Sailor 0.5B Chat is an open-source language model by sail. Features: 0.5b LLM, VRAM: 1.2GB, Context: 32K, License: apache-2.0, Instruction-Based, HF Score: 33.5, LLM Explorer Score: 0.17, Arc: 30.4, HellaSwag: 45.5, MMLU: 26.7, TruthfulQA: 39.9, WinoGrande: 56.5, GSM8K: 1.8.
Arxiv:2404.03608 Base model:finetune:sail/sailo... Base model:sail/sailor-0.5b Chat Conversational Dataset:cohereforai/aya collec... Dataset:cohereforai/aya datase... Dataset:open-orca/openorca En Id Instruct Instruction Lo Ms Multilingual Qwen2 Region:us Safetensors Sailor Sea Sft Th Vi Zh
Model Card on HF 🤗: https://huggingface.co/sail/Sailor-0.5B-Chat
Sailor 0.5B Chat Benchmarks
ARC: 30.38 vs 96.7 (so35)-68.6%
HellaSwag: 45.51 vs 95.3 (gpt4)-52.2%
MMLU: 26.73 vs 88.3 (so35)-69.7%
TruthfulQA: 39.85 vs 59 (gpt4)-32.5%
WinoGrande: 56.51 vs 87.5 (gpt4)-35.4%
GSM8K: 1.82 vs 96.4 (so35)-98.1%
LLME Score: 0.17165
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Sailor 0.5B Chat Parameters and Internals
| LLM Name | Sailor 0.5B Chat |
| Repository 🤗 | https://huggingface.co/sail/Sailor-0.5B-Chat |
| Base Model(s) | |
| Model Size | 0.5b |
| Required VRAM | 1.2 GB |
| Updated | 2026-03-30 |
| Maintainer | sail |
| Model Type | qwen2 |
| Instruction-Based | Yes |
| Model Files | |
| Supported Languages | en zh id th vi ms lo |
| Model Architecture | Qwen2ForCausalLM |
| License | apache-2.0 |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Transformers Version | 4.37.2 |
| Tokenizer Class | Qwen2Tokenizer |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 151936 |
| Torch Data Type | bfloat16 |
| Errors | replace |
Best Alternatives to Sailor 0.5B Chat
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Qwen2 0.5B Abyme Merge3 | 128K / 1.3 GB | 5 | 1 |
| Qwen2 0.5B Abyme Merge2 | 128K / 0.3 GB | 29 | 0 |
| Qwen2.5 0.5B Instruct | 64K / 1 GB | 7 | 0 |
| Qwen2.5 0.5B Instruct | 32K / 1 GB | 5717441 | 494 |
| Qwen2.5 0.5B Instruct | 32K / 1 GB | 3570029 | 22 |
| Qwen2 0.5B Instruct | 32K / 1 GB | 548631 | 200 |
| Qwen2.5 Coder 0.5B Instruct | 32K / 1 GB | 27326 | 52 |
| Tanzania 0.5B | 32K / 2 GB | 24 | 0 |
| ...etrain Mix Low Tweet 1M En GPT | 32K / 1 GB | 21 | 0 |
| ...l Pretrain Control Tweet 1M En | 32K / 1 GB | 20 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than sail/Sailor-0.5B-Chat .
Rank the Sailor 0.5B Chat 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! 🌟
What open-source LLMs or SLMs are you in search of? 52721 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a