casperhansen/deepseek-r1-distill-qwen-14b-awq

Source: 🤗At risk🏴‍☠️ Rescued by: Alogotronmit14.8B params56 GBsafetensors✓ 3 checksumsupdated 637d ago

No README published on Hugging Face for this revision.

CommentsShare your thoughts and shape the future of sovereign AI.

@casperhansen is unclaimedSign in with the matching Hugging Face account or an organization member account to verify this handle.Claim @casperhansen

Files and sources

SourceAvailable on Hugging Face

Checksums3 recorded

TorrentFiles match Hugging Face

Swarm0 seeders reported

Source revision: bc43ec1bbf08…
Witnessed: 14 Oct 2026, 01:35 UTC
Swarm checked: 7 Nov 2026, 19:18 UTC
Weight files: 3
Model size: 56 GB
Downloads · last month: 3,759
Publisher: Unclaimed

Speed on real hardware

LocalMaxxing

Output speed for one request at a time, measured by people running this model on their own machines.

CMP 170HX✓vllm · AWQ · 2 runs97.4 tok/s

vllm serve casperhansen/deepseek-r1-distill-qwen-14b-awq --max-model-len 131072 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code

View run on LocalMaxxing

RTX 3090✓vllm · AWQ · 8 runs73.4 tok/s

vllm serve casperhansen/deepseek-r1-distill-qwen-14b-awq --max-model-len 131072 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --kv-cache-dtype int4_per_token_head --attention-backend TRITON_ATTN

View run on LocalMaxxing

Median of 10 community runs on 2 setups, from LocalMaxxing.

Run it yourself: setup guide

View provenance entry #28992 →

Speed on real hardware

LocalMaxxing

Output speed for one request at a time, measured by people running this model on their own machines.

CMP 170HX✓vllm · AWQ · 2 runs97.4 tok/s

vllm serve casperhansen/deepseek-r1-distill-qwen-14b-awq --max-model-len 131072 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code

View run on LocalMaxxing

RTX 3090✓vllm · AWQ · 8 runs73.4 tok/s

vllm serve casperhansen/deepseek-r1-distill-qwen-14b-awq --max-model-len 131072 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --kv-cache-dtype int4_per_token_head --attention-backend TRITON_ATTN

View run on LocalMaxxing

Median of 10 community runs on 2 setups, from LocalMaxxing.

Run it yourself: setup guide