{ "architectures": [ "E1ForMaskedLM" ], "attn_backend": "sdpa", "auto_map": { "AutoConfig": "modeling_fastplms.E1Config", "AutoModel": "modeling_fastplms.E1Model", "AutoModelForMaskedLM": "modeling_fastplms.E1ForMaskedLM", "AutoModelForSequenceClassification": "modeling_fastplms.E1ForSequenceClassification", "AutoModelForTokenClassification": "modeling_fastplms.E1ForTokenClassification" }, "bos_token_id": 1, "clip_qkv": 8, "dtype": "bfloat16", "eos_token_id": 2, "fastplms_checkpoint_hash": "c29bdfd0241a0cd9685ed347bdd627018ecacff451baf0cd20674b64f44483f9", "fastplms_checkpoint_repo_id": "Synthyra/Profluent-E1-300M", "fastplms_checkpoint_revision": "5ef52c0ad2ae2578f40622696b763523810e8e26", "fastplms_model_id": "e1_300m", "fastplms_release_tool_revision": "e6dd397a9ad368c998d714f6bd64d40b533d1ed1", "fastplms_release_tool_sha256": "6d335c05aa49a232086a816deb25d248d1490529e5783acc3355b9bc6f03e0c2", "fastplms_runtime_bundle_sha256": "a8dda20d6a4a55e6808c7c2f2c9477c113edaa7d6ec6ef57f4a898cd47b73d48", "fastplms_runtime_revision": "e6dd397a9ad368c998d714f6bd64d40b533d1ed1", "fastplms_source_tree_sha256": "9e3cbab448ffde4fbbba79f17cf0321645584acd534c5f96c78b1017d9f3806b", "fastplms_weights_revision": "5ef52c0ad2ae2578f40622696b763523810e8e26", "gated_mlp": true, "global_attention_every_n_layers": 3, "gradient_checkpointing": false, "hidden_act": "silu", "hidden_size": 1024, "initializer_range": 0.02, "intermediate_size": 3072, "max_num_positions_global": 65536, "max_num_positions_within_seq": 8192, "max_num_sequences": 512, "model_type": "E1", "no_ffn_gradient_checkpointing": true, "num_attention_heads": 16, "num_hidden_layers": 20, "num_key_value_heads": 16, "pad_token_id": 0, "rms_norm_eps": 1e-05, "rope_theta_global": 500000.0, "rope_theta_within_seq": 10000.0, "tie_word_embeddings": false, "transformers_version": "4.57.6", "use_cache": true, "vocab_size": 34 }