SentenceTransformer based on google/embeddinggemma-300m

This is a sentence-transformers model finetuned from google/embeddinggemma-300m. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: google/embeddinggemma-300m
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 dimensions
Similarity Function: Cosine Similarity

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False, 'architecture': 'Gemma3TextModel'})
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Dense({'in_features': 768, 'out_features': 3072, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (3): Dense({'in_features': 3072, 'out_features': 768, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (4): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("XX/embedinggemma_arkts")
# Run inference
queries = [
    "Transform an array of points with all matrices. VERY IMPORTANT: Keep\nmatrix order \"value-touch-offset\" when transforming.\n\n@param pts",
]
documents = [
    "public pointValuesToPixel(pts: number[]) {\n    this.mMatrixValueToPx.mapPoints(pts);\n    this.mViewPortHandler.getMatrixTouch().mapPoints(pts);\n    this.mMatrixOffset.mapPoints(pts);\n  }",
    'makeNode(uiContext: UIContext): FrameNode {\n    this.rootNode = new FrameNode(uiContext);\n    if (this.rootNode !== null) {\n      this.rootRenderNode = this.rootNode.getRenderNode();\n    }\n    return this.rootNode;\n  }',
    'export interface OnlineLunarYear {\n  year: number;\n  zodiac: string;\n  ganzhi: string;\n  leapMonth: number;\n  isLeapYear: boolean;\n  leapMonthDays?: number;\n  solarTerms: SolarTermInfo[];\n  festivals: LunarFestival[];\n}',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[ 0.8923,  0.0264, -0.0212]])

Evaluation Results

On arkts-code-docstring dataset split test

Model	Params	MRR	NDCG@5	Recall@1	Recall@5
embedinggemma_arkts	308M	0.7788	0.8034	0.7142	0.8769
QWEN3-Embedding-0.6B	596M	0.6776	0.7015	0.6141	0.7723
embeddinggemma-300m	308M	0.6399	0.6654	0.5740	0.7416
BGE-M3	567M	0.5283	0.5603	0.4464	0.6558
BGE-base-zh-v1.5	110M	0.3598	0.3903	0.2841	0.4816
BGE-base-en-v1.5	110M	0.3439	0.3637	0.2935	0.4227
E5-base-v2	110M	0.3073	0.3261	0.2596	0.3823
BM25 (jieba)	–	0.2043	0.2204	0.1643	0.2690

Training Details

Training Dataset

Dataset: XX

Size: 39,122 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1
type string string
details
min: 3 tokens
mean: 97.17 tokens
max: 512 tokens

min: 3 tokens
mean: 94.4 tokens
max: 512 tokens

	sentence_0	sentence_1
type	string	string
details	min: 3 tokens mean: 97.17 tokens max: 512 tokens	min: 3 tokens mean: 94.4 tokens max: 512 tokens

Samples:

sentence_0	sentence_1
`移除登录状态监听`	`public removeLoginStateListener(listener: (isLoggedIn: boolean) => void) {\n const index = this.loginStateListeners.indexOf(listener);\n if (index !== -1) {\n this.loginStateListeners.splice(index, 1);\n }\n }`
`PUT请求`	`static put(url: string, data?: Object, config: RequestConfig = {}): Promise> { const putConfig: RequestConfig = { method: http.RequestMethod.PUT, headers: config.headers, timeout: config.timeout, data: data }; return HttpUtil.request(url, putConfig); }`
`Transform an array of points with all matrices. VERY IMPORTANT: Keep\nmatrix order "value-touch-offset" when transforming.\n\n@param pts`	`public pointValuesToPixel(pts: number[]) {\n this.mMatrixValueToPx.mapPoints(pts);\n this.mViewPortHandler.getMatrixTouch().mapPoints(pts);\n this.mMatrixOffset.mapPoints(pts);\n }`

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim",
    "gather_across_devices": false
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 32
per_device_eval_batch_size: 32
num_train_epochs: 2
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 32
per_device_eval_batch_size: 32
gradient_accumulation_steps: 1
eval_accumulation_steps: None
torch_empty_cache_steps: None
learning_rate: 5e-05
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1
num_train_epochs: 2
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: None
warmup_ratio: None
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
enable_jit_checkpoint: False
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
use_cpu: False
seed: 42
data_seed: None
bf16: False
fp16: False
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: -1
ddp_backend: None
debug: []
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: False
ignore_data_skip: False
fsdp: []
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
parallelism_config: None
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch_fused
optim_args: None
group_by_length: False
length_column_name: length
project: huggingface
trackio_space_id: trackio
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
dataloader_pin_memory: True
dataloader_persistent_workers: False
skip_memory_metrics: True
push_to_hub: False
resume_from_checkpoint: None
hub_model_id: None
hub_strategy: every_save
hub_private_repo: None
hub_always_push: False
hub_revision: None
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_for_metrics: []
eval_do_concat_batches: True
auto_find_batch_size: False
full_determinism: False
ddp_timeout: 1800
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
include_num_input_tokens_seen: no
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
eval_on_start: False
use_liger_kernel: False
liger_kernel_config: None
eval_use_gather_object: False
average_tokens_across_devices: True
use_cache: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss
0.4088	500	0.3798
0.8177	1000	0.2489
1.2265	1500	0.1308
1.6353	2000	0.0877

Framework Versions

Python: 3.10.19
Sentence Transformers: 5.2.2
Transformers: 5.0.0
PyTorch: 2.9.1
Accelerate: 1.12.0
Datasets: 4.5.0
Tokenizers: 0.22.2

Downloads last month: 93

Safetensors

Model size

0.3B params

Tensor type

F32

Model tree for hreyulog/embedinggemma_arkts

Base model

google/embeddinggemma-300m

Finetuned

(240)

this model