Instructions to use ram-lexsi/agenttune-testrun-vector-memory with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ram-lexsi/agenttune-testrun-vector-memory with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ram-lexsi/agenttune-testrun-vector-memory", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download training_stats.json from ram-lexsi/agenttune-testrun-vector-memory: direct link, hf CLI and curl.
- Browser
- Download file 2.11 kB
-
https://huggingface.co/ram-lexsi/agenttune-testrun-vector-memory/resolve/main/training_stats.json
- Command line
-
hf download hf://ram-lexsi/agenttune-testrun-vector-memory/training_stats.json
-
curl -L -o training_stats.json https://huggingface.co/ram-lexsi/agenttune-testrun-vector-memory/resolve/main/training_stats.json
2.11 kB
| { | |
| "model": "HuggingFaceTB/SmolLM2-360M-Instruct", | |
| "dataset_name": "unknown", | |
| "output_dir": "./out_vector-memory", | |
| "agentic_mode": true, | |
| "tools": [ | |
| "search_kb" | |
| ], | |
| "train_size": 40, | |
| "eval_size": 0, | |
| "final_loss": -0.1025804548036831, | |
| "total_steps": 10, | |
| "training_history": [], | |
| "metrics": { | |
| "train_runtime": 252.8961, | |
| "train_samples_per_second": 0.158, | |
| "train_steps_per_second": 0.04, | |
| "total_flos": 0.0, | |
| "train_loss": -0.1025804548036831, | |
| "epoch": 1.0 | |
| }, | |
| "config_kwargs": { | |
| "model": "HuggingFaceTB/SmolLM2-360M-Instruct", | |
| "train_dataset": "Dataset({\n features: ['prompt', 'answer'],\n num_rows: 40\n})", | |
| "reward_funcs": "<callable: combined_reward(reward_tool_used+reward_correct_answer)>", | |
| "tools": [ | |
| "<function search_kb at 0x79adf0a83ec0>" | |
| ], | |
| "system_prompt": "You are the Dallas DC policy recall agent. Reply with the short answer only.", | |
| "output_dir": "./out_vector-memory", | |
| "num_train_epochs": 1, | |
| "per_device_train_batch_size": 2, | |
| "gradient_accumulation_steps": 4, | |
| "learning_rate": 1e-06, | |
| "warmup_ratio": 0.03, | |
| "warmup_steps": 10, | |
| "weight_decay": 0.0, | |
| "lr_scheduler_type": "linear", | |
| "max_grad_norm": 1.0, | |
| "seed": 42, | |
| "data_seed": 42, | |
| "optim": "adamw_torch", | |
| "num_generations": 2, | |
| "max_prompt_length": 256, | |
| "max_completion_length": 128, | |
| "temperature": 0.7, | |
| "top_p": 0.95, | |
| "beta": 0.04, | |
| "use_vllm": false, | |
| "vllm_mode": "colocate", | |
| "log_completions": true, | |
| "chat_template_kwargs": { | |
| "enable_thinking": false | |
| }, | |
| "max_steps_per_turn": 8, | |
| "rollout_backend": null, | |
| "engine_kwargs": {}, | |
| "peft_config": { | |
| "r": 8, | |
| "lora_alpha": 16, | |
| "lora_dropout": 0.0, | |
| "bias": "none", | |
| "task_type": "CAUSAL_LM" | |
| }, | |
| "bf16": true, | |
| "fp16": false, | |
| "gradient_checkpointing": true, | |
| "dataloader_num_workers": 0, | |
| "logging_steps": 1, | |
| "save_steps": 500, | |
| "report_to": "none", | |
| "push_to_hub": false, | |
| "eval_strategy": "no", | |
| "remove_unused_columns": false | |
| } | |
| } |