Upload 19 files

LongWriter-Qwen+LongDPO/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

LongWriter-Qwen+LongDPO/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/models/LongWriter-Qwen2.5-7B-Instruct/",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.44.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

LongWriter-Qwen+LongDPO/generation_config.json

@@ -0,0 +1,12 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": 151645,
+  "max_new_tokens": 2048,
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.44.2"
+}

LongWriter-Qwen+LongDPO/pytorch_model-00001-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:640cd89818dd28c117fdb0395f7cae057a06eae3e4f81a263c883b74a8dab427
+size 1886428918

LongWriter-Qwen+LongDPO/pytorch_model-00002-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f10b3a4d51c594e3cc1993fec13776e84b14f78d51e620b39f1e7ed153fcf66e
+size 1864478730

LongWriter-Qwen+LongDPO/pytorch_model-00003-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c32d4ab3a5c163baee84c91917a5a378800d4fe2f81938718f1fdf5f3fde274f
+size 1864478730

LongWriter-Qwen+LongDPO/pytorch_model-00004-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62a3fc068bed22e255356f0bf16031012878b64d3c3347022ba923153e38f372
+size 1864478794

LongWriter-Qwen+LongDPO/pytorch_model-00005-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d8d96c7bed42b613f6f89f5f00f5472156218d4fb4fcf49fe7752932a7a0167
+size 1864478794

LongWriter-Qwen+LongDPO/pytorch_model-00006-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83a816a1ac998486fb55b1aedfdb1d85f01f98de86f7f8cee9463ab5b23f3c5c
+size 1864478794

LongWriter-Qwen+LongDPO/pytorch_model-00007-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf18ddc797fc09e3146f799f3bd74bf800ea473c1520149621643f8c8e4b7664
+size 1864478794

LongWriter-Qwen+LongDPO/pytorch_model-00008-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f56ef707d3366054f70a5185f769beb18ecf2356d8fa6d05a8680a38d027a41
+size 1068053090

LongWriter-Qwen+LongDPO/pytorch_model-00009-of-00009.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b98bf497ce8ad56412cafaeef3592d899613e1073a4ee96c8f57d18d6cb559d2
+size 1089996165

LongWriter-Qwen+LongDPO/pytorch_model.bin.index.json

@@ -0,0 +1,346 @@
+{
+  "metadata": {
+    "total_size": 15231233024
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00009-of-00009.bin",
+    "model.embed_tokens.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.k_proj.bias": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.q_proj.bias": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.v_proj.bias": "pytorch_model-00001-of-00009.bin",
+    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.input_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.1.mlp.down_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.1.self_attn.k_proj.bias": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.self_attn.q_proj.bias": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.self_attn.v_proj.bias": "pytorch_model-00001-of-00009.bin",
+    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00009.bin",
+    "model.layers.10.input_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.mlp.down_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.mlp.up_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.k_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.q_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.v_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.input_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.mlp.down_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.mlp.up_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.k_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.q_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.v_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.input_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.mlp.down_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.mlp.up_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.k_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.q_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.v_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.input_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.13.mlp.down_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.mlp.up_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.13.self_attn.k_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.self_attn.q_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.self_attn.v_proj.bias": "pytorch_model-00004-of-00009.bin",
+    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.14.input_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.mlp.down_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.mlp.up_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.k_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.q_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.v_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.input_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.mlp.down_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.mlp.up_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.k_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.q_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.v_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.input_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.mlp.down_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.mlp.up_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.k_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.q_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.v_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.input_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.17.mlp.down_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.mlp.up_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.17.self_attn.k_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.self_attn.q_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.self_attn.v_proj.bias": "pytorch_model-00005-of-00009.bin",
+    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00005-of-00009.bin",
+    "model.layers.18.input_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.mlp.down_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.mlp.up_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.k_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.q_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.v_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.input_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.mlp.down_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.mlp.up_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.k_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.q_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.v_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.2.input_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.mlp.down_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.mlp.up_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.k_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.q_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.v_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.20.input_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.mlp.down_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.mlp.up_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.k_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.q_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.v_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.input_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.21.mlp.down_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.mlp.up_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.21.self_attn.k_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.self_attn.q_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.self_attn.v_proj.bias": "pytorch_model-00006-of-00009.bin",
+    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00006-of-00009.bin",
+    "model.layers.22.input_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.mlp.down_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.mlp.up_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.k_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.q_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.v_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.input_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.mlp.down_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.mlp.up_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.k_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.q_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.v_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.input_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.mlp.down_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.mlp.up_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.k_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.q_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.v_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.input_layernorm.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.25.mlp.down_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.mlp.up_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.25.self_attn.k_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.self_attn.q_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.self_attn.v_proj.bias": "pytorch_model-00007-of-00009.bin",
+    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00007-of-00009.bin",
+    "model.layers.26.input_layernorm.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.mlp.down_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.mlp.up_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.k_proj.bias": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.q_proj.bias": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.v_proj.bias": "pytorch_model-00008-of-00009.bin",
+    "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.input_layernorm.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.mlp.down_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.mlp.up_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.k_proj.bias": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.q_proj.bias": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.v_proj.bias": "pytorch_model-00008-of-00009.bin",
+    "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00008-of-00009.bin",
+    "model.layers.3.input_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.mlp.down_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.mlp.up_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.k_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.q_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.v_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.input_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.mlp.down_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.mlp.up_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.k_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.q_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.v_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.input_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.5.mlp.down_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.mlp.up_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.5.self_attn.k_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.self_attn.q_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.self_attn.v_proj.bias": "pytorch_model-00002-of-00009.bin",
+    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00002-of-00009.bin",
+    "model.layers.6.input_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.mlp.down_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.mlp.up_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.k_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.q_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.v_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.input_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.mlp.down_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.mlp.up_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.k_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.q_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.v_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.input_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.mlp.down_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.mlp.up_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.k_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.q_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.v_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.input_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.9.mlp.down_proj.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.mlp.up_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00004-of-00009.bin",
+    "model.layers.9.self_attn.k_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.self_attn.q_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.self_attn.v_proj.bias": "pytorch_model-00003-of-00009.bin",
+    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00003-of-00009.bin",
+    "model.norm.weight": "pytorch_model-00008-of-00009.bin"
+  }
+}

LongWriter-Qwen+LongDPO/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

LongWriter-Qwen+LongDPO/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

LongWriter-Qwen+LongDPO/xtuner_config.py

@@ -0,0 +1,247 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import torch
+from datasets import load_dataset
+from mmengine.dataset import DefaultSampler
+from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
+                            LoggerHook, ParamSchedulerHook)
+from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
+from peft import LoraConfig
+from torch.optim import AdamW
+from transformers import (AutoModelForCausalLM, AutoTokenizer,
+                          BitsAndBytesConfig)
+
+from xtuner.dataset.collate_fns.preference_collate_fn import \
+    preference_collate_fn
+from xtuner.dataset.preference_dataset import (build_preference_dataset,
+                                               orpo_dpo_mix_40k_map_fn,ultrafeedback_dpo_map_fn,load_jsonl_dataset)
+from xtuner.engine.hooks import (DatasetInfoHook, EvaluateChatHook,
+                                 VarlenAttnArgsToMessageHubHook)
+from xtuner.engine.runner import TrainLoop
+from xtuner.model.dpo import DPO
+from xtuner.parallel.sequence import SequenceParallelSampler
+from xtuner.utils import PROMPT_TEMPLATE, SYSTEM_TEMPLATE
+#######################################################################
+#                          PART 1  Settings                           #
+#######################################################################
+# Model
+pretrained_model_name_or_path = '/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/models/LongWriter-Qwen2.5-7B-Instruct/'
+use_varlen_attn = True
+dpo_loss_type = 'sigmoid'  # One of ['sigmoid', 'hinge', 'ipo', 'kto_pair', 'sppo_hard', 'nca_pair', 'robust']  # noqa: E501
+loss_beta = 0.1
+label_smoothing = 0.0
+
+# Data
+# prompt_template = PROMPT_TEMPLATE.llama3_chat
+prompt_template = PROMPT_TEMPLATE.qwen_chat # llama2_chat
+max_length = 32768
+max_packed_length = max_length
+
+# parallel
+sequence_parallel_size = 2
+
+# Scheduler & Optimizer
+batch_size = 1  # per_device
+accumulative_counts = 1
+accumulative_counts *= sequence_parallel_size
+dataloader_num_workers = 0
+max_epochs = 2
+optim_type = AdamW
+lr = 1e-6  # refer to alignment handbook
+betas = (0.9, 0.999)
+weight_decay = 0
+max_norm = 1  # grad clip
+warmup_ratio = 0.1 # 0.03
+
+# Save
+save_steps = 100
+save_total_limit = -1  # Maximum checkpoints to keep (-1 means unlimited)
+
+# Evaluate the generation performance during the training
+evaluation_freq = 500
+SYSTEM = SYSTEM_TEMPLATE.alpaca
+evaluation_inputs = [
+    'What famous British author, known for his tales of mystery and the macabre, shares his initials with a common abbreviation for "rest in peace"?',  # noqa: E501
+    'Please tell me five scenic spots in Shanghai',
+    '890729 - 425663? Only respond with math and no words.'
+]
+
+#######################################################################
+#                      PART 2  Model & Tokenizer                      #
+#######################################################################
+tokenizer = dict(
+    type=AutoTokenizer.from_pretrained,
+    pretrained_model_name_or_path=pretrained_model_name_or_path,
+    trust_remote_code=True,
+    padding_side='right')
+
+model = dict(
+    type=DPO,
+    loss_type=dpo_loss_type,
+    use_varlen_attn=use_varlen_attn,
+    beta=loss_beta,
+    label_smoothing=label_smoothing,
+    llm=dict(
+        type=AutoModelForCausalLM.from_pretrained,
+        pretrained_model_name_or_path=pretrained_model_name_or_path,
+        trust_remote_code=True,
+        torch_dtype=torch.bfloat16,
+        ),
+    ref_llm=dict(  ##### initialization of ref_llm #######
+        type=AutoModelForCausalLM.from_pretrained,
+        pretrained_model_name_or_path=pretrained_model_name_or_path,
+        trust_remote_code=True,
+        torch_dtype=torch.bfloat16,
+        ),
+)
+    # llm=dict(
+    #     type=AutoModelForCausalLM.from_pretrained,
+    #     pretrained_model_name_or_path=pretrained_model_name_or_path,
+    #     trust_remote_code=True,
+    #     torch_dtype=torch.float16,
+    #     quantization_config=dict(
+    #         type=BitsAndBytesConfig,
+    #         load_in_4bit=True,
+    #         load_in_8bit=False,
+    #         llm_int8_threshold=6.0,
+    #         llm_int8_has_fp16_weight=False,
+    #         bnb_4bit_compute_dtype=torch.float16,
+    #         bnb_4bit_use_double_quant=True,
+    #         bnb_4bit_quant_type='nf4')),
+    # lora=dict(
+    #     type=LoraConfig,
+    #     r=64,
+    #     lora_alpha=16,
+    #     lora_dropout=0.1,
+    #     bias='none',
+    #     task_type='CAUSAL_LM'))
+#######################################################################
+#                      PART 3  Dataset & Dataloader                   #
+#######################################################################
+sampler = SequenceParallelSampler \
+    if sequence_parallel_size > 1 else DefaultSampler
+
+train_dataset = dict(
+    type=build_preference_dataset,
+    dataset=dict(type=load_jsonl_dataset,data_files=["/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/qwen_step_wise_fix_bug/merged/wild_chat_4000_32000.jsonl"]),
+    # dataset=dict(type=load_jsonl_dataset,data_files=["/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/qwen_step_wise_fix_bug/merged/wild_chat_500_2000_res.jsonl","/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/ultrafeedback_binarized.jsonl","/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/qwen_step_wise_fix_bug/merged/wild_chat_4000_32000.jsonl"]),
+    # dataset=dict(type=load_jsonl_dataset,data_files=["/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/qwen_step_wise_fix_bug/range_2_4k_res_filtered.jsonl","/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/ultrafeedback_binarized.jsonl","/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/qwen_step_wise_fix_bug/range_4_16k_res_filtered.jsonl","/mnt/gemininjceph2/geminicephfs/pr-others-prctrans/pingbowen/workspace/MCTS_DPO/MCTS-dpo/data/qwen_step_wise_fix_bug/range_16_32k_res.jsonl"]),
+    # dataset=dict(type=load_dataset, path='llamafactory/ultrafeedback_binarized'), # mlabonne/orpo-dpo-mix-40k
+    tokenizer=tokenizer,
+    max_length=max_length,
+    dataset_map_fn=ultrafeedback_dpo_map_fn,
+    is_dpo=True,
+    is_reward=False,
+    reward_token_id=-1,
+    num_proc=32,
+    use_varlen_attn=use_varlen_attn,
+    max_packed_length=max_packed_length,
+    shuffle_before_pack=True,
+)
+
+train_dataloader = dict(
+    batch_size=batch_size,
+    num_workers=dataloader_num_workers,
+    dataset=train_dataset,
+    sampler=dict(type=sampler, shuffle=True),
+    collate_fn=dict(
+        type=preference_collate_fn, use_varlen_attn=use_varlen_attn))
+
+#######################################################################
+#                    PART 4  Scheduler & Optimizer                    #
+#######################################################################
+# optimizer
+optim_wrapper = dict(
+    type=AmpOptimWrapper,
+    optimizer=dict(
+        type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
+    clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
+    accumulative_counts=accumulative_counts,
+    loss_scale='dynamic',
+    dtype='bfloat16')
+
+# learning policy
+# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md  # noqa: E501
+param_scheduler = [
+    dict(
+        type=LinearLR,
+        start_factor=1e-5,
+        by_epoch=True,
+        begin=0,
+        end=warmup_ratio * max_epochs,
+        convert_to_iter_based=True),
+    dict(
+        type=CosineAnnealingLR,
+        eta_min=0.0,
+        by_epoch=True,
+        begin=warmup_ratio * max_epochs,
+        end=max_epochs,
+        convert_to_iter_based=True)
+]
+
+# train, val, test setting
+train_cfg = dict(type=TrainLoop, max_epochs=max_epochs)
+
+#######################################################################
+#                           PART 5  Runtime                           #
+#######################################################################
+# Log the dialogue periodically during the training process, optional
+
+custom_hooks = [
+    dict(type=DatasetInfoHook, tokenizer=tokenizer),
+    # dict(
+    #     type=EvaluateChatHook,
+    #     tokenizer=tokenizer,
+    #     every_n_iters=evaluation_freq,
+    #     evaluation_inputs=evaluation_inputs,
+    #     system=SYSTEM,
+    #     prompt_template=prompt_template)
+]
+
+if use_varlen_attn:
+    custom_hooks += [dict(type=VarlenAttnArgsToMessageHubHook)]
+
+# configure default hooks
+default_hooks = dict(
+    # record the time of every iteration.
+    timer=dict(type=IterTimerHook),
+    # print log every 10 iterations.
+    logger=dict(type=LoggerHook, log_metric_by_epoch=False, interval=10),
+    # enable the parameter scheduler.
+    param_scheduler=dict(type=ParamSchedulerHook),
+    # save checkpoint per `save_steps`.
+    checkpoint=dict(
+        type=CheckpointHook,
+        by_epoch=False,
+        interval=save_steps,
+        max_keep_ckpts=save_total_limit),
+    # set sampler seed in distributed evrionment.
+    sampler_seed=dict(type=DistSamplerSeedHook),
+)
+
+# configure environment
+env_cfg = dict(
+    # whether to enable cudnn benchmark
+    cudnn_benchmark=False,
+    # set multi process parameters
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
+    # set distributed parameters
+    dist_cfg=dict(backend='nccl'),
+)
+
+# set visualizer
+visualizer = None
+
+# set log level
+log_level = 'INFO'
+
+# load from which checkpoint
+load_from = None
+
+# whether to resume training from the loaded checkpoint
+resume = False
+
+# Defaults to use random seed and disable `deterministic`
+randomness = dict(seed=42, deterministic=False)
+
+# set log processor
+log_processor = dict(by_epoch=False)