| { |
| "model_type": "slm", |
| "model_name": "slm", |
| "architectures": [ |
| "TinyGPTForCausalLM" |
| ], |
| "auto_map": { |
| "AutoConfig": "configuration_tiny_gpt.TinyGPTConfig", |
| "AutoModel": "modeling_tiny_gpt.TinyGPTModel", |
| "AutoModelForCausalLM": "modeling_tiny_gpt.TinyGPTForCausalLM" |
| }, |
| "vocab_size": 32000, |
| "ctx_len": 1536, |
| "max_position_embeddings": 1536, |
| "n_layer": 8, |
| "num_hidden_layers": 8, |
| "n_head": 6, |
| "num_attention_heads": 6, |
| "n_embd": 384, |
| "hidden_size": 384, |
| "dropout": 0.0, |
| "attention_backend": "torch", |
| "available_attention_backends": [ |
| "sage", |
| "torch", |
| "flash2", |
| "flash3" |
| ], |
| "trained_attention_backend": "flash2", |
| "torch_fallback": true, |
| "positional_encoding": "rope", |
| "trained_positional_encoding": "rope", |
| "rope_base": 10000.0, |
| "torch_dtype": "bfloat16", |
| "transformers_version": "custom", |
| "pad_token_id": 0, |
| "bos_token_id": 5, |
| "eos_token_id": 6, |
| "sep_token_id": 2, |
| "unk_token_id": 1 |
| } |
|
|
