Contoh resep evaluasi

Resep-resep ini memungkinkan Anda untuk mengevaluasi kemampuan dasar model Amazon Nova di seluruh rangkaian tolok ukur khusus teks yang komprehensif. Mereka disediakan dalam formatxxx_general_text_benchmark_eval.yaml.

Resep ini memungkinkan Anda untuk membawa dataset Anda sendiri untuk benchmarking dan membandingkan output model untuk referensi jawaban menggunakan berbagai jenis metrik. Mereka disediakan dalam formatxxx_bring_your_own_dataset_eval.yaml.

Berikut ini adalah persyaratan dataset Anda sendiri:

Berikut ini adalah entri contoh kumpulan data Anda sendiri

{
   "system":"You are a english major with top marks in class who likes to give minimal word responses: ",
   "query":"What is the symbol that ends the sentence as a question",
   "response":"?"
}
{
   "system":"You are a pattern analysis specialist that provides succinct answers: ",
   "query":"What is the next number in this series? 1, 2, 4, 8, 16, ?",
   "response":"32"
}
{
   "system":"You have great attention to detail that follows instructions accurately: ",
   "query":"Repeat only the last two words of the following: I ate a hamburger today and it was kind of dry",
   "response":"of dry"
}

Untuk menggunakan kumpulan data kustom Anda, ubah resep evaluasi Anda dengan bidang wajib berikut, jangan ubah konten apa pun:

evaluation:
  task: gen_qa
  strategy: gen_qa
  metric: all

Batasan berikut berlaku:

Amazon Nova LLM sebagai Hakim adalah fitur evaluasi model yang memungkinkan pelanggan membandingkan kualitas respons dari satu model ke respons model dasar pada kumpulan data khusus. Dibutuhkan kumpulan data dengan petunjuk, respons dasar, dan respons penantang, dan menggunakan model Nova Judge untuk memberikan metrik winrate berdasarkan probabilitas Bradley-Terry dengan perbandingan berpasangan.

Resep disediakan dalam formatxxx_llm_judge_eval.yaml.

Berikut ini adalah LLM sebagai persyaratan Hakim:

Berikut adalah LLM sebagai entri contoh juri

{
"prompt": "What is the most effective way to combat climate change?",
"response_A": "The most effective way to combat climate change is through a combination of transitioning to renewable energy sources and implementing strict carbon pricing policies. This creates economic incentives for businesses to reduce emissions while promoting clean energy adoption.",
"response_B": "We should focus on renewable energy. Solar and wind power are good. People should drive electric cars. Companies need to pollute less."
}
{
"prompt": "Explain how a computer's CPU works",
"response_A": "CPU is like brain of computer. It does math and makes computer work fast. Has lots of tiny parts inside.",
"response_B": "A CPU (Central Processing Unit) functions through a fetch-execute cycle, where instructions are retrieved from memory, decoded, and executed through its arithmetic logic unit (ALU). It coordinates with cache memory and registers to process data efficiently using binary operations."
}
{
"prompt": "How does photosynthesis work?",
"response_A": "Plants do photosynthesis to make food. They use sunlight and water. It happens in leaves.",
"response_B": "Photosynthesis is a complex biochemical process where plants convert light energy into chemical energy. They utilize chlorophyll to absorb sunlight, combining CO2 and water to produce glucose and oxygen through a series of chemical reactions in chloroplasts."
}

Untuk menggunakan kumpulan data kustom Anda, ubah resep evaluasi Anda dengan bidang wajib berikut, jangan ubah konten apa pun:

evaluation:
  task: llm_judge
  strategy: judge
  metric: all

Batasan berikut berlaku:

Nova LLM Judge for multi-modal (image), kependekan dari Nova MM_LLM Judge, adalah fitur evaluasi model yang memungkinkan Anda membandingkan kualitas respons dari satu model dengan respons model dasar menggunakan kumpulan data khusus. Ini menerima kumpulan data yang berisi petunjuk, respons dasar, dan respons penantang, dan gambar dalam bentuk string yang dikodekan Base64, kemudian menggunakan model Nova Judge untuk memberikan metrik tingkat kemenangan berdasarkan probabilitas Bradley-Terry melalui perbandingan berpasangan. Format resep:xxx_mm_llm_judge _eval.yaml.

Persyaratan set data Nova LLM

Format berkas:

Format skema - Setiap baris dalam .jsonl file harus berupa objek JSON dengan bidang berikut.

Entri contoh

Untuk keterbacaan, contoh berikut mencakup baris dan lekukan baru, tetapi dalam kumpulan data yang sebenarnya, setiap catatan harus berada pada satu baris.

{
  "prompt": "what is in the image?",
  "images": [
    {
      "data": "data:image/jpeg;Base64,/9j/2wBDAAQDAwQDAwQEAwQFBAQFBgo..."
    }
  ],
  "response_A": "a dog.",
  "response_B": "a cat.",
}
{
  "prompt": "how many animals in echo of the images?",
  "images": [
    {
      "data": "data:image/jpeg;Base64,/9j/2wBDAAQDAwQDAwQEAwQFBAQFBgo..."
    },
    {
      "data": "data:image/jpeg;Base64,/DKEafe3gihn..."
    }
  ],
  "response_A": "The first image contains one cat and the second image contains one dog",
  "response_B": "The first image has one aminal and the second has one animal",
}           

Untuk menggunakan kumpulan data kustom Anda, ubah resep evaluasi Anda dengan bidang wajib berikut, jangan ubah konten apa pun:

evaluation:
  task: mm_llm_judge
  strategy: judge
  metric: all

Batasan

Rubric Judge adalah model LLM-as-a-judge evaluasi yang disempurnakan yang dibangun di atas Nova 2.0 Lite. Berbeda dengan model hakim asli yang hanya memberikan putusan preferensi (A>B, B>A, atau dasi), Rubric Judge secara dinamis menghasilkan kriteria evaluasi khusus yang disesuaikan dengan setiap prompt dan memberikan skor granular di berbagai dimensi.

Kemampuan kunci:

Contoh kriteria yang dihasilkan oleh model:

price_validation:
  description: "The response includes validation to ensure price is a positive value."
  type: "scale"
  weight: 0.3

Model mengevaluasi kedua tanggapan terhadap semua kriteria yang dihasilkan, kemudian menggunakan skor tingkat kriteria ini untuk menginformasikan keputusan preferensi akhirnya.

Konfigurasi resep

Resep Rubrik Hakim

Aktifkan Rubric Judge dengan mengatur task: rubric_llm_judge resep Anda:

run:
  name: nova-eval-job-name                              # [MODIFIABLE] Unique identifier for your evaluation job
  model_type: amazon.nova-2-lite-v1:0:256k              # [FIXED] Rubric Judge model type
  model_name_or_path: "nova-lite-2/prod"                # [FIXED] Path to model checkpoint or identifier
  replicas: 1                                           # [MODIFIABLE] Number of replicas for SageMaker Training job
  data_s3_path: ""                                      # [FIXED] Leave empty for SageMaker Training job
  output_s3_path: ""                                    # [FIXED] Leave empty for SageMaker Training job

evaluation:
  task: rubric_llm_judge                                # [FIXED] Evaluation task - enables Rubric Judge
  strategy: judge                                       # [FIXED] Evaluation strategy
  metric: all                                           # [FIXED] Metric calculation method

inference:
  max_new_tokens: 12000                                 # [MODIFIABLE] Maximum tokens to generate
  top_k: -1                                             # [MODIFIABLE] Top-k sampling parameter
  top_p: 1.0                                            # [MODIFIABLE] Nucleus sampling parameter
  temperature: 0                                        # [MODIFIABLE] Sampling temperature (0 = deterministic)

LLM asli sebagai resep Hakim (untuk perbandingan)

Model juri asli menggunakantask: llm_judge:

run:
  name: eval-job-name                                   # [MODIFIABLE] Unique identifier for your evaluation job
  model_type: amazon.nova-micro-v1:0:128k               # [FIXED] Model type
  model_name_or_path: "nova-micro/prod"                 # [FIXED] Path to model checkpoint or identifier
  replicas: 1                                           # [MODIFIABLE] Number of replicas for SageMaker Training job
  data_s3_path: ""                                      # [FIXED] Leave empty for SageMaker Training job
  output_s3_path: ""                                    # [FIXED] Leave empty for SageMaker Training job

evaluation:
  task: llm_judge                                       # [FIXED] Original judge task
  strategy: judge                                       # [FIXED] Evaluation strategy
  metric: all                                           # [FIXED] Metric calculation method

inference:
  max_new_tokens: 12000                                 # [MODIFIABLE] Maximum tokens to generate
  top_k: -1                                             # [MODIFIABLE] Top-k sampling parameter
  top_p: 1.0                                            # [MODIFIABLE] Nucleus sampling parameter
  temperature: 0                                        # [MODIFIABLE] Sampling temperature (0 = deterministic)

Format dataset masukan

Format dataset input identik dengan model juri asli:

Bidang yang diperlukan:

Contoh kumpulan data (format JSONL):

{"prompt": "What is the most effective way to combat climate change?", "response_A": "The most effective way to combat climate change is through a combination of transitioning to renewable energy sources and implementing strict carbon pricing policies. This creates economic incentives for businesses to reduce emissions while promoting clean energy adoption.", "response_B": "We should focus on renewable energy. Solar and wind power are good. People should drive electric cars. Companies need to pollute less."}
{"prompt": "Explain how a computer's CPU works", "response_A": "CPU is like brain of computer. It does math and makes computer work fast. Has lots of tiny parts inside.", "response_B": "A CPU (Central Processing Unit) functions through a fetch-execute cycle, where instructions are retrieved from memory, decoded, and executed through its arithmetic logic unit (ALU). It coordinates with cache memory and registers to process data efficiently using binary operations."}
{"prompt": "How does photosynthesis work?", "response_A": "Plants do photosynthesis to make food. They use sunlight and water. It happens in leaves.", "response_B": "Photosynthesis is a complex biochemical process where plants convert light energy into chemical energy. They utilize chlorophyll to absorb sunlight, combining CO2 and water to produce glucose and oxygen through a series of chemical reactions in chloroplasts."}

Persyaratan format:

Keluaran evaluasi

Struktur keluaran

Rubric Judge menghasilkan metrik evaluasi yang ditingkatkan dibandingkan dengan model juri asli:

{
  "config_general": {
    "lighteval_sha": "string",
    "num_fewshot_seeds": "int",
    "max_samples": "int | null",
    "job_id": "int",
    "start_time": "float",
    "end_time": "float",
    "total_evaluation_time_secondes": "string",
    "model_name": "string",
    "model_sha": "string",
    "model_dtype": "string | null",
    "model_size": "string"
  },
  "results": {
    "custom|rubric_llm_judge_judge|0": {
      "a_scores": "float",
      "a_scores_stderr": "float",
      "b_scores": "float",
      "b_scores_stderr": "float",
      "ties": "float",
      "ties_stderr": "float",
      "inference_error": "float",
      "inference_error_stderr": "float",
      "score": "float",
      "score_stderr": "float",
      "weighted_score_A": "float",
      "weighted_score_A_stderr": "float",
      "weighted_score_B": "float",
      "weighted_score_B_stderr": "float",
      "score_margin": "float",
      "score_margin_stderr": "float",
      "winrate": "float",
      "lower_rate": "float",
      "upper_rate": "float"
    }
  },
  "versions": {
    "custom|rubric_llm_judge_judge|0": "int"
  }
}

Metrik baru di Rubric Judge

Enam metrik berikut unik untuk Rubric Judge dan memberikan penilaian kualitas terperinci:

Memahami metrik skor tertimbang

Tujuan: Skor tertimbang memberikan pengukuran kualitas berkelanjutan yang melengkapi putusan preferensi biner, memungkinkan wawasan yang lebih dalam tentang kinerja model.

Perbedaan utama dari hakim asli:

Menafsirkan score_margin:

Kasus penggunaan:

Metodologi perhitungan

Skor tertimbang dihitung melalui proses berikut:

Contoh: Jika response_A memiliki jumlah tertimbang 0,65 dan response_B memiliki 0,78, score_margin maka akan menjadi -0,13, menunjukkan response_B adalah 13 poin persentase lebih tinggi dalam kualitas di semua kriteria tertimbang.

Dukungan model penalaran

Dukungan model penalaran memungkinkan evaluasi dengan model Nova berkemampuan penalaran yang melakukan penalaran internal eksplisit sebelum menghasilkan tanggapan akhir. Fitur ini menggunakan kontrol tingkat API melalui reasoning_effort parameter untuk mengaktifkan atau menonaktifkan fungsionalitas penalaran secara dinamis, berpotensi meningkatkan kualitas respons untuk tugas analitis yang kompleks.

Model yang didukung:

Konfigurasi resep

Aktifkan penalaran dengan menambahkan reasoning_effort parameter ke inference bagian resep Anda:

run:
  name: eval-job-name                                    # [MODIFIABLE] Unique identifier for your evaluation job
  model_type: amazon.nova-2-lite-v1:0:256k               # [FIXED] Must be a reasoning-supported model
  model_name_or_path: nova-lite-2/prod                   # [FIXED] Path to model checkpoint or identifier
  replicas: 1                                            # [MODIFIABLE] Number of replicas for SageMaker Training job
  data_s3_path: ""                                       # [MODIFIABLE] Leave empty for SageMaker Training job; optional for  job
  output_s3_path: ""                                     # [MODIFIABLE] Output path for  job (not compatible with SageMaker Training jobs)

evaluation:
  task: mmlu                                             # [MODIFIABLE] Evaluation task
  strategy: generate                                     # [MODIFIABLE] Evaluation strategy
  metric: all                                            # [MODIFIABLE] Metric calculation method

inference:
  reasoning_effort: high                                 # [MODIFIABLE] Enables reasoning mode; options: low/medium/high or null to disable
  max_new_tokens: 200                                    # [MODIFIABLE] Maximum tokens to generate
  top_k: 50                                              # [MODIFIABLE] Top-k sampling parameter
  top_p: 1.0                                             # [MODIFIABLE] Nucleus sampling parameter
  temperature: 0                                         # [MODIFIABLE] Sampling temperature (0 = deterministic)

Menggunakan parameter reasing_effort

reasoning_effortParameter mengontrol perilaku penalaran untuk model berkemampuan penalaran.

Prasyarat:

Pilihan yang tersedia:

Kapan mengaktifkan penalaran

Gunakan mode penalaran (low,medium, atauhigh) untuk:

Gunakan mode non-penalaran (hilangkan parameter) untuk:

Pemecahan masalah

Kesalahan: “Mode penalaran diaktifkan tetapi model tidak mendukung penalaran”

Penyebab: reasoning_effort Parameter diatur ke nilai non-null, tetapi yang ditentukan model_type tidak mendukung penalaran.

Resolusi: