diff --git a/README.md b/README.md index 5bd4444..acc3e65 100644 --- a/README.md +++ b/README.md @@ -24,4 +24,6 @@ `llama-server -hf unsloth/gemma-4-E4B-it-GGUF:Q3_K_XL --reasoning off -fa on -ngl 99 -b 2048 -ub 2048 -c 4096 --temp 1.0 --top-p 0.95 --top-k 64 --no-mmap -t 4` -> 126 t/s -`llama-server -hf unsloth/gemma-4-26B-A4B-it-qat-GGUF:UD-Q4_K_XL --reasoning off -fa on -ngl 99 -b 2048 -ub 2048 -c 4096 --temp 1.0 --top-p 0.95 --top-k 64 --no-mmap -t 4` -> 20 t/s \ No newline at end of file +`llama-server -hf unsloth/gemma-4-26B-A4B-it-qat-GGUF:UD-Q4_K_XL --reasoning off -fa on -ngl 99 -b 2048 -ub 2048 -c 4096 --temp 1.0 --top-p 0.95 --top-k 64 --no-mmap -t 4` -> 20 t/s + +`llama-server -hf unsloth/gemma-4-E4B-it-GGUF:Q3_K_XL --reasoning off -fa on -ngl 99 -b 2048 -ub 2048 -c 4096 --temp 1.0 --top-p 0.95 --top-k 64 --no-mmap -t 4 --spec-type draft-mtp --spec-draft-n-max 3 --port 8081` -> 130t/s \ No newline at end of file diff --git a/web/generator.php b/web/generator.php index 6f32053..59a8e4b 100644 --- a/web/generator.php +++ b/web/generator.php @@ -7,7 +7,7 @@ ini_set('error_log', __DIR__ . '/../errors.log'); ini_set('display_errors', '0'); error_reporting(E_ALL); -define('AI_SITE_LLM_API_BASE_URI', 'https://ai.haschek.at'); +define('AI_SITE_LLM_API_BASE_URI', 'http://localhost:8080/v1'); $pexelsKey = 'Clyqp9i9pewaxiiAdqVya0yZuuedjHtUtJdjRipNltp7wf0dY8K10h4h'; $apiKey = 'sk-Y1FrlvtAV4PWwK40TSVq-Q';