recipes: set parent_model for all quant/fork repos (GLM 5.1/5.2/5.3/5.3-Flash, Qwen3.8 27B+Flash-Next, MiniMax-M3, MiMo-V2.5, Solar-Open2, DSV4-vision env-driven) — report medal/tree grouping now ranks quants inside the base model's category

recipes/dsv4-vision-aiden-373/recipe.toml

@@ -35,6 +35,7 @@ engine = "vllm"
 name = "DeepSeek-V4-Flash-Vision (aiden prod-3.73-vision img, 2x Spark)"
 docker_image = "aidendle94/sparkrun-vllm-ds4-gb10:production-3.73-vision"
 model = ""
+parent_model = "deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"
 port = 30000
 required_nodes = 2
 tensor_parallel_size = 2

recipes/glm51/recipe.toml

@@ -2,6 +2,7 @@ engine = "vllm"
 name = "GLM-5.1 NVFP8"
 docker_image = "vllm/vllm-openai:glm51-cu130"
 model = "lukealonso/GLM-5.1-NVFP4"
+parent_model = "zai-org/GLM-5.1"
 
 required_nodes = 4
 distributed_backend = "mp"

recipes/glm52-hybrid-fp8-mxfp4/recipe.toml

@@ -28,6 +28,7 @@ engine = "vllm"
 name = "GLM-5.2 Hybrid FP8-MXFP4 (vLLM)"
 docker_image = "sparkrun-vllm-ds4-gb10:glm52"
 model = "aidendle94/GLM-5.2-Hybrid-FP8-MXFP4"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm52-int4-int8/recipe.toml

@@ -20,6 +20,7 @@ engine = "vllm"
 name = "GLM-5.2 Int4-Int8Mix (vLLM, DS4 image)"
 docker_image = "aidendle94/sparkrun-vllm-ds4-gb10:production-hybrid-1.6"
 model = "QuantTrio/GLM-5.2-Int4-Int8Mix"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm52-mxfp4-experts-gptq/recipe.toml

@@ -25,6 +25,7 @@ engine = "vllm"
 name = "GLM-5.2 MXFP4-Experts-GPTQ (vLLM)"
 docker_image = "sparkrun-vllm-ds4-gb10:glm52"
 model = "aidendle94/GLM-5.2-MXFP4-Experts-GPTQ"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm52-nf3-hybrid-context/recipe.toml

@@ -42,6 +42,7 @@ engine = "vllm"
 name = "GLM-5.2 NF3-Hybrid CONTEXT LANE (vLLM)"
 docker_image = "aidendle94/sparkrun-vllm-ds4-gb10:production-hybrid-1.6"
 model = "madeby561/GLM-5.2-MXFP8-NVFP4-NF3-Hybrid"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm52-nf3-hybrid-fast/recipe.toml

@@ -42,6 +42,7 @@ engine = "vllm"
 name = "GLM-5.2 NF3-Hybrid FAST LANE (vLLM)"
 docker_image = "aidendle94/sparkrun-vllm-ds4-gb10:production-hybrid-1.6"
 model = "madeby561/GLM-5.2-MXFP8-NVFP4-NF3-Hybrid"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm52-nf4-lloyd-experts-gptq/recipe.toml

@@ -44,6 +44,7 @@ engine = "vllm"
 name = "GLM-5.2 NF4 Lloyd-Experts-GPTQ (vLLM)"
 docker_image = "sparkrun-vllm-ds4-gb10:glm52"
 model = "aidendle94/GLM-5.2-MXFP4-Experts-GPTQ"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm52/recipe.toml

@@ -2,6 +2,7 @@ engine = "vllm"
 name = "GLM-5.2 NVFP4 (vLLM)"
 docker_image = "vllm/vllm-openai:glm52"
 model = "Mapika/GLM-5.2-NVFP4"
+parent_model = "zai-org/GLM-5.2"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-aiden-2x/recipe.toml

@@ -38,6 +38,7 @@ engine = "vllm"
 name = "GLM-5.3-Flash aiden image (vLLM, 2x Spark, DFlash2-7)"
 docker_image = "aidendle94/sparkrun-vllm-glm53-gb10:production-1.2"
 model = "LibertAIDAI/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-aiden-4x-labdrafter/recipe.toml

@@ -40,6 +40,7 @@ name = "GLM-5.3-Flash aiden image (vLLM, 4x Spark, LAB DFlash2-7, NVFP4)"
 docker_image = "sparkrun-vllm-glm53-flash-aiden:glm53"
 docker_image_dir = "glm53-flash-aiden"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-dflash2-mxfp8/recipe.toml

@@ -80,6 +80,7 @@ docker_image_dir = "glm53-flash-dflash2"
 # TARGET model — the DFlash2 draft model is referenced only via
 # --speculative-config below, never as the top-level model.
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-eugr-2x/recipe.toml

@@ -27,6 +27,7 @@ engine = "vllm"
 name = "GLM-5.3-Flash NVFP4-Spark (eugr b12x img, 2x)"
 docker_image = "eugr/spark-vllm-b12x:nightly-20260910"  # prebuilt (no build: no docker_image_dir)
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-exl3/recipe.toml

@@ -73,6 +73,7 @@ name = "GLM-5.3-Flash EXL3-4bpw + DFlash2 (Entrpi image, 2x Spark, TP=2)"
 docker_image = "glm53-flash-exl3:local"
 docker_image_dir = "glm53-flash-exl3"
 model = "brandonmusic/GLM-5.3-Flash-tr3-4bpw"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-intel-w4a16/recipe.toml

@@ -42,6 +42,7 @@ name = "GLM-5.3-Flash W4A16-AutoRound Intel (vLLM, 2x Spark, r12 fork)"
 docker_image = "glm53-jovian-r12:local"
 docker_image_dir = "glm53-jovian-r12"
 model = "Intel/GLM-5.3-Flash-W4A16-AutoRound"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-4p67/recipe.toml

@@ -43,6 +43,7 @@ name = "GLM-5.3-Flash NVFP4-4p67 (vLLM, 2x Spark, MTP-3)"
 docker_image = "glm53-flash-nvfp4-4p67:local"
 docker_image_dir = "glm53-flash-nvfp4-4p67"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-4p67"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-eugr-4x/recipe.toml

@@ -63,6 +63,7 @@ engine = "vllm"
 name = "GLM-5.3-Flash NVFP4 eugr B12X 4x (vLLM, 4x Spark, MTP k=5, RoCE)"
 docker_image = "eugr/spark-vllm-b12x:nightly-20260904"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-eugr/recipe.toml

@@ -59,6 +59,7 @@ engine = "vllm"
 name = "GLM-5.3-Flash NVFP4 eugr B12X (vLLM, 2x Spark, MTP k=5, RoCE)"
 docker_image = "eugr/spark-vllm-b12x:nightly-20260904"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-lab-0rand/recipe.toml

@@ -49,6 +49,7 @@ name = "GLM-5.3-Flash NVFP4-lab - 0rand winning config (vLLM, 2x Spark, MTP-3)"
 docker_image = "glm53-flash-nvfp4-lab:local"
 docker_image_dir = "glm53-flash-nvfp4-lab"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-lab-dflash2/recipe.toml

@@ -66,6 +66,7 @@ name = "GLM-5.3-Flash NVFP4-lab (vLLM, 2x Spark, LAB DFlash2-7)"
 docker_image = "glm53-flash-nvfp4-lab:local"
 docker_image_dir = "glm53-flash-nvfp4-lab"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-lab/recipe.toml

@@ -48,6 +48,7 @@ name = "GLM-5.3-Flash NVFP4-lab (vLLM, 2x Spark, MTP-3)"
 docker_image = "glm53-flash-nvfp4-lab:local"
 docker_image_dir = "glm53-flash-nvfp4-lab"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r17abi/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r12"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-2x-nvfp4kv-spark/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-2x-nvfp4kv/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-4x-dcp4/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-4x-nvfp4kv-spark/recipe.toml

@@ -86,6 +86,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-4x-nvfp4kv/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-4x/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18-dflash2/recipe.toml

@@ -88,6 +88,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-r18/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-redhat/recipe.toml

@@ -56,6 +56,7 @@ name = "GLM-5.3-Flash NVFP4-redhat (vLLM, 2x Spark, MTP k=5)"
 docker_image = "glm53-flash-nvfp4-lab:local"
 docker_image_dir = "glm53-flash-nvfp4-lab"
 model = "RedHatAI/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4-spark/recipe.toml

@@ -79,6 +79,7 @@ docker_image_dir = "glm53-jovian-r12"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-nvfp4/recipe.toml

@@ -49,6 +49,7 @@ name = "GLM-5.3-Flash NVFP4 (vLLM, 2x Spark)"
 docker_image = "glm53-flash-nvfp4:local"
 docker_image_dir = "glm53-flash-nvfp4"
 model = "LibertAIDAI/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-r18-2x-nvfp4kv-spark-eager/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r18"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-r27-4x-dcp4/recipe.toml

@@ -85,6 +85,7 @@ docker_image_dir = "glm53-jovian-r27"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-r27-spark/recipe.toml

@@ -81,6 +81,7 @@ docker_image_dir = "glm53-jovian-r27"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-r28-4x-dcp4-b12xload/recipe.toml

@@ -107,6 +107,7 @@ build_profile = "glm53-r28"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-r28-4x-dcp4-split2048/recipe.toml

@@ -107,6 +107,7 @@ build_profile = "glm53-r28"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-r28-4x-dcp4/recipe.toml

@@ -107,6 +107,7 @@ build_profile = "glm53-r28"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-r28-dflash2-2x/recipe.toml

@@ -27,6 +27,7 @@ docker_image_dir = "glm53-jovian-r27"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-r28-spark/recipe.toml

@@ -82,6 +82,7 @@ build_profile = "glm53-r28"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-r31-2x/recipe.toml

@@ -82,6 +82,7 @@ build_profile = "glm53-r31"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-flash-r31-4x-dcp4/recipe.toml

@@ -107,6 +107,7 @@ build_profile = "glm53-r31"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-r35-4x-dcp4/recipe.toml

@@ -107,6 +107,7 @@ build_profile = "glm53-r35"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-flash-sparkring-2x/recipe.toml

@@ -13,6 +13,7 @@ engine = "vllm"
 name = "GLM-5.3-Flash NVFP4-Spark (SparkRing shared img, TP2 RoCEnante profile)"
 docker_image = "ghcr.io/fujitsupolycom/sparkring:glm53-shared-8b32ef2"  # prebuilt (no docker_image_dir); the 4774a7c5 memory-guard build (2026-09-11 re-run)
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 port = 8000
 required_nodes = 2
 tensor_parallel_size = 2

recipes/glm53-flash-sparkring-4x/recipe.toml

@@ -13,6 +13,7 @@ name = "GLM-5.3-Flash NVFP4-Spark (SparkRing shared img, TP4 switched profile)"
 docker_image = "sparkring-shared:local"
 docker_image_dir = "sparkring-shared"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3-Flash"
 port = 8000
 required_nodes = 4
 tensor_parallel_size = 4

recipes/glm53-flash-sparkrun/recipe.toml

@@ -35,6 +35,7 @@ engine = "vllm"
 name = "GLM-5.3-Flash NVFP4 on sparkrun image (vLLM, 2x Spark, MTP-3)"
 docker_image = "aidendle94/sparkrun-vllm-glm53-gb10:production-1.0"
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4"
+parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/glm53-full-nvfp4-spark-aiden/recipe.toml

@@ -26,6 +26,7 @@ engine = "vllm"
 name = "GLM-5.3 (full) NVFP4-Spark, aiden image (vLLM, 4x Spark)"
 docker_image = "aidendle94/sparkrun-vllm-glm53-gb10:production-1.2"
 model = "local-inference-lab/GLM-5.3-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-full-nvfp4-spark-lab/recipe.toml

@@ -23,6 +23,7 @@ name = "GLM-5.3 (full) NVFP4-Spark-lab (vLLM, 4x Spark)"
 docker_image = "glm53-flash-nvfp4-lab:local"
 docker_image_dir = "glm53-flash-nvfp4-lab"
 model = "local-inference-lab/GLM-5.3-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/glm53-spark-nvfp4-4x/recipe.toml

@@ -107,6 +107,7 @@ build_profile = "glm53-r28"
 # until the clean rebuild).
 entrypoint = ["vllm", "serve"]
 model = "local-inference-lab/GLM-5.3-NVFP4-Spark"
+parent_model = "zai-org/GLM-5.3"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/mimo-v2.5-nvfp4/recipe.toml

@@ -2,6 +2,7 @@ engine = "vllm"
 name = "MiMo-V2.5 Omni NVFP4 (vLLM)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.25.1-gb10.1"
 model = "lukealonso/MiMo-V2.5-NVFP4"
+parent_model = "XiaomiMiMo/MiMo-V2.5"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/minimax-m3-mxfp8/recipe.toml

@@ -2,6 +2,7 @@ engine = "vllm"
 name = "MiniMax M3 MXFP8 (vLLM)"
 docker_image = "vllm/vllm-openai:minimax-m3"
 model = "MiniMaxAI/MiniMax-M3-MXFP8"
+parent_model = "MiniMaxAI/MiniMax-M3"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/minimax-m3-nvfp4/recipe.toml

@@ -2,6 +2,7 @@ engine = "vllm"
 name = "MiniMax M3 NVFP (vLLM)"
 docker_image = "vllm/vllm-openai:minimax-m3"
 model = "Mapika/MiniMax-M3-NVFP4"
+parent_model = "MiniMaxAI/MiniMax-M3"
 required_nodes = 4
 tensor_parallel_size = 4
 pipeline_parallel_size = 1

recipes/minimax-m3/recipe.toml

@@ -3,6 +3,7 @@ name = "Minimax-M3-v0 NVFP4"
 docker_image = "scitrera/dgx-spark-sglang-mm:v0"
 #model = "sparkarena/Minimax-M3-v0-NVFP4"
 model = "Mapika/MiniMax-M3-NVFP4"
+parent_model = "MiniMaxAI/MiniMax-M3"
 required_nodes = 4
 tensor_parallel_size = 4
 distributed_backend = "mp"

recipes/qwen38-27b-fp8-dspark/recipe.toml

@@ -29,6 +29,7 @@ engine = "vllm"
 name = "Qwen3.8-27B (FP8 + DSpark)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.27.1-gb10.2"
 model = "Qwen/Qwen3.8-27B-FP8"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-fp8/recipe.toml

@@ -22,6 +22,7 @@ engine = "vllm"
 name = "Qwen3.8-27B (FP8)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.27.1-gb10.2"
 model = "Qwen/Qwen3.8-27B-FP8"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-int4/recipe.toml

@@ -24,6 +24,7 @@ engine = "vllm"
 name = "Qwen3.8-27B (Int4 AutoRound)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.27.1-gb10.2"
 model = "Pilcothink/Qwen3.8-27B-MixedInt4-AutoRound"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-nvfp4-sglang-dflash2/recipe.toml

@@ -32,6 +32,7 @@ engine = "sglang"
 name = "Qwen3.8-27B (RadixArk NVFP4, SGLang DFlash2)"
 docker_image = "lmsysorg/sglang:qwen38-27b-dflash2-minoverlay"
 model = "RadixArk/Qwen3.8-27B-NVFP4"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-nvfp4-sglang-dspark/recipe.toml

@@ -34,6 +34,7 @@ engine = "sglang"
 name = "Qwen3.8-27B (RadixArk NVFP4, SGLang DSpark)"
 docker_image = "lmsysorg/sglang:qwen38-27b"
 model = "RadixArk/Qwen3.8-27B-NVFP4"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-nvfp4-unsloth-dspark/recipe.toml

@@ -28,6 +28,7 @@ engine = "vllm"
 name = "Qwen3.8-27B (NVFP4 unsloth + DSpark)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.27.1-gb10.2"
 model = "unsloth/Qwen3.8-27B-NVFP4"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-nvfp4-unsloth/recipe.toml

@@ -28,6 +28,7 @@ engine = "vllm"
 name = "Qwen3.8-27B (NVFP4, unsloth)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.27.1-gb10.2"
 model = "unsloth/Qwen3.8-27B-NVFP4"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-27b-nvfp4/recipe.toml

@@ -29,6 +29,7 @@ engine = "vllm"
 name = "Qwen3.8-27B (NVFP4, RedHatAI)"
 docker_image = "ghcr.io/timothystewart6/vllm-gb10:v0.27.1-gb10.2"
 model = "RedHatAI/Qwen3.8-27B-NVFP4"
+parent_model = "Qwen/Qwen3.8-27B"
 required_nodes = 1
 tensor_parallel_size = 1
 pipeline_parallel_size = 1

recipes/qwen38-flash-next-fp8/recipe.toml

@@ -27,6 +27,7 @@ engine = "vllm"
 name = "Qwen3.8-Flash-Next FP8 (vLLM expert-parallel, 2x Spark)"
 docker_image = "vllm/vllm-openai:qwen38-flash-next"
 model = "Qwen/Qwen3.8-Flash-Next-FP8"
+parent_model = "Qwen/Qwen3.8-Flash-Next"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/qwen38-flash-next-nvfp4/recipe.toml

@@ -31,6 +31,7 @@ name = "Qwen3.8-Flash-Next NVFP4 (SGLang, 2× Spark)"
 docker_image = "qwen38-flashnext-nvfp4:local"
 docker_image_dir = "qwen38-flashnext-nvfp4"
 model = "RadixArk/Qwen3.8-Flash-Next-NVFP4"
+parent_model = "Qwen/Qwen3.8-Flash-Next"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1

recipes/solar-open2-nvfp4/recipe.toml

@@ -25,6 +25,7 @@ engine = "vllm"
 name = "Solar Open2 250B Nota NVFP4 (vLLM)"
 docker_image = "solar-open2-nvfp4:probe"
 model = "nota-ai/Solar-Open2-250B-Nota-NVFP4"
+parent_model = "upstage/Solar-Open2-250B"
 required_nodes = 2
 tensor_parallel_size = 2
 pipeline_parallel_size = 1