LLM EXPLORER 63,407 MODELS INDEXED

Zephyr 7B Beta GPTQ by TheBloke

By TheBloke · 215 downloads
HF Repo ↗ Copy URL

Zephyr 7B Beta GPTQ is an open-source language model by TheBloke. Features: 7b LLM, VRAM: 4.2GB, Context: 32K, License: mit, Quantized, LLM Explorer Score: 0.09.

  Arxiv:2305.18290   Arxiv:2310.16944   4-bit Base model:huggingfaceh4/zephy... Base model:quantized:huggingfa...   Conversational Dataset:huggingfaceh4/ultracha... Dataset:huggingfaceh4/ultrafee...   Deploy:azure   En   Generated from trainer   Gptq   Mistral   Quantized   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/TheBloke/zephyr-7B-beta-GPTQ 

Zephyr 7B Beta GPTQ Parameters and Internals

Model Type 
GPT-like, fine-tuned
Use Cases 
Primary Use Cases:
chat-based applications, testing via demo
Limitations:
problematic outputs possible, educational and research use recommended
Considerations:
not aligned to human preferences with techniques like RLHF
Supported Languages 
en (primary)
Training Details 
Data Sources:
HuggingFaceH4/ultrachat_200k, HuggingFaceH4/ultrafeedback_binarized
Methodology:
Direct Preference Optimization
Hardware Used:
multi-GPU (16 devices)
Input Output 
Input Format:
<|system|> ~~ <|user|> {prompt}~~ <|assistant|>
LLM NameZephyr 7B Beta GPTQ
Repository 🤗https://huggingface.co/TheBloke/zephyr-7B-beta-GPTQ 
Model NameZephyr 7B Beta
Model CreatorHugging Face H4
Base Model(s)  Zephyr 7B Beta   HuggingFaceH4/zephyr-7b-beta
Model Size7b
Required VRAM4.2 GB
Updated2026-07-30
MaintainerTheBloke
Model Typemistral
Model Files  4.2 GB
Supported Languagesen
GPTQ QuantizationYes
Quantization Typegptq
Model ArchitectureMistralForCausalLM
Licensemit
Context Length32768
Model Max Length32768
Transformers Version4.34.1
Tokenizer ClassLlamaTokenizer
Padding Token</s>
Vocabulary Size32000
Torch Data Typefloat16

Quantized Models of the Zephyr 7B Beta GPTQ

Model
Likes
Downloads
VRAM
Zephyr Finetuned Legal120 GB
Cesar Ptbr060 GB

Best Alternatives to Zephyr 7B Beta GPTQ

Best Alternatives
Context / RAM
Downloads
Likes
...enHermes 2.5 Mistral 7B Marlin32K / 4.1 GB62
Zephyr 7B Beta Marlin32K / 4.1 GB160
Mistral 7B Instruct V0.3 GPTQ32K / 4.2 GB1138811
...ral 7B Instruct V0.3 GPTQ 4bit32K / 4.2 GB294618
...ral 7B Instruct V0.3 GPTQ 4bit32K / 4.2 GB265825
Cosmosage V232K / 4.2 GB84
...istral 7B Pruned50 GPTQ Marlin32K / 4 GB50
Mistral 7B Unsloth Gptq 8bit32K / 7.7 GB70
Mistral 7B Instruct V0.2 GPTQ32K / 4.2 GB674356
...phyr 7B Beta Assistant V1 Gptq32K / 4.2 GB21
Note: green Score (e.g. "73.2") means that the model is better than TheBloke/zephyr-7B-beta-GPTQ.