SFT pada Nova 2.0

Amazon Nova Lite 2.0 menghadirkan kemampuan yang disempurnakan untuk fine-tuning yang diawasi, termasuk mode penalaran lanjutan, pemahaman multimodal yang ditingkatkan, dan penanganan konteks yang diperluas. SFT di Nova 2.0 memungkinkan Anda untuk menyesuaikan kemampuan canggih ini dengan kasus penggunaan spesifik Anda sambil mempertahankan kinerja superior model pada tugas-tugas kompleks.

Fitur utama SFT pada Nova 2.0 meliputi:

run:
  name: my-full-rank-sft-run
  model_type: amazon.nova-2-lite-v1:0:256k
  model_name_or_path: nova-lite-2/prod
  data_s3_path: s3://my-bucket-name/train.jsonl  #  only and not compatible with SageMaker Training Jobs
  replicas: 4                                     # Number of compute instances for training, allowed values are 4, 8, 16, 32
  output_s3_path: s3://my-bucket-name/outputs/    # Output artifact path (HyperPod job-specific; not compatible with standard SageMaker Training Jobs)
  mlflow_tracking_uri: ""                         # Required for MLFlow
  mlflow_experiment_name: "my-full-rank-sft-experiment"  # Optional for MLFlow. Note: leave this field non-empty
  mlflow_run_name: "my-full-rank-sft-run"         # Optional for MLFlow. Note: leave this field non-empty

training_config:
  max_steps: 100                    # Maximum training steps. Minimal is 4.
  save_steps: ${oc.select:training_config.max_steps}  # How many training steps the checkpoint will be saved
  save_top_k: 5                     # Keep top K best checkpoints. Note supported only for  jobs. Minimal is 1.
  max_length: 32768                 # Sequence length (options: 8192, 16384, 32768 [default], 65536)
  global_batch_size: 32             # Global batch size (options: 32, 64, 128)
  reasoning_enabled: true           # If data has reasoningContent, set to true; otherwise False

  lr_scheduler:
    warmup_steps: 15                # Learning rate warmup steps. Recommend 15% of max_steps
    min_lr: 1e-6                    # Minimum learning rate, must be between 0.0 and 1.0

  optim_config:                     # Optimizer settings
    lr: 1e-5                        # Learning rate, must be between 0.0 and 1.0
    weight_decay: 0.0               # L2 regularization strength, must be between 0.0 and 1.0
    adam_beta1: 0.9                  # Exponential decay rate for first-moment estimates
    adam_beta2: 0.95                 # Exponential decay rate for second-moment estimates

  peft:                             # Parameter-efficient fine-tuning (LoRA)
    peft_scheme: "null"             # Disable LoRA for PEFT

Pemilihan Mode Penalaran (Hanya Nova 2.0)

Amazon Nova 2.0 mendukung mode penalaran untuk kemampuan analitis yang ditingkatkan:

Alat memanggil format data

SFT mendukung model pelatihan untuk menggunakan alat (panggilan fungsi). Di bawah ini adalah contoh format input untuk pemanggilan alat:

Masukan sampel:

{
  "schemaVersion": "bedrock-conversation-2024",
  "system": [
    {
      "text": "You are an expert in composing function calls."
    }
  ],
  "toolConfig": {
    "tools": [
      {
        "toolSpec": {
          "name": "getItemCost",
          "description": "Retrieve the cost of an item from the catalog",
          "inputSchema": {
            "json": {
              "type": "object",
              "properties": {
                "item_name": {
                  "type": "string",
                  "description": "The name of the item to retrieve cost for"
                },
                "item_id": {
                  "type": "string",
                  "description": "The ASIN of item to retrieve cost for"
                }
              },
              "required": [
                "item_id"
              ]
            }
          }
        }
      },
      {
        "toolSpec": {
          "name": "getItemAvailability",
          "description": "Retrieve whether an item is available in a given location",
          "inputSchema": {
            "json": {
              "type": "object",
              "properties": {
                "zipcode": {
                  "type": "string",
                  "description": "The zipcode of the location to check in"
                },
                "quantity": {
                  "type": "integer",
                  "description": "The number of items to check availability for"
                },
                "item_id": {
                  "type": "string",
                  "description": "The ASIN of item to check availability for"
                }
              },
              "required": [
                "item_id", "zipcode"
              ]
            }
          }
        }
      }
    ]
  },
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "text": "I need to check whether there are twenty pieces of the following item available. Here is the item ASIN on Amazon: id-123. Please check for the zipcode 94086"
        }
      ]
    },
    {
      "role": "assistant",
      "content": [
        {
          "reasoningContent": {
            "reasoningText": {
              "text": "The user wants to check how many pieces of the item with ASIN id-123 are available in the zipcode 94086"
            }
          }
        },
        {
          "toolUse": {
            "toolUseId": "getItemAvailability_0",
            "name": "getItemAvailability",
            "input": {
              "zipcode": "94086",
              "quantity": 20,
              "item_id": "id-123"
            }
          }
        }
      ]
    },
    {
      "role": "user",
      "content": [
        {
          "toolResult": {
            "toolUseId": "getItemAvailability_0",
            "content": [
              {
                "text": "[{\"name\": \"getItemAvailability\", \"results\": {\"availability\": true}}]"
              }
            ]
          }
        }
      ]
    },
    {
      "role": "assistant",
      "content": [
        {
          "text": "Yes, there are twenty pieces of item id-123 available at 94086. Would you like to place an order or know the total cost?"
        }
      ]
    }
  ]
}

Pertimbangan penting untuk data panggilan alat:

Format data pemahaman dokumen

SFT mendukung model pelatihan tentang tugas pemahaman dokumen. Di bawah ini adalah contoh format input:

Masukan sampel

{
  "schemaVersion": "bedrock-conversation-2024",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "text": "What are the ways in which a customer can experience issues during checkout on Amazon?"
        },
        {
          "document": {
            "format": "pdf",
            "source": {
              "s3Location": {
                "uri": "s3://my-bucket-name/path/to/documents/customer_service_debugging.pdf",
                "bucketOwner": "123456789012"
              }
            }
          }
        }
      ]
    },
    {
      "role": "assistant",
      "content": [
        {
          "text": "Customers can experience issues with 1. Data entry, 2. Payment methods, 3. Connectivity while placing the order. Which one would you like to dive into?"
        }
      ],
      "reasoning_content": [
        {
          "text": "I need to find the relevant section in the document to answer the question.",
          "type": "text"
        }
      ]
    }
  ]
}

Pertimbangan penting untuk pemahaman dokumen:

Pemahaman Video untuk SFT

SFT mendukung model fine-tuning untuk tugas pemahaman video. Di bawah ini adalah contoh format input:

Masukan sampel

{
  "schemaVersion": "bedrock-conversation-2024",
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "text": "What are the ways in which a customer can experience issues during checkout on Amazon?"
        },
        {
          "video": {
            "format": "mp4",
            "source": {
              "s3Location": {
                "uri": "s3://my-bucket-name/path/to/videos/customer_service_debugging.mp4",
                "bucketOwner": "123456789012"
              }
            }
          }
        }
      ]
    },
    {
      "role": "assistant",
      "content": [
        {
          "text": "Customers can experience issues with 1. Data entry, 2. Payment methods, 3. Connectivity while placing the order. Which one would you like to dive into?"
        }
      ],
      "reasoning_content": [
        {
          "text": "I need to find the relevant section in the video to answer the question.",
          "type": "text"
        }
      ]
    }
  ]
}

Pertimbangan penting untuk pemahaman video:

Instruksi Unggah Data

Unggah kumpulan data pelatihan dan validasi ke bucket S3. Tentukan lokasi ini di run blok resep:

## Run config
run:
  ...
  data_s3_path: "s3://<bucket-name>/<training-directory>/<training-file>.jsonl"
  

Catatan: Ganti<bucket-name>,<training-directory>,<validation-directory>,<training-file>, dan <validation-file> dengan jalur S3 yang sebenarnya.

Catatan: Kumpulan data validasi saat ini tidak didukung untuk SFT dengan Amazon Nova 2.0. Jika dataset validasi disediakan, itu akan diabaikan.

Membuat Fine-Tuning Job

Tentukan model dasar menggunakan model_type dan model_name_or_path bidang di run blok:

## Run config
run:
  ...
  model_type: amazon.nova-2-lite-v1:0:256k
  model_name_or_path: nova-lite-2/prod
  ...

Parameter Penyetelan SFT

Parameter yang tersedia untuk penyetelan dengan SFT meliputi:

Jalankan Konfigurasi

Konfigurasi Pelatihan

Penjadwal Tingkat Pembelajaran

Konfigurasi Pengoptimal

Konfigurasi PEFT

LoRa Tuning (ketika peft_scheme adalah 'lora')

Panduan Hyperparameter

Gunakan hiperparameter yang direkomendasikan berikut berdasarkan pendekatan pelatihan:

Pelatihan Peringkat Penuh

LoRa (Adaptasi Peringkat Rendah)

Catatan: Sesuaikan nilai-nilai ini berdasarkan ukuran dataset dan kinerja validasi. Pantau metrik pelatihan untuk mencegah overfitting.