Add initial model files

configuration_hymba.py

@@ -0,0 +1,122 @@
+import math
+from transformers.configuration_utils import PretrainedConfig
+
+
+class HymbaConfig(PretrainedConfig):
+
+    model_type = "hymba"
+    keys_to_ignore_at_inference = ["past_key_values"]
+
+    def __init__(
+            self,
+            vocab_size=65536,
+            tie_word_embeddings=False,
+            hidden_size=4096,
+            intermediate_size=14336,
+            num_hidden_layers=32,
+            num_attention_heads=32,
+            num_key_value_heads=8,
+            hidden_act="silu",
+            initializer_range=0.02,
+            rms_norm_eps=1e-6,
+            attn_only_wo_proj=False,
+            use_cache=True,
+            calc_logits_for_entire_prompt=False,
+            output_router_logits=False,
+            router_aux_loss_coef=0.001,
+            pad_token_id=0,
+            bos_token_id=1,
+            eos_token_id=2,
+            sliding_window=None,
+            max_position_embeddings=262144,
+            orig_max_position_embeddings=None,
+            attention_dropout=0.0,
+            num_experts_per_tok=2,
+            num_experts=16,
+            use_mamba_kernels=True,
+            mamba_d_state=16,
+            mamba_d_conv=4,
+            mamba_expand=2,
+            mamba_dt_rank="auto",
+            mamba_conv_bias=True,
+            mamba_proj_bias=False,
+            mamba_inner_layernorms=True,
+            kv_reuse_every_i_layer=-1,
+            kv_reuse_group=None,
+            kv_weight_reuse=False,
+            global_attn_idx=None,
+            num_mamba=1,
+            pure_attn=False,
+            attn_implementation_new='sdpa',
+            rope_type=None,
+            attn_factor=0.5,
+            **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.tie_word_embeddings = tie_word_embeddings
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.sliding_window = sliding_window
+        self.max_position_embeddings = max_position_embeddings
+        self.orig_max_position_embeddings = orig_max_position_embeddings
+        self.attention_dropout = attention_dropout
+        self.attn_only_wo_proj = attn_only_wo_proj
+
+        if num_key_value_heads is None:
+            num_key_value_heads = num_attention_heads
+
+        self.num_key_value_heads = num_key_value_heads
+        self.hidden_act = hidden_act
+        self.initializer_range = initializer_range
+        self.rms_norm_eps = rms_norm_eps
+
+        self.use_cache = use_cache
+        self.calc_logits_for_entire_prompt = calc_logits_for_entire_prompt
+        self.output_router_logits = output_router_logits
+        self.router_aux_loss_coef = router_aux_loss_coef
+
+        self.num_experts_per_tok = num_experts_per_tok
+        self.num_experts = num_experts
+
+        self.use_mamba_kernels = use_mamba_kernels
+        self.mamba_d_state = mamba_d_state
+        self.mamba_d_conv = mamba_d_conv
+        self.mamba_expand = mamba_expand
+        self.mamba_dt_rank = math.ceil(self.hidden_size / 16) if mamba_dt_rank == "auto" else mamba_dt_rank
+        self.mamba_conv_bias = mamba_conv_bias
+        self.mamba_proj_bias = mamba_proj_bias
+        self.mamba_inner_layernorms = mamba_inner_layernorms
+
+        self.attn_hidden_size = kwargs.pop("attn_hidden_size", -1)
+        self.attn_factor = attn_factor
+        self.kq_head_dim = kwargs.pop("kq_head_dim", -1)
+        self.v_head_dim = kwargs.pop("v_head_dim", -1)
+        self.kq_norm = kwargs.pop("kq_norm", None)
+        self.rope = kwargs.pop("rope", False)
+        self.rope_theta = kwargs.pop("rope_theta", 10000.0)
+        self.num_memory_tokens = kwargs.pop("num_memory_tokens", 0)
+        self.memory_tokens_interspersed_every = kwargs.pop("memory_tokens_interspersed_every", 0)
+
+        self.kv_reuse_every_i_layer = kv_reuse_every_i_layer
+        self.kv_reuse_group = kv_reuse_group
+        self.kv_weight_reuse = kv_weight_reuse
+
+        self.global_attn_idx = global_attn_idx
+
+        self.num_mamba = num_mamba
+
+        self.attn_implementation_new = attn_implementation_new
+
+        self.rope_type = rope_type
+        
+        self.pure_attn = pure_attn
+
+        super().__init__(
+            pad_token_id=pad_token_id,
+            bos_token_id=bos_token_id,
+            eos_token_id=eos_token_id,
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs,
+        )