spark lanes: aiden production-1.2 image (r35 mm_mxfp8 n>=128 rejects the kda gemms)

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

@@ -73,15 +73,21 @@
 
 engine = "vllm"
 name = "GLM-5.3-Flash NVFP4-Spark (vLLM, 2x Spark, MTP-5)"
-# karmic-kraken-beta — the lab's unified GLM/DS image (2026-09-19, per the
-# rtx6kpro glm-5.3-flash-spark-tp2 doc): CUDA 13.4.1, B12X backends, the
-# NVFP4-Spark loader + MTP head support (the old r12 image is gone from the
-# repo as non-rebuildable, commit 17a91a9; the prebuilt lab image replaces
-# it). Prebuilt external image — no docker_image_dir.
-docker_image = "ghcr.io/local-inference-lab/vllm:karmic-kraken-beta"
+# Image: glm53-jovian-r35:local — our r35 anchor image for the GLM-5.3-Flash
+# NVFP4 family (the 433 GB 4x lane ran on it at 163/176 with MTP3 — the
+# Spark checkpoint is the same Glm5Next family, different quant variant).
+# The lab's newer karmic-kraken-beta (rtx6kpro spark-tp2 doc) is amd64-only
+# (their RTX 6000 x86 platform; "Exec format error" on our aarch64 nodes),
+# so it cannot serve the Spark checkpoint here; the r12 image (the recipe's
+# original) is gone from the repo as non-rebuildable (commit 17a91a9).
+# Aiden's production GLM-5.3-Flash image for GB10 (31.2 GB; the glm53-flash-aiden
+# lane ran on it at 100/100 agentic — the GDN+NVFP4 path routes through B12X,
+# which handles the Spark checkpoint's small KDA GEMMs that r35's flashinfer
+# mm_mxfp8 (n,k >= 128) rejects. Prebuilt external image — no image_dir.
+docker_image = "aidendle94/sparkrun-vllm-glm53-gb10:production-1.2"
 # Required: the clean image follows the lab contract (ENTRYPOINT []
 # + CMD bash), so the recipe carries the entrypoint.
-entrypoint = ["vllm", "serve"]
+entrypoint = []
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
 parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 2

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

@@ -73,15 +73,21 @@
 
 engine = "vllm"
 name = "GLM-5.3-Flash NVFP4-Spark (vLLM, 4x, MTP-5)"
-# karmic-kraken-beta — the lab's unified GLM/DS image (2026-09-19, per the
-# rtx6kpro glm-5.3-flash-spark-tp2 doc): CUDA 13.4.1, B12X backends, the
-# NVFP4-Spark loader + MTP head support (the old r12 image is gone from the
-# repo as non-rebuildable, commit 17a91a9; the prebuilt lab image replaces
-# it). Prebuilt external image — no docker_image_dir.
-docker_image = "ghcr.io/local-inference-lab/vllm:karmic-kraken-beta"
+# Image: glm53-jovian-r35:local — our r35 anchor image for the GLM-5.3-Flash
+# NVFP4 family (the 433 GB 4x lane ran on it at 163/176 with MTP3 — the
+# Spark checkpoint is the same Glm5Next family, different quant variant).
+# The lab's newer karmic-kraken-beta (rtx6kpro spark-tp2 doc) is amd64-only
+# (their RTX 6000 x86 platform; "Exec format error" on our aarch64 nodes),
+# so it cannot serve the Spark checkpoint here; the r12 image (the recipe's
+# original) is gone from the repo as non-rebuildable (commit 17a91a9).
+# Aiden's production GLM-5.3-Flash image for GB10 (31.2 GB; the glm53-flash-aiden
+# lane ran on it at 100/100 agentic — the GDN+NVFP4 path routes through B12X,
+# which handles the Spark checkpoint's small KDA GEMMs that r35's flashinfer
+# mm_mxfp8 (n,k >= 128) rejects. Prebuilt external image — no image_dir.
+docker_image = "aidendle94/sparkrun-vllm-glm53-gb10:production-1.2"
 # Required: the clean image follows the lab contract (ENTRYPOINT []
 # + CMD bash), so the recipe carries the entrypoint.
-entrypoint = ["vllm", "serve"]
+entrypoint = []
 model = "local-inference-lab/GLM-5.3-Flash-NVFP4-Spark"
 parent_model = "zai-org/GLM-5.3-Flash"
 required_nodes = 4