diff --git a/apps/llama/configs/config.yaml b/apps/llama/configs/config.yaml index 06e40a7..5e37eb6 100644 --- a/apps/llama/configs/config.yaml +++ b/apps/llama/configs/config.yaml @@ -497,11 +497,14 @@ models: --port ${PORT} --chat-template-kwargs "{\"enable_thinking\": false}" + # The "heretic" version does not provide the mmproj + # so providing url to the one from the non-heretic version. "Qwen3.5-35B-A3B-heretic-GGUF:Q4_K_M": ttl: 600 cmd: | /app/llama-server -hf mradermacher/Qwen3.5-35B-A3B-heretic-GGUF:Q4_K_M + --mmproj-url https://huggingface.co/unsloth/Qwen3.5-35B-A3B-GGUF/resolve/main/mmproj-F16.gguf --ctx-size 16384 --temp 1.0 --min-p 0.00 @@ -515,6 +518,7 @@ models: cmd: | /app/llama-server -hf mradermacher/Qwen3.5-35B-A3B-heretic-GGUF:Q4_K_M + --mmproj-url https://huggingface.co/unsloth/Qwen3.5-35B-A3B-GGUF/resolve/main/mmproj-F16.gguf --ctx-size 16384 --temp 1.0 --min-p 0.00