Issues

• Multinode launch: copy each node's pipeline stage to job-private /dev/shm before vllm serve

  Center-for-AI-Innovation/LLMHub#183 · Filter by author dadamsncsa opened on Nov 10, 2026

• Model eviction: race with a launch of the same model, and gated IDs containing /

  Center-for-AI-Innovation/LLMHub#182 · Filter by author joshfactorial opened on Nov 9, 2026

• launch-test: log the queue give-up, and fail when vLLM answers without its key

  Center-for-AI-Innovation/LLMHub#181 · Filter by author dadamsncsa opened on Nov 9, 2026

• Check if ALWAYS_ON_VLLM_MODEL is a chat model

  Bug

  Center-for-AI-Innovation/LLMHub#179 · Filter by author minump opened on Nov 6, 2026
  1

• Images: bake the Slingshot rendezvous settings, drop FI_LOG_LEVEL=warn, gate promotion on a 2-node correctness check

  Center-for-AI-Innovation/LLMHub#176 · Filter by author dadamsncsa opened on Nov 6, 2026

• Validate a launch's parallel layout before submitting, and refuse shapes the cluster can't run

  Center-for-AI-Innovation/LLMHub#175 · Filter by author dadamsncsa opened on Nov 6, 2026

• Kit: multinode config for Delta and DeltaAI (fabric env, DeltaAI catalog, 2-node launch-test)

  Center-for-AI-Innovation/LLMHub#174 · Filter by author dadamsncsa opened on Nov 6, 2026
  

• Multinode vLLM: launch num_nodes > 1 as one srun step with vLLM's native multi-node flags

  Center-for-AI-Innovation/LLMHub#173 · Filter by author dadamsncsa opened on Nov 6, 2026
  

• Have apptainer def file in github

  Task

  Center-for-AI-Innovation/LLMHub#171 · Filter by author minump opened on Nov 5, 2026
  

• Removed but reference models linger

  Bug

  Center-for-AI-Innovation/LLMHub#170 · Filter by author lmarini opened on Nov 5, 2026
  

• Docs: update for the embeddings UI and the model card Type chip (#152)

  documentation
  Center-for-AI-Innovation/LLMHub#166 · Filter by author joshfactorial opened on Nov 3, 2026 · 1.2
  

• Docs: document the public embeddings endpoint (#137)

  documentation
  Center-for-AI-Innovation/LLMHub#165 · Filter by author joshfactorial opened on Nov 3, 2026 · 1.2
  

• Show only chat models in the chat page model picker

  Bug

  Center-for-AI-Innovation/LLMHub#164 · Filter by author minump opened on Nov 3, 2026
  

• Telemetry setup

  Task

  Center-for-AI-Innovation/LLMHub#162 · Filter by author rohan-uiuc opened on Nov 3, 2026

• Delta kit: named staging instances, and refuse to share another stack's Postgres, ports or directories

  enhancement
  Center-for-AI-Innovation/LLMHub#161 · Filter by author joshfactorial opened on Nov 3, 2026

• Fit estimator: calibrate vLLM overhead per model architecture

  enhancement
  Center-for-AI-Innovation/LLMHub#160 · Filter by author joshfactorial opened on Nov 3, 2026 · 1.3

• Fit estimator: validate-config and fit-estimate size the KV cache for 1,024 full-length sequences

  enhancement
  Center-for-AI-Innovation/LLMHub#159 · Filter by author joshfactorial opened on Nov 3, 2026 · 1.2

• Local development should have the smtp server disabled

  Bug

  Center-for-AI-Innovation/LLMHub#156 · Filter by author minump opened on Oct 29, 2026 · 1.3

• frontend model filtering by type

  Center-for-AI-Innovation/LLMHub#154 · Filter by author minump opened on Oct 29, 2026 · 1.3
  1

  

• Have ModelSize (small, medium, large) be uniform across the codebase

  Center-for-AI-Innovation/LLMHub#151 · Filter by author minump opened on Oct 26, 2026

• Follow-ups to #147: derive vLLM keys, drop the env lock, fix shim rebuild check

  enhancement
  Center-for-AI-Innovation/LLMHub#148 · Filter by author joshfactorial opened on Oct 23, 2026 · 1.3

• Launching a model that isn't in the synced catalog returns 500 instead of "not synced yet"

  Bug

  Center-for-AI-Innovation/LLMHub#146 · Filter by author dadamsncsa opened on Oct 22, 2026
  1

• Create llmhub project in lumen for enabling always on model

  Center-for-AI-Innovation/LLMHub#145 · Filter by author minump opened on Oct 22, 2026 · 1.3
  

• User is able to go to model library without login

  Bug

  Center-for-AI-Innovation/LLMHub#144 · Filter by author rohan-uiuc opened on Oct 22, 2026 · 1.2
  

• Keep failed deployments visible on Your Active Models

  enhancement
  Center-for-AI-Innovation/LLMHub#132 · Filter by author joshfactorial opened on Oct 19, 2026 · 1.2