LLM EXPLORER 59,358 MODELS INDEXED

Qwen2.5 0.5B Instruct SFT MDPO 1epoch V1 by JayHyeon

By JayHyeon · 7 downloads

Qwen2.5 0.5B Instruct SFT MDPO 1epoch V1 is an open-source language model by JayHyeon. Features: 0.5b LLM, VRAM: 2GB, Context: 32K, License: mit, Instruction-Based, LLM Explorer Score: 0.18.

  Arxiv:1910.09700   Endpoints compatible   Feature-extraction   Instruct   Qwen2   Region:us   Safetensors   Text-embeddings-inference

Qwen2.5 0.5B Instruct SFT MDPO 1epoch V1 Parameters and Internals

LLM NameQwen2.5 0.5B Instruct SFT MDPO 1epoch V1
Repository πŸ€—https://huggingface.co/JayHyeon/Qwen2.5-0.5B-Instruct-SFT-MDPO-1epoch_v1 
Model Size0.5b
Required VRAM2 GB
Updated2026-08-02
MaintainerJayHyeon
Model Typeqwen2
Instruction-BasedYes
Model Files  2.0 GB
Model ArchitectureQwen2Model
Licensemit
Context Length32768
Model Max Length32768
Transformers Version4.47.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|im_end|>
Vocabulary Size151936
Torch Data Typefloat32
Errorsreplace

Best Alternatives to Qwen2.5 0.5B Instruct SFT MDPO 1epoch V1

Best Alternatives
Context / RAM
Downloads
Likes
...5B Instruct SFT IRPO 1epoch V132K / 2 GB70
....5B Instruct SFT DPO 1epoch V132K / 2 GB220
Note: green Score (e.g. "73.2") means that the model is better than JayHyeon/Qwen2.5-0.5B-Instruct-SFT-MDPO-1epoch_v1.