| { |
| "activation_function": "swiglu", |
| "attn_pdrop": 0.0, |
| "bos_token_id": "<bos>", |
| "embd_pdrop": 0.0, |
| "eos_token_id": "<eos>", |
| "fused_bias_fc": false, |
| "fused_dropout_add_ln": false, |
| "fused_mlp": false, |
| "initializer_range": 0.02, |
| "intermediate_size": 2560, |
| "layer_norm_epsilon": 1e-06, |
| "max_seq_length": 64, |
| "mlp_fc1_bias": false, |
| "mlp_fc2_bias": false, |
| "model_name_or_path": "llama_small_FA", |
| "model_type": "gpt2", |
| "n_embd": 640, |
| "n_head": 10, |
| "n_head_kv": 10, |
| "n_inner": 2560, |
| "n_layer": 24, |
| "n_positions": 0, |
| "out_proj_bias": false, |
| "qkv_proj_bias": false, |
| "reorder_and_upcast_attn": false, |
| "resid_pdrop": 0.0, |
| "residual_in_fp32": true, |
| "rms_norm": true, |
| "rotary_emb_base": 10000.0, |
| "rotary_emb_fraction": 1.0, |
| "rotary_emb_interleaved": true, |
| "scale_attn_by_inverse_layer_idx": false, |
| "scale_attn_weights": true, |
| "summary_activation": null, |
| "summary_first_dropout": 0.1, |
| "summary_proj_to_labels": true, |
| "summary_type": "cls_index", |
| "summary_use_proj": true, |
| "tie_word_embeddings": false, |
| "transformers_version": "4.34.0", |
| "use_cache": true, |
| "use_flash_attn": true, |
| "vocab_size": 94 |
| } |
|
|
