Downloads · 30 days
0
0% of all-time downloads
hovor-gemma-3-4b-it-coreml-int4
dreamer1cc/hovor-gemma-3-4b-it-coreml-int4
hovor-gemma-3-4b-it-coreml-int4 is a machine learning model from dreamer1cc. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as gemma.
Converted from google/gemma-3-4b-it for on-device inference on iOS 18+ / macOS 15+ via Apple Neural Engine (ANE).
All-time downloads
16
Public
Repo size
10.8 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin2.1 GB · 98%
At a glance
- License
- gemma
- Access
- Public
- Created
- Jun 8, 2026
- Updated
- Jun 8, 2026
- SHA
- 7aaceb18
Model card
From the Hugging Face model README
Gemma 3 4B IT — CoreML int4 (Apple Neural Engine)
Converted from google/gemma-3-4b-it
for on-device inference on iOS 18+ / macOS 15+ via Apple Neural Engine (ANE).
Used in Hovor iOS quality comparison benchmarks.
Files
| File | Purpose |
|---|---|
model.mlpackage/ | CoreML stateful transformer (int4 palettized, CPU+ANE) |
model_config.json | Metadata for the Swift inference engine |
tokenizer.json, tokenizer_config.json, tokenizer.model | Tokenizer files |
special_tokens_map.json | Special token definitions |
Chat template
Gemma 3 uses the <start_of_turn> / <end_of_turn> turn marker format:
<bos><start_of_turn>user
{user message}<end_of_turn>
<start_of_turn>model
{model response}<end_of_turn>
With generation prompt (inference):
<bos><start_of_turn>user
{user message}<end_of_turn>
<start_of_turn>model
Attribution
Built with Gemma. Subject to the Gemma Terms of Use.
- License
- gemma
- Created
- Jun 8, 2026
- Updated
- Jun 8, 2026