bartowski / text-generation updated 1 year ago

Meta-Llama-3.1-8B-Instruct-GGUF

- 30-07-2024: Updated chat template to fix small bug with tool usage being undefined, if you don't use the built-in chat template it shouldn't change anything

Params
Context
Downloads 30d
302 K
Likes
388
Commercial use: conditional · llama3.1 Not gated GGUF 8 languages View on Hugging Face ↗

Download history

daily snapshots · 10 days
302 K288 K
Jul 28Jul 31Aug 3Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

FileQuantSizeEst. VRAMVerdict on RTX 4090 · 24 GB
Meta-Llama-3.1-8B-Instruct-IQ2_M.gguf IQ2_M 2.9 GB 3.7 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-Q2_K.gguf Q2_K 3.2 GB 4.0 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-IQ3_XS.gguf IQ3_XS 3.5 GB 4.4 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-Q2_K_L.gguf Q2_K_L 3.7 GB 4.6 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-IQ3_M.gguf IQ3_M 3.8 GB 4.7 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-Q3_K_L.gguf Q3_K_L 4.3 GB 5.3 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-IQ4_XS.gguf IQ4_XS 4.4 GB 5.4 GB ✅ Runs comfortably
Meta-Llama-3.1-8B-Instruct-IQ4_NL.gguf IQ4_NL 4.7 GB 5.6 GB ✅ Runs comfortably
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub
~ · ollama · IQ2_M
$ ollama run meta-llama-3-1-8b-instruct-gguf

# pin the quantization explicitly
$ ollama run meta-llama-3-1-8b-instruct-gguf-iq2_m
est. VRAM 3.7 GBon RTX 4090 · 24 GBJSON API →

Specifications

Licence
llama3.1
First seen on the Hub
2024-07-23
Training datasets
undisclosed
Added to our catalog
2026-07-28
Sponsored · GPU cloud
Not enough VRAM?
Spin up a 24 GB L4 instance in 40 seconds. $0.44/hr.