Furhter changes for remote execution

Author: AlexejPenner

llm-complete-guide/README.md

@@ -43,16 +43,16 @@ environment and install the dependencies using the following command:
 pip install -r requirements.txt
 ```
 
-Depending on your setup you may run into some issues when running the `pip install` command with the
-`flash_attn` package. In that case running `FLASH_ATTENTION_SKIP_CUDA_BUILD=TRUE pip install flash-attn --no-build-isolation` could help you.
+Depending on your hardware you may run into some issues when running the `pip install` command with the
+`flash_attn` package. In that case running `FLASH_ATTENTION_SKIP_CUDA_BUILD=TRUE pip install flash-attn --no-build-isolation` 
+could help you.
 
 In order to use the default LLM for this query, you'll need an account and an
-API key from OpenAI specified as another environment variable:
-
-zenml secret create supabase_postgres_db --password="YOUR_PASSWORD" --user="YOU_USER" --host="YOUR_HOST" --port="YOUR_PORT"
+API key from OpenAI specified as a ZenML secret:
 
 ```shell
 zenml secret create openai --api_key=<your-openai-api-key>
+export ZENML_OPENAI_SECRET_NAME=openai
 ```
 
 ### Setting up Supabase
@@ -69,14 +69,15 @@ isn't stored in plaintext. You can do this by running the following command:
 
 ```shell
 zenml secret create supabase_postgres_db --password="YOUR_PASSWORD" --user="YOU_USER" --host="YOUR_HOST" --port="YOUR_PORT"
+export ZENML_SUPABASE_SECRET_NAME=supabase_postgres_db
 ```
 
 You can get the user, host and port for this database instance by getting the connection
 string from the Supabase dashboard.
 
 ![](.assets/supabase-connection-string.png)
 
-Alternatively you can use a different database as the backend. 
+In case supabase is not an option for you, you can use a different database as the backend. 
 
 ### Running the RAG pipeline
 

llm-complete-guide/configs/embeddings.yaml

@@ -1,3 +1,5 @@
+enable_cache: False
+
 # environment configuration
 settings:
   docker:
@@ -14,7 +16,7 @@ settings:
       - pandas
       - numpy
       - sentence-transformers>=3
-      - transformers
+      - transformers[torch]
       - litellm
       - ollama
       - tiktoken

llm-complete-guide/configs/rag_local_dev.yaml

@@ -12,19 +12,12 @@ settings:
       - tiktoken
       - ratelimit
     environment:
-      ZENML_SUPABASE_SECRET_NAME: alexej_supabase_postgres_db
-      ZENML_OPENAI_SECRET_NAME: alexej_openai
+      ZENML_SUPABASE_SECRET_NAME: supabase_postgres_db
+      ZENML_OPENAI_SECRET_NAME: openai
       ZENML_ENABLE_RICH_TRACEBACK: FALSE
       ZENML_LOGGING_VERBOSITY: INFO
 
 steps:
   url_scraper:
     parameters:
       docs_url: https://docs.zenml.io/stack-components/orchestrators
-#  generate_embeddings:
-#    step_operator: "terraform-gcp-6c0fd52233ca"
-#    settings:
-#      step_operator.vertex:
-#        accelerator_type: "NVIDIA_TESLA_P100"
-#        accelerator_count: 1
-#        machine_type: "n1-standard-8"

llm-complete-guide/configs/synthetic.yaml

@@ -1,3 +1,5 @@
+# enable_cache: False
+
 # environment configuration
 settings:
   docker:
@@ -24,3 +26,7 @@ settings:
       - rerankers[flashrank]
       - datasets
       - torch
+      - distilabel
+    environment:
+      ZENML_SUPABASE_SECRET_NAME: supabase_postgres_db
+      ZENML_OPENAI_SECRET_NAME: openai

llm-complete-guide/steps/eval_e2e.py

@@ -25,6 +25,8 @@
 from utils.llm_utils import process_input_with_retrieval
 from zenml import step
 
+from utils.openai_utils import get_openai_api_key
+
 logging.getLogger().setLevel(logging.WARNING)
 
 logger = logging.getLogger(__name__)
@@ -208,7 +210,8 @@ def llm_judged_test_e2e(
     """
     logging.debug("Prompt created.")
     response = completion(
-        model="gpt-4-turbo", messages=[{"content": prompt, "role": "user"}]
+        model="gpt-4-turbo", messages=[{"content": prompt, "role": "user"}], api_key=get_openai_api_key()
+
     )
 
     json_output = response["choices"][0]["message"]["content"].strip()

llm-complete-guide/steps/generate_questions.py

@@ -23,6 +23,8 @@
 from zenml import log_artifact_metadata, step
 from zenml.logger import get_logger
 
+from utils.openai_utils import get_openai_api_key
+
 logger = get_logger(__name__)
 LOCAL_MODEL = "ollama/mixtral"
 
@@ -58,6 +60,7 @@ def generate_question(
                     }
                 ],
                 api_base="http://localhost:11434" if local else None,
+                api_key=get_openai_api_key()
             )
             return response.choices[0].message.content
         except (Timeout, APIConnectionError) as e:

llm-complete-guide/steps/push_to_hf.py

@@ -11,6 +11,7 @@
 #  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
 #  or implied. See the License for the specific language governing
 #  permissions and limitations under the License.
+from zenml.client import Client
 
 from constants import DATASET_NAME_DISTILABEL
 from datasets import Dataset, DatasetDict
@@ -19,7 +20,12 @@
 
 @step
 def push_to_hf(train_dataset: Dataset, test_dataset: Dataset):
+    zenml_client = Client()
+
     combined_dataset = DatasetDict(
         {"train": train_dataset, "test": test_dataset}
     )
-    combined_dataset.push_to_hub(DATASET_NAME_DISTILABEL)
+    combined_dataset.push_to_hub(
+        DATASET_NAME_DISTILABEL,
+        token=zenml_client.get_secret("argilla_secrets").secret_values["hf_token"]
+    )

llm-complete-guide/steps/synthetic_data.py

@@ -26,6 +26,8 @@
 from zenml import ArtifactConfig, step
 from zenml.client import Client
 
+from utils.openai_utils import get_openai_api_key
+
 logger = logging.getLogger(__name__)
 
 LOCAL_MODEL = "ollama/mixtral"
@@ -50,6 +52,7 @@ def generate_question(chunk: str, local: bool = False) -> str:
             }
         ],
         api_base="http://localhost:11434" if local else None,
+        api_key=get_openai_api_key()
     )
     return response.choices[0].message.content
 

llm-complete-guide/utils/llm_utils.py

@@ -23,6 +23,8 @@
 
 from zenml.cli import secret
 
+from utils.openai_utils import get_openai_api_key
+
 # Configure logging levels for specific modules
 logging.getLogger("pytorch").setLevel(logging.CRITICAL)
 logging.getLogger("sentence-transformers").setLevel(logging.CRITICAL)
@@ -34,7 +36,7 @@
 
 import os
 import re
-from typing import Dict, List, Tuple
+from typing import List, Tuple
 
 import litellm
 import numpy as np
@@ -377,6 +379,7 @@ def get_completion_from_messages(
         messages=messages,
         temperature=temperature,
         max_tokens=max_tokens,
+        api_key=get_openai_api_key()
     )
     return completion_response.choices[0].message.content