Azure OpenAI-utvärdering (förhandsversion)

Skapa utvärdering

1. Välj Azure OpenAI Evaluation (PREVIEW) i Azure AI Foundry-portalen. Om du vill se den här vyn som ett alternativ väljer du först en befintlig Azure OpenAI-resurs i en region som stöds.

2. Välj + Ny utvärdering

   

3. Välj hur du vill tillhandahålla testdata för utvärdering. Du kan importera lagrade chattavslutningar, skapa data med hjälp av tillhandahållna standardmallar eller ladda upp dina egna data. Nu ska vi gå igenom hur du laddar upp dina egna data.

   

4. Välj dina utvärderingsdata som ska vara i .jsonl format. Om du redan har en befintlig data kan du välja en eller ladda upp en ny data.

   

   När du laddar upp nya data visas de tre första raderna i filen som en förhandsgranskning till höger:

   

   Om du behöver en exempeltestfil kan du använda den här exempeltexten .jsonl . Det här exemplet innehåller meningar av olika delar av tekniskt innehåll, och vi kommer att utvärdera semantisk likhet mellan dessa meningar.

   {"input": [{"role": "system", "content": "Provide a clear and concise summary of the technical content, highlighting key concepts and their relationships. Focus on the main ideas and practical implications."}, {"role": "user", "content": "Tokenization is a key step in preprocessing for natural language processing, involving the division of text into smaller components called tokens. These can be words, subwords, or characters, depending on the method chosen. Word tokenization divides text at word boundaries, while subword techniques like Byte Pair Encoding (BPE) or WordPiece can manage unknown words by breaking them into subunits. Character tokenization splits text into individual characters, useful for multiple languages and misspellings. The tokenization method chosen greatly affects model performance and its capacity to handle various languages and vocabularies."}], "output": "Tokenization divides text into smaller units (tokens) for NLP applications, using word, subword (e.g., BPE), or character methods. Each has unique benefits, impacting model performance and language processing capabilities."}      
   {"input": [{"role": "system", "content": "Create a comprehensive yet concise summary that captures the essential technical details and significance of the topic."}, {"role": "user", "content": "Self-attention mechanisms are vital in modern transformer models, allowing them to evaluate the relevance of different parts of an input sequence when processing each element. In self-attention, each position in a sequence learns to focus on all other positions, forming a weighted sum of their values. Attention weights are derived using queries, keys, and values, with compatibility between a query and key determining attention to each value. This mechanism enables capturing long-range dependencies and relationships within a sequence, making it effective for tasks needing context understanding and inter-part relationships."}], "output": "Self-attention is crucial in transformers, enabling each sequence position to evaluate and collect data from all others. It uses queries, keys, and values for attention weights, effectively capturing long-range dependencies and contextual links."}      
   {"input": [{"role": "system", "content": "Synthesize the technical information into a brief, accurate summary that maintains the core concepts and their significance."}, {"role": "user", "content": "The transformer architecture significantly changed natural language processing by introducing a new method for sequence processing. Unlike recurrent neural networks, transformers process whole sequences at once using self-attention. The architecture consists of an encoder and decoder, each with multiple layers of multi-head attention, feed-forward networks, and normalization layers. The encoder processes the input sequence, and the decoder produces the output sequence. Positional encodings are added to retain sequence order information since the model lacks inherent sequence position awareness. This architecture underlies many leading language models like BERT and GPT."}], "output": "Transformers are a transformative neural architecture processing sequences with self-attention instead of recurrence. Their encoder-decoder design, with multi-head attention and position encodings, is fundamental to modern language models like BERT and GPT."}      
   {"input": [{"role": "system", "content": "Distill the technical content into a clear, concise summary that preserves the most important concepts and their practical applications."}, {"role": "user", "content": "Cosine similarity is a metric for measuring similarity between vectors by calculating the cosine of the angle between them. The outcome ranges from -1 to 1, with 1 indicating identical orientation, 0 orthogonality, and -1 opposite orientation. In machine learning and NLP, cosine similarity is useful for comparing document vectors, word embeddings, or any high-dimensional representations. It's often preferred over Euclidean distance because it's invariant to vector magnitude, focusing instead on orientation. This makes it particularly useful for comparing documents or embeddings of various lengths or scales."}], "output": "Cosine similarity evaluates vector similarity by computing their angle's cosine, with values from -1 to 1. It's widely used in ML and NLP for comparing document vectors and embeddings due to its magnitude-invariant nature."}      
   {"input": [{"role": "system", "content": "Provide a succinct summary that captures the key technical concepts and their relationships while maintaining clarity and accuracy."}, {"role": "user", "content": "Neural network activation functions are crucial for introducing non-linearity to deep learning models. Commonly used activation functions include ReLU (Rectified Linear Unit), which outputs the input directly if positive and zero otherwise, sigmoid, which compresses values between 0 and 1, and tanh, which normalizes values between -1 and 1. Each function has specific benefits and applications: ReLU avoids vanishing gradients and enables sparse activation, sigmoid is useful for binary classification, and tanh is often used in hidden layers for its zero-centered output. The activation function choice greatly affects model training dynamics and ultimate performance."}], "output": "Activation functions introduce non-linearity to neural networks, with ReLU, sigmoid, and tanh being the most common. Each serves distinct purposes: ReLU prevents vanishing gradients, sigmoid is suitable for binary classification, and tanh offers zero-centered output, all affecting model performance differently."}      
   {"input": [{"role": "system", "content": "Create a focused summary that emphasizes the main technical concepts and their practical implications while maintaining accuracy."}, {"role": "user", "content": "Batch normalization is a technique that stabilizes and accelerates deep neural network training by normalizing each layer's inputs. It normalizes activations using the mean and variance computed over the current data batch. This normalization reduces internal covariate shift, where the input distribution changes during training. Batch normalization includes learnable parameters (gamma and beta) allowing the network to reverse normalization if necessary. Benefits include enabling higher learning rates, lessening the importance of precise parameter initialization, and serving as a regularizer, potentially removing the need for dropout."}], "output": "Batch normalization stabilizes neural network training by normalizing layer inputs using batch statistics. It addresses internal covariate shift with learnable parameters, allowing higher learning rates and improved training dynamics, while offering regularization benefits."}      
   {"input": [{"role": "system", "content": "Summarize the technical content while preserving the essential concepts and their interconnections, ensuring clarity and precision."}, {"role": "user", "content": "Gradient descent optimization algorithms are foundational to training machine learning models. The basic version updates parameters in the steepest descent direction of the loss function. Variants like Stochastic Gradient Descent (SGD) use random samples to estimate gradients, enhancing computational efficiency. Advanced optimizers like Adam combine momentum (helping overcome local minima) and RMSprop (adapting learning rates for each parameter). These optimizers maintain running averages of gradients and their squares, enabling adaptive learning rates and more efficient training. The optimizer choice can significantly influence model convergence speed and final performance."}], "output": "Gradient descent algorithms optimize ML models by updating parameters based on loss gradients. Variants like SGD use sampling for efficiency, while advanced optimizers like Adam combine momentum with adaptive learning rates for improved training performance."}      
   {"input": [{"role": "system", "content": "Deliver a concise yet comprehensive summary that highlights the key technical aspects and their practical significance."}, {"role": "user", "content": "Cross-validation is a resampling method used to evaluate machine learning models on limited data. The most common form, k-fold cross-validation, involves dividing the dataset into k equal parts, training the model on k-1 folds while validating on the remaining fold, and rotating through all combinations. This provides a more robust model performance assessment than a single train-test split, helping detect overfitting and estimate model generalization to unseen data. Variations include stratified k-fold (maintaining class distribution), leave-one-out (k equals dataset size), and time series cross-validation for temporal data."}], "output": "Cross-validation evaluates ML models by training and testing on different data splits, typically using k-fold methodology. This approach offers better performance assessment than single splits, with variations for different data types and requirements."}      
   {"input": [{"role": "system", "content": "Generate a clear and focused summary that captures the essential technical details while maintaining their relationships and significance."}, {"role": "user", "content": "Transfer learning is a machine learning method where a model developed for one task is reused as the starting point for a model on a second task. This approach is powerful in deep learning, where pre-trained models on large datasets (like ImageNet for computer vision or BERT for NLP) are fine-tuned on specific downstream tasks. Transfer learning reduces the need for large amounts of task-specific training data and computational resources, as the model has already learned useful features from the source domain. Common strategies include feature extraction (freezing pre-trained layers) and fine-tuning (updating all or some pre-trained weights)."}], "output": "Transfer learning reuses models trained on one task for different tasks, particularly effective in deep learning. It leverages pre-trained models through feature extraction or fine-tuning, reducing data and computational needs for new tasks."}      
   {"input": [{"role": "system", "content": "Provide a precise and informative summary that distills the key technical concepts while maintaining their relationships and practical importance."}, {"role": "user", "content": "Ensemble methods combine multiple machine learning models to create a more robust and accurate predictor. Common techniques include bagging (training models on random data subsets), boosting (sequentially training models to correct earlier errors), and stacking (using a meta-model to combine base model predictions). Random Forests, a popular bagging method, create multiple decision trees using random feature subsets. Gradient Boosting builds trees sequentially, with each tree correcting the errors of previous ones. These methods often outperform single models by reducing overfitting and variance while capturing different data aspects."}], "output": "Ensemble methods enhance prediction accuracy by combining multiple models through techniques like bagging, boosting, and stacking. Popular implementations include Random Forests (using multiple trees with random features) and Gradient Boosting (sequential error correction), offering better performance than single models."}
   

5. Om du vill skapa nya svar med hjälp av indata från dina testdata kan du välja Generera nya svar. Detta matar in indatafälten från vår utvärderingsfil i enskilda frågor för en modell som du väljer för att generera utdata.

Du väljer valfri modell. Om du inte har någon modell kan du skapa en ny modellimplementering. Den valda modellen tar indata och genererar sina egna unika utdata, som i det här fallet lagras i en variabel med namnet {{sample.output_text}}. Vi använder sedan utdata senare som en del av våra testkriterier. Du kan också ange ett eget anpassat systemmeddelande och enskilda meddelandeexempel manuellt.

6. Om du vill skapa ett testvillkor väljer du Lägg till. För den exempelfil som vi angav kommer vi att utvärdera semantisk likhet. Välj Modellbedömare, som innehåller testvillkorsförinställningar för semantisk likhet.

   

   Välj Semantisk likhet överst. Rulla längst ned och i User avsnittet anger du {{item.output}} som Ground truthoch anger {{sample.output_text}} som Output. Detta tar de ursprungliga referensutdata från utvärderingsfilen .jsonl (den angivna exempelfilen) och jämför den med de utdata som genereras av den modell som du valde i föregående steg.

   

7. Välj Lägg till för att lägga till det här testvillkoren. Om du vill lägga till ytterligare testkriterier kan du lägga till dem i det här steget.

8. Du är redo att skapa din utvärdering. Ange ditt utvärderingsnamn, granska att allt ser korrekt ut och Skicka för att skapa utvärderingsuppgiften. Du kommer till en statussida för utvärderingsjobbet, som visar statusen Väntar.

9. När utvärderingsjobbet har skapats kan du välja jobbet för att visa fullständig information om jobbet:

10. För semantisk likhet visar utdatainformation en JSON-representation som du kan kopiera och klistra in av dina godkända tester.

11. Du kan också lägga till fler Eval-körningar genom att välja knappen + Lägg till körning i det övre vänstra hörnet på utvärderingsjobbsidan.

Skapa utvärdering

Du kan skapa en utvärdering genom att ange en konfiguration av datakällan och kriterierna för utvärderingstestning. Nedan visas ett av många sätt att definiera en konfiguration av datakällor. Du kan också ange ett eller flera testvillkor.

import asyncio
import json
import requests

async def create_eval():
    response = await asyncio.to_thread(
        requests.post,
        f'{API_ENDPOINT}/openai/v1/evals',
        headers={
            'api-key': API_KEY,
            'aoai-evals': 'preview'
        },
        json={
            'name': 'My Evaluation',
            'data_source_config': {
                'type': 'custom',
                'item_schema': {
                'type': 'object',
                'properties': {
                    'question': {
                    'type': 'string'
                    },
                    'subject': {
                    'type': 'string'
                    },
                    'A': {
                    'type': 'string'
                    },
                    'B': {
                    'type': 'string'
                    },
                    'C': {
                    'type': 'string'
                    },
                    'D': {
                    'type': 'string'
                    },
                    'answer': {
                    'type': 'string'
                    },
                    'completion': {
                    'type': 'string'
                    }
                }
                }
            },
            'testing_criteria': [
                {
                'type': 'string_check',
                'reference': '{{item.completion}}',
                'input': '{{item.answer}}',
                'operation': 'eq',
                'name': 'string check'
                }
            ]
        })

    print(response.status_code)
    print(json.dumps(response.json(), indent=2))

Skapa en enskild körning

Med Azure OpenAI Evaluation kan du skapa flera körningar under ett utvärderingsjobb. Om du vill lägga till en enskild utvärderingsomgång i en befintlig utvärdering kan du specificera eval_id för den befintliga utvärderingen.

import asyncio
import requests
import json


response = await asyncio.to_thread(
    requests.post,
    f'{API_ENDPOINT}/openai/v1/evals/{eval_id}/runs',
    headers={
        'api-key': API_KEY,
        'aoai-evals': 'preview'
    },
    json={
        "name": "No sample",
        "metadata": {
            "sample_generation": "off",
            "file_format": "jsonl"
        },
        "data_source": {
            "type": "jsonl",
            "source": {
            "type": "file_id",
            "id": "file-75099d8d4b5b4abca7cc91e9eca7bba1"
            }
        }
    })

print(response.status_code)
print(json.dumps(response.json(), indent=2))

Uppdatera befintlig utvärdering

import asyncio
import requests
import json

async def update_eval():
    response = await asyncio.to_thread(
        requests.post,
        f'{API_ENDPOINT}/openai/v1/evals/{eval_id}',
        headers={
            'api-key': API_KEY,
            'aoai-evals': 'preview'
        },
        json={
            "name": "Updated Eval Name",
            "metadata": {
                "sample_generation": "off",
                "file_format": "jsonl",
                "updated": "metadata"
            }
        })

    print(response.status_code)
    print(json.dumps(response.json(), indent=2))

Utvärderingsresultat

När utvärderingen är klar kan du hämta utvärderingsresultatet för utvärderingsjobbet genom att specificera eval_id.

import asyncio
import requests

async def get_eval():
    response = await asyncio.to_thread(
        requests.get,
        f'{API_ENDPOINT}/openai/v1/evals/{eval_id}',
        headers={
            'api-key': API_KEY,
            'aoai-evals': 'preview'
        })

    print(response.status_code)
    print(response.json())

Resultat av en utvärderingskörning

Precis som hur du kan skapa en utvärderingskörning under ett befintligt utvärderingsjobb kan du också hämta resultatet för en enda körning:

import asyncio
import requests
import json

async def get_eval_run():
    response = await asyncio.to_thread(
        requests.get,
        f'{API_ENDPOINT}/openai/v1/evals/eval_67fd95c864f08190817f0dff5f42f49e/runs/evalrun_67fe987a6c548190ba6f33f7cd89343d',
        headers={
            'api-key': API_KEY,
            'aoai-evals': 'preview'
        })

    print(response.status_code)
    print(json.dumps(response.json(), indent=2))

Förutom parametrarna i exemplen ovan kan du lägga till dessa parametrar för mer specifika detaljnivå i utvärderingsresultaten:

Utvärderingslista

Så här ser du listan över alla utvärderingsjobb som har skapats:

import asyncio
import requests
import json

async def get_eval_list():
    response = await asyncio.to_thread(
        requests.get,
        f'{API_ENDPOINT}/openai/v1/evals',
        headers={
            'api-key': API_KEY,
            'aoai-evals': 'preview'
        })

    print(response.status_code)
    print(json.dumps(response.json(), indent=2))

Utdatainformation för en körning

Du kan visa de enskilda utdata som genereras från utvärderare för en enda utvärderingsomgång:

import asyncio
import requests
import json

async def get_eval_output_item_list():
    response = await asyncio.to_thread(
        requests.get,
        f'{API_ENDPOINT}/openai/v1/evals/eval_67fd95c864f08190817f0dff5f42f49e/runs/evalrun_67fe987a6c548190ba6f33f7cd89343d/output_items',
        headers={
            'api-key': API_KEY,
            'aoai-evals': 'preview'
        })

    print(response.status_code)
    print(json.dumps(response.json(), indent=2))

Om du har ett visst utdataresultat som du vill se kan du ange utdataobjektets ID:

import asyncio
import requests
import json

async def get_eval_output_item():
    response = await asyncio.to_thread(
        requests.get,
        f'{API_ENDPOINT}/openai/v1/evals/eval_67fd95c864f08190817f0dff5f42f49e/runs/evalrun_67fe987a6c548190ba6f33f7cd89343d/output_items/outputitem_67fe988369308190b50d805120945deb',
        headers={'api-key': API_KEY})

    print(response.status_code)
    print(json.dumps(response.json(), indent=2))

Skapa utvärdering

Du kan skapa en utvärdering genom att ange en konfiguration av datakällan och kriterierna för utvärderingstestning. Nedan visas ett av många sätt att definiera en konfiguration av datakällor. Du kan också ange ett eller flera testvillkor.

curl -X POST "$AZURE_OPENAI_ENDPOINT/openai/v1/evals" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" \
  -H "aoai-evals: preview" \
  -d '{
    "name": "Math Quiz",
    "data_source_config": {
      "type": "custom",
      "include_sample_schema": true,
      "item_schema": {
        "type": "object",
        "properties": {
          "question": { "type": "string" },
          "A": { "type": "string" },
          "B": { "type": "string" },
          "C": { "type": "string" },
          "D": { "type": "string" },
          "answer": { "type": "string" }
        }
      }
    },
    "testing_criteria": [
      {
        "type": "string_check",
        "reference": "{{item.answer}}",
        "input": "{{sample.output_text}}",
        "operation": "eq",
        "name": "string check"
      }
    ]
  }'

Skapa en enskild körning

Med Azure OpenAI Evaluation kan du skapa flera körningar under ett utvärderingsjobb. Du kan lägga till nya utvärderingskörningar i det utvärderingsjobb som du skapade i föregående steg genom att ange eval-id.

curl -X POST "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id}/runs" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" 
  -H "aoai-evals: preview" \

Uppdatera befintlig utvärdering

curl -X POST "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id} \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" 
  -H "aoai-evals: preview" \

Utvärderingsresultat

När utvärderingen är klar kan du hämta det utvärderingsresultatet för utvärderingsjobbet genom att ange eval_id.

curl -X GET "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id}" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" 
  -H "aoai-evals: preview" \

Resultat av en utvärderingskörning

Precis som hur du kan skapa en utvärderingskörning under ett befintligt utvärderingsjobb kan du också hämta resultatet för en enda körning:

curl -X GET "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id}/runs/{run-id}" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" 
  -H "aoai-evals: preview" \

Förutom parametrarna i exemplen ovan kan du lägga till dessa parametrar för mer specifika detaljnivå i utvärderingsresultaten:

Utvärderingslista

Så här ser du listan över alla utvärderingsjobb som har skapats:

curl -X GET "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id}/runs" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" 
  -H "aoai-evals: preview" \

Utdatainformation för en körning

Du kan visa de enskilda resultaten som genereras från utvärderare för en enda utvärderingskörning.

curl -X GET "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id}/runs/{run-id}/output_items" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY" 
  -H "aoai-evals: preview" \

Om du har ett visst utdataresultat som du vill se kan du ange utdataobjektets ID:

curl -X GET "$AZURE_OPENAI_ENDPOINT/openai/v1/evals/{eval-id}/runs/{run-id}/output_items/{output-item-id}" \
  -H "Content-Type: application/json" \
  -H "api-key: $AZURE_OPENAI_API_KEY"
  -H "aoai-evals: preview" \