Add fp16 model, converted to safetensors

config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "EleutherAI/polyglot-ko-12.8b",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "classifier_dropout": 0.1,
+  "eos_token_id": 0,
+  "hidden_act": "gelu",
+  "hidden_size": 5120,
+  "initializer_range": 0.02,
+  "intermediate_size": 20480,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 40,
+  "num_hidden_layers": 40,
+  "num_steps": "global_step301000",
+  "rotary_emb_base": 10000,
+  "rotary_pct": 0.5,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.29.2",
+  "use_cache": true,
+  "use_parallel_residual": true,
+  "vocab_size": 30080
+}

generation_config.json

@@ -0,0 +1 @@
+{"_from_model_config": true, "bos_token_id": 0, "eos_token_id": 2, "transformers_version": "4.29.2"}

model-00001-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1077f2f2f2d1a4adda1ba69524723a3343e1c9af0c9d839e0818487700964a3
+size 945730380

model-00002-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf66e85dfaa55d13d00c37bc84acde2290a387c1345290c18dbd2d34dd711d03
+size 843231522

model-00003-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4725b4a4ca9469b1063fc393b0350ae368694a2c24a852a9b1ed44d98e2c72cf
+size 843231290

model-00004-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b46da00378d9570cd8969063f680b6db2c2a045f778776652446c75258a6261
+size 1004754108

model-00005-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:947d464386804caf29f0b5e05cdc583eae20d44145ecd188405e0e0925d52291
+size 895670546

model-00006-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11407a250b137bba535c36858f8a449532d364649bf526e8d5bbb312a1e3dac2
+size 1004754108

model-00007-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c136a2a37e220233430fd98ef08f1a6b065c7de168be74a7cd33e5070276ce9
+size 895670546

model-00008-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f134b18eaafcd76b161903985b5c0ab7ea1e9fd00fbbb5428a4c08975a8b5920
+size 1004754132

model-00009-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b62e7815e79e7408bb0a675690941e528498da56f044393100d44b1d361623bf
+size 895670570

model-00010-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f151e74235be0f549323cba93a6003556b7e13d8480e27626d7db540b9db6bc6
+size 1004754140

model-00011-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9c3d6eec5f29e8e1939d4ab754341023d25692d816185be3fbac9b9901d97ae
+size 895670570

model-00012-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdfd157f5e29d53c0ed3ecad7ce30721b2dafde9c2092328facc5df5eff5fde3
+size 1004754140

model-00013-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd724af69f86a500164cc1780478b7104afe3910ecc8c95cd516f550397f1c55
+size 895670570

model-00014-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11b110a8f23c20c659195f1b146c1b82e4717ea4955ea242a4318be3e7ae2025
+size 1004754140

model-00015-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8323b729fb2df1177e357f7ecd96b25b142f0d024e920d21cf873fd6021326c1
+size 895670570

model-00016-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08ee1e2e803c1272e60f3eae7cccf9c68647608d52b79315ee52edf8a059b945
+size 1004754140

model-00017-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cb2de4ff8946b29a9776a2e0360f81d9cf02d6efbae6db26496d62cd01d87be
+size 895670570

model-00018-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba8337215210384ad9ac82b3efd260bc6c4c6a5fe1e51f4ba76f293e8844be1c
+size 1004754140

model-00019-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:027f7e1ced944fb2180a95cb42cbe5fc9ae98537c7d2500166688f15526307a7
+size 895670570

model-00020-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cac26f449546e6926433479c6acae25731adc9cb741f888050e5fc9562c28d1c
+size 1004754140

model-00021-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5bf23c027ad7d34408b03592bf1a6d4719b557ca7646096fe3e0523fda00e56
+size 895670570

model-00022-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53a5e1e407bf5a1c7dcfe71856b05a9eb7146a46012254c71a41c09de03d599c
+size 1004754140

model-00023-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41993efaf3f39c8f244ce8fba2b199a9f3a46b524619dfe3c940a5211f4581ea
+size 895670570

model-00024-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89b41c91e41e3c4bee9d8a1474381e3e78c388fe889cfba2640f6f5265104164
+size 1004754140

model-00025-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6074ebecce3abe57fa02f8f71c389945095831bc2795ce5e494f1be7a89fd35f
+size 895670570

model-00026-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec350b1755ef7ac78bc353115e2eb15cb2e358a5505ccfedd9f2277cb4776548
+size 1004754140

model-00027-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91a9f7671a47acdfaa6a17f7492271a7b4339b7344aaed47429ade0bdf7f6e15
+size 895670570

model-00028-of-00028.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7ea0188f126c4cd893bfdf40fbb5fb222da030f321ce595dfcf855887d72ead
+size 517765688

model.safetensors.index.json

@@ -0,0 +1,611 @@
+{
+  "metadata": {
+    "total_size": 25808184400.0
+  },
+  "weight_map": {
+    "embed_out.weight": "model-00028-of-00028.safetensors",
+    "gpt_neox.embed_in.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.final_layer_norm.bias": "model-00028-of-00028.safetensors",
+    "gpt_neox.final_layer_norm.weight": "model-00028-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.dense.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.dense.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.masked_bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.query_key_value.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.query_key_value.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.attention.rotary_emb.inv_freq": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.input_layernorm.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.input_layernorm.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.mlp.dense_4h_to_h.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.mlp.dense_4h_to_h.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.mlp.dense_h_to_4h.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.mlp.dense_h_to_4h.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.post_attention_layernorm.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.0.post_attention_layernorm.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.dense.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.dense.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.masked_bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.query_key_value.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.query_key_value.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.attention.rotary_emb.inv_freq": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.input_layernorm.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.input_layernorm.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.mlp.dense_4h_to_h.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.mlp.dense_4h_to_h.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.mlp.dense_h_to_4h.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.mlp.dense_h_to_4h.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.1.post_attention_layernorm.bias": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.1.post_attention_layernorm.weight": "model-00001-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.dense.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.dense.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.masked_bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.query_key_value.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.query_key_value.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.attention.rotary_emb.inv_freq": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.input_layernorm.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.input_layernorm.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.mlp.dense_4h_to_h.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.mlp.dense_4h_to_h.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.mlp.dense_h_to_4h.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.mlp.dense_h_to_4h.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.post_attention_layernorm.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.10.post_attention_layernorm.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.dense.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.dense.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.masked_bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.query_key_value.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.query_key_value.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.attention.rotary_emb.inv_freq": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.input_layernorm.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.input_layernorm.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.mlp.dense_4h_to_h.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.11.mlp.dense_4h_to_h.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.11.mlp.dense_h_to_4h.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.11.mlp.dense_h_to_4h.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.11.post_attention_layernorm.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.11.post_attention_layernorm.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.dense.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.dense.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.masked_bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.query_key_value.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.query_key_value.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.attention.rotary_emb.inv_freq": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.input_layernorm.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.input_layernorm.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.mlp.dense_4h_to_h.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.12.mlp.dense_4h_to_h.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.12.mlp.dense_h_to_4h.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.mlp.dense_h_to_4h.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.post_attention_layernorm.bias": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.12.post_attention_layernorm.weight": "model-00009-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.dense.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.dense.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.masked_bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.query_key_value.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.query_key_value.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.attention.rotary_emb.inv_freq": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.input_layernorm.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.input_layernorm.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.mlp.dense_4h_to_h.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.mlp.dense_4h_to_h.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.mlp.dense_h_to_4h.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.mlp.dense_h_to_4h.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.post_attention_layernorm.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.13.post_attention_layernorm.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.dense.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.dense.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.masked_bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.query_key_value.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.query_key_value.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.attention.rotary_emb.inv_freq": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.input_layernorm.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.input_layernorm.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.mlp.dense_4h_to_h.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.14.mlp.dense_4h_to_h.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.14.mlp.dense_h_to_4h.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.14.mlp.dense_h_to_4h.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.14.post_attention_layernorm.bias": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.14.post_attention_layernorm.weight": "model-00010-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.dense.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.dense.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.masked_bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.query_key_value.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.query_key_value.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.attention.rotary_emb.inv_freq": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.input_layernorm.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.input_layernorm.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.mlp.dense_4h_to_h.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.15.mlp.dense_4h_to_h.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.15.mlp.dense_h_to_4h.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.mlp.dense_h_to_4h.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.post_attention_layernorm.bias": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.15.post_attention_layernorm.weight": "model-00011-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.dense.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.dense.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.masked_bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.query_key_value.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.query_key_value.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.attention.rotary_emb.inv_freq": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.input_layernorm.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.input_layernorm.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.mlp.dense_4h_to_h.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.mlp.dense_4h_to_h.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.mlp.dense_h_to_4h.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.mlp.dense_h_to_4h.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.post_attention_layernorm.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.16.post_attention_layernorm.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.dense.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.dense.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.masked_bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.query_key_value.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.query_key_value.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.attention.rotary_emb.inv_freq": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.input_layernorm.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.input_layernorm.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.mlp.dense_4h_to_h.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.17.mlp.dense_4h_to_h.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.17.mlp.dense_h_to_4h.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.17.mlp.dense_h_to_4h.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.17.post_attention_layernorm.bias": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.17.post_attention_layernorm.weight": "model-00012-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.dense.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.dense.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.masked_bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.query_key_value.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.query_key_value.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.attention.rotary_emb.inv_freq": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.input_layernorm.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.input_layernorm.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.mlp.dense_4h_to_h.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.18.mlp.dense_4h_to_h.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.18.mlp.dense_h_to_4h.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.mlp.dense_h_to_4h.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.post_attention_layernorm.bias": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.18.post_attention_layernorm.weight": "model-00013-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.dense.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.dense.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.masked_bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.query_key_value.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.query_key_value.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.attention.rotary_emb.inv_freq": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.input_layernorm.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.input_layernorm.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.mlp.dense_4h_to_h.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.mlp.dense_4h_to_h.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.mlp.dense_h_to_4h.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.mlp.dense_h_to_4h.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.post_attention_layernorm.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.19.post_attention_layernorm.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.dense.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.dense.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.masked_bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.query_key_value.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.query_key_value.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.attention.rotary_emb.inv_freq": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.input_layernorm.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.input_layernorm.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.mlp.dense_4h_to_h.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.2.mlp.dense_4h_to_h.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.2.mlp.dense_h_to_4h.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.2.mlp.dense_h_to_4h.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.2.post_attention_layernorm.bias": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.2.post_attention_layernorm.weight": "model-00002-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.dense.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.dense.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.masked_bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.query_key_value.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.query_key_value.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.attention.rotary_emb.inv_freq": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.input_layernorm.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.input_layernorm.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.mlp.dense_4h_to_h.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.20.mlp.dense_4h_to_h.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.20.mlp.dense_h_to_4h.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.20.mlp.dense_h_to_4h.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.20.post_attention_layernorm.bias": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.20.post_attention_layernorm.weight": "model-00014-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.dense.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.dense.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.masked_bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.query_key_value.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.query_key_value.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.attention.rotary_emb.inv_freq": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.input_layernorm.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.input_layernorm.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.mlp.dense_4h_to_h.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.21.mlp.dense_4h_to_h.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.21.mlp.dense_h_to_4h.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.mlp.dense_h_to_4h.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.post_attention_layernorm.bias": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.21.post_attention_layernorm.weight": "model-00015-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.dense.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.dense.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.masked_bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.query_key_value.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.query_key_value.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.attention.rotary_emb.inv_freq": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.input_layernorm.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.input_layernorm.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.mlp.dense_4h_to_h.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.mlp.dense_4h_to_h.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.mlp.dense_h_to_4h.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.mlp.dense_h_to_4h.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.post_attention_layernorm.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.22.post_attention_layernorm.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.dense.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.dense.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.masked_bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.query_key_value.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.query_key_value.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.attention.rotary_emb.inv_freq": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.input_layernorm.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.input_layernorm.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.mlp.dense_4h_to_h.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.23.mlp.dense_4h_to_h.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.23.mlp.dense_h_to_4h.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.23.mlp.dense_h_to_4h.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.23.post_attention_layernorm.bias": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.23.post_attention_layernorm.weight": "model-00016-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.dense.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.dense.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.masked_bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.query_key_value.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.query_key_value.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.attention.rotary_emb.inv_freq": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.input_layernorm.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.input_layernorm.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.mlp.dense_4h_to_h.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.24.mlp.dense_4h_to_h.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.24.mlp.dense_h_to_4h.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.mlp.dense_h_to_4h.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.post_attention_layernorm.bias": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.24.post_attention_layernorm.weight": "model-00017-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.dense.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.dense.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.masked_bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.query_key_value.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.query_key_value.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.attention.rotary_emb.inv_freq": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.input_layernorm.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.input_layernorm.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.mlp.dense_4h_to_h.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.mlp.dense_4h_to_h.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.mlp.dense_h_to_4h.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.mlp.dense_h_to_4h.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.post_attention_layernorm.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.25.post_attention_layernorm.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.dense.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.dense.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.masked_bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.query_key_value.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.query_key_value.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.attention.rotary_emb.inv_freq": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.input_layernorm.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.input_layernorm.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.mlp.dense_4h_to_h.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.26.mlp.dense_4h_to_h.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.26.mlp.dense_h_to_4h.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.26.mlp.dense_h_to_4h.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.26.post_attention_layernorm.bias": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.26.post_attention_layernorm.weight": "model-00018-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.dense.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.dense.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.masked_bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.query_key_value.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.query_key_value.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.attention.rotary_emb.inv_freq": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.input_layernorm.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.input_layernorm.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.mlp.dense_4h_to_h.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.27.mlp.dense_4h_to_h.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.27.mlp.dense_h_to_4h.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.mlp.dense_h_to_4h.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.post_attention_layernorm.bias": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.27.post_attention_layernorm.weight": "model-00019-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.dense.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.dense.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.masked_bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.query_key_value.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.query_key_value.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.attention.rotary_emb.inv_freq": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.input_layernorm.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.input_layernorm.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.mlp.dense_4h_to_h.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.mlp.dense_4h_to_h.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.mlp.dense_h_to_4h.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.mlp.dense_h_to_4h.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.post_attention_layernorm.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.28.post_attention_layernorm.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.dense.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.dense.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.masked_bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.query_key_value.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.query_key_value.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.attention.rotary_emb.inv_freq": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.input_layernorm.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.input_layernorm.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.mlp.dense_4h_to_h.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.29.mlp.dense_4h_to_h.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.29.mlp.dense_h_to_4h.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.29.mlp.dense_h_to_4h.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.29.post_attention_layernorm.bias": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.29.post_attention_layernorm.weight": "model-00020-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.dense.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.dense.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.masked_bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.query_key_value.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.query_key_value.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.attention.rotary_emb.inv_freq": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.input_layernorm.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.input_layernorm.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.mlp.dense_4h_to_h.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.3.mlp.dense_4h_to_h.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.3.mlp.dense_h_to_4h.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.mlp.dense_h_to_4h.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.post_attention_layernorm.bias": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.3.post_attention_layernorm.weight": "model-00003-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.dense.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.dense.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.masked_bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.query_key_value.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.query_key_value.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.attention.rotary_emb.inv_freq": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.input_layernorm.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.input_layernorm.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.mlp.dense_4h_to_h.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.30.mlp.dense_4h_to_h.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.30.mlp.dense_h_to_4h.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.mlp.dense_h_to_4h.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.post_attention_layernorm.bias": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.30.post_attention_layernorm.weight": "model-00021-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.dense.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.dense.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.masked_bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.query_key_value.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.query_key_value.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.attention.rotary_emb.inv_freq": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.input_layernorm.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.input_layernorm.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.mlp.dense_4h_to_h.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.mlp.dense_4h_to_h.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.mlp.dense_h_to_4h.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.mlp.dense_h_to_4h.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.post_attention_layernorm.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.31.post_attention_layernorm.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.dense.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.dense.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.masked_bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.query_key_value.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.query_key_value.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.attention.rotary_emb.inv_freq": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.input_layernorm.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.input_layernorm.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.mlp.dense_4h_to_h.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.32.mlp.dense_4h_to_h.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.32.mlp.dense_h_to_4h.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.32.mlp.dense_h_to_4h.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.32.post_attention_layernorm.bias": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.32.post_attention_layernorm.weight": "model-00022-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.dense.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.dense.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.masked_bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.query_key_value.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.query_key_value.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.attention.rotary_emb.inv_freq": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.input_layernorm.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.input_layernorm.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.mlp.dense_4h_to_h.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.33.mlp.dense_4h_to_h.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.33.mlp.dense_h_to_4h.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.mlp.dense_h_to_4h.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.post_attention_layernorm.bias": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.33.post_attention_layernorm.weight": "model-00023-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.dense.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.dense.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.masked_bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.query_key_value.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.query_key_value.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.attention.rotary_emb.inv_freq": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.input_layernorm.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.input_layernorm.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.mlp.dense_4h_to_h.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.mlp.dense_4h_to_h.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.mlp.dense_h_to_4h.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.mlp.dense_h_to_4h.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.post_attention_layernorm.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.34.post_attention_layernorm.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.dense.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.dense.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.masked_bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.query_key_value.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.query_key_value.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.attention.rotary_emb.inv_freq": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.input_layernorm.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.input_layernorm.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.mlp.dense_4h_to_h.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.35.mlp.dense_4h_to_h.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.35.mlp.dense_h_to_4h.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.35.mlp.dense_h_to_4h.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.35.post_attention_layernorm.bias": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.35.post_attention_layernorm.weight": "model-00024-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.dense.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.dense.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.masked_bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.query_key_value.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.query_key_value.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.attention.rotary_emb.inv_freq": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.input_layernorm.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.input_layernorm.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.mlp.dense_4h_to_h.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.36.mlp.dense_4h_to_h.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.36.mlp.dense_h_to_4h.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.mlp.dense_h_to_4h.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.post_attention_layernorm.bias": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.36.post_attention_layernorm.weight": "model-00025-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.dense.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.dense.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.masked_bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.query_key_value.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.query_key_value.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.attention.rotary_emb.inv_freq": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.input_layernorm.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.input_layernorm.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.mlp.dense_4h_to_h.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.mlp.dense_4h_to_h.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.mlp.dense_h_to_4h.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.mlp.dense_h_to_4h.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.post_attention_layernorm.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.37.post_attention_layernorm.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.dense.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.dense.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.masked_bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.query_key_value.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.query_key_value.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.attention.rotary_emb.inv_freq": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.input_layernorm.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.input_layernorm.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.mlp.dense_4h_to_h.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.38.mlp.dense_4h_to_h.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.38.mlp.dense_h_to_4h.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.38.mlp.dense_h_to_4h.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.38.post_attention_layernorm.bias": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.38.post_attention_layernorm.weight": "model-00026-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.dense.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.dense.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.masked_bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.query_key_value.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.query_key_value.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.attention.rotary_emb.inv_freq": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.input_layernorm.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.input_layernorm.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.mlp.dense_4h_to_h.bias": "model-00028-of-00028.safetensors",
+    "gpt_neox.layers.39.mlp.dense_4h_to_h.weight": "model-00028-of-00028.safetensors",
+    "gpt_neox.layers.39.mlp.dense_h_to_4h.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.mlp.dense_h_to_4h.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.post_attention_layernorm.bias": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.39.post_attention_layernorm.weight": "model-00027-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.dense.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.dense.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.masked_bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.query_key_value.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.query_key_value.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.attention.rotary_emb.inv_freq": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.input_layernorm.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.input_layernorm.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.mlp.dense_4h_to_h.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.mlp.dense_4h_to_h.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.mlp.dense_h_to_4h.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.mlp.dense_h_to_4h.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.post_attention_layernorm.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.4.post_attention_layernorm.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.dense.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.dense.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.masked_bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.query_key_value.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.query_key_value.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.attention.rotary_emb.inv_freq": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.input_layernorm.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.input_layernorm.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.mlp.dense_4h_to_h.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.5.mlp.dense_4h_to_h.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.5.mlp.dense_h_to_4h.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.5.mlp.dense_h_to_4h.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.5.post_attention_layernorm.bias": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.5.post_attention_layernorm.weight": "model-00004-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.dense.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.dense.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.masked_bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.query_key_value.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.query_key_value.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.attention.rotary_emb.inv_freq": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.input_layernorm.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.input_layernorm.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.mlp.dense_4h_to_h.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.6.mlp.dense_4h_to_h.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.6.mlp.dense_h_to_4h.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.mlp.dense_h_to_4h.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.post_attention_layernorm.bias": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.6.post_attention_layernorm.weight": "model-00005-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.dense.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.dense.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.masked_bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.query_key_value.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.query_key_value.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.attention.rotary_emb.inv_freq": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.input_layernorm.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.input_layernorm.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.mlp.dense_4h_to_h.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.mlp.dense_4h_to_h.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.mlp.dense_h_to_4h.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.mlp.dense_h_to_4h.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.post_attention_layernorm.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.7.post_attention_layernorm.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.dense.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.dense.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.masked_bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.query_key_value.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.query_key_value.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.attention.rotary_emb.inv_freq": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.input_layernorm.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.input_layernorm.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.mlp.dense_4h_to_h.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.8.mlp.dense_4h_to_h.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.8.mlp.dense_h_to_4h.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.8.mlp.dense_h_to_4h.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.8.post_attention_layernorm.bias": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.8.post_attention_layernorm.weight": "model-00006-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.dense.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.dense.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.masked_bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.query_key_value.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.query_key_value.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.attention.rotary_emb.inv_freq": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.input_layernorm.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.input_layernorm.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.mlp.dense_4h_to_h.bias": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.9.mlp.dense_4h_to_h.weight": "model-00008-of-00028.safetensors",
+    "gpt_neox.layers.9.mlp.dense_h_to_4h.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.mlp.dense_h_to_4h.weight": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.post_attention_layernorm.bias": "model-00007-of-00028.safetensors",
+    "gpt_neox.layers.9.post_attention_layernorm.weight": "model-00007-of-00028.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,11 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<|sep|>",
+    "<|acc|>",
+    "<|tel|>",
+    "<|rrn|>"
+  ],
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>"
+}

tokenizer_config.json

@@ -0,0 +1,7 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}