unsloth / text-generation updated 1 year ago

DeepSeek-R1-Distill-Llama-70B-GGUF

Or you can view more detailed instructions here: unsloth.ai/blog/deepseek-r1 1. Do not forget about and tokens! - Or use a chat template formatter 2. Obtain the latest llama.cpp at https://github.com/ggerganov/llama.cpp 3. Example with Q80 K quantized cache Notice -no-cnv disables auto conversation mode bash ./llama.cpp/llama-cli \ --model unsloth/DeepSeek-R1-Distill-Llama-70B-GGUF/DeepSeek-R1-Distill-Llama-70B-Q4KM....

Params
Context
131,072
Downloads 30d
208 K
Likes
121
Commercial use: conditional · llama3.3 Not gated GGUF 1 languages View on Hugging Face ↗

Download history

daily snapshots · 16 days
208 K191 K
Sep 6Sep 11Sep 16Sep 21

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
DeepSeek-R1-Distill-Llama-70B-Q2_K.gguf Q2_K 26.4 GB 29.5 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-Q2_K_L.gguf Q2_K_L 26.6 GB 29.8 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-Q3_K_S.gguf Q3_K_S 30.9 GB 34.5 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-Q3_K_M.gguf Q3_K_M 34.3 GB 38.2 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-IQ4_XS.gguf IQ4_XS 37.9 GB 42.2 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-IQ4_NL.gguf IQ4_NL 40.1 GB 44.6 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-Q4_0.gguf Q4_0 40.1 GB 44.6 GB ❌ Won’t fit
DeepSeek-R1-Distill-Llama-70B-BF16-00001-of-00003.gguf GGUF 49.9 GB 55.4 GB ❌ Won’t fit
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 · Q4_0
$ ollama run deepseek-r1-distill-llama-70b-gguf

# pin the quantization explicitly
$ ollama run deepseek-r1-distill-llama-70b-gguf-q4_0
est. VRAM 44.6 GBon RTX 4090 · 24 GBJSON API →

Specifications

Architecture
LlamaForCausalLM
Context length
131,072
Vocabulary
128,256
Layers / heads
80 / 64
Licence
llama3.3
First seen on the Hub
2025-01-20
Base model
DeepSeek-R1-Distill-Llama-70B
Training datasets
undisclosed
Added to our catalog
2026-09-06

Family

Base model and the most-downloaded derivatives in the catalog.

Compare with any text-generation model