LLM EXPLORER 63,892 MODELS INDEXED

Atria Dawn Preview MLX Q4i by inferencerlabs

By inferencerlabs · 400 downloads
HF Repo ↗ Copy URL

Atria Dawn Preview MLX Q4i is an open-source language model by inferencerlabs. Features: LLM, VRAM: 198.7GB, Context: 1024K, MoE, LLM Explorer Score: 0.33.

Base model:internlm/atria-dawn... Base model:quantized:internlm/...   Conversational   En   Glm moe dsa   Mlx   Moe   Quantized   Region:us   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/inferencerlabs/Atria-Dawn-Preview-MLX-Q4i 

Atria Dawn Preview MLX Q4i Parameters and Internals

LLM NameAtria Dawn Preview MLX Q4i
Repository 🤗https://huggingface.co/inferencerlabs/Atria-Dawn-Preview-MLX-Q4i 
Base Model(s)  Atria Dawn Preview   internlm/Atria-Dawn-Preview
Required VRAM198.7 GB
Updated2026-10-12
Maintainerinferencerlabs
Model Typeglm_moe_dsa
Model Files  5.0 GB: 1-of-103   4.3 GB: 2-of-103   5.3 GB: 3-of-103   4.1 GB: 4-of-103   4.5 GB: 5-of-103   4.0 GB: 6-of-103   4.2 GB: 7-of-103   4.8 GB: 8-of-103   4.5 GB: 9-of-103   4.2 GB: 10-of-103   4.6 GB: 11-of-103   3.4 GB: 12-of-103   4.0 GB: 13-of-103   5.3 GB: 14-of-103   4.0 GB: 15-of-103   5.4 GB: 16-of-103   3.8 GB: 17-of-103   4.0 GB: 18-of-103   4.5 GB: 19-of-103   5.2 GB: 20-of-103   3.8 GB: 21-of-103   5.3 GB: 22-of-103   4.1 GB: 23-of-103   3.6 GB: 24-of-103   3.8 GB: 25-of-103   5.3 GB: 26-of-103   4.0 GB: 27-of-103   5.3 GB: 28-of-103   5.3 GB: 29-of-103   3.8 GB: 30-of-103   5.3 GB: 31-of-103   5.4 GB: 32-of-103   5.4 GB: 33-of-103   5.4 GB: 34-of-103   5.3 GB: 35-of-103   4.3 GB: 36-of-103   3.9 GB: 37-of-103   4.0 GB: 38-of-103   4.1 GB: 39-of-103   4.0 GB: 40-of-103   5.4 GB: 41-of-103   4.5 GB: 42-of-103   3.8 GB: 43-of-103   4.5 GB: 44-of-103
Supported Languagesen
Model ArchitectureGlmMoeDsaForCausalLM
Context Length1048576
Model Max Length1048576
Transformers Version5.2.0
Vocabulary Size154880

Best Alternatives to Atria Dawn Preview MLX Q4i

Best Alternatives
Context / RAM
Downloads
Likes
Glm 5.3 Exlr8 K2 K3 K41024K /  GB11764912
...Colibri Int4 G64 With Int8 Mtp1024K / 2.8 GB1052643
GLM 5.3 NVFP4 CSF1024K /  GB1331
GLM 5.3 Int4 Int8Mix1024K / 62.2 GB2192
GLM 5.3 Colibri Int4 G641024K / 2.8 GB251
GLM 5.2 NVFP4 TR3 Hybrid1024K / 0.8 GB3383
GLM 5 FP8198K / 243 GB545
Note: green Score (e.g. "73.2") means that the model is better than inferencerlabs/Atria-Dawn-Preview-MLX-Q4i.