GPT Oss 20B BF16 is an open-source language model by unsloth. Features: 20b LLM, VRAM: 41.6GB, Context: 128K, License: apache-2.0, LLM Explorer Score: 0.3.
| LLM Name | GPT Oss 20B BF16 |
| Repository 🤗 | https://huggingface.co/unsloth/gpt-oss-20b-BF16 |
| Base Model(s) | |
| Model Size | 20b |
| Required VRAM | 41.6 GB |
| Updated | 2025-09-23 |
| Maintainer | unsloth |
| Model Type | gpt_oss |
| Model Files | |
| Model Architecture | GptOssForCausalLM |
| License | apache-2.0 |
| Context Length | 131072 |
| Model Max Length | 131072 |
| Transformers Version | 4.56.0.dev0 |
| Tokenizer Class | PreTrainedTokenizerFast |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 201088 |
| Torch Data Type | bfloat16 |
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| GPT Oss 20B | 128K / 13.8 GB | 6499172 | 4492 |
| OTel LLM 20B Reasoning | 128K / 83.8 GB | 86621 | 0 |
| Context 1 | 128K / 41.8 GB | 4317 | 381 |
| GPT Oss 20B | 128K / 13.8 GB | 41873 | 35 |
| GPT Oss Safeguard 20B | 128K / 13.8 GB | 47089 | 206 |
| GPT Oss 20B Heretic Ara V4 | 128K / 13.8 GB | 462 | 11 |
| GPT Oss 20B Abliterated | 128K / 41.6 GB | 332 | 2 |
| GPT Oss 20B Heretic | 128K / 41.6 GB | 3612 | 103 |
| GPT Oss 20B Ja Fin Thinking | 128K / 41.6 GB | 405 | 8 |
| ...ecurity 20B Merged Heretic Ara | 128K / 41.6 GB | 228 | 1 |
🆘 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! 🌟