auto-coder 0.1.185__tar.gz → 0.1.188__tar.gz

{auto_coder-0.1.185 → auto_coder-0.1.188}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.185
+Version: 0.1.188
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
@@ -26,7 +26,7 @@ Requires-Dist: tabulate
 Requires-Dist: jupyter_client
 Requires-Dist: prompt-toolkit
 Requires-Dist: tokenizers
-Requires-Dist: byzerllm[saas]>=0.1.137
+Requires-Dist: byzerllm[saas]>=0.1.138
 Requires-Dist: patch
 Requires-Dist: diff_match_patch
 Requires-Dist: GitPython

{auto_coder-0.1.185 → auto_coder-0.1.188}/src/auto_coder.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.185
+Version: 0.1.188
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
@@ -26,7 +26,7 @@ Requires-Dist: tabulate
 Requires-Dist: jupyter_client
 Requires-Dist: prompt-toolkit
 Requires-Dist: tokenizers
-Requires-Dist: byzerllm[saas]>=0.1.137
+Requires-Dist: byzerllm[saas]>=0.1.138
 Requires-Dist: patch
 Requires-Dist: diff_match_patch
 Requires-Dist: GitPython

{auto_coder-0.1.185 → auto_coder-0.1.188}/src/auto_coder.egg-info/requires.txt

@@ -16,7 +16,7 @@ tabulate
 jupyter_client
 prompt-toolkit
 tokenizers
-byzerllm[saas]>=0.1.137
+byzerllm[saas]>=0.1.138
 patch
 diff_match_patch
 GitPython

{auto_coder-0.1.185 → auto_coder-0.1.188}/src/autocoder/auto_coder_rag.py

@@ -256,6 +256,11 @@ def main(input_args: Optional[List[str]] = None):
         action="store_true",
         help="Enable deep thought in inference mode",
     )
+    serve_parser.add_argument(
+        "--inference_slow_without_deep_thought",
+        action="store_true",
+        help="Enable slow inference without deep thought",
+    )
 
     serve_parser.add_argument(
         "--enable_hybrid_index",
@@ -269,6 +274,12 @@ def main(input_args: Optional[List[str]] = None):
         help="The maximum number of tokens in the output. This is only used when enable_hybrid_index is true.",
     )
 
+    serve_parser.add_argument(
+        "--without_contexts",
+        action="store_true", 
+        help="Whether to return responses without contexts. only works when pro plugin is installed",
+    )
+
     # Tools command
     tools_parser = subparsers.add_parser("tools", help="Various tools")
     tools_subparsers = tools_parser.add_subparsers(dest="tool", help="Available tools")

{auto_coder-0.1.185 → auto_coder-0.1.188}/src/autocoder/common/__init__.py

@@ -316,6 +316,8 @@ class AutoCoderArgs(pydantic.BaseModel):
 
     disable_inference_enhance: Optional[bool] = False
     inference_deep_thought: Optional[bool] = False
+    inference_slow_without_deep_thought: Optional[bool] = False
+    without_contexts: Optional[bool] = False
 
     class Config:
         protected_namespaces = ()

{auto_coder-0.1.185 → auto_coder-0.1.188}/src/autocoder/rag/long_context_rag.py

@@ -343,6 +343,30 @@ class LongContextRAG:
                 )
                 return (chunk[0] for chunk in chunks), context
 
+            if self.args.without_contexts and LLMComputeEngine is not None:
+                llm_compute_engine = LLMComputeEngine(
+                    llm=self.llm,
+                    inference_enhance=not self.args.disable_inference_enhance,
+                    inference_deep_thought=self.args.inference_deep_thought,
+                    inference_slow_without_deep_thought=self.args.inference_slow_without_deep_thought                    
+                )
+                conversations = conversations[:-1]
+                new_conversations = llm_compute_engine.process_conversation(
+                    conversations, query, []
+                )
+
+                return (
+                    llm_compute_engine.stream_chat_oai(
+                        conversations=new_conversations,
+                        model=model,
+                        role_mapping=role_mapping,
+                        llm_config=llm_config,
+                        delta_mode=True,
+                    ),
+                    context,
+                )
+
+
             only_contexts = False
             try:
                 v = json.loads(query)
@@ -500,7 +524,7 @@ class LongContextRAG:
                 f"Start to send to model {target_model} with {request_tokens} tokens"
             )
 
-            if LLMComputeEngine is not None:
+            if LLMComputeEngine is not None and not self.args.disable_inference_enhance:
                 llm_compute_engine = LLMComputeEngine(
                     llm=self.llm,
                     inference_enhance=not self.args.disable_inference_enhance,

auto_coder-0.1.188/src/autocoder/version.py

@@ -0,0 +1 @@
+__version__ = "0.1.188"

auto_coder-0.1.185/src/autocoder/version.py

@@ -1 +0,0 @@
-__version__ = "0.1.185"