Llama 2 70B Chat 4bit Japanese V1 by doshisha-mil

 ยป  All LLMs  ยป  doshisha-mil  ยป  Llama 2 70B Chat 4bit Japanese V1   URL Share it on

  4bit   Adapter   Facebook   Finetuned   Ja   Llama2   Lora   Meta   Peft   Pytorch   Quantized   Region:us

Llama 2 70B Chat 4bit Japanese V1 Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Llama 2 70B Chat 4bit Japanese V1 (doshisha-mil/llama-2-70b-chat-4bit-japanese-v1)

Llama 2 70B Chat 4bit Japanese V1 Parameters and Internals

Model Type 
text-generation
Additional Notes 
The model is fine-tuned using the Japanese version of the alpaca dataset.
Supported Languages 
ja (Japanese)
Training Details 
Data Sources:
alpaca_ja dataset (https://github.com/shi3z/alpaca_ja)
Methodology:
fine-tuning
Input Output 
Input Format:
Text input as a string with structure - questions and answers
Accepted Modalities:
text
Output Format:
Text output as a generated response
LLM NameLlama 2 70B Chat 4bit Japanese V1
Repository ๐Ÿค—https://huggingface.co/doshisha-mil/llama-2-70b-chat-4bit-japanese-v1 
Model Size70b
Required VRAM3.3 GB
Updated2024-12-22
Maintainerdoshisha-mil
Model Files  3.3 GB
Supported Languagesja
Quantization Type4bit
Model ArchitectureAdapter
Licensellama2
Is Biasednone
PEFT TypeLORA
LoRA ModelYes
PEFT Target Modulesdown_proj|o_proj|k_proj|gate_proj|q_proj|up_proj|v_proj
LoRA Alpha16
LoRA Dropout0.1
R Param64

Best Alternatives to Llama 2 70B Chat 4bit Japanese V1

Best Alternatives
Context / RAM
Downloads
Likes
Llama3 70b Ft0K / 39.9 GB60
...LM 2Bit QLoRA Function Calling0K / 0.1 GB137
Llama 2 70B Chat 4bit Japanese0K / 3.3 GB25
...aiga Llama3 70b Sft M1 D5 Lora0K / 5.9 GB01
Llama 3 70B Instruct Spider0K / 141.9 GB60
Airoboros 70B 3.3 Peft0K / 0.4 GB02
Llama3v10K / 0.1 GB50
Xwin LM 70B V0.1 LORA0K / 1.7 GB01
Euryale 1.3 L2 70B LORA0K / 1.7 GB31
Miqu 1 70B Hermes2.5 Qlora0K / 4.8 GB04
Note: green Score (e.g. "73.2") means that the model is better than doshisha-mil/llama-2-70b-chat-4bit-japanese-v1.

Rank the Llama 2 70B Chat 4bit Japanese V1 Capabilities

๐Ÿ†˜ 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! ๐ŸŒŸ

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? 40123 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241217