synth-ai 0.2.9.dev17__tar.gz → 0.2.12__tar.gz

{synth_ai-0.2.9.dev17/synth_ai.egg-info → synth_ai-0.2.12}/PKG-INFO

@@ -1,12 +1,10 @@
 Metadata-Version: 2.4
 Name: synth-ai
-Version: 0.2.9.dev17
-Summary: Python SDK and CLI for Synth — RL-as-a-service
+Version: 0.2.12
+Summary: RL as a service SDK - Core AI functionality and tracing
 Author-email: Synth AI <josh@usesynth.ai>
 License-Expression: MIT
-Project-URL: Homepage, https://usesynth.ai
-Project-URL: Documentation, https://docs.usesynth.ai/sdk/get-started
-Project-URL: Changelog, https://docs.usesynth.ai/changelog
+Project-URL: Homepage, https://github.com/synth-laboratories/synth-ai
 Project-URL: Repository, https://github.com/synth-laboratories/synth-ai
 Project-URL: Issues, https://github.com/synth-laboratories/synth-ai/issues
 Requires-Python: >=3.11
@@ -77,7 +75,7 @@ Provides-Extra: analytics
 Requires-Dist: pandas>=2.2.3; extra == "analytics"
 Dynamic: license-file
 
-# Synth-AI SDK
+# Synth-AI
 
 [![Python](https://img.shields.io/badge/python-3.11+-blue)](https://www.python.org/)
 [![License](https://img.shields.io/badge/license-MIT-green)](LICENSE)
@@ -86,106 +84,53 @@ Dynamic: license-file
 ![Tests](https://img.shields.io/badge/tests-37%2F38%20passing-brightgreen)
 ![Blacksmith CI](https://img.shields.io/badge/CI-Blacksmith%20Worker-blue)
 
-> **Synth-AI** — Reinforcement Learning-as-a-Service for agents.  
-> **Docs:** [Get Started →](https://docs.usesynth.ai/sdk/get-started)
+Docs: [Synth‑AI Documentation](https://docs.usesynth.ai/welcome/introduction)
 
----
+Fast and effective reinforcement learning for agents, via an API
 
-## 🚀 Install
+> Latest: 0.2.10 published to PyPI (uv publish)
 
-```bash
-pip install synth-ai
-# or
-uv add synth-ai
-```
+## Highlights
 
-**Import:**
+- Easily scale gpu topologies - train on 3 a10gs or 8 H100s (multi-node available upon request)
+- Requires only a thin fastapi wrapper to integrate with existing agent software.
+- Supports the best OSS models like Qwen3. (gpt-oss available upon request, GA soon)
+- Own your trained models
 
-```python
-import synth_ai
-```
+## Getting Started
 
-**CLI (with uvx):**
+synth-ai comes with a built-in RL example tailored for training Qwen/Qwen3-0.6B to succeed at Math.
+
+Please create an account at [Synth](https://usesynth.ai) and [Modal](https://modal.com) for the Math hello‑world test run. Then run:
 
 ```bash
-uvx synth-ai setup
 uvx synth-ai demo
+uvx synth-ai setup
 uvx synth-ai deploy
 uvx synth-ai run
 ```
 
-> Full quickstart: [https://docs.usesynth.ai/sdk/get-started](https://docs.usesynth.ai/sdk/get-started)
-
----
-
-## 🧩 About
-
-Fast and effective reinforcement learning for agents, via an API.  
-Easily scale GPU topologies, train multi-node, and integrate with existing agent software.
-
-### Highlights
-
-- Scale GPU topologies (A10Gs, H100s, multi-node available on request)
-- Thin FastAPI wrapper integration
-- Supports OSS models like **Qwen3** (GPT-OSS GA soon)
-- Own your trained models
-
----
+To walk through kicking off your first RL run, see the [Synth‑AI Documentation](https://docs.usesynth.ai/welcome/introduction).
 
-## ⚙️ Getting Started
+### What `setup` does now
 
-Synth-AI ships with a built-in RL example: training **Qwen3-0.6B** on math reasoning.
+When you run `uvx synth-ai setup` (or the legacy `uvx synth-ai rl_demo setup`), the SDK opens your browser to the Synth dashboard for a one‑time pairing (handshake) with your signed‑in session. The SDK will automatically:
 
-1. Create accounts at [Synth](https://usesynth.ai) and [Modal](https://modal.com)
-2. Then run:
+- Detect your current user and organization
+- Ensure both API keys exist for that user+org
+- Write the keys to your project’s `.env` file as `SYNTH_API_KEY` and `ENVIRONMENT_API_KEY`
 
-   ```bash
-   uvx synth-ai demo
-   uvx synth-ai setup
-   uvx synth-ai deploy
-   uvx synth-ai run
-   ```
+No keys are printed or requested interactively. You’ll see a confirmation like:
 
-3. To walk through your first RL run, see  
-   👉 [Synth-AI SDK Docs](https://docs.usesynth.ai/sdk/get-started)
-
----
-
-## 🔐 SDK → Dashboard Pairing
-
-When you run `uvx synth-ai setup` (or legacy `uvx synth-ai rl_demo setup`):
-
-- The SDK opens your browser to the Synth dashboard to pair your SDK with your signed-in session.
-- Automatically detects your **user + organization**
-- Ensures both **API keys** exist
-- Writes them to your project’s `.env` as:
-
-  ```
-  SYNTH_API_KEY=
-  ENVIRONMENT_API_KEY=
-  ```
-
-✅ No keys printed or requested interactively — all handled via browser pairing.
-
-### Environment overrides
-
-- `SYNTH_CANONICAL_ORIGIN` → override dashboard base URL (default: https://www.usesynth.ai/dashboard)
-- `SYNTH_CANONICAL_DEV` → `1|true|on` to use local dashboard (http://localhost:3000)
-
----
-
-## 📚 Documentation
-
-- **SDK Docs:** [https://docs.usesynth.ai/sdk/get-started](https://docs.usesynth.ai/sdk/get-started)
-- **CLI Reference:** [https://docs.usesynth.ai/cli](https://docs.usesynth.ai/cli)
-- **API Reference:** [https://docs.usesynth.ai/api](https://docs.usesynth.ai/api)
-- **Changelog:** [https://docs.usesynth.ai/changelog](https://docs.usesynth.ai/changelog)
+```
+Connecting SDK to your browser session…
+Connected to Acme Labs via browser.
+```
 
----
+If your browser isn’t already signed in, sign in when prompted and the pairing completes automatically. The dashboard’s welcome modal will reflect a successful pairing.
 
-## 🧠 Meta
+Environment variables:
 
-- Package: [`synth-ai`](https://pypi.org/project/synth-ai)
-- Import: `synth_ai`
-- Source: [github.com/synth-laboratories/synth-ai](https://github.com/synth-laboratories/synth-ai)
-- License: MIT
+- `SYNTH_CANONICAL_ORIGIN` (optional): override the dashboard base URL the SDK uses for the handshake (defaults to `https://www.usesynth.ai/dashboard`).
+- `SYNTH_CANONICAL_DEV` (optional): set to `1`, `true`, `yes`, or `on` to target the local dashboard at `http://localhost:3000`.
+- Keys are stored only in your project’s `.env` file, not exported to your shell.

synth_ai-0.2.12/README.md

@@ -0,0 +1,59 @@
+# Synth-AI
+
+[![Python](https://img.shields.io/badge/python-3.11+-blue)](https://www.python.org/)
+[![License](https://img.shields.io/badge/license-MIT-green)](LICENSE)
+[![PyPI](https://img.shields.io/badge/PyPI-0.2.10-orange)](https://pypi.org/project/synth-ai/)
+![Coverage](https://img.shields.io/badge/coverage-9.09%25-red)
+![Tests](https://img.shields.io/badge/tests-37%2F38%20passing-brightgreen)
+![Blacksmith CI](https://img.shields.io/badge/CI-Blacksmith%20Worker-blue)
+
+Docs: [Synth‑AI Documentation](https://docs.usesynth.ai/welcome/introduction)
+
+Fast and effective reinforcement learning for agents, via an API
+
+> Latest: 0.2.10 published to PyPI (uv publish)
+
+## Highlights
+
+- Easily scale gpu topologies - train on 3 a10gs or 8 H100s (multi-node available upon request)
+- Requires only a thin fastapi wrapper to integrate with existing agent software.
+- Supports the best OSS models like Qwen3. (gpt-oss available upon request, GA soon)
+- Own your trained models
+
+## Getting Started
+
+synth-ai comes with a built-in RL example tailored for training Qwen/Qwen3-0.6B to succeed at Math.
+
+Please create an account at [Synth](https://usesynth.ai) and [Modal](https://modal.com) for the Math hello‑world test run. Then run:
+
+```bash
+uvx synth-ai demo
+uvx synth-ai setup
+uvx synth-ai deploy
+uvx synth-ai run
+```
+
+To walk through kicking off your first RL run, see the [Synth‑AI Documentation](https://docs.usesynth.ai/welcome/introduction).
+
+### What `setup` does now
+
+When you run `uvx synth-ai setup` (or the legacy `uvx synth-ai rl_demo setup`), the SDK opens your browser to the Synth dashboard for a one‑time pairing (handshake) with your signed‑in session. The SDK will automatically:
+
+- Detect your current user and organization
+- Ensure both API keys exist for that user+org
+- Write the keys to your project’s `.env` file as `SYNTH_API_KEY` and `ENVIRONMENT_API_KEY`
+
+No keys are printed or requested interactively. You’ll see a confirmation like:
+
+```
+Connecting SDK to your browser session…
+Connected to Acme Labs via browser.
+```
+
+If your browser isn’t already signed in, sign in when prompted and the pairing completes automatically. The dashboard’s welcome modal will reflect a successful pairing.
+
+Environment variables:
+
+- `SYNTH_CANONICAL_ORIGIN` (optional): override the dashboard base URL the SDK uses for the handshake (defaults to `https://www.usesynth.ai/dashboard`).
+- `SYNTH_CANONICAL_DEV` (optional): set to `1`, `true`, `yes`, or `on` to target the local dashboard at `http://localhost:3000`.
+- Keys are stored only in your project’s `.env` file, not exported to your shell.

synth_ai-0.2.12/examples/dev/qwen3_32b_qlora_4xh100.toml

@@ -0,0 +1,40 @@
+[job]
+model = "Qwen/Qwen3-32B"
+# Optionally set here; you can also pass --dataset
+# data = "/abs/path/to/train.jsonl"
+
+[compute]
+gpu_type = "H100"
+gpu_count = 4
+nodes = 1
+
+[data]
+# Optional; forwarded into metadata.effective_config.data.topology
+topology = { container_count = 4 }
+
+[training]
+mode = "sft_offline"
+use_qlora = true
+
+[training.validation]
+enabled = true
+evaluation_strategy = "steps"
+eval_steps = 20
+save_best_model_at_end = true
+metric_for_best_model = "val.loss"
+greater_is_better = false
+
+[hyperparameters]
+n_epochs = 1
+per_device_batch = 1
+gradient_accumulation_steps = 64
+sequence_length = 4096
+learning_rate = 5e-6
+warmup_ratio = 0.03
+
+[hyperparameters.parallelism]
+use_deepspeed = true
+deepspeed_stage = 2
+bf16 = true
+fp16 = false
+fsdp = false

synth_ai-0.2.12/examples/multi_step/crafter_rl_lora.md

@@ -0,0 +1,29 @@
+# Crafter RL LoRA (10-step runs)
+
+This walkthrough shows how to fine-tune the Crafter task app with our 10-step RL LoRA config.
+
+1. **Start the Crafter task app on Modal (with tracing + text-only prompts)**
+
+   ```bash
+   BACKEND_BASE_URL=https://agent-learning.onrender.com/api \
+   uvx synth-ai modal-serve grpo-crafter \
+     --env-file examples/warming_up_to_rl/.env \
+     --name grpo-crafter-task-app
+   ```
+
+   * Deploys the Modal task app with the tracing/text-only fixes baked in.*
+
+2. **Launch the RL job using the updated LoRA config**
+
+   ```bash
+   uvx synth-ai train --type rl \
+     --config tests/artifacts/configs/rl.lora.small.toml \
+     --backend https://agent-learning.onrender.com/api \
+     --env-file .env \
+     --no-poll
+   ```
+
+   * This config forces 10 agent turns per rollout, reduces batch size to avoid OOMs, and enforces Crafter-specific defaults.*
+
+  INFO - 🎉 Training completed successfully!
+  INFO - All batch rewards: [0.0625, 0.0625, 0.125, 0.0625, 0.0625, 0.3125, 0.375, 0.4375, 0.5, 0.9375]