Model: Correct exl3 generation, add concurrency, and cleanup

Fixes application of sampler parameters by adding a new sampler builder
interface. Also expose the generator class-wide and add wait_for_jobs.

Finally, allow inline loading to specify the backend.

Signed-off-by: kingbri <8082010+kingbri1@users.noreply.github.com>

common/config_models.py

@@ -163,8 +163,10 @@ class ModelConfig(BaseConfigModel):
             "Example: ['max_seq_len', 'cache_mode']."
         ),
     )
+
+    # Defaults to exllamav2 in common/model.py
     backend: Optional[str] = Field(
-        "exllamav2",
+        None,
         description=(
             "Backend to use for this model (default: exllamav2)\n"
             "Options: exllamav2, exllamav3",