CodeBERT Fine-tuned on CrewAI

This is a sentence-transformers model finetuned from microsoft/codebert-base. 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: microsoft/codebert-base
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity

• Language: en
• License: apache-2.0

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': 'RobertaModel'})
  (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})
)

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("itsanan/codebert-embed-crewai-base")
# Run inference
sentences = [
    'Best practices for handle_a2a_polling_started',
    'def handle_a2a_polling_started(\n        self,\n        task_id: str,\n        polling_interval: float,\n        endpoint: str,\n    ) -> None:\n        """Handle A2A polling started event with panel display."""\n        content = Text()\n        content.append("A2A Polling Started\\n", style="cyan bold")\n        content.append("Task ID: ", style="white")\n        content.append(f"{task_id[:8]}...\\n", style="cyan")\n        content.append("Interval: ", style="white")\n        content.append(f"{polling_interval}s\\n", style="cyan")\n\n        self.print_panel(content, "⏳ A2A Polling", "cyan")',
    'def test_agent_with_knowledge_sources_generate_search_query():\n    content = "Brandon\'s favorite color is red and he likes Mexican food."\n    string_source = StringKnowledgeSource(content=content)\n\n    with (\n        patch("crewai.knowledge") as mock_knowledge,\n        patch(\n            "crewai.knowledge.storage.knowledge_storage.KnowledgeStorage"\n        ) as mock_knowledge_storage,\n        patch(\n            "crewai.knowledge.source.base_knowledge_source.KnowledgeStorage"\n        ) as mock_base_knowledge_storage,\n        patch("crewai.rag.chromadb.client.ChromaDBClient") as mock_chromadb,\n    ):\n        mock_knowledge_instance = mock_knowledge.return_value\n        mock_knowledge_instance.sources = [string_source]\n        mock_knowledge_instance.query.return_value = [{"content": content}]\n\n        mock_storage_instance = mock_knowledge_storage.return_value\n        mock_storage_instance.sources = [string_source]\n        mock_storage_instance.query.return_value = [{"content": content}]\n        mock_storage_instance.save.return_value = None\n\n        mock_chromadb_instance = mock_chromadb.return_value\n        mock_chromadb_instance.add_documents.return_value = None\n\n        mock_base_knowledge_storage.return_value = mock_storage_instance\n\n        agent = Agent(\n            role="Information Agent with extensive role description that is longer than 80 characters",\n            goal="Provide information based on knowledge sources",\n            backstory="You have access to specific knowledge sources.",\n            llm=LLM(model="gpt-4o-mini"),\n            knowledge_sources=[string_source],\n        )\n\n        task = Task(\n            description="What is Brandon\'s favorite color?",\n            expected_output="The answer to the question, in a format like this: `{{name: str, favorite_color: str}}`",\n            agent=agent,\n        )\n\n        crew = Crew(agents=[agent], tasks=[task])\n        result = crew.kickoff()\n\n        # Updated assertion to check the JSON content\n        assert "Brandon" in str(agent.knowledge_search_query)\n        assert "favorite color" in str(agent.knowledge_search_query)\n\n        assert "red" in result.raw.lower()',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.7350, 0.6480],
#         [0.7350, 1.0000, 0.8133],
#         [0.6480, 0.8133, 1.0000]])

Evaluation

Metrics

Information Retrieval

• Dataset: dim_768
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 768
  }
  

Information Retrieval

• Dataset: dim_512
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 512
  }
  

Information Retrieval

• Dataset: dim_256
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 256
  }
  

Information Retrieval

• Dataset: dim_128
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 128
  }
  

Information Retrieval

• Dataset: dim_64
• Evaluated with InformationRetrievalEvaluator with these parameters:

  {
      "truncate_dim": 64
  }
  

Training Details

Training Dataset

Unnamed Dataset

• Size: 900 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 900 samples:

  	anchor	positive
  type	string	string
  details	min: 6 tokensmean: 13.96 tokensmax: 141 tokens	min: 20 tokensmean: 254.94 tokensmax: 512 tokens

• Samples:

  anchor	positive
  Example usage of DeeplyNestedFlow	class DeeplyNestedFlow(Flow): @start() def a(self): execution_order.append("a") @start() def b(self): execution_order.append("b") @start() def c(self): execution_order.append("c") @start() def d(self): execution_order.append("d") # Nested: (a AND b) OR (c AND d) @listen(or_(and_(a, b), and_(c, d))) def result(self): execution_order.append("result")
  Explain the test_agent_with_knowledge_sources_generate_search_query logic	def test_agent_with_knowledge_sources_generate_search_query(): content = "Brandon's favorite color is red and he likes Mexican food." string_source = StringKnowledgeSource(content=content) with ( patch("crewai.knowledge") as mock_knowledge, patch( "crewai.knowledge.storage.knowledge_storage.KnowledgeStorage" ) as mock_knowledge_storage, patch( "crewai.knowledge.source.base_knowledge_source.KnowledgeStorage" ) as mock_base_knowledge_storage, patch("crewai.rag.chromadb.client.ChromaDBClient") as mock_chromadb, ): mock_knowledge_instance = mock_knowledge.return_value mock_knowledge_instance.sources = [string_source] mock_knowledge_instance.query.return_value = [{"content": content}] mock_storage_instance = mock_knowledge_storage.return_value mock_storage_instance.sources = [string_source] mock_storage_instance.query.return_value = [{"content": content}]...
  Example usage of agent	def agent(self) -> Agent | None: """Get the current agent associated with this memory.""" return self._agent

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 4
• per_device_eval_batch_size: 4
• gradient_accumulation_steps: 16
• learning_rate: 2e-05
• num_train_epochs: 4
• lr_scheduler_type: cosine
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True
• optim: adamw_torch
• batch_sampler: no_duplicates

All Hyperparameters

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: epoch
• prediction_loss_only: True
• per_device_train_batch_size: 4
• per_device_eval_batch_size: 4
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 16
• 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: 4
• max_steps: -1
• lr_scheduler_type: cosine
• lr_scheduler_kwargs: None
• 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: True
• 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: True
• 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
• 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: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.12.12
• Sentence Transformers: 5.2.2
• Transformers: 4.57.6
• PyTorch: 2.9.0+cu126
• Accelerate: 1.12.0
• Datasets: 4.0.0
• Tokenizers: 0.22.2

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}