TeleChat 12B V2 by Tele-AI

 »  All LLMs  »  Tele-AI  »  TeleChat 12B V2   URL Share it on

  Arxiv:1910.07467   Arxiv:2002.05202   Arxiv:2104.09864   Arxiv:2401.03804   Autotrain compatible   Custom code   Pytorch   Region:us   Telechat

TeleChat 12B V2 Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
TeleChat 12B V2 (Tele-AI/TeleChat-12B-v2)

TeleChat 12B V2 Parameters and Internals

Model Type 
chat model, language model
Additional Notes 
Supports deepspeed-based training and multi-round data construction with enhanced long text generation capabilities.
Training Details 
Data Sources:
网页, 书籍, 官方媒体
Data Volume:
1.5万亿 Tokens for 7B and 3万亿 Tokens for 12B
Model Architecture:
Decoder-only
Input Output 
Performance Tips:
Supports multi-round abilities and better focus on multi-round answers.
Release Notes 
Version:
12B-v2
Date:
2024.5.17
Notes:
Release of 12B-v2 version chat model with quantized versions.
Version:
12B
Date:
2024.3.20
Notes:
Release of 12B version chat model with quantized versions.
Version:
1T
Date:
2024.1.11
Notes:
Release of 1T Chinese dataset.
Version:
7B
Date:
2024.1.10
Notes:
Release of 7B version chat model with quantized versions.
LLM NameTeleChat 12B V2
Repository 🤗https://huggingface.co/Tele-AI/TeleChat-12B-v2 
Model Size12b
Required VRAM5.9 GB
Updated2025-01-22
MaintainerTele-AI
Model Typetelechat
Model Files  5.9 GB: 1-of-4   5.9 GB: 2-of-4   5.9 GB: 3-of-4   7.1 GB: 4-of-4
Model ArchitectureTelechatForCausalLM
Model Max Length100000000
Transformers Version4.24.0
Tokenizer ClassTelechatTokenizer
End of Sentence Token<_end>
Unk Token<_end>
Vocabulary Size120000

Best Alternatives to TeleChat 12B V2

Best Alternatives
Context / RAM
Downloads
Likes
TeleChat 12B0K / 5.9 GB59711
TeleChat 12B Int80K / 13.9 GB192
TeleChat 12B Int40K / 8.3 GB92
Note: green Score (e.g. "73.2") means that the model is better than Tele-AI/TeleChat-12B-v2.

Rank the TeleChat 12B V2 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? 41728 in total.

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