Downloads · 30 days
101
39% of all-time downloads
heretic-nvidia-Q8_0-GGUF
charlinoss/heretic-nvidia-Q8_0-GGUF
heretic-nvidia-Q8_0-GGUF is a text generation model from charlinoss. Use it when you need the model to write or continue text. It is set up for llama.cpp.
Q80 GGUF conversion of charlinoss/heretic-nvidia, revision 33fc0f65c7adb86280c54bcb759bcefb49b6949e. The source is the Heretic Trial 98 derivative of NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16.
All-time downloads
260
Public
Repo size
35 GB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.gguf35 GB · 100%
At a glance
- Task
- Text Generation
- Library
- llama.cpp
- Access
- Public
- Created
- Sep 18, 2026
- Updated
- Sep 18, 2026
- SHA
- 1417d9af
Try a prompt
Base models
Model card
From the Hugging Face model README
Heretic NVIDIA — Trial 98 — Q8_0 GGUF
Q8_0 GGUF conversion of charlinoss/heretic-nvidia, revision 33fc0f65c7adb86280c54bcb759bcefb49b6949e.
The source is the Heretic Trial 98 derivative of NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16.
- File:
heretic-nvidia-Q8_0.gguf - Size: 35,004,320,320 bytes (approximately 35 GB)
- Converter: llama.cpp commit
4fea119de30f6a923992780f6fd5ccb0bee5d47d - Conversion:
convert_hf_to_gguf.py source --outfile output/heretic-nvidia-Q8_0.gguf --outtype q8_0 - Transformers: 5.6.2
- MTP tensors retained by the converter; source includes the preserved original MTP weights.
Basic loading and short generation were tested with the same llama.cpp build on CPU. No comprehensive quality or BF16-versus-Q8 benchmark has been performed. Ollama compatibility has not been tested.
Refer to the original NVIDIA model and source repository for applicable license and use terms.
- Task
- Text Generation
- Library
- llama.cpp
- Created
- Sep 18, 2026
- Updated
- Sep 18, 2026