Curated and trained by Alican Kiraz

Links:

• Medium: https://alican-kiraz1.medium.com/
• Linkedin: https://tr.linkedin.com/in/alican-kiraz
• X: https://x.com/AlicanKiraz0
• YouTube: https://youtube.com/@alicankiraz0

With the release of the new DeepSeek-R1, I quickly began training SenecaLLM v1.3 based on this model. During training:

• About 20 hours on BF16 with 4×H200
• About 10 hours on BF16 with 8×A100
• About 12 hours on FP32 with 8×H200

It does not pursue any profit.

Thanks to DeepSeek R1’s Turkish support capability and the dataset used in SenecaLLM v1.3, it can now provide Turkish support! With the new dataset I’ve prepared, it can produce quite good outputs in the following areas:

• Information Security v1.4
• Incident Response v1.3
• Threat Hunting v1.3
• Ethical Exploit Development v1.2
• Purple Team Tactics v1.2
• Reverse Engineering v1.0

"Those who shed light on others do not remain in darkness..."

AlicanKiraz0/Seneca-x-DeepSeek-R1-Distill-Qwen-32B-v1.3-Safe-Q2_K-GGUF

This model was converted to GGUF format from AlicanKiraz0/Seneca-x-DeepSeek-R1-Distill-Qwen-32B-v1.3-Safe using llama.cpp via the ggml.ai's GGUF-my-repo space. Refer to the original model card for more details on the model.

Use with llama.cpp

Install llama.cpp through brew (works on Mac and Linux)

brew install llama.cpp

Invoke the llama.cpp server or the CLI.

CLI:

llama-cli --hf-repo AlicanKiraz0/Seneca-x-DeepSeek-R1-Distill-Qwen-32B-v1.3-Safe-Q2_K-GGUF --hf-file seneca-x-deepseek-r1-distill-qwen-32b-v1.3-safe-q2_k.gguf -p "The meaning to life and the universe is"

Server:

llama-server --hf-repo AlicanKiraz0/Seneca-x-DeepSeek-R1-Distill-Qwen-32B-v1.3-Safe-Q2_K-GGUF --hf-file seneca-x-deepseek-r1-distill-qwen-32b-v1.3-safe-q2_k.gguf -c 2048

Note: You can also use this checkpoint directly through the usage steps listed in the Llama.cpp repo as well.

Step 1: Clone llama.cpp from GitHub.

git clone https://github.com/ggerganov/llama.cpp

Step 2: Move into the llama.cpp folder and build it with LLAMA_CURL=1 flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).

cd llama.cpp && LLAMA_CURL=1 make

Step 3: Run inference through the main binary.

./llama-cli --hf-repo AlicanKiraz0/Seneca-x-DeepSeek-R1-Distill-Qwen-32B-v1.3-Safe-Q2_K-GGUF --hf-file seneca-x-deepseek-r1-distill-qwen-32b-v1.3-safe-q2_k.gguf -p "The meaning to life and the universe is"

or

./llama-server --hf-repo AlicanKiraz0/Seneca-x-DeepSeek-R1-Distill-Qwen-32B-v1.3-Safe-Q2_K-GGUF --hf-file seneca-x-deepseek-r1-distill-qwen-32b-v1.3-safe-q2_k.gguf -c 2048