update

Config.py

@@ -26,11 +26,9 @@ MODELS = [
 
 RANDOM_SEED = 42
 
-
-
-
-# DEFAULT_SUMMARIZER = "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO"
-DEFAULT_SUMMARIZER = 'mistralai/Mistral-7B-Instruct-v0.2'
+DEFAULT_SUMMARIZER = "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO"
+# DEFAULT_SUMMARIZER = 'NousResearch/Nous-Hermes-2-Mistral-7B-DPO'
+# DEFAULT_SUMMARIZER = 'mistralai/Mistral-7B-Instruct-v0.2'
 DEFAULT_DATASET = "mmlu"
 DEFAULT_TOPIC = random.choice(TOPICS[DEFAULT_DATASET])
 

models.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import json
 import random
 import re
+import os
 from abc import ABC, abstractmethod
 from typing import List, Dict, Union, Optional
 
@@ -10,12 +11,15 @@ from huggingface_hub import InferenceClient
 from tenacity import retry, stop_after_attempt, wait_random_exponential
 from transformers import AutoTokenizer
 
+# from config import *
+
 ROLE_SYSTEM = 'system'
 ROLE_USER = 'user'
 ROLE_ASSISTANT = 'assistant'
 
 SUPPORTED_MISTRAL_MODELS = ['mistralai/Mixtral-8x7B-Instruct-v0.1', 'mistralai/Mistral-7B-Instruct-v0.2']
-SUPPORTED_NOUS_MODELS = ['NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO']
+SUPPORTED_NOUS_MODELS = ['NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO',
+                         'NousResearch/Nous-Hermes-2-Mistral-7B-DPO']
 SUPPORTED_LLAMA_MODELS = ['meta-llama/Llama-2-70b-chat-hf',
                           'meta-llama/Llama-2-13b-chat-hf',
                           'meta-llama/Llama-2-7b-chat-hf']
@@ -93,7 +97,8 @@ class HFAPIModel(Model):
 
     @retry(stop=stop_after_attempt(5), wait=wait_random_exponential(max=10), reraise=True)  # retry if exception
     def get_response(self, temperature: float, use_json: bool, timeout: float, cache: bool) -> str:
-        client = InferenceClient(model=self.name, timeout=timeout)
+        # hf_api_token =
+        client = InferenceClient(model=self.name, token=os.getenv('HF_API_TOKEN'), timeout=timeout)
         # client = InferenceClient(model=self.name, token=random.choice(HF_API_TOKENS), timeout=timeout)
         if not cache:
             client.headers["x-use-cache"] = "0"
@@ -156,7 +161,7 @@ class MistralModel(HFAPIModel):
 class NousHermesModel(HFAPIModel):
 
     def __init__(self, system_prompt: str, model_name: str = 'NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO') -> None:
-        assert model_name in ['NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO'], 'Model not supported'
+        assert model_name in SUPPORTED_NOUS_MODELS, 'Model not supported'
         super().__init__(model_name, system_prompt)
 
     def format_messages(self) -> str:
@@ -200,4 +205,4 @@ class LlamaModel(HFAPIModel):
                 r += f'{content}</s>'
             else:
                 raise ValueError
-        return r
+        return r