Mallam 1.1B 4096 — Hardware Requirements & GPU Compatibility
Chat
Mallam 1.1B 4096 is a 1.1B-parameter open language model from mesolitica. It supports a context window of up to 32,768 tokens. At BF16 it needs about 2.64 GB of VRAM — see which GPUs and Macs can run it below.
375 downloads
14 likesSpecifications
- Publisher
- mesolitica
- Parameters
- 1.1B
- Architecture
- MistralForCausalLM
- Context Length
- 32,768 tokens
- Vocabulary Size
- 32,000
- Release Date
- 2024-10-07
Get Started
HuggingFace
How Much VRAM Does Mallam 1.1B 4096 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 2.6 GB | 4.0 GB | 2.25 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Mallam 1.1B 4096?
BF16 · 2.6 GBMallam 1.1B 4096 (BF16) requires 2.6 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 4+ GB is recommended. Using the full 33K context window can add up to 1.4 GB, bringing total usage to 4.0 GB. 35 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run Mallam 1.1B 4096?
BF16 · 2.6 GB33 devices with unified memory can run Mallam 1.1B 4096, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does Mallam 1.1B 4096 need?
Mallam 1.1B 4096 requires 2.6 GB of VRAM at BF16. Full 33K context adds up to 1.4 GB (4.0 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 1.1B × 16 bits ÷ 8 = 2.2 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 1.8 GB (at full 33K context)
VRAM usage by quantization
BF162.6 GBBF16 + full context4.0 GB- Can I run Mallam 1.1B 4096 on a Mac?
Mallam 1.1B 4096 requires at least 2.6 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Mallam 1.1B 4096 locally?
Yes — Mallam 1.1B 4096 can run locally on consumer hardware. At BF16 quantization it needs 2.6 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Mallam 1.1B 4096?
At BF16, Mallam 1.1B 4096 can reach ~1104 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~248 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 2.6 × 0.55 = ~1104 tok/s
Estimated speed at BF16 (2.6 GB)
~1104 tok/s~248 tok/s~825 tok/s~683 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Mallam 1.1B 4096?
At BF16, the download is about 2.25 GB.
- Which GPUs can run Mallam 1.1B 4096?
35 consumer GPUs can run Mallam 1.1B 4096 at BF16 (2.6 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 35 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Mallam 1.1B 4096?
33 devices with unified memory can run Mallam 1.1B 4096 at BF16 (2.6 GB), including Mac Mini M4 (16 GB), Mac Mini M4 (32 GB), Mac Mini M4 Pro (24 GB), Mac Mini M4 Pro (48 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.