Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

bridge_state.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b119cfd562228dcd8fde6e71a95e9af1fdb1812af708380f4103a1627bc76b8
+size 6293646

chat_template.jinja

@@ -0,0 +1,85 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set content = message.content %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in message.content %}
+                {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
+                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,153 @@
+{
+  "auto_map": {
+    "AutoConfig": "configuration_hybrid.HybridConfig",
+    "AutoModelForCausalLM": "modeling_hybrid.HybridForCausalLM"
+  },
+  "minicpm_dict": {
+    "_name_or_path": "openbmb/MiniCPM5-1B",
+    "architectures": [
+      "LlamaForCausalLM"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 0,
+    "chunk_size_feed_forward": 0,
+    "dtype": "bfloat16",
+    "eos_token_id": [
+      1,
+      130073
+    ],
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 1536,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 4608,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "max_position_embeddings": 131072,
+    "mlp_bias": false,
+    "model_type": "llama",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "num_key_value_heads": 2,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "pad_token_id": 1,
+    "pretraining_tp": 1,
+    "problem_type": null,
+    "return_dict": true,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "rope_theta": 5000000,
+      "rope_type": "default"
+    },
+    "tie_word_embeddings": false,
+    "transformers_version": "5.10.2",
+    "use_cache": true,
+    "vocab_size": 130560
+  },
+  "minicpm_model_name": "openbmb/MiniCPM5-1B",
+  "model_type": "true_hybrid_reasoning_agent",
+  "qwen_dict": {
+    "_name_or_path": "DavidAU/Qwen3-Zero-Coder-Reasoning-V2-0.8B",
+    "architectures": [
+      "Qwen3ForCausalLM"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 151643,
+    "chunk_size_feed_forward": 0,
+    "dtype": "float16",
+    "eos_token_id": 151645,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 1024,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 3072,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_types": [
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention",
+      "full_attention"
+    ],
+    "max_position_embeddings": 40960,
+    "max_window_layers": 28,
+    "model_type": "qwen3",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 42,
+    "num_key_value_heads": 8,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "pad_token_id": null,
+    "problem_type": null,
+    "return_dict": true,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "rope_theta": 1000000,
+      "rope_type": "default"
+    },
+    "sliding_window": null,
+    "tie_word_embeddings": true,
+    "transformers_version": "5.10.2",
+    "use_cache": true,
+    "use_sliding_window": false,
+    "vocab_size": 151936
+  },
+  "qwen_model_name": "DavidAU/Qwen3-Zero-Coder-Reasoning-V2-0.8B",
+  "transformers_version": "5.10.2",
+  "use_cache": false
+}

configuration_hybrid.py

@@ -0,0 +1,11 @@
+from transformers import PretrainedConfig
+
+class HybridConfig(PretrainedConfig):
+    model_type = "true_hybrid_reasoning_agent"
+
+    def __init__(self, qwen_dict=None, minicpm_dict=None, qwen_model_name=None, minicpm_model_name=None, **kwargs):
+        super().__init__(**kwargs)
+        self.qwen_dict = qwen_dict or {}
+        self.minicpm_dict = minicpm_dict or {}
+        self.qwen_model_name = qwen_model_name or "DavidAU/Qwen3-Zero-Coder-Reasoning-V2-0.8B"
+        self.minicpm_model_name = minicpm_model_name or "openbmb/MiniCPM5-1B"

modeling_hybrid.py

@@ -0,0 +1,56 @@
+import torch
+import torch.nn as nn
+from transformers import PreTrainedModel, AutoModelForCausalLM, AutoConfig, GenerationMixin
+from transformers.modeling_outputs import CausalLMOutputWithPast
+from configuration_hybrid import HybridConfig
+
+class HybridForCausalLM(PreTrainedModel, GenerationMixin):
+    config_class = HybridConfig
+    base_model_prefix = "hybrid"
+
+    def __init__(self, config, load_weights=False):
+        super().__init__(config)
+
+        qwen_obj_config = AutoConfig.from_pretrained(config.qwen_model_name)
+        minicpm_obj_config = AutoConfig.from_pretrained(config.minicpm_model_name)
+
+        if load_weights:
+            self.qwen = AutoModelForCausalLM.from_pretrained(config.qwen_model_name, torch_dtype=torch.bfloat16, low_cpu_mem_usage=True)
+            self.minicpm = AutoModelForCausalLM.from_pretrained(config.minicpm_model_name, torch_dtype=torch.bfloat16, low_cpu_mem_usage=True)
+        else:
+            self.qwen = AutoModelForCausalLM.from_config(qwen_obj_config)
+            self.minicpm = AutoModelForCausalLM.from_config(minicpm_obj_config)
+
+        self.qwen.eval()
+        self.minicpm.eval()
+        for p in self.qwen.parameters(): p.requires_grad = False
+        for p in self.minicpm.parameters(): p.requires_grad = False
+
+        self.embed_projection = nn.Linear(qwen_obj_config.hidden_size, minicpm_obj_config.hidden_size, bias=False, dtype=torch.bfloat16)
+        self.minicpm_to_qwen = nn.Linear(minicpm_obj_config.hidden_size, qwen_obj_config.hidden_size, bias=False, dtype=torch.bfloat16)
+        self.mix_gate = nn.Parameter(torch.tensor(0.0, dtype=torch.bfloat16))
+        self.config.use_cache = False
+
+    def forward(self, input_ids=None, attention_mask=None, labels=None, past_key_values=None, use_cache=None, **kwargs):
+        with torch.no_grad():
+            qwen_outputs = self.qwen(input_ids=input_ids, attention_mask=attention_mask, use_cache=False, output_hidden_states=True, return_dict=True, **kwargs)
+            qwen_hidden = qwen_outputs.hidden_states[-1]
+            qwen_embeds = self.qwen.get_input_embeddings()(input_ids)
+
+        minicpm_embeds = self.embed_projection(qwen_embeds)
+        minicpm_outputs = self.minicpm(inputs_embeds=minicpm_embeds, attention_mask=attention_mask, use_cache=False, output_hidden_states=True, return_dict=True, **kwargs)
+        minicpm_hidden = minicpm_outputs.hidden_states[-1]
+
+        fused_hidden = qwen_hidden + torch.sigmoid(self.mix_gate) * self.minicpm_to_qwen(minicpm_hidden)
+        lm_head = self.qwen.get_output_embeddings() or self.qwen.lm_head
+        final_logits = lm_head(fused_hidden)
+
+        loss = None
+        if labels is not None:
+            loss_fct = nn.CrossEntropyLoss(ignore_index=-100)
+            loss = loss_fct(final_logits[:, :-1, :].contiguous().reshape(-1, final_logits.size(-1)), labels[:, 1:].contiguous().reshape(-1))
+
+        return CausalLMOutputWithPast(loss=loss, logits=final_logits)
+
+    def prepare_inputs_for_generation(self, input_ids, past_key_values=None, attention_mask=None, **kwargs):
+        return {"input_ids": input_ids[:, -1:] if past_key_values is not None else input_ids, "attention_mask": attention_mask, "past_key_values": None}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6b7653762283a57ee2d7fcfb04c357f85e469f53d9ca456d4cd61914124f712
+size 11422928

tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}