Spaces:
Sleeping
Sleeping
Zwea Htet
commited on
Commit
·
a550aaa
1
Parent(s):
e594eb9
update llama custom
Browse files- models/llamaCustom.py +19 -33
models/llamaCustom.py
CHANGED
@@ -21,8 +21,8 @@ from llama_index import (
|
|
21 |
load_index_from_storage,
|
22 |
)
|
23 |
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
24 |
-
|
25 |
-
|
26 |
|
27 |
load_dotenv()
|
28 |
# openai.api_key = os.getenv("OPENAI_API_KEY")
|
@@ -104,36 +104,10 @@ class OurLLM(CustomLLM):
|
|
104 |
def stream_complete(self, prompt: str, **kwargs: Any) -> CompletionResponse:
|
105 |
raise NotImplementedError()
|
106 |
|
107 |
-
# def _call(self, prompt: str, stop: Optional[List[str]] = None) -> str:
|
108 |
-
# prompt_length = len(prompt)
|
109 |
-
# response = self.pipeline(prompt, max_new_tokens=525)[0]["generated_text"]
|
110 |
-
|
111 |
-
# # only return newly generated tokens
|
112 |
-
# return response[prompt_length:]
|
113 |
-
|
114 |
-
# @property
|
115 |
-
# def _identifying_params(self) -> Mapping[str, Any]:
|
116 |
-
# return {"name_of_model": self.model_name}
|
117 |
-
|
118 |
-
# @property
|
119 |
-
# def _llm_type(self) -> str:
|
120 |
-
# return "custom"
|
121 |
-
|
122 |
class LlamaCustom:
|
123 |
-
# define llm
|
124 |
-
# llm_predictor = LLMPredictor(llm=OurLLM())
|
125 |
-
# service_context = ServiceContext.from_defaults(
|
126 |
-
# llm_predictor=llm_predictor, prompt_helper=prompt_helper
|
127 |
-
# )
|
128 |
def __init__(self, model_name: str) -> None:
|
129 |
-
pipe = load_model(mode_name=model_name)
|
130 |
-
llm = OurLLM(model_name=model_name, model_pipeline=pipe)
|
131 |
-
self.service_context = ServiceContext.from_defaults(
|
132 |
-
llm=llm, prompt_helper=prompt_helper
|
133 |
-
)
|
134 |
self.vector_index = self.initialize_index(model_name=model_name)
|
135 |
|
136 |
-
@st.cache_resource
|
137 |
def initialize_index(_self, model_name: str):
|
138 |
index_name = model_name.split("/")[-1]
|
139 |
|
@@ -151,11 +125,26 @@ class LlamaCustom:
|
|
151 |
# index = pickle.loads(file.readlines())
|
152 |
return index
|
153 |
else:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
154 |
# documents = prepare_data(r"./assets/regItems.json")
|
155 |
documents = SimpleDirectoryReader(input_dir="./assets/pdf").load_data()
|
156 |
|
157 |
index = GPTVectorStoreIndex.from_documents(
|
158 |
-
documents, service_context=
|
159 |
)
|
160 |
|
161 |
# local write access
|
@@ -168,10 +157,7 @@ class LlamaCustom:
|
|
168 |
|
169 |
def get_response(self, query_str):
|
170 |
print("query_str: ", query_str)
|
171 |
-
|
172 |
-
query_engine = self.vector_index.as_query_engine(
|
173 |
-
text_qa_template=text_qa_template, refine_template=refine_template
|
174 |
-
)
|
175 |
response = query_engine.query(query_str)
|
176 |
print("metadata: ", response.metadata)
|
177 |
return str(response)
|
|
|
21 |
load_index_from_storage,
|
22 |
)
|
23 |
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
24 |
+
from llama_index.llms import CompletionResponse, CustomLLM, LLMMetadata
|
25 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
26 |
|
27 |
load_dotenv()
|
28 |
# openai.api_key = os.getenv("OPENAI_API_KEY")
|
|
|
104 |
def stream_complete(self, prompt: str, **kwargs: Any) -> CompletionResponse:
|
105 |
raise NotImplementedError()
|
106 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
107 |
class LlamaCustom:
|
|
|
|
|
|
|
|
|
|
|
108 |
def __init__(self, model_name: str) -> None:
|
|
|
|
|
|
|
|
|
|
|
109 |
self.vector_index = self.initialize_index(model_name=model_name)
|
110 |
|
|
|
111 |
def initialize_index(_self, model_name: str):
|
112 |
index_name = model_name.split("/")[-1]
|
113 |
|
|
|
125 |
# index = pickle.loads(file.readlines())
|
126 |
return index
|
127 |
else:
|
128 |
+
prompt_helper = PromptHelper(
|
129 |
+
context_window=CONTEXT_WINDOW,
|
130 |
+
num_output=NUM_OUTPUT,
|
131 |
+
chunk_overlap_ratio=CHUNK_OVERLAP_RATION,
|
132 |
+
)
|
133 |
+
|
134 |
+
# define llm
|
135 |
+
pipe = load_model(mode_name=model_name)
|
136 |
+
llm = OurLLM(model_name=model_name, model_pipeline=pipe)
|
137 |
+
|
138 |
+
llm_predictor = LLMPredictor(llm=llm)
|
139 |
+
service_context = ServiceContext.from_defaults(
|
140 |
+
llm_predictor=llm_predictor, prompt_helper=prompt_helper
|
141 |
+
)
|
142 |
+
|
143 |
# documents = prepare_data(r"./assets/regItems.json")
|
144 |
documents = SimpleDirectoryReader(input_dir="./assets/pdf").load_data()
|
145 |
|
146 |
index = GPTVectorStoreIndex.from_documents(
|
147 |
+
documents, service_context=service_context
|
148 |
)
|
149 |
|
150 |
# local write access
|
|
|
157 |
|
158 |
def get_response(self, query_str):
|
159 |
print("query_str: ", query_str)
|
160 |
+
query_engine = self.vector_index.as_query_engine()
|
|
|
|
|
|
|
161 |
response = query_engine.query(query_str)
|
162 |
print("metadata: ", response.metadata)
|
163 |
return str(response)
|