---
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- dense
- generated_from_trainer
- dataset_size:1979
- loss:MultipleNegativesRankingLoss
base_model: google/embeddinggemma-300m
widget:
- source_sentence: The iPhone Bluetooth connection drops unexpectedly during use.
sentences:
- The customer is asking how to update apps on their iPhone automatically.
- The device loses power quickly despite showing a 100% battery level earlier.
- Bluetooth devices disconnect frequently when paired with the iPhone.
- source_sentence: The crash in the cloud service happened when the authentication
module failed to process tokens correctly.
sentences:
- During the parade, the crash of a float startled the spectators but caused no
injuries.
- Cross-platform contact synchronization failed in the application.
- Engineers fixed the crash by patching the token validation algorithm causing the
server to terminate.
- source_sentence: Atlassian Confluence is often used for internal company wikis.
sentences:
- Many organizations rely on Confluence to maintain up-to-date internal knowledge
bases.
- Threading in Slack helps prevent important messages from getting lost.
- Confluence users sometimes face performance issues when integrating with third-party
plugins, causing delays in page loading and editing.
- source_sentence: Windows supports file system formats like NTFS and FAT32 for data
storage management.
sentences:
- Network-stored PDFs cause Adobe Acrobat to shut down unexpectedly.
- Windows users often face network connectivity issues that prevent access to shared
drives, unrelated to file system formats like NTFS or FAT32.
- Converting drives between NTFS and FAT32 is common in Windows for compatibility
reasons.
- source_sentence: Tableau can be extended with Python or R for advanced analytics.
sentences:
- Data scientists integrate Tableau with external scripts to enhance visualizations.
- By using ServiceNow, companies reduce manual work through automation.
- The system administrator troubleshoots network latency issues affecting server
performance.
pipeline_tag: sentence-similarity
library_name: sentence-transformers
---
# SentenceTransformer based on google/embeddinggemma-300m
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [google/embeddinggemma-300m](https://huggingface.co/google/embeddinggemma-300m) on the json dataset. 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](https://huggingface.co/google/embeddinggemma-300m)
- **Maximum Sequence Length:** 2048 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Cosine Similarity
- **Training Dataset:**
- json
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 2048, '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:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("KameronB/IT-embeddinggemma")
# Run inference
queries = [
"Tableau can be extended with Python or R for advanced analytics.",
]
documents = [
'Data scientists integrate Tableau with external scripts to enhance visualizations.',
'The system administrator troubleshoots network latency issues affecting server performance.',
'By using ServiceNow, companies reduce manual work through automation.',
]
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([[ 0.7530, -0.0187, 0.1202]])
```
## Training Details
### Training Dataset
#### json
* Dataset: json
* Size: 1,979 training samples
* Columns: anchor, positive, and negative
* Approximate statistics based on the first 1000 samples:
| | anchor | positive | negative |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
| type | string | string | string |
| details | - min: 5 tokens
- mean: 14.95 tokens
- max: 25 tokens
| - min: 6 tokens
- mean: 14.98 tokens
- max: 25 tokens
| - min: 5 tokens
- mean: 16.91 tokens
- max: 31 tokens
|
* Samples:
| anchor | positive | negative |
|:-----------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------|
| He uses part of his workday to conduct quality assurance tests on products. | She checks product functionality and identifies defects as part of her role. | The violinist practices scales and pieces late into the evening. |
| Gmail search function does not return relevant results. | He struggles to find emails using Gmail's search bar. | She watches YouTube videos on email organization tips. |
| Software installation hangs midway | Setup process freezes during installation | Desktop icons are rearranged randomly |
* Loss: [MultipleNegativesRankingLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
```json
{
"scale": 20.0,
"similarity_fct": "cos_sim",
"gather_across_devices": false
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `per_device_train_batch_size`: 64
- `learning_rate`: 2e-05
- `num_train_epochs`: 1
- `warmup_ratio`: 0.1
- `prompts`: task: sentence similarity | query:
#### All Hyperparameters
Click to expand
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: no
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 64
- `per_device_eval_batch_size`: 8
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 2e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 1
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `bf16`: False
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `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
- `adafactor`: False
- `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
- `use_legacy_prediction_loop`: False
- `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_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `include_tokens_per_second`: False
- `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
- `prompts`: task: sentence similarity | query:
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: proportional
- `router_mapping`: {}
- `learning_rate_mapping`: {}