Spaces:
Runtime error
Runtime error
Commit
·
b299ae3
1
Parent(s):
8091023
added files
Browse files- pages/🦙_Alpaca.py +36 -6
pages/🦙_Alpaca.py
CHANGED
@@ -8,10 +8,11 @@
|
|
8 |
#### Loading needed Python libraries
|
9 |
import streamlit as st
|
10 |
#from llamacpypy import Llama
|
11 |
-
import llamacpp
|
12 |
-
from llama_cpp import Llama
|
13 |
import os
|
14 |
import subprocess
|
|
|
15 |
|
16 |
|
17 |
|
@@ -134,7 +135,7 @@ def llama(
|
|
134 |
|
135 |
|
136 |
|
137 |
-
### Python Wrapper (functions above
|
138 |
#text = []
|
139 |
#for token in llama(prompt = 'What is your purpose?', repeat_penalty = 1.05, skip_prompt = False, interactive = False):
|
140 |
# print(token, end = '', flush = True)
|
@@ -161,11 +162,40 @@ def llama(
|
|
161 |
#model = llamacpp.PyLLAMA(model_path, params)
|
162 |
#text = model.predict("Hello, I'm a llama.", 10)
|
163 |
#st.write(text)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
164 |
|
165 |
|
166 |
|
167 |
### Llama cpp
|
168 |
-
llm = Llama(model_path="models/7B/ggml-model-q4_0.bin")
|
169 |
-
output = llm("Q: Name the planets in the solar system? A: ", max_tokens = 32, stop = ["Q:", "\n"], echo = True)
|
170 |
-
st.write(output)
|
171 |
|
|
|
8 |
#### Loading needed Python libraries
|
9 |
import streamlit as st
|
10 |
#from llamacpypy import Llama
|
11 |
+
#import llamacpp
|
12 |
+
#from llama_cpp import Llama
|
13 |
import os
|
14 |
import subprocess
|
15 |
+
import sys
|
16 |
|
17 |
|
18 |
|
|
|
135 |
|
136 |
|
137 |
|
138 |
+
### Python Wrapper (functions above)
|
139 |
#text = []
|
140 |
#for token in llama(prompt = 'What is your purpose?', repeat_penalty = 1.05, skip_prompt = False, interactive = False):
|
141 |
# print(token, end = '', flush = True)
|
|
|
162 |
#model = llamacpp.PyLLAMA(model_path, params)
|
163 |
#text = model.predict("Hello, I'm a llama.", 10)
|
164 |
#st.write(text)
|
165 |
+
#params = llamacpp.gpt_params('./models/7B/ggml-model-q4_0.bin', # model,
|
166 |
+
# 512, # ctx_size
|
167 |
+
# 100, # n_predict
|
168 |
+
# 40, # top_k
|
169 |
+
# 0.95, # top_p
|
170 |
+
# 0.85, # temp
|
171 |
+
# 1.30, # repeat_penalty
|
172 |
+
# -1, # seed
|
173 |
+
# 8, # threads
|
174 |
+
# 64, # repeat_last_n
|
175 |
+
# 8, # batch_size
|
176 |
+
#)
|
177 |
+
#model = llamacpp.PyLLAMA(params)
|
178 |
+
#model.add_bos() # Adds "beginning of string" token
|
179 |
+
#model.update_input("A llama is a")
|
180 |
+
#model.print_startup_stats()
|
181 |
+
#model.prepare_context()
|
182 |
+
|
183 |
+
#model.ingest_all_pending_input(True)
|
184 |
+
#while not model.is_finished():
|
185 |
+
# text, is_finished = model.infer_text()
|
186 |
+
# print(text, end="")
|
187 |
+
|
188 |
+
# if is_finished:
|
189 |
+
# break
|
190 |
+
|
191 |
+
# Flush stdout
|
192 |
+
#sys.stdout.flush()
|
193 |
+
#model.print_end_stats()
|
194 |
|
195 |
|
196 |
|
197 |
### Llama cpp
|
198 |
+
#llm = Llama(model_path="models/7B/ggml-model-q4_0.bin")
|
199 |
+
#output = llm("Q: Name the planets in the solar system? A: ", max_tokens = 32, stop = ["Q:", "\n"], echo = True)
|
200 |
+
#st.write(output)
|
201 |
|