bugfix

llmengine/base_chat_llm.py

@@ -3,6 +3,7 @@ import asyncio
 import json
 import torch
 from time import time
+from aiostream import stream
 from transformers import TextIteratorStreamer
 from appPublic.log import debug
 from appPublic.worker import awaitify
@@ -26,18 +27,6 @@ class BaseChatLLM:
 			device = torch.device("mps")
 			self.model = self.model.to(device)
 
-	def get_session_key(self):
-		return self.model_id + ':messages'
-
-	def _get_session_messages(self, session):
-		key = self.get_session_key()
-		messages = session.get(key) or []
-		return messages
-
-	def _set_session_messages(self, session, messages):
-		key = self.get_session_key()
-		session[key] = messages
-
 	def get_streamer(self):
 		return TextIteratorStreamer(
 			tokenizer=self.tokenizer,
@@ -49,7 +38,7 @@ class BaseChatLLM:
 		all_txt = ''
 		t1 = time()
 		i = 0
-		id = f'chatllm-{getID}'
+		id = f'chatllm-{getID()}'
 		for txt in streamer:
 			if txt == '':
 				continue
@@ -60,15 +49,15 @@ class BaseChatLLM:
 			yield {
 				"id":id,
 				"object":"chat.completion.chunk",
-				"created":time(),
+				"created": t1,
 				"model":self.model_id,
 				"choices":[
 					{
 						"index":0,
 						"delta":{
+							"role": "assistant",
 							"content":txt
 						},
-						"logprobs":None,
 						"finish_reason":None
 					}
 				]
@@ -80,7 +69,7 @@ class BaseChatLLM:
 		yield {
 			"id":id,
 			"object":"chat.completion.chunk",
-			"created":time(),
+			"created": t1,
 			"model":self.model_id,
 			"response_time": t2 - t1,
 			"finish_time": t3 - t1,
@@ -91,69 +80,11 @@ class BaseChatLLM:
 					"delta":{
 						"content":""
 					},
-					"logprobs":None,
 					"finish_reason":"stop"
 				}
 			]
 		}
 		
-	def _generator(self, session, prompt, image_path, video_path, audio_path, sys_prompt):
-		messages = self._get_session_messages(session)
-		if sys_prompt:
-			messages.append(self._build_sys_message(sys_prompt))
-		messages.append(self._build_user_message(prompt, image_path=image_path))
-		# debug(f'{messages=}')
-		all_txt = ''
-		for d in self._gen(messages):
-			if d['choices'][0]['finish_reason'] == 'stop':
-				messages.append(self._build_assistant_message(all_txt))
-			else:
-				all_txt += d['choices'][0]['delta']['content']
-			yield d
-		self._set_session_messages(session, messages)
-
-	async def _async_generator(self, session, prompt, image_path, video_path, audio_path, sys_prompt):
-		for d in self._generator(session, prompt, image_path, video_path, audio_path, sys_prompt):
-			await asyncio.sleep(0)
-			yield d
-
-	def generate(self, session, prompt,
-						image_path=None, 
-						video_path=None, 
-						audio_path=None, 
-						sys_prompt=None):
-		for d in self._generator(session, prompt, image_path, video_path, audio_path, sys_prompt):
-			if d['choices'][0]['finish_reason'] == 'stop':
-				return d
-	def stream_generate(self, session, prompt,
-						image_path=None, 
-						video_path=None, 
-						audio_path=None, 
-						sys_prompt=None):
-		for d in self._generator(session, prompt, image_path, video_path, audio_path, sys_prompt):
-			s = f'data: {json.dumps(d)}\n'
-			yield s
-
-	async def async_generate(self, session, prompt,
-                        image_path=None, 
-                        video_path=None,
-                        audio_path=None,
-                        sys_prompt=None):
-		async for d in self._async_generator(session, prompt, image_path, video_path, audio_path, sys_prompt):
-			await asyncio.sleep(0)
-			if d['choices'][0]['finish_reason'] == 'stop':
-				return d
-
-	async def async_stream_generate(self, session, prompt,
-                        image_path=None,
-                        video_path=None,
-                        audio_path=None,
-                        sys_prompt=None):
-		async for d in self._async_generator(session, prompt, image_path, video_path, audio_path, sys_prompt):
-			s = f'data: {json.dumps(d)}\n'
-			yield s
-		yield 'data: [DONE]'
-
 	def build_kwargs(self, inputs, streamer):
 		generate_kwargs = dict(
 			**inputs,
@@ -165,6 +96,7 @@ class BaseChatLLM:
 		return generate_kwargs
 
 	def _messages2inputs(self, messages):
+		debug(f'{messages=}')
 		return self.processor.apply_chat_template(
 			messages, add_generation_prompt=True, 
 			tokenize=True,
@@ -184,63 +116,70 @@ class BaseChatLLM:
 				d['input_tokens'] = input_len
 			yield d
 
-class T2TChatLLM(BaseChatLLM):
+	async def async_gen(self, messages):
-	def _build_assistant_message(self, prompt):
+		async for d in stream.iterate(self._gen(messages)):
-		return {
+			yield d
-			"role":"assistant",
-			"content":prompt
-		}
 	
-	def _build_sys_message(self, prompt):
+	async def chat_completion_stream(self, messages):
-		return {
+		async for d in self.async_gen(messages):
-			"role":"system",
+			if d['choices'][0]['finish_reason']:
-			"content": prompt
+				d['usage'] = {
-		}
+					'prompt_tokens': d['input_tokens'],
+					'completion_tokens': d['output_tokens'],
+					'total_tokens': d['input_tokens'] + d['output_tokens']
+				}
+			s = f'data: {json.dumps(d)}\n'
+			yield s
+		yield 'data: [DONE]\n'
 	
-	def _build_user_message(self, prompt, **kw):
+	def reference(self, messages):
-		return {
+		t1 = time()
-			"role":"user",
+		inputs = self._messages2inputs(messages)
-			"content": prompt
+		input_len = inputs["input_ids"].shape[-1]
-		}
+		streamer = self.get_streamer()
+		kwargs = self.build_kwargs(inputs, streamer)
+		thread = threading.Thread(target=self.model.generate, 
+						kwargs=kwargs)
+		thread.start()
+		txt = ''
+		i = 0
+		for d in self.output_generator(streamer):
+			if i == 0:
+				i = 1
+				t1 = time()
+			if d['choices'][0]['finish_reason'] != 'stop':
+				txt += d['choices'][0]['delta']['content']
+			else:
+				i_tokens = d['input_tokens']
+				o_tokens = d['output_tokens']
 
-class MMChatLLM(BaseChatLLM):
+		t2 = time()
-	""" multiple modal chat LLM """
-	def _build_assistant_message(self, prompt):
 		return {
-			"role":"assistant",
+			'id': f'chatcmpl-{getID()}',
-			"content":[{"type": "text", "text": prompt}]
+			"object":"chat.completion",
-		}
+			"created":t1,
-
+			"model":self.model_id,
-	def _build_sys_message(self, prompt):
+			"response_time": t2 - t1,
-		return {
+			"finish_time": t3 - t1,
-			"role":"system",
+			"output_token": output_tokens,
-			"content":[{"type": "text", "text": prompt}]
+			"choices":[
-		}
+				{
-
+					"index":0,
-	def _build_user_message(self, prompt, image_path=None, 
+					"message":{
-					video_path=None, audio_path=None):
+						"role": "assistant",
-		contents = [
+						"content": txt
-			{
+					},
-				"type":"text", "text": prompt
+					"finish_reason":"stop"
+				}
+			],
+			"usage": {
+				"prompt_tokens": i_tokens,
+				"completion_tokens": o_tokens,
+				"total_tokens": i_tokens + o_tokens
 			}
-		]
-		if image_path:
-			contents.append({
-				"type": "image",
-				"image": image_path
-			})
-		if video_path:
-			contents.append({
-				"type": "video",
-				"video":video_path
-			})
-		if audio_path:
-			contents.append({
-				"tyoe": "audio",
-				"audio": audio_path
-			})
-		return {
-				"role": "user",
-				"content": contents
 		}
 
+	async def chat_completion(self, messages):
+		f = awaitify(self.reference)
+		return await f(messages)
+

llmengine/base_embedding.py

@@ -1,4 +1,5 @@
 import torch
+import numpy as np
 
 model_pathMap = {
 }
@@ -20,14 +21,19 @@ class BaseEmbedding:
 			device = torch.device("mps")
 			self.model = self.model.to(device)
 
-	def embeddings(self, input):
+	def encode(self, input):
 		es = self.model.encode(input)
+
+	def embeddings(self, input):
+		es = self.encode(input)
 		data = []
 		for i, e in enumerate(es):
+			if isinstance(e, np.ndarray):
+				r = e.tolist()
 			d = {
 				"object": "embedding",
 				"index": i,
-				"embedding": e.tolist()
+				"embedding": e
 			}
 			data.append(d)
 		return {

llmengine/embedding.py

@@ -3,6 +3,7 @@ import os
 import sys
 import argparse
 from llmengine.qwen3embedding import *
+from llmengine.bgeembedding import *
 from llmengine.base_embedding import get_llm_class
 
 from appPublic.registerfunction import RegisterFunction

llmengine/gemma3_it.py

@@ -9,9 +9,9 @@ from transformers import AutoProcessor, Gemma3ForConditionalGeneration, TextIter
 from PIL import Image
 import requests
 import torch
-from llmengine.base_chat_llm import MMChatLLM, llm_register
+from llmengine.base_chat_llm import BaseChatLLM, llm_register
 
-class Gemma3LLM(MMChatLLM):
+class Gemma3LLM(BaseChatLLM):
 	def __init__(self, model_id):
 		self.model = Gemma3ForConditionalGeneration.from_pretrained(
 			model_id, device_map="auto"
@@ -24,6 +24,10 @@ class Gemma3LLM(MMChatLLM):
 llm_register("gemma-3", Gemma3LLM)
 
 if __name__ == '__main__':
+	def get_stream_text(chunk):
+		chunk = chunk[6:]
+		d = json.loads(chunk)
+		return d['choices'][0]['delta']['content']
 	gemma3 = Gemma3LLM('/share/models/google/gemma-3-4b-it')
 	session = {}
 	while True:
@@ -35,7 +39,7 @@ if __name__ == '__main__':
 			print('input image path')
 			imgpath=input()
 			for d in gemma3.stream_generate(session, p, image_path=imgpath):
-				if not d['done']:
+				if not d['DONE']:
 					print(d['text'], end='', flush=True)
 				else:
 					x = {k:v for k,v in d.items() if k != 'text'}

llmengine/medgemma3_it.py

@@ -4,11 +4,11 @@ from transformers import AutoProcessor, AutoModelForImageTextToText
 from PIL import Image
 import requests
 import torch
-from llmengine.base_chat_llm import MMChatLLM, llm_register
+from llmengine.base_chat_llm import BaseChatLLM, llm_register
 
 model_id = "google/medgemma-4b-it"
 
-class MedgemmaLLM(MMChatLLM):
+class MedgemmaLLM(BaseChatLLM):
 	def __init__(self, model_id):
 		self.model = AutoModelForImageTextToText.from_pretrained(
 			model_id,

llmengine/qwen3.py

@@ -4,12 +4,12 @@
 from appPublic.worker import awaitify
 from appPublic.log import debug
 from ahserver.serverenv import get_serverenv
-from transformers import AutoModelForCausalLM, AutoTokenizer
 from PIL import Image
 import torch
-from llmengine.base_chat_llm import BaseChatLLM, T2TChatLLM, llm_register
+from llmengine.base_chat_llm import BaseChatLLM, llm_register
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
-class Qwen3LLM(T2TChatLLM):
+class Qwen3LLM(BaseChatLLM):
 	def __init__(self, model_id):
 		self.tokenizer = AutoTokenizer.from_pretrained(model_id)
 		self.model = AutoModelForCausalLM.from_pretrained(
@@ -17,9 +17,6 @@ class Qwen3LLM(T2TChatLLM):
 			torch_dtype="auto",
 			device_map="auto"
 		)
-		if torch.backends.mps.is_available():
-			device = torch.device("mps")
-			self.model = self.model.to(device)
 		self.model_id = model_id
 
 	def build_kwargs(self, inputs, streamer):
@@ -33,7 +30,7 @@ class Qwen3LLM(T2TChatLLM):
 		return generate_kwargs
 
 	def _messages2inputs(self, messages):
-		debug(f'{messages=}')
+		debug(f'-----------{messages=}-----------')
 		text = self.tokenizer.apply_chat_template(
 			messages, 
 			tokenize=False,
@@ -43,26 +40,3 @@ class Qwen3LLM(T2TChatLLM):
 		return self.tokenizer([text], return_tensors="pt").to(self.model.device)
 
 llm_register("Qwen/Qwen3", Qwen3LLM)
-
-if __name__ == '__main__':
-	import sys
-	model_path = sys.argv[1]
-	q3 = Qwen3LLM(model_path)
-	session = {}
-	while True:
-		print('input prompt')
-		p = input()
-		if p:
-			if p == 'q':
-				break;
-			for d in q3.stream_generate(session, p):
-				print(d)
-				"""
-				if not d['done']:
-					print(d['text'], end='', flush=True)
-				else:
-					x = {k:v for k,v in d.items() if k != 'text'}
-					print(f'\n{x}\n')
-				"""
-					
-

llmengine/server.py

@@ -21,23 +21,17 @@ def init():
 	rf.register('chat_completions', chat_completions)
 
 async def chat_completions(request, params_kw, *params, **kw):
+	se = ServerEnv()
+	engine = se.engine
 	async def gor():
-		se = ServerEnv()
+		async for d in engine.chat_completion_stream(params_kw.messages):
-		engine = se.chat_engine
-		session = await get_session(request)
-		kwargs = {
-		}
-		if params_kw.image_path:
-			kwargs['image_path'] = fs.reapPath(params_kw.image_path)
-		if params_kw.video_path:
-			kwargs['video_path'] = fs.reapPath(params_kw.video_path)
-		if params_kw.audio_path:
-			kwargs['audio_path'] = fs.reapPath(params_kw.audio_path)
-		async for d in engine.async_stream_generate(session, params_kw.prompt, **kwargs):
 			debug(f'{d=}')
 			yield d
 
-	return await stream_response(request, gor)
+	if params_kw.stream:
+		return await stream_response(request, gor)
+	else:
+		return await engine.chat_completion(params_kw.messages)
 
 def main():
 	parser = argparse.ArgumentParser(prog="Sage")

pyproject.toml

@@ -1,23 +1,3 @@
-[project]
-name="llmengine"
-version = "0.0.1"
-description = "Your project description"
-authors = [{ name = "yu moqing", email = "yumoqing@gmail.com" }]
-readme = "README.md"
-requires-python = ">=3.8"
-license = {text = "MIT"}
-dependencies = [
-	"torch",
-	"transformers",
-	"sentence-transformers>=2.7.0",
-	# "flash_attention_2",
-	"mistral-common",
-	"accelerate"
-]
-
-[project.optional-dependencies]
-dev = ["pytest", "black", "mypy"]
-
 [build-system]
 requires = ["setuptools>=61", "wheel"]
 build-backend = "setuptools.build_meta"