Skip to content

docs: remove duplicate info froms README, remove flash run from instructions#43

Closed
muhsinking wants to merge 4 commits into
mainfrom
readme-duplication
Closed

docs: remove duplicate info froms README, remove flash run from instructions#43
muhsinking wants to merge 4 commits into
mainfrom
readme-duplication

Conversation

@muhsinking

Copy link
Copy Markdown
Contributor

Users in the UX study reported that it was unnecessary/confusing to have the conceptual information repeated in the flash examples readme.

Also restructured it to put "what is flash" after the getting started steps.

@muhsinking muhsinking requested a review from deanq March 11, 2026 14:59
@muhsinking muhsinking changed the title docs: remove duplicate info from Flash examples readme docs: remove duplicate info froms README, remove flash run from instructions Mar 11, 2026
@muhsinking muhsinking requested review from KAJdev March 11, 2026 17:20

@KAJdev KAJdev left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Okay, I got halfway through using individual comments, but the underlying theme is that there's a mismatch of python script.py and uv run script.py. We should prefer uv run in all cases.

That's the main thing I found. everything else looks okay.

Comment thread 01_getting_started/01_hello_world/gpu_worker.py Outdated
Comment thread 01_getting_started/01_hello_world/README.md Outdated
Comment thread 01_getting_started/01_hello_world/README.md Outdated
Comment thread 01_getting_started/02_cpu_worker/cpu_worker.py Outdated
Comment thread 01_getting_started/02_cpu_worker/README.md Outdated
Comment thread 03_advanced_workers/05_load_balancer/cpu_lb.py Outdated
Comment thread 03_advanced_workers/05_load_balancer/gpu_lb.py Outdated
Comment thread 03_advanced_workers/05_load_balancer/README.md Outdated
Comment thread 04_scaling_performance/01_autoscaling/cpu_worker.py Outdated
Comment thread 04_scaling_performance/01_autoscaling/gpu_worker.py Outdated
Address review feedback on #43 (python script.py vs uv run mismatch):

- Worker scripts now use `uv run <script>.py` (was `python <script>.py`)
  across all example READMEs and worker file header comments.
- Local dev server uses `flash dev` (the public CLI command; `flash run`
  is a hidden deprecated alias). CLI commands kept bare; add
  `uv tool install runpod-flash` to setup steps so the CLI is on PATH.
@capy-ai

capy-ai Bot commented Jun 29, 2026

Copy link
Copy Markdown

Capy auto-review is paused for this organization because the usage-cycle auto-review limit has been reached. Increase the limit or turn it off in billing settings to resume automatic reviews.

@deanq

deanq commented Jun 29, 2026

Copy link
Copy Markdown
Member

Addressed the python script.py vs uv run inconsistency in 030bd33. Two notes on the approach, since it differs slightly from "prefer uv run in all cases":

  • Worker scripts now use uv run <script>.py everywhere (READMEs + file header comments) — these need the project venv, so uv run is required. This covers all the inline comments.
  • flash CLI is kept bare (flash dev, flash login, flash deploy) rather than uv run flash …, on the assumption the CLI is installed globally. To make that real, setup steps now include uv tool install runpod-flash so flash is on PATH.
  • Also switched flash runflash dev, since run is now a hidden/deprecated alias of dev in the CLI.

Happy to flip the CLI commands to uv run flash … instead if you'd rather standardize on the project-dependency model.

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Updates the Flash examples documentation to reduce duplicated conceptual content and shift the “getting started” flow toward running individual worker scripts directly, while moving the “What is Flash?” section after setup steps.

Changes:

  • Removed duplicate conceptual/CLI sections from the root README and reorganized “What is Flash?” below Quick Start.
  • Updated multiple example READMEs to run workers via uv run <worker>.py and adjusted local HTTP testing instructions.
  • Refreshed top-of-file comments in example workers/pipeline to match the new recommended dev/test flow.

Reviewed changes

Copilot reviewed 25 out of 25 changed files in this pull request and generated 15 comments.

Show a summary per file
File Description
README.md Removes duplicated conceptual content; updates Quick Start and prerequisites; reintroduces “What is Flash?” later.
06_real_world/README.md Updates development pattern docs to run workers directly and run a full HTTP app locally.
05_data_workflows/01_network_volumes/README.md Updates quick start to run GPU worker directly; adjusts CPU HTTP testing instructions.
05_data_workflows/01_network_volumes/gpu_worker.py Updates header comments to match the new run/testing guidance.
05_data_workflows/01_network_volumes/cpu_worker.py Updates header comments clarifying HTTP-route testing workflow.
04_scaling_performance/01_autoscaling/README.md Switches to running workers directly; updates load-test invocation to uv run.
04_scaling_performance/01_autoscaling/gpu_worker.py Updates header comments to match the new run/testing guidance.
04_scaling_performance/01_autoscaling/cpu_worker.py Updates header comments to match the new run/testing guidance.
03_advanced_workers/05_load_balancer/README.md Reworks quick start to test workers directly and separately run HTTP routes locally.
03_advanced_workers/05_load_balancer/gpu_lb.py Updates header comments to match new worker-vs-HTTP testing guidance.
03_advanced_workers/05_load_balancer/cpu_lb.py Updates header comments to match new worker-vs-HTTP testing guidance.
02_ml_inference/01_text_to_speech/README.md Switches to running the worker directly; keeps an option for HTTP API testing.
02_ml_inference/01_text_to_speech/gpu_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/04_dependencies/README.md Updates run instructions to execute workers directly; adds optional HTTP API testing.
01_getting_started/04_dependencies/mixed_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/04_dependencies/gpu_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/04_dependencies/cpu_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/03_mixed_workers/README.md Updates workflow to test workers directly and separately run the full HTTP pipeline.
01_getting_started/03_mixed_workers/pipeline.py Adds a sys.path tweak for sibling imports; updates header comments and keeps local test harness.
01_getting_started/03_mixed_workers/gpu_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/03_mixed_workers/cpu_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/02_cpu_worker/README.md Switches to running the worker directly; retains optional HTTP API testing guidance.
01_getting_started/02_cpu_worker/cpu_worker.py Updates header comments to match the new run/testing guidance.
01_getting_started/01_hello_world/README.md Switches to running the worker directly; retains optional HTTP API testing guidance.
01_getting_started/01_hello_world/gpu_worker.py Updates header comments to match the new run/testing guidance.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread README.md
**Key features:**
- **`@Endpoint` decorator**: Mark any async function to run on serverless infrastructure
- **Auto-scaling**: Scale to zero when idle, scale up under load
- **Local development**: `flash dev` starts a local server with hot reload
Comment thread 06_real_world/README.md

Or run the full app with HTTP routes:
```bash
flash dev
curl -X POST http://localhost:8888/gpu_worker/runsync \
-H "Content-Type: application/json" \
-d '{"prompt": "a sunset over mountains"}'
flash dev
curl -X POST http://localhost:8888/gpu_worker/runsync \
-H "Content-Type: application/json" \
-d '{"matrix_size": 512}'
flash dev
uv run load_test.py --endpoint /cpu_worker/runsync --requests 100 --concurrency 50
```

**Requires:** `pip install aiohttp`
The pipeline endpoint (`/classify`) orchestrates multiple workers via HTTP. To test it:

```bash
flash dev

Visit **http://localhost:8888/docs** for interactive API documentation. QB endpoints are auto-generated by `flash run` based on your `@Endpoint` functions.
```bash
flash dev
### HTTP API Testing (Optional)
```bash
flash run
flash dev

Visit **http://localhost:8888/docs** for interactive API documentation. QB endpoints are auto-generated by `flash run` based on your `@Endpoint` functions.
```bash
flash dev
### HTTP API Testing (Optional)
```bash
flash run
flash dev
@deanq

deanq commented Jun 29, 2026

Copy link
Copy Markdown
Member

Closing this as superseded by what has since landed on main:

  • fix: replace flash run with flash dev, remove standalone python references #45 (fix: replace flash run with flash dev, remove standalone python references) already covers the run-command goal of this PR — using uv run flash dev / uv run flash login and dropping the standalone python <script>.py references.
  • The "remove duplicate info" goal also no longer applies the way this PR approached it: main deliberately keeps the What is Flash? / Prerequisites / CLI Commands / Key Concepts sections.

Rebasing readme-duplication onto current main leaves an empty diff (the one unique code change, the pipeline.py sibling-import fix, is already upstream). Nothing left to merge, so closing. If the README dedup is still wanted, it should be a fresh PR on top of main using main's uv run flash dev convention.

@deanq deanq closed this Jun 29, 2026
@deanq deanq deleted the readme-duplication branch June 29, 2026 21:55
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants