LLM EXPLORER 60,897 MODELS INDEXED

Seed Oss 36B Instruct Fp16 Mlx by dnakov

By dnakov · 27 downloads
HF Repo ↗ Copy URL

Seed Oss 36B Instruct Fp16 Mlx is an open-source language model by dnakov. Features: 36b LLM, VRAM: 72.2GB, Context: 512K, License: apache-2.0, Quantized, Instruction-Based, LLM Explorer Score: 0.19.

Base model:bytedance-seed/seed... Base model:finetune:bytedance-...   Conversational   En   Fp16   Instruct   Mlx   Quantized   Region:us   Safetensors   Seed oss   Sharded   Tensorflow   Vllm   Zh
Model Card on HF 🤗: https://huggingface.co/dnakov/seed-oss-36b-instruct-fp16-mlx 

Seed Oss 36B Instruct Fp16 Mlx Parameters and Internals

LLM NameSeed Oss 36B Instruct Fp16 Mlx
Repository 🤗https://huggingface.co/dnakov/seed-oss-36b-instruct-fp16-mlx 
Base Model(s)  Seed OSS 36B Instruct   ByteDance-Seed/Seed-OSS-36B-Instruct
Model Size36b
Required VRAM72.2 GB
Updated2026-07-17
Maintainerdnakov
Model Typeseed_oss
Instruction-BasedYes
Model Files  5.3 GB: 1-of-14   5.1 GB: 2-of-14   5.3 GB: 3-of-14   5.3 GB: 4-of-14   5.1 GB: 5-of-14   5.1 GB: 6-of-14   5.1 GB: 7-of-14   5.3 GB: 8-of-14   5.3 GB: 9-of-14   5.1 GB: 10-of-14   5.1 GB: 11-of-14   5.1 GB: 12-of-14   5.3 GB: 13-of-14   4.7 GB: 14-of-14
Supported Languagesen zh
Quantization Typefp16
Model ArchitectureSeedOssForCausalLM
Licenseapache-2.0
Context Length524288
Model Max Length524288
Transformers Version4.55.0
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<seed:pad>
Vocabulary Size155136
Torch Data Typebfloat16

Best Alternatives to Seed Oss 36B Instruct Fp16 Mlx

Best Alternatives
Context / RAM
Downloads
Likes
Seed OSS 36B Instruct 4bit512K / 20.2 GB603
Seed OSS 36B Instruct512K / 72.2 GB45727505
Seed OSS 36B Instruct512K / 72.2 GB3823
Seed OSS 36B Instruct AWQ512K / 21.1 GB3208
...6B Base Instruct Karcher Merge512K / 72.1 GB382
...eed OSS 36B Instruct GPTQ Int4512K / 21.1 GB215
Note: green Score (e.g. "73.2") means that the model is better than dnakov/seed-oss-36b-instruct-fp16-mlx.