Models

yashagrawal/ReAct-Deepseek-R1-Distill-Llama-8B-bnb-4bit

richardyoung/deepseek-r1-0528-qwen3-8b-heretic

Uncensored DeepSeek-R1-0528-Qwen3-8B via Heretic: 4/100 refusals (from 89), KL 0.041. Reasoning model, MIT.

brsilvapimentel/DeepSeek-R1-0528-Qwen3-8B

DeepSeek-R1-0528-Qwen3-8B

lcz2026/DeepSeek-R1-8B

rockn/DeepSeek-R1-0528-Qwen3-8B-IQ4_NL

DeepSeek-R1-0528-Qwen3-8B-IQ4_NL

okamototk/deepseek-r1

DeepSeek R1 0528 Qwen3 8B with tool calling/MCP support

MHKetbi/DeepSeek-R1-Distill-Llama-8B-NexaQuant

DeepSeek R1 Distilled model to one-fourth its original file size—without losing any accuracy.

lucasmg/deepseek-r1-8b-0528-qwen3-q4_K_M-tool-true

DeepSeek R1 0528 Qwen3 8B Q4 with tool calling

dengcao/DeepSeek-R1-0528-Qwen3-8B

DeepSeek-R1-0528-Qwen3-8B，包含2个量化模型：Q5_K_M，Q8_0

cyberuser42/DeepSeek-R1-Distill-Llama-8B

fzkun/deepseek-r1-medical

unsloth微调DeepSeek-R1-Distill-Llama-8B

lumolabs/Lumo-DeepSeek-R1-8B

Trained on Dataset: https://huggingface.co/datasets/lumolabs-ai/Lumo-Iris-DS-Instruct

bsahane/DeepSeek-R1-Distill-Llama-8B

lsm03624/deepseek-r1

DeepSeek-R1-0528 仍然使用 2024 年 12 月所发布的 DeepSeek V3 Base 模型作为基座，但在后训练过程中投入了更多算力，显著提升了模型的思维深度与推理能力。这个8B精馏版本编程能力都爆表！

geradeluxer/deepSeek-r1-distill-llama-8b-4-bits

SIGJNF/deepseek-r1-671b-1.58bit

Unsloth's DeepSeek-R1 1.58-bit, I just merged the thing and uploaded it here. This is the full 671b model, albeit dynamically quantized to 1.58bits.

dabl/deepseek-r1-8b-120k-24GB

dabl/deepseek-r1-8b-32k-8GB

dabl/deepseek-r1-8b-70k-12GB

dabl/deepseek-r1-8b-112k-16GB