Downloads · 30 days
180
22% of all-time downloads
gemma-4-E4B-it-oQ4-mtp
mlx-works/gemma-4-E4B-it-oQ4-mtp
gemma-4-E4B-it-oQ4-mtp is a machine learning model from mlx-works. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for mlx.
This model was quantized using oQ (oMLX v0.5.5) mixed-precision quantization.
All-time downloads
817
Public
Parameters
8.1B
5.6 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors5.5 GB · 99%
Parameter types
How the weights are stored.
U327.5B · 93%
Model card
From the Hugging Face model README
gemma-4-E4B-it-oQ4-mtp
This model was quantized using oQ (oMLX v0.5.5) mixed-precision quantization.
Base model: google/gemma-4-E4B-it
MTP head: google/gemma-4-E4B-it-assistant (merged into single checkpoint)
Quantization details
- Model type: gemma4
- Bits: 4
- Group size: 64
- Format: MLX safetensors
- MTP: Merged from assistant checkpoint (Lightning MTP)
Environment
- Hardware: M5 MacBook Air 32GB
- Inference Framework: oMLX v0.5.5
- Max Concurrent Requests: 4
- Settings:
- Thinking: Disabled
- TurboQuant KV Cache: Enabled (4-bit)
- Lightning MTP: Enabled (key speed improvement)
Performance Benchmarks
Note: Results are for reference only and may vary depending on hardware, software configuration, and workload.
Single Request Results
| Test | TTFT(ms) | TPOT(ms) | pp TPS | tg TPS | E2E(s) | Throughput | Peak Mem |
|---|---|---|---|---|---|---|---|
| pp1024/tg128 | 678.6 | 25.77 | 1508.9 tok/s | 39.1 tok/s | 3.967 | 290.4 tok/s | 4.65 GB |
| pp4096/tg128 | 2871.6 | 27.07 | 1426.4 tok/s | 37.2 tok/s | 6.340 | 666.3 tok/s | 4.72 GB |
Continuous Batching (pp1024 / tg128)
| Batch | tg TPS | Speedup | pp TPS | pp TPS/req | TTFT(ms) | E2E(s) |
|---|---|---|---|---|---|---|
| 1x | 39.1 tok/s | 1.00x | 1508.9 tok/s | 1508.9 tok/s | 678.6 | 3.967 |
| 2x | 70.0 tok/s | 1.79x | 1265.9 tok/s | 633.0 tok/s | 1617.6 | 5.275 |
| 4x | 119.4 tok/s | 3.05x | 1699.1 tok/s | 424.8 tok/s | 2225.1 | 6.699 |
Intelligence Benchmark
Note: Each benchmark round tests only 30 questions. Results are for reference only.
| Benchmark | Accuracy | Correct | Total | Time(s) | Think |
|---|---|---|---|---|---|
| MMLU | 56.7% | 17 | 30 | 32.9 | No |
| TRUTHFULQA | 60.0% | 18 | 30 | 8.6 | No |
| GSM8K | 83.3% | 25 | 30 | 150.3 | No |
| MATHQA | 26.7% | 8 | 30 | 56.8 | No |
| HUMANEVAL | 83.3% | 25 | 30 | 215.2 | No |
- Library
- mlx
- Type
- gemma4
- Created
- Aug 5, 2026
- Updated
- Aug 16, 2026