feat: make snapshot tests run on x86 Ubuntu CI without GPU

MLX already supports x86 CPU via mlx[cpu] and the Dockerfile has the
GCC workaround for CPU JIT. The only barriers were the 'slow' markers
causing tests to be skipped in CI.

Changes:
- Remove 'slow' marker from all snapshot tests so they run by default
- Make snapshots architecture-aware (snapshots/{arch}/{name}.json) since
  floating-point results differ between x86_64 and arm64
- Store architecture in snapshot metadata
- Increase CI timeout from 30 to 45 minutes for model download + CPU inference
- Update docstrings to remove Apple Silicon requirement

Co-Authored-By: Claude Opus 4.6 <[email protected]>

.github/workflows/e2e.yml

@@ -10,7 +10,7 @@ on:
 jobs:
   e2e:
     runs-on: ubuntu-latest
-    timeout-minutes: 30
+    timeout-minutes: 45
     steps:
       - name: Free up disk space
         run: |

e2e/snapshot.py

@@ -3,14 +3,19 @@
 Provides deterministic regression testing by comparing inference output
 against saved snapshots. On first run, snapshots are created automatically.
 Set UPDATE_SNAPSHOTS=1 to regenerate snapshots when output intentionally changes.
+
+Snapshots are stored per-architecture (e.g. snapshots/x86_64/, snapshots/arm64/)
+since floating-point results differ between CPU architectures.
 """
 
 import difflib
 import json
 import os
+import platform
 from pathlib import Path
 
-SNAPSHOTS_DIR = Path(__file__).parent / "snapshots"
+ARCH = platform.machine()
+SNAPSHOTS_DIR = Path(__file__).parent / "snapshots" / ARCH
 
 
 def assert_snapshot(
@@ -21,7 +26,7 @@ def assert_snapshot(
     """Compare content against a saved snapshot, or create one if missing.
 
     Args:
-        name: Snapshot identifier (used as filename: snapshots/{name}.json).
+        name: Snapshot identifier (used as filename: snapshots/{arch}/{name}.json).
         content: The actual inference output to compare.
         metadata: Additional context stored alongside content (model, seed, etc.).
                   Not used for comparison -- purely documentary.
@@ -49,16 +54,16 @@ def assert_snapshot(
                 )
             )
             raise AssertionError(
-                f"Snapshot mismatch for '{name}'!\n\n"
+                f"Snapshot mismatch for '{name}' on {ARCH}!\n\n"
                 f"{diff}\n\n"
                 f"Expected: {expected!r}\n"
                 f"Actual:   {content!r}\n\n"
-                f"To update: UPDATE_SNAPSHOTS=1 python3 e2e/run_all.py --slow"
+                f"To update: UPDATE_SNAPSHOTS=1 python3 e2e/run_all.py"
             )
-        print(f"  Output matches snapshot ({snapshot_file.name})")
+        print(f"  Output matches snapshot ({ARCH}/{snapshot_file.name})")
     else:
         SNAPSHOTS_DIR.mkdir(parents=True, exist_ok=True)
-        snapshot_data = {**metadata, "content": content}
+        snapshot_data = {**metadata, "arch": ARCH, "content": content}
         snapshot_file.write_text(json.dumps(snapshot_data, indent=2) + "\n")
         action = "Updated" if update else "Created"
-        print(f"  {action} snapshot: {snapshot_file}")
+        print(f"  {action} snapshot: {ARCH}/{snapshot_file.name}")

e2e/test_inference_snapshot.py

@@ -1,12 +1,10 @@
 """Test: Deterministic inference output (snapshot test).
-slow
 
 Sends a chat completion request with a fixed seed,
 then verifies the output matches a known-good snapshot. This ensures
 inference produces consistent results across runs.
 
-Requires a machine that can run MLX inference at reasonable speed (Apple Silicon).
-Run with: python3 e2e/run_all.py --slow  or  E2E_SLOW=1 python3 e2e/run_all.py
+Uses MLX CPU backend in Docker on x86 Linux.
 """
 
 import asyncio

e2e/test_snapshot_code_gen.py

@@ -1,8 +1,6 @@
 """Test: Code generation snapshot.
-slow
 
 Verifies deterministic output for a code generation prompt.
-Run with: python3 e2e/run_all.py --slow  or  E2E_SLOW=1 python3 e2e/run_all.py
 """
 
 import asyncio

e2e/test_snapshot_edge.py

@@ -1,9 +1,7 @@
 """Test: Edge case snapshots.
-slow
 
-Verifies deterministic output for edge-case prompts: empty-ish input,
-very short input, and special characters.
-Run with: python3 e2e/run_all.py --slow  or  E2E_SLOW=1 python3 e2e/run_all.py
+Verifies deterministic output for edge-case prompts: single word input,
+special characters, and unicode.
 """
 
 import asyncio

e2e/test_snapshot_long_output.py

@@ -1,8 +1,6 @@
 """Test: Longer output snapshot.
-slow
 
 Verifies deterministic output with a higher max_tokens (128).
-Run with: python3 e2e/run_all.py --slow  or  E2E_SLOW=1 python3 e2e/run_all.py
 """
 
 import asyncio

e2e/test_snapshot_reasoning.py

@@ -1,8 +1,6 @@
 """Test: Reasoning/math snapshot.
-slow
 
 Verifies deterministic output for a simple reasoning prompt.
-Run with: python3 e2e/run_all.py --slow  or  E2E_SLOW=1 python3 e2e/run_all.py
 """
 
 import asyncio