5a086353ab
Add end-to-end embedding/sentence-transformer training pipeline using FastSentenceTransformer, SentenceTransformerTrainer, and MultipleNegativesRankingLoss with BatchSamplers.NO_DUPLICATES. Backend: - Add is_embedding_model() detection via HF tags + pipeline_tag - Add /check-embedding/ API route and EmbeddingCheckResponse - Extend derive_model_type() to return "embeddings" - Add _run_embedding_training() in worker.py with progress callbacks, stop handling, LoRA (task_type=FEATURE_EXTRACTION), and model saving - Add is_embedding field to TrainingStartRequest and ModelDetails - Add YAML configs for 5 models: all-MiniLM-L6-v2, bge-m3, embeddinggemma-300m, gte-modernbert-base, Qwen3-Embedding-0.6B Frontend: - Wire isEmbeddingModel flag through store, API types, and mappers - Force packing=false, train_on_completions=false, warmup_ratio=0.03 - Hide packing and train_on_completions checkboxes for embedding models - Auto-set modelType to "embeddings" from backend model_type response
119 lines
2.6 KiB
Python
119 lines
2.6 KiB
Python
# SPDX-License-Identifier: AGPL-3.0-only - See /studio/LICENSE.AGPL-3.0
|
|
# Copyright © 2025 Unsloth AI
|
|
|
|
"""
|
|
Pydantic models for API request/response schemas
|
|
"""
|
|
from .training import (
|
|
TrainingStartRequest,
|
|
TrainingJobResponse,
|
|
TrainingStatus,
|
|
TrainingProgress,
|
|
)
|
|
from .models import (
|
|
CheckpointInfo,
|
|
ModelCheckpoints,
|
|
CheckpointListResponse,
|
|
ModelDetails,
|
|
LocalModelInfo,
|
|
LocalModelListResponse,
|
|
LoRAInfo,
|
|
LoRAScanResponse,
|
|
ModelListResponse,
|
|
)
|
|
from .auth import (
|
|
AuthSetupRequest,
|
|
AuthLoginRequest,
|
|
RefreshTokenRequest,
|
|
AuthStatusResponse,
|
|
)
|
|
from .export import (
|
|
LoadCheckpointRequest,
|
|
ExportStatusResponse,
|
|
ExportOperationResponse,
|
|
ExportMergedModelRequest,
|
|
ExportBaseModelRequest,
|
|
ExportGGUFRequest,
|
|
ExportLoRAAdapterRequest,
|
|
)
|
|
from .users import Token
|
|
from .datasets import (
|
|
CheckFormatRequest,
|
|
CheckFormatResponse,
|
|
)
|
|
from .inference import (
|
|
LoadRequest,
|
|
UnloadRequest,
|
|
GenerateRequest,
|
|
LoadResponse,
|
|
UnloadResponse,
|
|
InferenceStatusResponse,
|
|
)
|
|
from .responses import (
|
|
TrainingStopResponse,
|
|
TrainingMetricsResponse,
|
|
LoRABaseModelResponse,
|
|
VisionCheckResponse,
|
|
EmbeddingCheckResponse,
|
|
)
|
|
from .data_recipe import (
|
|
RecipePayload,
|
|
PreviewResponse,
|
|
ValidateError,
|
|
ValidateResponse,
|
|
JobCreateResponse,
|
|
)
|
|
|
|
__all__ = [
|
|
# Training schemas
|
|
"TrainingStartRequest",
|
|
"TrainingJobResponse",
|
|
"TrainingStatus",
|
|
"TrainingProgress",
|
|
# Model management schemas
|
|
"ModelDetails",
|
|
"LocalModelInfo",
|
|
"LocalModelListResponse",
|
|
"LoRAInfo",
|
|
"LoRAScanResponse",
|
|
"ModelListResponse",
|
|
# Auth schemas
|
|
"AuthSetupRequest",
|
|
"AuthLoginRequest",
|
|
"RefreshTokenRequest",
|
|
"AuthStatusResponse",
|
|
# Export schemas
|
|
"CheckpointInfo",
|
|
"ModelCheckpoints",
|
|
"CheckpointListResponse",
|
|
"LoadCheckpointRequest",
|
|
"ExportStatusResponse",
|
|
"ExportOperationResponse",
|
|
"ExportMergedModelRequest",
|
|
"ExportBaseModelRequest",
|
|
"ExportGGUFRequest",
|
|
"ExportLoRAAdapterRequest",
|
|
"Token",
|
|
# Dataset schemas
|
|
"CheckFormatRequest",
|
|
"CheckFormatResponse",
|
|
# Inference schemas
|
|
"LoadRequest",
|
|
"UnloadRequest",
|
|
"GenerateRequest",
|
|
"LoadResponse",
|
|
"UnloadResponse",
|
|
"InferenceStatusResponse",
|
|
# Response schemas
|
|
"TrainingStopResponse",
|
|
"TrainingMetricsResponse",
|
|
"LoRABaseModelResponse",
|
|
"VisionCheckResponse",
|
|
"EmbeddingCheckResponse",
|
|
# Data recipe
|
|
"RecipePayload",
|
|
"PreviewResponse",
|
|
"ValidateError",
|
|
"ValidateResponse",
|
|
"JobCreateResponse",
|
|
]
|