fix(hf-cpu): slim requirements.txt for Spaces; requirements-train.txt for PyTorch/TRL; suggested_hardware cpu-basic

Dockerfile

@@ -13,8 +13,9 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
     curl \
     && rm -rf /var/lib/apt/lists/*
 
-COPY requirements.txt ./
-RUN pip install --upgrade pip && pip install -r requirements.txt
+COPY requirements.txt requirements-train.txt ./
+RUN pip install --upgrade pip && \
+    pip install -r requirements.txt -r requirements-train.txt
 
 COPY . .
 

README.md

@@ -9,6 +9,8 @@ sdk_version: 5.25.0
 python_version: 3.11
 app_file: app.py
 pinned: false
+# Hint for duplicators; does not upgrade your Space by itself. CPU-friendly demo (no PyTorch in requirements.txt).
+suggested_hardware: cpu-basic
 ---
 
 # FORGE-v4 — Adversarial self-improvement for robust code generation
@@ -35,11 +37,18 @@ pinned: false
 | **Command / security cheat sheet** | [guide.md](guide.md) |
 | **Video / slides** | Optional. Current submission uses the mini-blog requirement via [MINI_BLOG.md](MINI_BLOG.md). |
 
+### Hugging Face Space (CPU-only)
+
+- This repo’s **Space README** sets **`suggested_hardware: cpu-basic`** (see [Spaces config](https://huggingface.co/docs/hub/spaces-config-reference)). **Pick CPU hardware in the Space Settings UI** if you are not on GPU.
+- **`requirements.txt`** is intentionally **light** (no PyTorch / bitsandbytes) so CPU Spaces **build and start** reliably.
+- For a stable demo on CPU, set Space secret **`CODE_PROVIDER_MODE=mock`** (or use **NIM** / **OpenRouter** keys so the router never loads local `custom_hf`). Loading **`Qwen2.5-Coder-1.5B` + LoRA** on free CPU is likely to **OOM or time out**.
+- Full training stack: install **[`requirements-train.txt`](requirements-train.txt)** on **Colab** or locally (see Quickstart).
+
 ### NOTE 1 — Non‑negotiable submission requirements (checklist)
 
 | # | Requirement | FORGE-v4 |
 | :--: | :-- | :-- |
-| 1 | **OpenEnv (latest):** build on the framework | **`openenv-core>=0.2.3`** in [`requirements.txt`](requirements.txt). Wrapper: [`env_openenv.py`](env_openenv.py) (`FORGEOpenEnvironment` subclasses `openenv.core.Environment`). Core loop remains [`env.py`](env.py) [`FORGEEnv`]. |
+| 1 | **OpenEnv (latest):** build on the framework | **`openenv-core>=0.2.3`** in [`requirements.txt`](requirements.txt). Training extras in [`requirements-train.txt`](requirements-train.txt). Wrapper: [`env_openenv.py`](env_openenv.py). Core: [`env.py`](env.py). |
 | 2 | **Training:** Unsloth or TRL (or other RL stack) + **Colab** | [`train_unsloth.py`](train_unsloth.py) (Unsloth + TRL), [`train_colab.py`](train_colab.py), [`FORGE_Training_Colab.ipynb`](FORGE_Training_Colab.ipynb), Colab links in the table above. |
 | 3 | **Evidence of training:** loss + reward plots (real run) | Committed: [`outputs/reward_curve.png`](outputs/reward_curve.png), [`outputs/loss_curve.png`](outputs/loss_curve.png), [`outputs/pass_rate.png`](outputs/pass_rate.png), [`outputs/final_report.json`](outputs/final_report.json). |
 | 4 | **Writeup / video:** mini-blog on HF *or* <2 min YouTube *etc.* | **[MINI_BLOG.md](MINI_BLOG.md)** linked here; add **public YouTube or slide URL** in the table row when published. |
@@ -277,6 +286,9 @@ Deployment note: as of the latest verification, the Space URL is serving the Gra
 
 ```bash
 pip install -r requirements.txt
+# GPU / training / HF custom local weights (skip on HF CPU Space):
+pip install -r requirements-train.txt
+
 cp .env.example .env   # then edit — do not commit .env
 python app.py          # Gradio :7860
 python api_server.py   # OpenEnv API :8000

requirements-train.txt

@@ -0,0 +1,12 @@
+# GPU / Colab / full Docker image — install on top of requirements.txt
+# Not used on Hugging Face CPU Spaces (use CODE_PROVIDER_MODE=mock there).
+
+torch>=2.3.0
+transformers>=4.41.0
+trl>=0.9.4
+peft>=0.11.1
+datasets>=2.19.0
+accelerate>=0.30.0
+bitsandbytes>=0.43.1
+
+# Unsloth: install in Colab per CUDA — https://github.com/unslothai/unsloth

requirements.txt

@@ -1,11 +1,11 @@
-# FORGE-v4: Adversarial Code Generation Environment
-# Finalized requirements for CLI, Colab, and Hugging Face Spaces.
+# FORGE-v4 — default install (Hugging Face CPU Spaces, Gradio demo, API).
+# For training / local_model / HF custom weights, also install: pip install -r requirements-train.txt
 
 # OpenEnv (official framework — Meta / HF ecosystem)
-# Needs websockets>=15.1; use Gradio >=5.18 (see README Space sdk_version) so HF Space resolves deps.
+# Needs websockets>=15.1; use Gradio >=5.18 (see README Space sdk_version).
 openenv-core>=0.2.3
 
-# Core Dependencies
+# Core
 tqdm>=4.66.0
 numpy>=1.26.0
 pandas>=2.2.0
@@ -13,22 +13,11 @@ matplotlib>=3.8.0
 PyYAML>=6.0.1
 requests>=2.32.0
 python-dotenv>=1.0.1
-# HF Spaces also installs gradio[oauth]==sdk_version from README YAML — keep in sync with sdk_version.
+
+# HF Spaces installs gradio[oauth]==sdk_version from README YAML — keep sdk_version in sync.
 gradio>=5.18.0,<6
 
-# API / Serving — Gradio 5.25+ requires fastapi>=0.115.2 (<1.0)
+# API — Gradio 5.x needs fastapi>=0.115.2
 fastapi>=0.115.2,<1.0
 uvicorn>=0.30.0
 pydantic>=2.7.0
-
-# RL Training Layer (Unsloth + TRL)
-torch>=2.3.0
-transformers>=4.41.0
-trl>=0.9.4
-peft>=0.11.1
-datasets>=2.19.0
-accelerate>=0.30.0
-bitsandbytes>=0.43.1
-
-# Note: unsloth requires specific installation based on CUDA version.
-# See: https://github.com/unslothai/unsloth

setup_colab.sh

@@ -9,8 +9,9 @@ echo "------------------------------------------------------------"
 echo "[*] Updating pip..."
 python -m pip install --upgrade pip
 
-echo "[*] Installing core requirements..."
+echo "[*] Installing core + training requirements..."
 pip install -r requirements.txt
+pip install -r requirements-train.txt
 
 echo "[*] Installing Unsloth (Colab Optimized)..."
 # Using the specific Unsloth Colab installation command