auto-coder 0.1.191__py3-none-any.whl → 0.1.192__py3-none-any.whl

{auto_coder-0.1.191.dist-info → auto_coder-0.1.192.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.191
+Version: 0.1.192
 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] >=0.1.138
+Requires-Dist: byzerllm[saas] >=0.1.139
 Requires-Dist: patch
 Requires-Dist: diff-match-patch
 Requires-Dist: GitPython

{auto_coder-0.1.191.dist-info → auto_coder-0.1.192.dist-info}/RECORD

@@ -1,13 +1,13 @@
 autocoder/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/auto_coder.py,sha256=tSNXFMJSrffagFi4egZJp8XZH9trSCwQjOdXKyHVqwo,37106
 autocoder/auto_coder_lang.py,sha256=Rtupq6N3_HT7JRhDKdgCBcwRaiAnyCOR_Gsp4jUomrI,3229
-autocoder/auto_coder_rag.py,sha256=GFPp02FE8x14pmmRxIpbkdx5ll-zYyPK3SACjhSeZ8A,16578
+autocoder/auto_coder_rag.py,sha256=H75N9rk5rB1u80oDWsPS_DYc2djQcFLOpjklmlGmag0,16579
 autocoder/auto_coder_server.py,sha256=qRY88mkBnqSGFDcwYE5gwpe2WPhIw1nEH6LdbjCQhQk,20306
 autocoder/chat_auto_coder.py,sha256=1jCx-J83mj_8JnojYSTfPjYide-thbmsFbr12E_kgcQ,81773
 autocoder/chat_auto_coder_lang.py,sha256=QYtu5gWEQmWKVovR_qUZ8plySZarNFX_Onk-1vN9IiA,8524
 autocoder/command_args.py,sha256=ftWw6HnFUZPiQPt1oV-SfpHQe69XN3knaFy1lpROBcU,26854
 autocoder/lang.py,sha256=e-07rYTgimpxS8sm-AxKSmH4kKQX4N05YFHJBg9trVs,12598
-autocoder/version.py,sha256=XJsMAfwtRQRLx1sRlYREnDsvZzeVXdAZOFWs0dVaf_I,24
+autocoder/version.py,sha256=_O_6jlMqTIlOgOmhHKPoMTN-W0CigzKhwb-eIbBGdIQ,24
 autocoder/agent/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/agent/auto_tool.py,sha256=DBzip-P_T6ZtT2eHexPcusmKYD0h7ufzp7TLwXAY10E,11554
 autocoder/agent/coder.py,sha256=dnITYHqkcOip8zV4lywbkYNH9w7Q3qyYaUArJ4WPrTs,866
@@ -17,7 +17,7 @@ autocoder/agent/project_reader.py,sha256=-MWRqsr7O4mvU0PIpAhOUBb29htZAvA37pa_GeE
 autocoder/chat/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/common/JupyterClient.py,sha256=O-wi6pXeAEYhAY24kDa0BINrLYvKS6rKyWe98pDClS0,2816
 autocoder/common/ShellClient.py,sha256=fM1q8t_XMSbLBl2zkCNC2J9xuyKN3eXzGm6hHhqL2WY,2286
-autocoder/common/__init__.py,sha256=CceR1pHKhPPl-y_vHgjf-p8pe4xgeOI_CTjkUrTN2hM,10510
+autocoder/common/__init__.py,sha256=7K74EPvyNyZehp5jA4NlMPpmdAkqMPcphcRnkuYYRMg,10511
 autocoder/common/anything2images.py,sha256=0ILBbWzY02M-CiWB-vzuomb_J1hVdxRcenAfIrAXq9M,25283
 autocoder/common/audio.py,sha256=Kn9nWKQddWnUrAz0a_ZUgjcu4VUU_IcZBigT7n3N3qc,7439
 autocoder/common/cleaner.py,sha256=NU72i8C6o9m0vXExab7nao5bstBUsfJFcj11cXa9l4U,1089
@@ -58,11 +58,11 @@ autocoder/index/index.py,sha256=6uakPXThpDWxAyOAP-7AbMuXaXJJkBKctL5RkNWGdGw,2248
 autocoder/index/symbols_utils.py,sha256=CjcjUVajmJZB75Ty3a7kMv1BZphrm-tIBAdOJv6uo-0,2037
 autocoder/pyproject/__init__.py,sha256=-2-ImQVw6e3NQZQOyDlHEP5b4xVs5ur2G5izB-JCa-A,13160
 autocoder/rag/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-autocoder/rag/api_server.py,sha256=zokIlDJlk7ucRorSLQm80uICO1mecfmn4J2zVqEBskE,6786
+autocoder/rag/api_server.py,sha256=dRbhAZVRAOlZ64Cnxf4_rKb4iJwHnrWS9Zr67IVORw0,7288
 autocoder/rag/doc_filter.py,sha256=Ha0Yae_G_hF72YzvrO7NoDZcG18K4hRcqGAEqfrIwAs,9330
 autocoder/rag/document_retriever.py,sha256=_jCbCEX0I-5UPWuHocESaWHatQcv1r_DqA0yOfOAiZ0,9092
-autocoder/rag/llm_wrapper.py,sha256=xRbTBpLUH43Ah5jplL8WWWU-kjKfNgEJoUntLGBq5F4,2484
-autocoder/rag/long_context_rag.py,sha256=0buBOS72OcMNY61fMxF13iiLoJ6W7IqdmHEmsjv7t68,22559
+autocoder/rag/llm_wrapper.py,sha256=sbDxCANiZyWb_ocqNgqu2oy3c2t8orPNRGleEs-Uwl8,2649
+autocoder/rag/long_context_rag.py,sha256=B9lBeiBF7p5RRRFV8Wv_JtqNTED7hWuBwY766ZPQI1c,23749
 autocoder/rag/rag_config.py,sha256=8LwFcTd8OJWWwi1_WY4IzjqgtT6RyE2j4PjxS5cCTDE,802
 autocoder/rag/rag_entry.py,sha256=V1RJ8RGqM30DNPmzymv64rZjNRGWn6kfc8sRy_LECg0,2451
 autocoder/rag/raw_rag.py,sha256=yS2Ur6kG0IRjhCj2_VonwxjY_xls_E62jO5Gz5j2nqE,2952
@@ -85,6 +85,9 @@ autocoder/rag/loaders/docx_loader.py,sha256=g6Ta8rMUbfgwB8N1qiajhyO6wpaWl7zygAZi
 autocoder/rag/loaders/excel_loader.py,sha256=Ue8YB1z_kBs8SjIPuBskyM08Q1JiONs_BJZPrzi59oo,896
 autocoder/rag/loaders/pdf_loader.py,sha256=CGfXOja7QZ7mHN-U5MsTiVMFzjP322rTj3dkYlVKKVU,264
 autocoder/rag/loaders/ppt_loader.py,sha256=7VEYc-bqgK8VHCoGC3DIUcqbpda-E5jQF9lYLqP256I,1681
+autocoder/rag/stream_event/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+autocoder/rag/stream_event/event_writer.py,sha256=l7kq_LnDDE8E5dZ-73C7J2MgzSL7WrozdXk0eV-k55Q,409
+autocoder/rag/stream_event/types.py,sha256=rtLwOE8rShmi1dJdxyBpAV5ZjLBGG9vptMiSzMxGuIA,318
 autocoder/regex_project/__init__.py,sha256=EBZeCL5ORyD_9_5u_UuG4s7XtpXOu0y1sWDmxWFtufE,6781
 autocoder/regexproject/__init__.py,sha256=ThuvVFdpw1EgWv4aIRkhg3ZclKPxMVharUKWppFpQ8o,8436
 autocoder/suffixproject/__init__.py,sha256=EaQoumMzZ2COxMiI_GnL3SG4LGzRj0Qw7UpqLfNLCw8,9823
@@ -101,9 +104,9 @@ autocoder/utils/request_event_queue.py,sha256=r3lo5qGsB1dIjzVQ05dnr0z_9Z3zOkBdP1
 autocoder/utils/request_queue.py,sha256=nwp6PMtgTCiuwJI24p8OLNZjUiprC-TsefQrhMI-yPE,3889
 autocoder/utils/rest.py,sha256=3tXA8KZG6jKz_tddHNLGx77Icee88WcUeesfNsgPno4,8790
 autocoder/utils/tests.py,sha256=BqphrwyycGAvs-5mhH8pKtMZdObwhFtJ5MC_ZAOiLq8,1340
-auto_coder-0.1.191.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
-auto_coder-0.1.191.dist-info/METADATA,sha256=CXw53iIxkBB4wKuLNiBdmkkdYbRuNnD-newVfg8aBic,2352
-auto_coder-0.1.191.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-auto_coder-0.1.191.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
-auto_coder-0.1.191.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
-auto_coder-0.1.191.dist-info/RECORD,,
+auto_coder-0.1.192.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
+auto_coder-0.1.192.dist-info/METADATA,sha256=BBKDDDbjH7IKFXSuveBZsDjn3soimiXniATh-8-_Rlk,2352
+auto_coder-0.1.192.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+auto_coder-0.1.192.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
+auto_coder-0.1.192.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
+auto_coder-0.1.192.dist-info/RECORD,,

autocoder/auto_coder_rag.py

@@ -264,7 +264,7 @@ def main(input_args: Optional[List[str]] = None):
     serve_parser.add_argument(
         "--inference_compute_precision",
         type=int,
-        default=6,
+        default=64,
         help="The precision of the inference compute",
     )
 

autocoder/common/__init__.py

@@ -317,7 +317,7 @@ class AutoCoderArgs(pydantic.BaseModel):
     disable_inference_enhance: Optional[bool] = False
     inference_deep_thought: Optional[bool] = False
     inference_slow_without_deep_thought: Optional[bool] = False
-    inference_compute_precision: int = 6
+    inference_compute_precision: int = 64
     without_contexts: Optional[bool] = False
 
     class Config:

autocoder/rag/api_server.py

@@ -19,10 +19,10 @@ from byzerllm.utils.client.entrypoints.openai.protocol import (
     ChatCompletionRequest,
     ErrorResponse,
     CompletionRequest,
-    Embeddings,
-    EmbeddingsOutput,
-    EmbeddingsData,
-    EmbeddingsUsage,
+    EmbeddingCompletionRequest,
+    EmbeddingResponse,
+    EmbeddingResponseData,
+    UsageInfo,
 )
 from pydantic import BaseModel
 from typing import List,Optional
@@ -122,34 +122,49 @@ async def create_chat_completion(
 
 
 @router_app.post("/v1/embeddings")
-async def embed(body: Embeddings):
-    """Given a prompt, the model will return one embedding.
-
+async def embed(body: EmbeddingCompletionRequest):
+    """Generate embeddings for given input text.
+    
+    Args:
+        body: The embedding request containing input text and parameters.
+        
     Returns:
-        A response object with an embedding.
+        EmbeddingResponse with embeddings and usage statistics.
     """
     embedding_id = f"embed-{random_uuid()}"
-
-    results_list = llm_client.emb(body.model, request=LLMRequest(instruction=body.input))
-    tokens = 0
-
-    return EmbeddingsOutput(
-        data=[
-            EmbeddingsData(
-                embedding=results.output,
-                index=i,
-                object="embedding",
-            )
-            for i, results in enumerate(results_list)
-        ],
-        id=embedding_id,
-        object="list",
-        created=int(time.time()),
+    
+    # Handle both string and list inputs
+    inputs = body.input if isinstance(body.input, list) else [body.input]
+    
+    # Generate embeddings for each input
+    results_list = []
+    for text in inputs:
+        result = llm_client.emb(body.model, request=LLMRequest(instruction=text))
+        results_list.extend(result)
+
+    # Build response data
+    data = [
+        EmbeddingResponseData(
+            embedding=result.output,
+            index=i,
+            object="embedding"
+        )
+        for i, result in enumerate(results_list)
+    ]
+    
+    # Calculate token usage (simplified)
+    token_count = sum(len(str(input).split()) for input in inputs)
+    
+    return EmbeddingResponse(
+        data=data,
         model=body.model,
-        usage=EmbeddingsUsage(
-            prompt_tokens=tokens,
-            total_tokens=tokens,
+        object="list",
+        usage=UsageInfo(
+            prompt_tokens=token_count,
+            total_tokens=token_count
         ),
+        created=int(time.time()),
+        id=embedding_id
     )
 
 class ServerArgs(BaseModel):

autocoder/rag/llm_wrapper.py

@@ -9,6 +9,7 @@ from byzerllm.utils.client import LLMResponse
 from byzerllm.utils.types import SingleOutputMeta
 from autocoder.rag.simple_rag import SimpleRAG
 from loguru import logger
+from byzerllm.utils.langutil import asyncfy_with_semaphore
 
 class LLWrapper:
 
@@ -31,7 +32,7 @@ class LLWrapper:
                  model:Optional[str] = None,
                  role_mapping=None,llm_config:Dict[str,Any]={}
                  )->Union[List[LLMResponse],List[LLMFunctionCallResponse],List[LLMClassResponse]]: 
-        res,contexts = self.rag.stream_chat_oai(conversations)
+        res,contexts = self.rag.stream_chat_oai(conversations,llm_config=llm_config)
         s = "".join(res)        
         return [LLMResponse(output=s,metadata={},input="")]
 
@@ -40,7 +41,7 @@ class LLWrapper:
                         role_mapping=None,
                         delta_mode=False,
                         llm_config:Dict[str,Any]={}): 
-        res,contexts = self.rag.stream_chat_oai(conversations)        
+        res,contexts = self.rag.stream_chat_oai(conversations,llm_config=llm_config)        
         for t in res:                        
             yield (t,SingleOutputMeta(0,0))
 
@@ -49,7 +50,7 @@ class LLWrapper:
                         role_mapping=None,
                         delta_mode=False,
                         llm_config:Dict[str,Any]={}): 
-        res,contexts = self.rag.stream_chat_oai(conversations)                
+        res,contexts = await asyncfy_with_semaphore(lambda: self.rag.stream_chat_oai(conversations,llm_config=llm_config))()                
         for t in res:                                    
             yield (t,SingleOutputMeta(0,0))
                          

autocoder/rag/long_context_rag.py

@@ -30,6 +30,7 @@ from autocoder.rag.token_limiter import TokenLimiter
 from tokenizers import Tokenizer
 from autocoder.rag.variable_holder import VariableHolder
 from importlib.metadata import version
+from autocoder.rag.stream_event import event_writer
 
 try:        
     from autocoder_pro.rag.llm_compute import LLMComputeEngine
@@ -346,6 +347,29 @@ class LongContextRAG:
                     delta_mode=True,
                 )
                 return (chunk[0] for chunk in chunks), context
+            
+            try:
+                request_params = json.loads(query)
+                if "request_id" in request_params:                    
+                    request_id = request_params["request_id"]
+                    index = request_params["index"]
+                    
+                    file_path = event_writer.get_event_file_path(request_id)                    
+                    logger.info(f"Get events for request_id: {request_id} index: {index} file_path: {file_path}")
+                    events = []
+                    if not os.path.exists(file_path):
+                        return [],context
+
+                    with open(file_path, "r") as f:
+                        for line in f:
+                            event = json.loads(line)
+                            if event["index"] >= index:
+                                events.append(event)
+                    return [json.dumps({
+                        "events": [event for event in events],                        
+                    },ensure_ascii=False)], context                
+            except json.JSONDecodeError:
+                pass
 
             if self.args.without_contexts and LLMComputeEngine is not None:
                 llm_compute_engine = LLMComputeEngine(

autocoder/rag/stream_event/event_writer.py

@@ -0,0 +1,12 @@
+import os
+from autocoder.rag.stream_event.types import Event
+
+def write_event(event: Event,base_path: str="events"):
+    os.makedirs(base_path, exist_ok=True)
+    with open(f"{base_path}/{event.request_id}.jsonl", "a") as f:
+        f.write(event.model_dump_json() + "\n")
+
+def get_event_file_path(request_id: str,base_path: str="events") -> str:    
+    return f"{base_path}/{request_id}.jsonl"
+        
+    

autocoder/rag/stream_event/types.py

@@ -0,0 +1,16 @@
+from typing import List, Dict, Any
+from enum import Enum
+from pydantic import BaseModel
+
+class EventType(Enum):
+    START = "start"
+    THOUGHT = "thought"
+    CHUNK = "chunk"
+    DONE = "done"
+    ERROR = "error"
+
+class Event(BaseModel):
+    request_id: str
+    event_type: EventType
+    content: str  
+    index: int

autocoder/version.py

@@ -1 +1 @@
-__version__ = "0.1.191"
+__version__ = "0.1.192"