Spaces:
Sleeping
Sleeping
gabrielaltay
commited on
Commit
•
eae1702
1
Parent(s):
fe9941d
together
Browse files- app.py +13 -5
- requirements.txt +17 -2
app.py
CHANGED
@@ -28,7 +28,7 @@ from langchain_community.embeddings import HuggingFaceBgeEmbeddings
|
|
28 |
from langchain_community.vectorstores.utils import DistanceStrategy
|
29 |
from langchain_openai import ChatOpenAI
|
30 |
from langchain_anthropic import ChatAnthropic
|
31 |
-
from
|
32 |
from langchain_pinecone import PineconeVectorStore
|
33 |
from pinecone import Pinecone
|
34 |
import streamlit as st
|
@@ -65,11 +65,19 @@ ANTHROPIC_CHAT_MODELS = [
|
|
65 |
"claude-3-5-sonnet-20240620",
|
66 |
"claude-3-opus-20240229",
|
67 |
]
|
68 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
69 |
|
70 |
PROVIDER_MODELS = {
|
71 |
"OpenAI": OPENAI_CHAT_MODELS,
|
72 |
"Anthropic": ANTHROPIC_CHAT_MODELS,
|
|
|
73 |
}
|
74 |
|
75 |
|
@@ -503,14 +511,14 @@ elif SS["model_name"] in ANTHROPIC_CHAT_MODELS:
|
|
503 |
top_p=SS["top_p"],
|
504 |
max_tokens_to_sample=SS["max_output_tokens"],
|
505 |
)
|
506 |
-
elif SS["model_name"] in
|
507 |
-
llm =
|
508 |
model=SS["model_name"],
|
509 |
temperature=SS["temperature"],
|
510 |
max_tokens=SS["max_output_tokens"],
|
511 |
top_p=SS["top_p"],
|
512 |
seed=SEED,
|
513 |
-
|
514 |
)
|
515 |
else:
|
516 |
raise ValueError()
|
|
|
28 |
from langchain_community.vectorstores.utils import DistanceStrategy
|
29 |
from langchain_openai import ChatOpenAI
|
30 |
from langchain_anthropic import ChatAnthropic
|
31 |
+
from langchain_together import ChatTogether
|
32 |
from langchain_pinecone import PineconeVectorStore
|
33 |
from pinecone import Pinecone
|
34 |
import streamlit as st
|
|
|
65 |
"claude-3-5-sonnet-20240620",
|
66 |
"claude-3-opus-20240229",
|
67 |
]
|
68 |
+
TOGETHER_CHAT_MODELS = [
|
69 |
+
"meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo",
|
70 |
+
"meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
|
71 |
+
"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo",
|
72 |
+
]
|
73 |
+
|
74 |
+
|
75 |
+
CHAT_MODELS = OPENAI_CHAT_MODELS + ANTHROPIC_CHAT_MODELS + TOGETHER_CHAT_MODELS
|
76 |
|
77 |
PROVIDER_MODELS = {
|
78 |
"OpenAI": OPENAI_CHAT_MODELS,
|
79 |
"Anthropic": ANTHROPIC_CHAT_MODELS,
|
80 |
+
"Together": TOGETHER_CHAT_MODELS,
|
81 |
}
|
82 |
|
83 |
|
|
|
511 |
top_p=SS["top_p"],
|
512 |
max_tokens_to_sample=SS["max_output_tokens"],
|
513 |
)
|
514 |
+
elif SS["model_name"] in TOGETHER_CHAT_MODELS:
|
515 |
+
llm = ChatTogether(
|
516 |
model=SS["model_name"],
|
517 |
temperature=SS["temperature"],
|
518 |
max_tokens=SS["max_output_tokens"],
|
519 |
top_p=SS["top_p"],
|
520 |
seed=SEED,
|
521 |
+
api_key=st.secrets["together_api_key"],
|
522 |
)
|
523 |
else:
|
524 |
raise ValueError()
|
requirements.txt
CHANGED
@@ -28,6 +28,7 @@ frozenlist==1.4.1
|
|
28 |
fsspec==2024.2.0
|
29 |
gitdb==4.0.11
|
30 |
GitPython==3.1.42
|
|
|
31 |
h11==0.14.0
|
32 |
httpcore==1.0.4
|
33 |
httpx==0.27.0
|
@@ -46,11 +47,12 @@ jsonschema-specifications==2023.12.1
|
|
46 |
langchain==0.2.5
|
47 |
langchain-anthropic==0.1.15
|
48 |
langchain-community==0.2.5
|
49 |
-
langchain-core==0.2.
|
50 |
langchain-nvidia-ai-endpoints==0.1.2
|
51 |
-
langchain-openai==0.1.
|
52 |
langchain-pinecone==0.1.1
|
53 |
langchain-text-splitters==0.2.1
|
|
|
54 |
langchainhub==0.1.15
|
55 |
langsmith==0.1.77
|
56 |
markdown-it-py==3.0.0
|
@@ -63,6 +65,18 @@ multidict==6.0.5
|
|
63 |
mypy-extensions==1.0.0
|
64 |
networkx==3.2.1
|
65 |
numpy==1.26.4
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
openai==1.34.0
|
67 |
orjson==3.10.0
|
68 |
packaging==23.2
|
@@ -115,6 +129,7 @@ tornado==6.4
|
|
115 |
tqdm==4.66.2
|
116 |
traitlets==5.14.1
|
117 |
transformers==4.38.1
|
|
|
118 |
types-requests==2.31.0.20240311
|
119 |
typing-inspect==0.9.0
|
120 |
typing_extensions==4.9.0
|
|
|
28 |
fsspec==2024.2.0
|
29 |
gitdb==4.0.11
|
30 |
GitPython==3.1.42
|
31 |
+
greenlet==3.0.3
|
32 |
h11==0.14.0
|
33 |
httpcore==1.0.4
|
34 |
httpx==0.27.0
|
|
|
47 |
langchain==0.2.5
|
48 |
langchain-anthropic==0.1.15
|
49 |
langchain-community==0.2.5
|
50 |
+
langchain-core==0.2.28
|
51 |
langchain-nvidia-ai-endpoints==0.1.2
|
52 |
+
langchain-openai==0.1.20
|
53 |
langchain-pinecone==0.1.1
|
54 |
langchain-text-splitters==0.2.1
|
55 |
+
langchain-together==0.1.5
|
56 |
langchainhub==0.1.15
|
57 |
langsmith==0.1.77
|
58 |
markdown-it-py==3.0.0
|
|
|
65 |
mypy-extensions==1.0.0
|
66 |
networkx==3.2.1
|
67 |
numpy==1.26.4
|
68 |
+
nvidia-cublas-cu12==12.1.3.1
|
69 |
+
nvidia-cuda-cupti-cu12==12.1.105
|
70 |
+
nvidia-cuda-nvrtc-cu12==12.1.105
|
71 |
+
nvidia-cuda-runtime-cu12==12.1.105
|
72 |
+
nvidia-cudnn-cu12==8.9.2.26
|
73 |
+
nvidia-cufft-cu12==11.0.2.54
|
74 |
+
nvidia-curand-cu12==10.3.2.106
|
75 |
+
nvidia-cusolver-cu12==11.4.5.107
|
76 |
+
nvidia-cusparse-cu12==12.1.0.106
|
77 |
+
nvidia-nccl-cu12==2.19.3
|
78 |
+
nvidia-nvjitlink-cu12==12.6.20
|
79 |
+
nvidia-nvtx-cu12==12.1.105
|
80 |
openai==1.34.0
|
81 |
orjson==3.10.0
|
82 |
packaging==23.2
|
|
|
129 |
tqdm==4.66.2
|
130 |
traitlets==5.14.1
|
131 |
transformers==4.38.1
|
132 |
+
triton==2.2.0
|
133 |
types-requests==2.31.0.20240311
|
134 |
typing-inspect==0.9.0
|
135 |
typing_extensions==4.9.0
|