unsloth
gemma 4 E2B it
gemma 4 E2B it is a 4.65B dense gemma4 model from unsloth. Configured context length is 131,072 tokens. License: apache-2.0. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters4.6B
Context128K
Attentionhybrid
Licenseapache-2.0
35 layers · 1 KV heads · head dim 256 · 28 sliding layers · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| BF16 | 8.67 GB | Hugging Face file bytes |
| UD-Q8_K_XL | 4.92 GB | Hugging Face file bytes |
| Q8_0 | 4.70 GB | Hugging Face file bytes |
| UD-Q6_K_XL | 4.39 GB | Hugging Face file bytes |
| Q6_K | 4.19 GB | Hugging Face file bytes |
| UD-Q5_K_XL | 4.00 GB | Hugging Face file bytes |
| Q5_K_M | 3.13 GB | Hugging Face file bytes |
| Q5_K_S | 3.09 GB | Hugging Face file bytes |
| UD-Q4_K_XL | 2.97 GB | Hugging Face file bytes |
| Q4_K_M | 2.89 GB | Hugging Face file bytes |
| Q4_K_S | 2.83 GB | Hugging Face file bytes |
| IQ4_NL | 2.83 GB | Hugging Face file bytes |
| IQ4_XS | 2.78 GB | Hugging Face file bytes |
| Q4_1 | 2.94 GB | Hugging Face file bytes |
| Q4_0 | 2.83 GB | Hugging Face file bytes |
| UD-Q3_K_XL | 2.72 GB | Hugging Face file bytes |
| Q3_K_M | 2.36 GB | Hugging Face file bytes |
| Q3_K_S | 2.28 GB | Hugging Face file bytes |
| UD-IQ3_XXS | 2.21 GB | Hugging Face file bytes |
| UD-Q2_K_XL | 2.24 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Runs | UD-Q8_K_XL | 28–39 |
| GeForce RTX 3060 12GB | Runs | BF16 | 21–28 |
| GeForce RTX 4060 Ti 16GB | Runs | BF16 | 16–22 |
| GeForce RTX 4090 | Runs great | BF16 | 58–79 |
| GeForce RTX 5090 | Runs great | BF16 | 103–140 |
| Apple M4 Pro | Runs great | BF16 | 16–21 |
| Apple M5 | Runs | BF16 | 8.8–12 |
| CPU only · DDR5-5600 | Runs on CPU | BF16 | 5.1–7.0 |
Smallest devices that fit it fully
- GeForce RTX 3050 6GB — Tight fit, Q8_0, 17–24 tok/s
- GeForce RTX 2060 6GB — Tight fit, Q8_0, 35–47 tok/s
- GeForce GTX 1660 Super — Tight fit, Q8_0, 35–47 tok/s
- GeForce GTX 1660 Ti — Tight fit, Q8_0, 30–40 tok/s
- Intel Arc A380 — Tight fit, Q8_0, 19–26 tok/s
- GeForce RTX 5060 Ti 8GB — Runs, UD-Q8_K_XL, 44–60 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (BF16).
llama.cpp
llama-server -hf unsloth/gemma-4-E2B-it-GGUF:BF16 -c 8192 -ngl 35
Ollama
ollama run gemma
LM Studio
lms get unsloth/gemma-4-E2B-it-GGUF # Then load quant BF16