Upload BloomForCausalLM

config.json

@@ -0,0 +1,32 @@
+{
+  "_name_or_path": "/mnt/nfs/algo/wlm/tigerbot-7b-sft-vanilla-simple-11600-1",
+  "apply_residual_connection_post_layernorm": false,
+  "architectures": [
+    "BloomForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "attention_softmax_in_fp32": true,
+  "bias_dropout_fusion": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_dropout": 0.0,
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "masked_softmax_fusion": true,
+  "model_type": "bloom",
+  "n_head": 32,
+  "n_inner": null,
+  "n_layer": 30,
+  "offset_alibi": 100,
+  "pad_token_id": 3,
+  "pretraining_tp": 1,
+  "skip_bias_add": true,
+  "skip_bias_add_qkv": false,
+  "slow_but_exact": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.1",
+  "unk_token_id": 0,
+  "use_cache": true,
+  "vocab_size": 250682
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 3,
+  "transformers_version": "4.28.1"
+}

pytorch_model-00001-of-00004.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49e19e673b0e2d014b547d5cdc783e0cf88bbe05f5648ad0a917c749cb0bbb4f
+size 9947282397

pytorch_model-00002-of-00004.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5ecba90e22e7fc87104098cb1e25b4c1fb04ebb0dba19972307c03b40cddd0f
+size 9733441515

pytorch_model-00003-of-00004.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0a5fac8b0674a5f1904ac8d6855b396525c2a1f00c39728e57603a2bebea810
+size 8592222589

pytorch_model-00004-of-00004.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfee01d09b48d739618a4cd3568c44a7d3155c605ac1a929b001e3aa15b1149c
+size 4107174826

pytorch_model.bin.index.json

@@ -0,0 +1,373 @@
+{
+  "metadata": {
+    "total_size": 32379994112
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00004-of-00004.bin",
+    "transformer.h.0.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.0.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.1.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.10.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.10.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.11.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.12.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.13.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.14.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.15.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.16.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.17.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.18.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.19.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.19.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.19.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.19.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.19.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.2.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.2.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.20.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.20.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.21.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.22.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.23.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.24.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.25.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.26.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.27.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.28.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.input_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.input_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.mlp.dense_4h_to_h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.mlp.dense_4h_to_h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.mlp.dense_h_to_4h.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.mlp.dense_h_to_4h.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.post_attention_layernorm.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.post_attention_layernorm.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.self_attention.dense.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.self_attention.dense.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.self_attention.query_key_value.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.29.self_attention.query_key_value.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.h.3.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.3.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.4.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.5.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.post_attention_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.self_attention.dense.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.self_attention.dense.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.6.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.7.input_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.7.input_layernorm.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.7.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.7.self_attention.query_key_value.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.7.self_attention.query_key_value.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.h.8.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.8.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.input_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.input_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.post_attention_layernorm.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.post_attention_layernorm.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.self_attention.dense.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.self_attention.dense.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.self_attention.query_key_value.bias": "pytorch_model-00002-of-00004.bin",
+    "transformer.h.9.self_attention.query_key_value.weight": "pytorch_model-00002-of-00004.bin",
+    "transformer.ln_f.bias": "pytorch_model-00003-of-00004.bin",
+    "transformer.ln_f.weight": "pytorch_model-00003-of-00004.bin",
+    "transformer.word_embeddings.weight": "pytorch_model-00001-of-00004.bin",
+    "transformer.word_embeddings_layernorm.bias": "pytorch_model-00001-of-00004.bin",
+    "transformer.word_embeddings_layernorm.weight": "pytorch_model-00001-of-00004.bin"
+  }
+}