Skip to content

fix: listing copy names the model it serves and links only to pages that exist #5

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
danstotts-ops merged 1 commit into from
Oct 9, 2026
Merged

fix: listing copy names the model it serves and links only to pages that exist #5

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions .runpod/hub.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
{
"title": "Holo4 (Holo4-27B-GGUF, Holo4-35B-A3B-GGUF)",
"description": "Serve Holo4 (Holo4-27B-GGUF, Holo4-35B-A3B-GGUF) on Runpod Serverless with vLLM. Computer use agents are a fast growing self-host workload, and teams want them on their own GPUs for privacy and cost control over long screenshot-heavy loops. Two sizes cover different rentals: the 27B dense fits a single 48GB to 80GB card, while the 35B-A3B MoE runs cheap per token and scales well for parallel agent sessions. GGUF weights mean day-one llama.cpp and Ollama deploys with the hai-agents harness on Runpod.",
"title": "Holo4-35B-A3B",
"description": "Serve Holo4-35B-A3B (H Company) on Runpod Serverless with vLLM.",
"type": "serverless",
"category": "image",
"iconUrl": "https://www.runpod.io/favicon.ico",
Expand Down
31 changes: 9 additions & 22 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,16 +1,16 @@
# Holo4 (Holo4-27B-GGUF, Holo4-35B-A3B-GGUF) on Runpod Serverless
# Holo4-35B-A3B on Runpod Serverless

[![Runpod](https://api.runpod.io/badge/runpod-workers/worker-holo4)](https://console.runpod.io/hub/runpod-workers/worker-holo4)

Serve Holo4 (Holo4-27B-GGUF, Holo4-35B-A3B-GGUF) (H Company) on Runpod Serverless with vLLM.
Serve Holo4-35B-A3B (H Company) on Runpod Serverless with vLLM.
License: apache-2.0.

## Recipe

| Setting | Value |
|---|---|
| Engine | vllm |
| Image | `runpod/worker-v1-vllm:v2.27.0` |
| Image | this repo's `Dockerfile`, on `vllm/vllm-openai:v0.29.0` |
| GPU | 1x RTX 6000 Ada |
| Precision | fp8 |
| Max model length | 8192 |
Expand All @@ -22,6 +22,7 @@ License: apache-2.0.
| `GPU_MEMORY_UTILIZATION` | `0.90` |
| `MAX_CONCURRENCY` | `30` |
| `MAX_MODEL_LEN` | `8192` |
| `MAX_NUM_SEQS` | `32` |
| `MODEL_NAME` | `Hcompany/Holo4-35B-A3B` |
| `QUANTIZATION` | `fp8` |
| `TENSOR_PARALLEL_SIZE` | `1` |
Expand All @@ -33,26 +34,12 @@ License: apache-2.0.
| 1x RTX 6000 Ada fp8 | $0.84 | 362.6 | 351.9 | 123.7 | $1.89 |

Cost per 1M output tokens is the hourly rate divided by measured throughput. It
assumes one saturated worker and no idle time, so treat it as a floor.
assumes one saturated worker and no idle time, so treat it as a floor. Measured by Launch
Builder on `runpod/worker-v1-vllm:v2.27.0` with the same model, GPU, precision and context
length; this worker runs vLLM v0.29.0.

## Deploy links

- Deploy on Runpod: https://console.runpod.io/deploy?template=yye0eu6dtk&utm_source=hub&utm_medium=product&utm_campaign=202609_activation_ml-engineer_h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf&utm_content=readme
- Model page: https://www.runpod.io/models/h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf?utm_source=hub&utm_medium=product&utm_campaign=202609_activation_ml-engineer_h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf&utm_content=readme
- Docs: https://docs.runpod.io/public-endpoints/models/h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf?utm_source=hub&utm_medium=product&utm_campaign=202609_activation_ml-engineer_h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf&utm_content=readme
- Deploy on Runpod: https://console.runpod.io/hub/runpod-workers/worker-holo4?utm_source=hub&utm_medium=product&utm_campaign=202609_activation_ml-engineer_h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf&utm_content=readme

Every link carries `utm_campaign=202609_activation_ml-engineer_h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf`. Keep it intact when you copy a link anywhere else.

## Before merging this bundle

Run the suite against a live endpoint:

```bash
node hub-test-suite.mjs --repo <owner>/<name> --prefix h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf- --create
```

## Fast path while this is in review

```bash
runpodctl serverless create --hub-id <vllm listing> --model-reference hf://Hcompany/Holo4-35B-A3B
```
The link carries `utm_campaign=202609_activation_ml-engineer_h-company-holo4-holo4-27b-gguf-holo4-35b-a3b-gguf`. Keep it intact when you copy a link anywhere else.
Loading