Autotrain compatible Base model:finetune:sakalti/ta... Base model:sakalti/tara-3.8b-v... Conversational Dataset:hachiml/hachi-alpaca Dataset:yahma/alpaca-cleaned En Endpoints compatible Llama Region:us Safetensors Sft Sharded Tensorflow Trl Unsloth
Model Card on HF 🤗: https://huggingface.co/Sakalti/SJT-4B
SJT 4B Benchmarks
MMLU Pro: 25.35
GPQA: 5.93
MUSR: 19.44
BBH: 28.91
IFEval: 40.77 vs 88 (so35)-53.7%
MATH Lvl 5: 11.56
LLME Score: 0.28234
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)
SJT 4B Parameters and Internals
| LLM Name | SJT 4B |
| Repository 🤗 | https://huggingface.co/Sakalti/SJT-4B |
| Base Model(s) | |
| Model Size | 4b |
| Required VRAM | 7.6 GB |
| Updated | 2025-06-11 |
| Maintainer | Sakalti |
| Model Type | llama |
| Model Files | |
| Supported Languages | en |
| Model Architecture | LlamaForCausalLM |
| License | mit |
| Context Length | 150000 |
| Model Max Length | 150000 |
| Transformers Version | 4.47.1 |
| Tokenizer Class | LlamaTokenizer |
| Padding Token | <|placeholder6|> |
| Vocabulary Size | 32064 |
| Torch Data Type | float16 |
Best Alternatives to SJT 4B
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Xgen Small 4B Instruct R | 256K / 17.7 GB | 41 | 3 |
| ...lama 3.1 Nemotron Nano 4B V1.1 | 128K / 9 GB | 17290 | 88 |
| ...lama 3.1 Nemotron Nano 4B V1.1 | 128K / 9 GB | 1064 | 4 |
| Basqui R1 4B V1 | 128K / 6.5 GB | 309 | 1 |
| Loxa 4B | 128K / 16 GB | 158 | 0 |
| Nemotron W 4b MagLight 0.1 | 128K / 9.2 GB | 15 | 3 |
| CharGen V3 Mini | 128K / 9 GB | 50 | 6 |
| Nemotron W 4b Halo 0.1 | 128K / 9.2 GB | 19 | 3 |
| Aura 4B | 128K / 9 GB | 45 | 11 |
| ...ama 3.1 Minitron 4B Depth Base | 128K / 9.1 GB | 3341 | 21 |
Note: green Score (e.g. "73.2") means that the model is better than Sakalti/SJT-4B .
Rank the SJT 4B 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? 48104 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124