Meta Llama 3 8B Instruct Zh 10K by XavierSpycy

 ยป  All LLMs  ยป  XavierSpycy  ยป  Meta Llama 3 8B Instruct Zh 10K   URL Share it on

  Arxiv:2403.13372   Auto-gptq   Autoawq   Autotrain compatible Base model:adapter:meta-llama/... Base model:meta-llama/meta-lla...   Conversational Dataset:llamafactory/alpaca gp...   Dataset:llamafactory/alpaca zh   En   Endpoints compatible   Gptq   Instruct   Llama   Llama.cpp   Lora   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Zh

Meta Llama 3 8B Instruct Zh 10K Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Meta Llama 3 8B Instruct Zh 10K (XavierSpycy/Meta-Llama-3-8B-Instruct-zh-10k)

Meta Llama 3 8B Instruct Zh 10K Parameters and Internals

Model Type 
text-generation, transformers, lora
Use Cases 
Areas:
research, commercial applications
Primary Use Cases:
Text generation in Chinese, AI assistant
Limitations:
Needs further benchmarks for capability evaluation, Limited by quality and quantity of Chinese corpora used, Improvements needed in role-playing, mathematics, and handling complex tasks
Additional Notes 
The model is optimized for Chinese language operations utilizing LoRa technology for adaptation.
Supported Languages 
en (basic support), zh (enhanced support)
Training Details 
Data Sources:
llamafactory/alpaca_zh, llamafactory/alpaca_gpt4_zh, oaast_sft_zh
Data Volume:
10,000 examples
Methodology:
Fine-tuning with LoRa technology
Training Time:
12 hours
Hardware Used:
RTX 4090D x 1
Model Architecture:
Adapted Meta-Llama-3-8B-Instruct for better Chinese handling
Safety Evaluation 
Risk Categories:
bias, safety
Ethical Considerations:
Refer to Meta Llama 3's Ethical Considerations for information on bias monitoring, responsible usage guidelines, and model limitation transparency.
Responsible Ai Considerations 
Fairness:
Bias monitoring is important.
Transparency:
Transparency in model limitations is necessary.
Input Output 
Input Format:
text prompts
Accepted Modalities:
text
Output Format:
text responses
Performance Tips:
The model offers enhanced performance in Chinese compared to its predecessor.
LLM NameMeta Llama 3 8B Instruct Zh 10K
Repository ๐Ÿค—https://huggingface.co/XavierSpycy/Meta-Llama-3-8B-Instruct-zh-10k 
Base Model(s)  Meta Llama 3 8B Instruct   meta-llama/Meta-Llama-3-8B-Instruct
Model Size8b
Required VRAM16.1 GB
Updated2025-01-22
MaintainerXavierSpycy
Model Typellama
Instruction-BasedYes
Model Files  2.0 GB: 1-of-9   1.9 GB: 2-of-9   2.0 GB: 3-of-9   1.9 GB: 4-of-9   2.0 GB: 5-of-9   1.9 GB: 6-of-9   2.0 GB: 7-of-9   1.3 GB: 8-of-9   1.1 GB: 9-of-9
Supported Languagesen zh
GPTQ QuantizationYes
Quantization Typegptq
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length8192
Model Max Length8192
Transformers Version4.41.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|eot_id|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to Meta Llama 3 8B Instruct Zh 10K

Best Alternatives
Context / RAM
Downloads
Likes
...a 3 8B Instruct 262K 4bit GPTQ256K / 5.8 GB911
... 8B Instruct 262K 4bit GPTQ 02256K / 5.7 GB890
...lama 3.1 8B Instruct GPTQ INT4128K / 5.8 GB4151621
...Instruct 80K Qlora Merged GPTQ80K / 5.8 GB160
...oLeo Instruct 8B 32K V0.1 GPTQ64K / 5.7 GB170
Llama 3 8B Instruct GPTQ 4 Bit8K / 5.7 GB1021825
Meta Llama 3 8B Instruct GPTQ8K / 5.8 GB900
Meta Llama 3 8B Instruct GPTQ8K / 5.8 GB901
Meta Llama 3 8B Instruct GPTQ8K / 5.7 GB3723
...truct Abliterated V3 GPTQ 4bit8K / 5.8 GB970
Note: green Score (e.g. "73.2") means that the model is better than XavierSpycy/Meta-Llama-3-8B-Instruct-zh-10k.

Rank the Meta Llama 3 8B Instruct Zh 10K 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? 41728 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227