Add support for deepseek_r1_qwen based models

python/ctranslate2/converters/transformers.py

@@ -2448,6 +2448,119 @@ def get_falcon_spec(self, model):
         self._num_kv_attr = "num_kv_heads"
 
 
+@register_loader("Qwen2Config")
+class DeepSeekR1DistillQwen2Loader(ModelLoader):
+    @property
+    def architecture_name(self):
+        return "Qwen2ForCausalLM"
+
+    def get_model_spec(self, model):
+        num_layers = model.config.num_hidden_layers
+
+        num_heads = model.config.num_attention_heads
+        num_heads_kv = getattr(model.config, "num_key_value_heads", num_heads)
+        if num_heads_kv == num_heads:
+            num_heads_kv = None
+
+        sliding_window = getattr(model.config, "sliding_window", 0)
+        max_window_layers = getattr(model.config, "max_window_layers", None)
+
+        rope_scaling = getattr(model.config, "rope_scaling", None)
+        if rope_scaling:
+            rope_type = rope_scaling.get("type") or rope_scaling.get("rope_type")
+            rotary_scaling_type = _SUPPORTED_ROPE_SCALING.get(rope_type)
+            rotary_scaling_factor = rope_scaling["factor"]
+
+            if rotary_scaling_type is None:
+                raise NotImplementedError(
+                    "RoPE scaling type '%s' is not yet implemented. "
+                    "The following RoPE scaling types are currently supported: %s"
+                    % (rope_type, ", ".join(_SUPPORTED_ROPE_SCALING.keys()))
+                )
+        else:
+            rotary_scaling_type = None
+            rotary_scaling_factor = 1
+
+        spec = transformer_spec.TransformerDecoderModelSpec.from_config(
+            num_layers,
+            num_heads,
+            activation=common_spec.Activation.SWISH,
+            pre_norm=True,
+            ffn_glu=True,
+            rms_norm=True,
+            rotary_dim=0,
+            rotary_interleave=False,
+            rotary_scaling_type=rotary_scaling_type,
+            rotary_scaling_factor=rotary_scaling_factor,
+            rotary_base=getattr(model.config, "rope_theta", 10000),
+            num_heads_kv=num_heads_kv,
+            sliding_window=sliding_window if sliding_window > 0 else None,
+            head_dim=getattr(model.config, "head_dim", None),
+        )
+
+        self.set_decoder(spec.decoder, model.model)
+        self.set_linear(spec.decoder.projection, model.lm_head)
+        return spec
+
+    def get_vocabulary(self, model, tokenizer):
+        tokens = super().get_vocabulary(model, tokenizer)
+
+        extra_ids = model.config.vocab_size - len(tokens)
+        for i in range(extra_ids):
+            tokens.append("<extra_id_%d>" % i)
+        return tokens
+
+    def set_vocabulary(self, spec, tokens):
+        spec.register_vocabulary(tokens)
+
+    def set_config(self, config, model, tokenizer):
+        config.bos_token = (
+            tokenizer.bos_token
+            if tokenizer.bos_token is not None
+            else tokenizer.pad_token
+        )
+        config.eos_token = tokenizer.eos_token
+        config.unk_token = (
+            tokenizer.unk_token if tokenizer.unk_token is not None else ""
+        )
+        config.layer_norm_epsilon = model.config.rms_norm_eps
+
+    def set_layer_norm(self, spec, layer_norm):
+        spec.gamma = layer_norm.weight
+
+    def set_decoder(self, spec, module):
+        spec.scale_embeddings = False
+        self.set_embeddings(spec.embeddings, module.embed_tokens)
+        self.set_layer_norm(spec.layer_norm, module.norm)
+
+        for layer_spec, layer in zip(spec.layer, module.layers):
+            self.set_layer_norm(
+                layer_spec.self_attention.layer_norm, layer.input_layernorm
+            )
+            self.set_layer_norm(
+                layer_spec.ffn.layer_norm, layer.post_attention_layernorm
+            )
+
+            split_layers = [common_spec.LinearSpec() for _ in range(3)]
+            self.set_linear(split_layers[0], layer.self_attn.q_proj)
+            self.set_linear(split_layers[1], layer.self_attn.k_proj)
+            self.set_linear(split_layers[2], layer.self_attn.v_proj)
+
+            utils.fuse_linear(layer_spec.self_attention.linear[0], split_layers)
+            self.set_linear(
+                layer_spec.self_attention.linear[1],
+                layer.self_attn.o_proj,
+            )
+
+            self.set_linear(layer_spec.ffn.linear_0, layer.mlp.gate_proj)
+            self.set_linear(layer_spec.ffn.linear_0_noact, layer.mlp.up_proj)
+            self.set_linear(layer_spec.ffn.linear_1, layer.mlp.down_proj)
+
+            delattr(layer, "self_attn")
+            delattr(layer, "mlp")
+            gc.collect()
+
+
 @register_loader("DistilBertConfig")
 class DistilBertLoader(ModelLoader):
     @property