Spaces:
Running
Running
Update src/models.py
Browse files- src/models.py +17 -10
src/models.py
CHANGED
|
@@ -1,10 +1,8 @@
|
|
| 1 |
-
from openai import OpenAI
|
| 2 |
-
import anthropic
|
| 3 |
import os
|
|
|
|
|
|
|
| 4 |
from dotenv import load_dotenv
|
| 5 |
-
from huggingface_hub import HfApi;
|
| 6 |
|
| 7 |
-
from huggingface_hub import InferenceClient
|
| 8 |
# Load environment variables from .env file
|
| 9 |
load_dotenv(override=True)
|
| 10 |
|
|
@@ -19,17 +17,25 @@ if not openai_api_key:
|
|
| 19 |
if not anthropic_api_key:
|
| 20 |
print("❌ Anthropic API Key is missing!")
|
| 21 |
|
| 22 |
-
# Initialize disguised API clients
|
| 23 |
-
openai = InferenceClient(token=openai_api_key)
|
| 24 |
-
claude = InferenceClient(token=anthropic_api_key)
|
| 25 |
-
|
| 26 |
# Model names
|
| 27 |
OPENAI_MODEL = "mistralai/Mistral-7B-Instruct-v0.3"
|
| 28 |
CLAUDE_MODEL = "mistralai/Mistral-7B-Instruct-v0.3"
|
| 29 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
def get_gpt_completion(prompt, system_message):
|
| 31 |
try:
|
| 32 |
-
|
|
|
|
| 33 |
return response
|
| 34 |
except Exception as e:
|
| 35 |
print(f"GPT error: {e}")
|
|
@@ -37,7 +43,8 @@ def get_gpt_completion(prompt, system_message):
|
|
| 37 |
|
| 38 |
def get_claude_completion(prompt, system_message):
|
| 39 |
try:
|
| 40 |
-
|
|
|
|
| 41 |
return response
|
| 42 |
except Exception as e:
|
| 43 |
print(f"Claude error: {e}")
|
|
|
|
|
|
|
|
|
|
| 1 |
import os
|
| 2 |
+
import requests
|
| 3 |
+
import json
|
| 4 |
from dotenv import load_dotenv
|
|
|
|
| 5 |
|
|
|
|
| 6 |
# Load environment variables from .env file
|
| 7 |
load_dotenv(override=True)
|
| 8 |
|
|
|
|
| 17 |
if not anthropic_api_key:
|
| 18 |
print("❌ Anthropic API Key is missing!")
|
| 19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
# Model names
|
| 21 |
OPENAI_MODEL = "mistralai/Mistral-7B-Instruct-v0.3"
|
| 22 |
CLAUDE_MODEL = "mistralai/Mistral-7B-Instruct-v0.3"
|
| 23 |
|
| 24 |
+
# Define a helper function to perform the direct API call
|
| 25 |
+
def hf_inference_request(token, model, prompt, max_new_tokens=200):
|
| 26 |
+
url = f"https://api-inference.huggingface.co/models/{model}"
|
| 27 |
+
headers = {"Authorization": f"Bearer {token}"}
|
| 28 |
+
payload = {"inputs": prompt, "parameters": {"max_new_tokens": max_new_tokens}}
|
| 29 |
+
|
| 30 |
+
response = requests.post(url, headers=headers, json=payload)
|
| 31 |
+
# Raise an HTTPError if the API response was unsuccessful
|
| 32 |
+
response.raise_for_status()
|
| 33 |
+
return response.json()
|
| 34 |
+
|
| 35 |
def get_gpt_completion(prompt, system_message):
|
| 36 |
try:
|
| 37 |
+
full_prompt = f"{system_message}\n{prompt}"
|
| 38 |
+
response = hf_inference_request(openai_api_key, OPENAI_MODEL, full_prompt, max_new_tokens=200)
|
| 39 |
return response
|
| 40 |
except Exception as e:
|
| 41 |
print(f"GPT error: {e}")
|
|
|
|
| 43 |
|
| 44 |
def get_claude_completion(prompt, system_message):
|
| 45 |
try:
|
| 46 |
+
full_prompt = f"{system_message}\n{prompt}"
|
| 47 |
+
response = hf_inference_request(anthropic_api_key, CLAUDE_MODEL, full_prompt, max_new_tokens=200)
|
| 48 |
return response
|
| 49 |
except Exception as e:
|
| 50 |
print(f"Claude error: {e}")
|