add config

ChatYuan-large-v2/bert4torch_config.json

@@ -0,0 +1,17 @@
+{
+    "model": "mt5.1.1",
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 1024,
+    "intermediate_size": 2816,
+    "num_attention_heads": 16,
+    "attention_head_size": 64,
+    "num_hidden_layers": 24,
+    "vocab_size": 32128,
+    "relative_attention_num_buckets": 32,
+    "attention_scale":  false,
+    "is_dropout": true,
+    "max_position_embeddings": 768,
+    "segment_vocab_size": 0,
+    "logit_scale": false
+  }

Erlangshen-DeBERTa-v2-320M-Chinese/bert4torch_config.json

@@ -0,0 +1,25 @@
+{
+    "model": "deberta_v2",
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 1024,
+    "initializer_range": 0.02,
+    "intermediate_size": 4096,
+    "max_position_embeddings": 512,
+    "relative_attention": true,
+    "position_buckets": 256,
+    "norm_rel_ebd": "layer_norm",
+    "share_att_key": true,
+    "pos_att_type": "c2p|p2c",
+    "conv_kernel_size": 3,
+    "conv_act": "gelu",
+    "layer_norm_eps": 1e-7,
+    "max_relative_positions": -1,
+    "position_biased_input": false,
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "type_vocab_size": 0,
+    "vocab_size": 12800,
+    "num_labels": 119
+}

Erlangshen-DeBERTa-v2-710M-Chinese/bert4torch_config.json

@@ -0,0 +1,32 @@
+{
+    "model": "deberta_v2",
+    "attention_probs_dropout_prob": 0.1,
+    "attention_head_size": 64,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 1536,
+    "initializer_range": 0.02,
+    "intermediate_size": 6144,
+    "max_position_embeddings": 512,
+    "relative_attention": true,
+    "position_buckets": 256,
+    "norm_rel_ebd": "layer_norm",
+    "share_att_key": true,
+    "pos_att_type": [
+        "p2c",
+        "c2p"
+    ],
+    "conv_kernel_size": 3,
+    "pooler_dropout": 0,
+    "pooler_hidden_act": "gelu",
+    "pooler_hidden_size": 1536,
+    "conv_act": "gelu",
+    "layer_norm_eps": 1e-7,
+    "max_relative_positions": -1,
+    "position_biased_input": false,
+    "num_attention_heads": 24,
+    "num_hidden_layers": 24,
+    "type_vocab_size": 0,
+    "num_labels": 119,
+    "vocab_size": 12800
+}

Erlangshen-DeBERTa-v2-97M-Chinese/bert4torch_config.json

@@ -0,0 +1,24 @@
+{
+    "model": "deberta_v2",
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 768,
+    "initializer_range": 0.02,
+    "intermediate_size": 3072,
+    "max_position_embeddings": 512,
+    "relative_attention": true,
+    "position_buckets": 256,
+    "norm_rel_ebd": "layer_norm",
+    "share_att_key": true,
+    "pos_att_type": "c2p|p2c",
+    "conv_kernel_size": 3,
+    "conv_act": "gelu",
+    "layer_norm_eps": 1e-7,
+    "max_relative_positions": -1,
+    "position_biased_input": false,
+    "num_attention_heads": 12,
+    "num_hidden_layers": 12,
+    "type_vocab_size": 0,
+    "vocab_size": 12800
+}

chatglm-6b-v0.1.0/bert4torch_config.json

@@ -0,0 +1,22 @@
+{
+    "model": "glm",
+    "hidden_act": "gelu_new", 
+    "bos_token_id": 130004,
+    "eos_token_id": 130005,
+    "mask_token_id": 130000,
+    "gmask_token_id": 130001,
+    "pad_token_id": 3,
+    "hidden_size": 4096,
+    "intermediate_size": 16384,
+    "layer_norm_eps": 1e-05,
+    "max_sequence_length": 2048,
+    "num_attention_heads": 32,
+    "num_hidden_layers": 28,
+    "position_encoding_2d": true,
+    "torch_dtype": "float16",
+    "vocab_size": 130528,
+    "segment_vocab_size": 0,
+    "skip_init": true,
+    "rope_rank": "updown",
+    "tie_emb_prj_weight": false
+}

chinese-electra-base-discriminator/bert4torch_config.json

@@ -0,0 +1,18 @@
+{
+  "model": "electra",
+  "attention_probs_dropout_prob": 0.1,
+  "directionality": "bidi",
+  "embedding_size": 768,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "type_vocab_size": 2,
+  "vocab_size": 21128
+}

ernie-1.0-base-zh/bert4torch_config.json

@@ -0,0 +1,16 @@
+{
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_act": "relu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 768,
+    "initializer_range": 0.02,
+    "max_position_embeddings": 513,
+    "num_attention_heads": 12,
+    "num_hidden_layers": 12,
+    "type_vocab_size": 2,
+    "vocab_size": 18000,
+    "pad_token_id": 0,
+    "layer_norm_eps": 1e-05,
+    "model": "ernie",
+    "intermediate_size": 3072
+}

ernie-3.0-base-zh/bert4torch_config.json

@@ -0,0 +1,17 @@
+{
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 768,
+    "initializer_range": 0.02,
+    "max_position_embeddings": 2048,
+    "num_attention_heads": 12,
+    "num_hidden_layers": 12,
+    "task_type_vocab_size": 3,
+    "type_vocab_size": 4,
+    "use_task_id": true,
+    "vocab_size": 40000,
+    "layer_norm_eps": 1e-05,
+    "model": "ernie",
+    "intermediate_size": 3072
+}

gpt2-ml_15g_corpus/README.md

@@ -0,0 +1,3 @@
+# gpt2-ml
+- 项目链接(tf版本)：https://github.com/imcaspar/gpt2-ml
+- pytorch权重转换和下载：https://github.com/ghosthamlet/gpt2-ml-torch

gpt2-ml_15g_corpus/bert4torch_config.json

@@ -0,0 +1,14 @@
+{
+    "model": "gpt2_ml", 
+    "segment_vocab_size": 0,
+    "vocab_size": 21130,
+    "hidden_size": 1536,
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_dropout_prob": 0.1,
+    "hidden_act": "gelu",
+    "initializer_range": 0.014142135623731,
+    "intermediate_size": 6144,
+    "max_position_embeddings": 1024,
+    "num_attention_heads": 24,
+    "num_hidden_layers": 48
+  }

gpt2-ml_30g_corpus/README.md

@@ -0,0 +1,3 @@
+# gpt2-ml
+- 项目链接(tf版本)：https://github.com/imcaspar/gpt2-ml
+- pytorch权重转换和下载：https://github.com/ghosthamlet/gpt2-ml-torch

gpt2-ml_30g_corpus/bert4torch_config.json

@@ -0,0 +1,14 @@
+{
+    "model": "gpt2_ml", 
+    "segment_vocab_size": 0,
+    "vocab_size": 21130,
+    "hidden_size": 1536,
+    "attention_probs_dropout_prob": 0.1,
+    "hidden_dropout_prob": 0.1,
+    "hidden_act": "gelu",
+    "initializer_range": 0.014142135623731,
+    "intermediate_size": 6144,
+    "max_position_embeddings": 1024,
+    "num_attention_heads": 24,
+    "num_hidden_layers": 48
+  }