| { | |
| "model_type": "cubic_v11_long_context", | |
| "architectures": [ | |
| "CubicV11LongContext" | |
| ], | |
| "step": 6500, | |
| "parameter_count_state_entries": 156878861, | |
| "vocab_size": 16384, | |
| "hidden_size": 896, | |
| "num_hidden_layers": 14, | |
| "num_attention_heads": 14, | |
| "num_key_value_heads": 7, | |
| "depth_rank": 224, | |
| "local_window": 4096, | |
| "summary_block": 256, | |
| "rope_theta": 500000.0, | |
| "max_position_embeddings": 32768, | |
| "torch_dtype": "float32", | |
| "inference_autocast_dtype": "bfloat16", | |
| "tokenizer_file": "tokenizer.json", | |
| "weights_file": "model.safetensors", | |
| "generation_defaults": { | |
| "max_new_tokens": 4096, | |
| "temperature": 0.6, | |
| "top_p": 0.92, | |
| "top_k": 50, | |
| "repetition_penalty": 1.08 | |
| }, | |
| "source_dataset": "kabsis/AM-DeepSeek-R1-0528-Distilled" | |
| } | |