Arxiv:2404.03608 Autotrain compatible Base model:finetune:sail/sailo... Base model:sail/sailor-4b 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
Sailor 4B Chat Benchmarks
ARC: 45.05 vs 96.7 (so35)-53.4%
HellaSwag: 68.36 vs 95.3 (gpt4)-28.3%
MMLU: 43.96 vs 88.3 (so35)-50.2%
TruthfulQA: 42.09 vs 59 (gpt4)-28.7%
WinoGrande: 66.22 vs 87.5 (gpt4)-24.3%
GSM8K: 9.1 vs 96.4 (so35)-90.6%
LLME Score: 0.22706
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 4B Chat Parameters and Internals
| LLM Name | Sailor 4B Chat |
| Repository 🤗 | https://huggingface.co/sail/Sailor-4B-Chat |
| Base Model(s) | |
| Model Size | 4b |
| Required VRAM | 7.9 GB |
| Updated | 2024-10-18 |
| 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 4B Chat
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Nusantara 4B Indo Chat | 32K / 7.9 GB | 453 | 1 |
| Duloxetine 4B V1 | 32K / 7.9 GB | 18 | 3 |
| Qwen1.5 4B Chat Paraph | 32K / 7.9 GB | 15 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than sail/Sailor-4B-Chat .
Rank the Sailor 4B 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? 36966 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v2024072803