update model

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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

config.json

@@ -0,0 +1,37 @@
+{
+  "architectures": [
+    "QWenLMHeadModel"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_qwen.QWenConfig",
+    "AutoModelForCausalLM": "modeling_qwen.QWenLMHeadModel"
+  },
+  "attn_dropout_prob": 0.0,
+  "bf16": false,
+  "emb_dropout_prob": 0.0,
+  "fp16": false,
+  "fp32": false,
+  "hidden_size": 4096,
+  "intermediate_size": 22016,
+  "initializer_range": 0.02,
+  "kv_channels": 128,
+  "layer_norm_epsilon": 1e-06,
+  "max_position_embeddings": 8192,
+  "model_type": "qwen",
+  "no_bias": true,
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "onnx_safe": null,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 1.0,
+  "scale_attn_weights": true,
+  "seq_length": 8192,
+  "tie_word_embeddings": false,
+  "tokenizer_class": "QWenTokenizer",
+  "transformers_version": "4.32.0",
+  "use_cache": true,
+  "use_dynamic_ntk": true,
+  "use_flash_attn": "auto",
+  "use_logn_attn": true,
+  "vocab_size": 151936
+}

configuration_qwen.py

@@ -0,0 +1,65 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from transformers import PretrainedConfig
+
+
+class QWenConfig(PretrainedConfig):
+    model_type = "qwen"
+    keys_to_ignore_at_inference = ["past_key_values"]
+
+    def __init__(
+        self,
+        vocab_size=151936,
+        hidden_size=4096,
+        num_hidden_layers=32,
+        num_attention_heads=32,
+        emb_dropout_prob=0.0,
+        attn_dropout_prob=0.0,
+        layer_norm_epsilon=1e-6,
+        initializer_range=0.02,
+        max_position_embeddings=8192,
+        scale_attn_weights=True,
+        use_cache=True,
+        bf16=False,
+        fp16=False,
+        fp32=False,
+        kv_channels=128,
+        rotary_pct=1.0,
+        rotary_emb_base=10000,
+        use_dynamic_ntk=True,
+        use_logn_attn=True,
+        use_flash_attn="auto",
+        intermediate_size=22016,
+        no_bias=True,
+        tie_word_embeddings=False,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        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.emb_dropout_prob = emb_dropout_prob
+        self.attn_dropout_prob = attn_dropout_prob
+        self.layer_norm_epsilon = layer_norm_epsilon
+        self.initializer_range = initializer_range
+        self.scale_attn_weights = scale_attn_weights
+        self.use_cache = use_cache
+        self.max_position_embeddings = max_position_embeddings
+        self.bf16 = bf16
+        self.fp16 = fp16
+        self.fp32 = fp32
+        self.kv_channels = kv_channels
+        self.rotary_pct = rotary_pct
+        self.rotary_emb_base = rotary_emb_base
+        self.use_dynamic_ntk = use_dynamic_ntk
+        self.use_logn_attn = use_logn_attn
+        self.use_flash_attn = use_flash_attn
+        self.no_bias = no_bias
+        super().__init__(
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs
+        )

generation_config.json

@@ -0,0 +1,12 @@
+{
+    "chat_format": "raw",
+    "eos_token_id": 151643,
+    "pad_token_id": 151643,
+    "stop_words_ids": [[151643]],
+    "max_new_tokens": 512,
+    "do_sample": true,
+    "top_k": 0,
+    "top_p": 0.8,
+    "transformers_version": "4.31.0"
+}
+  

model-00001-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9dfd6266bcf80de9c3e5cd4e60300d839d03e459e48975b08d3e3b286044a306
+size 1964066488

model-00002-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dedac66034371aa3b284a7886e9ce0fde9245ebac60f507f089b33ef82a2912
+size 2023960808

model-00003-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81b25b14a58b62300d11b16c66933a8b631400bf846b27a2a5c5344629cd26e8
+size 2023960816

model-00004-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59a22cb822f9e6d0a6a8415a7bab7b8448ce9672fc71f09266db264afc26ce48
+size 2023960848

model-00005-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e61126f7ad8c520112c49808a73d59bee18c6da7693d9a963a4867f389c91e4a
+size 2023960848

model-00006-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ced56cc3265fe03ee73658f845dc23c713389b5d68087e918d24d0e2dea624f
+size 2023960848

model-00007-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0999d47ea087bf79a075ed10889aa3497caff2356200204b032fba208109517
+size 2023960848

model-00008-of-00008.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bc05473a78ade06d526cc206e4a01722563abe99099367cdcbb9b3bf670a5de
+size 1334845784

model.safetensors.index.json

@@ -0,0 +1,266 @@
+{
+  "metadata": {
+    "total_size": 15442649088
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00008-of-00008.safetensors",
+    "transformer.h.0.attn.c_attn.bias": "model-00001-of-00008.safetensors",
+    "transformer.h.0.attn.c_attn.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.0.attn.c_proj.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.0.ln_1.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.0.ln_2.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.0.mlp.c_proj.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.0.mlp.w1.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.0.mlp.w2.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.1.attn.c_attn.bias": "model-00001-of-00008.safetensors",
+    "transformer.h.1.attn.c_attn.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.1.attn.c_proj.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.1.ln_1.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.1.ln_2.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.1.mlp.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.1.mlp.w1.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.1.mlp.w2.weight": "model-00001-of-00008.safetensors",
+    "transformer.h.10.attn.c_attn.bias": "model-00003-of-00008.safetensors",
+    "transformer.h.10.attn.c_attn.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.10.attn.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.10.ln_1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.10.ln_2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.10.mlp.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.10.mlp.w1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.10.mlp.w2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.11.attn.c_attn.bias": "model-00003-of-00008.safetensors",
+    "transformer.h.11.attn.c_attn.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.11.attn.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.11.ln_1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.11.ln_2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.11.mlp.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.11.mlp.w1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.11.mlp.w2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.12.attn.c_attn.bias": "model-00004-of-00008.safetensors",
+    "transformer.h.12.attn.c_attn.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.12.attn.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.12.ln_1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.12.ln_2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.12.mlp.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.12.mlp.w1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.12.mlp.w2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.attn.c_attn.bias": "model-00004-of-00008.safetensors",
+    "transformer.h.13.attn.c_attn.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.attn.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.ln_1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.ln_2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.mlp.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.mlp.w1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.13.mlp.w2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.attn.c_attn.bias": "model-00004-of-00008.safetensors",
+    "transformer.h.14.attn.c_attn.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.attn.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.ln_1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.ln_2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.mlp.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.mlp.w1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.14.mlp.w2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.attn.c_attn.bias": "model-00004-of-00008.safetensors",
+    "transformer.h.15.attn.c_attn.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.attn.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.ln_1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.ln_2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.mlp.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.mlp.w1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.15.mlp.w2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.16.attn.c_attn.bias": "model-00004-of-00008.safetensors",
+    "transformer.h.16.attn.c_attn.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.16.attn.c_proj.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.16.ln_1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.16.ln_2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.16.mlp.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.16.mlp.w1.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.16.mlp.w2.weight": "model-00004-of-00008.safetensors",
+    "transformer.h.17.attn.c_attn.bias": "model-00005-of-00008.safetensors",
+    "transformer.h.17.attn.c_attn.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.17.attn.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.17.ln_1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.17.ln_2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.17.mlp.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.17.mlp.w1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.17.mlp.w2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.attn.c_attn.bias": "model-00005-of-00008.safetensors",
+    "transformer.h.18.attn.c_attn.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.attn.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.ln_1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.ln_2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.mlp.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.mlp.w1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.18.mlp.w2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.attn.c_attn.bias": "model-00005-of-00008.safetensors",
+    "transformer.h.19.attn.c_attn.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.attn.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.ln_1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.ln_2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.mlp.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.mlp.w1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.19.mlp.w2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.2.attn.c_attn.bias": "model-00002-of-00008.safetensors",
+    "transformer.h.2.attn.c_attn.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.2.attn.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.2.ln_1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.2.ln_2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.2.mlp.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.2.mlp.w1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.2.mlp.w2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.20.attn.c_attn.bias": "model-00005-of-00008.safetensors",
+    "transformer.h.20.attn.c_attn.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.20.attn.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.20.ln_1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.20.ln_2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.20.mlp.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.20.mlp.w1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.20.mlp.w2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.21.attn.c_attn.bias": "model-00005-of-00008.safetensors",
+    "transformer.h.21.attn.c_attn.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.21.attn.c_proj.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.21.ln_1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.21.ln_2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.21.mlp.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.21.mlp.w1.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.21.mlp.w2.weight": "model-00005-of-00008.safetensors",
+    "transformer.h.22.attn.c_attn.bias": "model-00006-of-00008.safetensors",
+    "transformer.h.22.attn.c_attn.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.22.attn.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.22.ln_1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.22.ln_2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.22.mlp.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.22.mlp.w1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.22.mlp.w2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.attn.c_attn.bias": "model-00006-of-00008.safetensors",
+    "transformer.h.23.attn.c_attn.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.attn.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.ln_1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.ln_2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.mlp.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.mlp.w1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.23.mlp.w2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.attn.c_attn.bias": "model-00006-of-00008.safetensors",
+    "transformer.h.24.attn.c_attn.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.attn.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.ln_1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.ln_2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.mlp.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.mlp.w1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.24.mlp.w2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.attn.c_attn.bias": "model-00006-of-00008.safetensors",
+    "transformer.h.25.attn.c_attn.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.attn.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.ln_1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.ln_2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.mlp.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.mlp.w1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.25.mlp.w2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.26.attn.c_attn.bias": "model-00006-of-00008.safetensors",
+    "transformer.h.26.attn.c_attn.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.26.attn.c_proj.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.26.ln_1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.26.ln_2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.26.mlp.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.26.mlp.w1.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.26.mlp.w2.weight": "model-00006-of-00008.safetensors",
+    "transformer.h.27.attn.c_attn.bias": "model-00007-of-00008.safetensors",
+    "transformer.h.27.attn.c_attn.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.27.attn.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.27.ln_1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.27.ln_2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.27.mlp.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.27.mlp.w1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.27.mlp.w2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.attn.c_attn.bias": "model-00007-of-00008.safetensors",
+    "transformer.h.28.attn.c_attn.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.attn.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.ln_1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.ln_2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.mlp.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.mlp.w1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.28.mlp.w2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.attn.c_attn.bias": "model-00007-of-00008.safetensors",
+    "transformer.h.29.attn.c_attn.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.attn.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.ln_1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.ln_2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.mlp.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.mlp.w1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.29.mlp.w2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.3.attn.c_attn.bias": "model-00002-of-00008.safetensors",
+    "transformer.h.3.attn.c_attn.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.3.attn.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.3.ln_1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.3.ln_2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.3.mlp.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.3.mlp.w1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.3.mlp.w2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.30.attn.c_attn.bias": "model-00007-of-00008.safetensors",
+    "transformer.h.30.attn.c_attn.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.30.attn.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.30.ln_1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.30.ln_2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.30.mlp.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.30.mlp.w1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.30.mlp.w2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.31.attn.c_attn.bias": "model-00007-of-00008.safetensors",
+    "transformer.h.31.attn.c_attn.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.31.attn.c_proj.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.31.ln_1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.31.ln_2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.31.mlp.c_proj.weight": "model-00008-of-00008.safetensors",
+    "transformer.h.31.mlp.w1.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.31.mlp.w2.weight": "model-00007-of-00008.safetensors",
+    "transformer.h.4.attn.c_attn.bias": "model-00002-of-00008.safetensors",
+    "transformer.h.4.attn.c_attn.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.4.attn.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.4.ln_1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.4.ln_2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.4.mlp.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.4.mlp.w1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.4.mlp.w2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.attn.c_attn.bias": "model-00002-of-00008.safetensors",
+    "transformer.h.5.attn.c_attn.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.attn.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.ln_1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.ln_2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.mlp.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.mlp.w1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.5.mlp.w2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.6.attn.c_attn.bias": "model-00002-of-00008.safetensors",
+    "transformer.h.6.attn.c_attn.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.6.attn.c_proj.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.6.ln_1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.6.ln_2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.6.mlp.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.6.mlp.w1.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.6.mlp.w2.weight": "model-00002-of-00008.safetensors",
+    "transformer.h.7.attn.c_attn.bias": "model-00003-of-00008.safetensors",
+    "transformer.h.7.attn.c_attn.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.7.attn.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.7.ln_1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.7.ln_2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.7.mlp.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.7.mlp.w1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.7.mlp.w2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.attn.c_attn.bias": "model-00003-of-00008.safetensors",
+    "transformer.h.8.attn.c_attn.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.attn.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.ln_1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.ln_2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.mlp.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.mlp.w1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.8.mlp.w2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.attn.c_attn.bias": "model-00003-of-00008.safetensors",
+    "transformer.h.9.attn.c_attn.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.attn.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.ln_1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.ln_2.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.mlp.c_proj.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.mlp.w1.weight": "model-00003-of-00008.safetensors",
+    "transformer.h.9.mlp.w2.weight": "model-00003-of-00008.safetensors",
+    "transformer.ln_f.weight": "model-00008-of-00008.safetensors",
+    "transformer.wte.weight": "model-00001-of-00008.safetensors"
+  }
+}

modeling_qwen.py

@@ -0,0 +1,1232 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import importlib
+import math
+from typing import TYPE_CHECKING, Optional, Tuple, Union, Callable, List, Any, Generator
+
+import torch
+import torch.nn.functional as F
+import torch.utils.checkpoint
+from torch.cuda.amp import autocast
+
+from torch.nn import CrossEntropyLoss
+from transformers import PreTrainedTokenizer, GenerationConfig, StoppingCriteriaList
+from transformers.generation.logits_process import LogitsProcessorList
+
+if TYPE_CHECKING:
+    from transformers.generation.streamers import BaseStreamer
+from transformers.generation.utils import GenerateOutput
+from transformers.modeling_outputs import (
+    BaseModelOutputWithPast,
+    CausalLMOutputWithPast,
+)
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import logging
+
+try:
+    from einops import rearrange
+except ImportError:
+    rearrange = None
+from torch import nn
+
+SUPPORT_CUDA = torch.cuda.is_available()
+SUPPORT_BF16 = SUPPORT_CUDA and torch.cuda.is_bf16_supported()
+SUPPORT_FP16 = SUPPORT_CUDA and torch.cuda.get_device_capability(0)[0] >= 7
+
+from .configuration_qwen import QWenConfig
+from .qwen_generation_utils import (
+    HistoryType,
+    make_context,
+    decode_tokens,
+    get_stop_words_ids,
+    StopWordsLogitsProcessor,
+)
+
+
+logger = logging.get_logger(__name__)
+
+_CHECKPOINT_FOR_DOC = "qwen"
+_CONFIG_FOR_DOC = "QWenConfig"
+
+QWen_PRETRAINED_MODEL_ARCHIVE_LIST = ["qwen-7b"]
+
+_ERROR_BAD_CHAT_FORMAT = """\
+We detect you are probably using the pretrained model (rather than chat model) for chatting, since the chat_format in generation_config is not "chatml".
+If you are directly using the model downloaded from Huggingface, please make sure you are using our "Qwen/Qwen-7B-Chat" Huggingface model (rather than "Qwen/Qwen-7B") when you call model.chat().
+我们检测到您可能在使用预训练模型（而非chat模型）进行多轮chat，因为您当前在generation_config指定的chat_format，并未设置为我们在对话中所支持的"chatml"格式。
+如果您在直接使用我们从Huggingface提供的模型，请确保您在调用model.chat()时，使用的是"Qwen/Qwen-7B-Chat"模型（而非"Qwen/Qwen-7B"预训练模型）。
+"""
+
+_SENTINEL = object()
+_ERROR_STREAM_IN_CHAT = """\
+Pass argument `stream` to model.chat() is buggy, deprecated, and marked for removal. Please use model.chat_stream(...) instead of model.chat(..., stream=True).
+向model.chat()传入参数stream的用法可能存在Bug，该用法已被废弃，将在未来被移除。请使用model.chat_stream(...)代替model.chat(..., stream=True)。
+"""
+
+_ERROR_INPUT_CPU_QUERY_WITH_FLASH_ATTN_ACTIVATED = """\
+We detect you have activated flash attention support, but running model computation on CPU. Please make sure that your input data has been placed on GPU. If you actually want to run CPU computation, please following the readme and set device_map="cpu" to disable flash attention when loading the model (calling AutoModelForCausalLM.from_pretrained).
+检测到您的模型已激活了flash attention支持，但正在执行CPU运算任务。如使用flash attention，请您确认模型输入已经传到GPU上。如果您确认要执行CPU运算，请您在载入模型（调用AutoModelForCausalLM.from_pretrained）时，按照readme说法，指定device_map="cpu"以禁用flash attention。
+"""
+
+apply_rotary_emb_func = None
+rms_norm = None
+flash_attn_unpadded_func = None
+
+
+def _import_flash_attn():
+    global apply_rotary_emb_func, rms_norm, flash_attn_unpadded_func
+    try:
+        from flash_attn.layers.rotary import apply_rotary_emb_func as __apply_rotary_emb_func
+        apply_rotary_emb_func = __apply_rotary_emb_func
+    except ImportError:
+        logger.warn(
+            "Warning: import flash_attn rotary fail, please install FlashAttention rotary to get higher efficiency "
+            "https://github.com/Dao-AILab/flash-attention/tree/main/csrc/rotary"
+        )
+
+    try:
+        from flash_attn.ops.rms_norm import rms_norm as __rms_norm
+        rms_norm = __rms_norm
+    except ImportError:
+        logger.warn(
+            "Warning: import flash_attn rms_norm fail, please install FlashAttention layer_norm to get higher efficiency "
+            "https://github.com/Dao-AILab/flash-attention/tree/main/csrc/layer_norm"
+        )
+
+    try:
+        import flash_attn
+        if not hasattr(flash_attn, '__version__'):
+            from flash_attn.flash_attn_interface import flash_attn_unpadded_func as __flash_attn_unpadded_func
+        else:
+            if int(flash_attn.__version__.split(".")[0]) >= 2:
+                from flash_attn.flash_attn_interface import flash_attn_varlen_func as __flash_attn_unpadded_func
+            else:
+                from flash_attn.flash_attn_interface import flash_attn_unpadded_func as __flash_attn_unpadded_func
+        flash_attn_unpadded_func = __flash_attn_unpadded_func
+    except ImportError:
+        logger.warn(
+            "Warning: import flash_attn fail, please install FlashAttention to get higher efficiency "
+            "https://github.com/Dao-AILab/flash-attention"
+        )
+
+
+class FlashSelfAttention(torch.nn.Module):
+    def __init__(
+        self,
+        causal=False,
+        softmax_scale=None,
+        attention_dropout=0.0,
+    ):
+        super().__init__()
+        assert flash_attn_unpadded_func is not None, (
+            "Please install FlashAttention first, " "e.g., with pip install flash-attn"
+        )
+        assert (
+            rearrange is not None
+        ), "Please install einops first, e.g., with pip install einops"
+        self.causal = causal
+        self.softmax_scale = softmax_scale
+        self.dropout_p = attention_dropout
+
+    def forward(self, q, k, v):
+        assert all((i.dtype in [torch.float16, torch.bfloat16] for i in (q, k, v)))
+        assert all((i.is_cuda for i in (q, k, v)))
+        batch_size, seqlen_q = q.shape[0], q.shape[1]
+        seqlen_k = k.shape[1]
+
+        q, k, v = [rearrange(x, "b s ... -> (b s) ...") for x in [q, k, v]]
+        cu_seqlens_q = torch.arange(
+            0,
+            (batch_size + 1) * seqlen_q,
+            step=seqlen_q,
+            dtype=torch.int32,
+            device=q.device,
+        )
+
+        if self.training:
+            assert seqlen_k == seqlen_q
+
+            is_causal = self.causal
+            cu_seqlens_k = cu_seqlens_q
+        else:
+            is_causal = seqlen_q == seqlen_k
+            cu_seqlens_k = torch.arange(
+                0,
+                (batch_size + 1) * seqlen_k,
+                step=seqlen_k,
+                dtype=torch.int32,
+                device=q.device,
+            )
+            self.dropout_p = 0
+
+        output = flash_attn_unpadded_func(
+            q,
+            k,
+            v,
+            cu_seqlens_q,
+            cu_seqlens_k,
+            seqlen_q,
+            seqlen_k,
+            self.dropout_p,
+            softmax_scale=self.softmax_scale,
+            causal=is_causal,
+        )
+
+        new_shape = (batch_size, output.shape[0] // batch_size) + output.shape[1:]
+        output = output.view(new_shape)
+        return output
+
+
+class QWenAttention(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+
+        self.register_buffer("masked_bias", torch.tensor(-1e4), persistent=False)
+        self.seq_length = config.seq_length
+
+        self.hidden_size = config.hidden_size
+        self.split_size = config.hidden_size
+        self.num_heads = config.num_attention_heads
+        self.head_dim = self.hidden_size // self.num_heads
+
+        self.use_flash_attn = config.use_flash_attn
+        self.scale_attn_weights = True
+
+        self.projection_size = config.kv_channels * config.num_attention_heads
+
+        assert self.projection_size % config.num_attention_heads == 0
+        self.hidden_size_per_attention_head = (
+            self.projection_size // config.num_attention_heads
+        )
+
+        self.c_attn = nn.Linear(config.hidden_size, 3 * self.projection_size)
+
+        self.c_proj = nn.Linear(
+            config.hidden_size, self.projection_size, bias=not config.no_bias
+        )
+
+        self.is_fp32 = not (config.bf16 or config.fp16)
+        if (
+            self.use_flash_attn
+            and flash_attn_unpadded_func is not None
+            and not self.is_fp32
+        ):
+            self.core_attention_flash = FlashSelfAttention(
+                causal=True, attention_dropout=config.attn_dropout_prob
+            )
+        self.bf16 = config.bf16
+
+        self.use_dynamic_ntk = config.use_dynamic_ntk
+        self.use_logn_attn = config.use_logn_attn
+
+        logn_list = [
+            math.log(i, self.seq_length) if i > self.seq_length else 1
+            for i in range(1, 32768)
+        ]
+        self.logn_tensor = torch.tensor(logn_list)[None, :, None, None]
+
+        self.attn_dropout = nn.Dropout(config.attn_dropout_prob)
+
+    def _attn(self, query, key, value, registered_causal_mask, attention_mask=None, head_mask=None):
+        attn_weights = torch.matmul(query, key.transpose(-1, -2))
+
+        if self.scale_attn_weights:
+            attn_weights = attn_weights / torch.full(
+                [],
+                value.size(-1) ** 0.5,
+                dtype=attn_weights.dtype,
+                device=attn_weights.device,
+            )
+
+        query_length, key_length = query.size(-2), key.size(-2)
+        causal_mask = registered_causal_mask[
+            :, :, key_length - query_length : key_length, :key_length
+        ]
+        mask_value = torch.finfo(attn_weights.dtype).min
+        mask_value = torch.full([], mask_value, dtype=attn_weights.dtype).to(
+            attn_weights.device
+        )
+        attn_weights = torch.where(
+            causal_mask, attn_weights.to(attn_weights.dtype), mask_value
+        )
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+
+        attn_weights = attn_weights.type(value.dtype)
+        attn_weights = self.attn_dropout(attn_weights)
+
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        attn_output = torch.matmul(attn_weights, value)
+        attn_output = attn_output.transpose(1, 2)
+
+        return attn_output, attn_weights
+
+    def _upcast_and_reordered_attn(
+        self, query, key, value, registered_causal_mask, attention_mask=None, head_mask=None
+    ):
+        bsz, num_heads, q_seq_len, dk = query.size()
+        _, _, k_seq_len, _ = key.size()
+
+        attn_weights = torch.empty(
+            bsz * num_heads,
+            q_seq_len,
+            k_seq_len,
+            dtype=torch.float32,
+            device=query.device,
+        )
+
+        scale_factor = 1.0
+        if self.scale_attn_weights:
+            scale_factor /= float(value.size(-1)) ** 0.5
+
+        with autocast(enabled=False):
+            q, k = query.reshape(-1, q_seq_len, dk), key.transpose(-1, -2).reshape(
+                -1, dk, k_seq_len
+            )
+            attn_weights = torch.baddbmm(
+                attn_weights, q.float(), k.float(), beta=0, alpha=scale_factor
+            )
+            attn_weights = attn_weights.reshape(bsz, num_heads, q_seq_len, k_seq_len)
+
+        query_length, key_length = query.size(-2), key.size(-2)
+        causal_mask = registered_causal_mask[
+            :, :, key_length - query_length : key_length, :key_length
+        ]
+        mask_value = torch.finfo(attn_weights.dtype).min
+        mask_value = torch.tensor(mask_value, dtype=attn_weights.dtype).to(
+            attn_weights.device
+        )
+        attn_weights = torch.where(causal_mask, attn_weights, mask_value)
+
+        if attention_mask is not None:
+            attn_weights = attn_weights + attention_mask
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+
+        if attn_weights.dtype != torch.float32:
+            raise RuntimeError(
+                "Error with upcasting, attn_weights does not have dtype torch.float32"
+            )
+        attn_weights = attn_weights.type(value.dtype)
+        attn_weights = self.attn_dropout(attn_weights)
+
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        attn_output = torch.matmul(attn_weights, value)
+
+        return attn_output, attn_weights
+
+    def _split_heads(self, tensor, num_heads, attn_head_size):
+        new_shape = tensor.size()[:-1] + (num_heads, attn_head_size)
+        tensor = tensor.view(new_shape)
+        return tensor
+
+    def _merge_heads(self, tensor, num_heads, attn_head_size):
+        tensor = tensor.contiguous()
+        new_shape = tensor.size()[:-2] + (num_heads * attn_head_size,)
+        return tensor.view(new_shape)
+
+    def forward(
+        self,
+        hidden_states: Optional[Tuple[torch.FloatTensor]],
+        rotary_pos_emb: Optional[List[torch.Tensor]] = None,
+        registered_causal_mask: Optional[torch.Tensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+    ):
+
+        mixed_x_layer = self.c_attn(hidden_states)
+
+        query, key, value = mixed_x_layer.split(self.split_size, dim=2)
+
+        query = self._split_heads(query, self.num_heads, self.head_dim)
+        key = self._split_heads(key, self.num_heads, self.head_dim)
+        value = self._split_heads(value, self.num_heads, self.head_dim)
+
+        if rotary_pos_emb is not None:
+            cur_len = query.shape[1]
+            rotary_pos_emb = [i[:, -cur_len:, :, :] for i in rotary_pos_emb]
+            rotary_pos_emb = (rotary_pos_emb,) * 2
+            q_pos_emb, k_pos_emb = rotary_pos_emb
+            # Slice the pos emb for current inference
+            query = apply_rotary_pos_emb(query, q_pos_emb)
+            key = apply_rotary_pos_emb(key, k_pos_emb)
+
+        if layer_past is not None:
+            past_key, past_value = layer_past[0], layer_past[1]
+            key = torch.cat((past_key, key), dim=1)
+            value = torch.cat((past_value, value), dim=1)
+
+        if use_cache:
+            present = (key, value)
+        else:
+            present = None
+
+        if self.use_logn_attn and not self.training:
+            if self.logn_tensor.device != query.device or self.logn_tensor.dtype != query.dtype:
+                self.logn_tensor = self.logn_tensor.to(query.device).type_as(query)
+            seq_start = key.size(1) - query.size(1)
+            seq_end = key.size(1)
+            logn_tensor = self.logn_tensor[:, seq_start:seq_end, :, :]
+            query = query * logn_tensor.expand_as(query)
+
+        if (
+            self.use_flash_attn
+            and flash_attn_unpadded_func is not None
+            and not self.is_fp32
+            and query.is_cuda
+        ):
+            q, k, v = query, key, value
+            context_layer = self.core_attention_flash(q, k, v)
+
+            # b s h d -> b s (h d)
+            context_layer = context_layer.flatten(2,3).contiguous()
+
+        else:
+            query = query.permute(0, 2, 1, 3)
+            key = key.permute(0, 2, 1, 3)
+            value = value.permute(0, 2, 1, 3)
+            if (
+                registered_causal_mask is None
+                and self.use_flash_attn
+                and flash_attn_unpadded_func is not None
+                and not self.is_fp32
+                and not query.is_cuda
+            ):
+                raise Exception(_ERROR_INPUT_CPU_QUERY_WITH_FLASH_ATTN_ACTIVATED)
+            attn_output, attn_weight = self._attn(
+                query, key, value, registered_causal_mask, attention_mask, head_mask
+            )
+            context_layer = self._merge_heads(
+                attn_output, self.num_heads, self.head_dim
+            )
+
+        attn_output = self.c_proj(context_layer)
+
+        outputs = (attn_output, present)
+        if output_attentions:
+            if (
+                self.use_flash_attn
+                and flash_attn_unpadded_func is not None
+                and not self.is_fp32
+            ):
+                raise ValueError("Cannot output attentions while using flash-attn")
+            else:
+                outputs += (attn_weight,)
+
+        return outputs
+
+
+class QWenMLP(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.w1 = nn.Linear(
+            config.hidden_size, config.intermediate_size // 2, bias=not config.no_bias
+        )
+        self.w2 = nn.Linear(
+            config.hidden_size, config.intermediate_size // 2, bias=not config.no_bias
+        )
+        ff_dim_in = config.intermediate_size // 2
+        self.c_proj = nn.Linear(ff_dim_in, config.hidden_size, bias=not config.no_bias)
+
+    def forward(self, hidden_states):
+        a1 = self.w1(hidden_states)
+        a2 = self.w2(hidden_states)
+        intermediate_parallel = a1 * F.silu(a2)
+        output = self.c_proj(intermediate_parallel)
+        return output
+
+class QWenBlock(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        hidden_size = config.hidden_size
+        self.bf16 = config.bf16
+
+        self.ln_1 = RMSNorm(
+            hidden_size,
+            eps=config.layer_norm_epsilon,
+        )
+        self.attn = QWenAttention(config)
+        self.ln_2 = RMSNorm(
+            hidden_size,
+            eps=config.layer_norm_epsilon,
+        )
+
+        self.mlp = QWenMLP(config)
+
+    def forward(
+        self,
+        hidden_states: Optional[Tuple[torch.FloatTensor]],
+        rotary_pos_emb: Optional[List[torch.Tensor]] = None,
+        registered_causal_mask: Optional[torch.Tensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+    ):
+        layernorm_output = self.ln_1(hidden_states)
+
+        attn_outputs = self.attn(
+            layernorm_output,
+            rotary_pos_emb,
+            registered_causal_mask=registered_causal_mask,
+            layer_past=layer_past,
+            at