| { |
| "architectures": [ |
| "GruyereForCausalLM" |
| ], |
| "auto_map": { |
| "AutoConfig": "configuration_gruyere.GruyereConfig", |
| "AutoModelForCausalLM": "modeling_gruyere.GruyereForCausalLM" |
| }, |
| "dtype": "float32", |
| "max_seq_len": 4096, |
| "mlp_hidden": 1024, |
| "model_type": "gouda-gruyere", |
| "n_embd": 384, |
| "n_head": 6, |
| "n_kv_head": 2, |
| "n_layer": 32, |
| "name": "Gruyre-2.0", |
| "qk_norm": true, |
| "rms_eps": 1e-06, |
| "rope_base": 10000.0, |
| "tie_word_embeddings": true, |
| "transformers_version": "5.17.0", |
| "vocab_size": 32768, |
| "z_loss": 0.0 |
| } |
|
|
