DPO Qwen2 by trl-lib

 ยป  All LLMs  ยป  trl-lib  ยป  DPO Qwen2   URL Share it on

  Arxiv:2305.18290   Autotrain compatible Base model:finetune:qwen/qwen2... Base model:qwen/qwen2-0.5b-ins...   Conversational Dataset:trl-lib/capybara-prefe...   Dpo   Endpoints compatible   Generated from trainer   Instruct   Qwen2   Region:us   Safetensors   Trl

DPO Qwen2 Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

DPO Qwen2 Parameters and Internals

LLM NameDPO Qwen2
Repository ๐Ÿค—https://huggingface.co/trl-lib/dpo-qwen2 
Model Namedpo-qwen2
Base Model(s)  Qwen/Qwen2-0.5B-Instruct   Qwen/Qwen2-0.5B-Instruct
Model Size0.5b
Required VRAM2 GB
Updated2024-10-01
Maintainertrl-lib
Model Typeqwen2
Instruction-BasedYes
Model Files  2.0 GB   0.0 GB
Model ArchitectureQwen2ForCausalLM
Context Length32768
Model Max Length32768
Transformers Version4.45.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size151936
Torch Data Typefloat32
Errorsreplace
DPO Qwen2 (trl-lib/dpo-qwen2)

Best Alternatives to DPO Qwen2

Best Alternatives
Context / RAM
Downloads
Likes
Qwen2 0.5B Abyme Merge3128K / 1.3 GB70
Qwen2 0.5B Abyme Merge2128K / 0.3 GB50
Qwen2.5 0.5B Instruct32K / 1 GB4833150
Qwen2 0.5B Instruct32K / 1 GB278170155
Replete LLM V2.5 Qwen 0.5B32K / 1.3 GB1802
Qwen2.5 0.5B Test Ft32K / 1 GB993
Qwen2.5 0.5B Instruct32K / 1 GB4171
Qwen2 0.5B DPO32K / 2 GB562
Qwen2.5 0.5B Instruct32K /  GB322
Qwama 0.5B Instruct32K / 0.9 GB119213
Note: green Score (e.g. "73.2") means that the model is better than trl-lib/dpo-qwen2.

Rank the DPO Qwen2 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? 36542 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v2024072803