a
  Support LLM Explorer